ICGSCE2014

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/300489393
Solvent Extraction of Castor Beans Oil: Experimental Optimization via

Response Surface Methodology
Chapter · June 2015

DOI: 10.1007/978-981-287-505-1_12
CITATION READS
1 1,982
3 authors:
Jibrin Mohammed Danlami Agus Bin Arsad

Universiti Teknologi Malaysia Universiti Teknologi Malaysia
11 PUBLICATIONS 233 CITATIONS 98 PUBLICATIONS 1,064 CITATIONS
SEE PROFILE SEE PROFILE
Muhammad Abbas Ahmad Zaini

Universiti Teknologi Malaysia
221 PUBLICATIONS 3,087 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Original Article View project
Investigation on Silver Nanoparticles in Water Environment View project
All content following this page was uploaded by Muhammad Abbas Ahmad Zaini on 15 March 2019.
The user has requested enhancement of the downloaded file.

Md Amin Hashim Editor
ICGSCE
2014
Proceedings of the International
Conference on Global Sustainability and
Chemical Engineering
ICGSCE 2014
[email protected]
Md Amin Hashim
Editor
ICGSCE 2014
Proceedings of the International Conference
on Global Sustainability and Chemical
Engineering
123
[email protected]
Editor
Md Amin Hashim
Faculty of Chemical Engineering
Universiti Teknologi MARA (UiTM)
Selangor
Malaysia
ISBN 978-981-287-504-4 ISBN 978-981-287-505-1 (eBook)

DOI 10.1007/978-981-287-505-1
Library of Congress Control Number: 2015939164
Springer Singapore Heidelberg New York Dordrecht London

© Springer Science+Business Media Singapore 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.
Printed on acid-free paper
Springer Science+Business Media Singapore Pte Ltd. is part of Springer Science+Business Media
(www.springer.com)
[email protected]
Preface
On behalf of the editorial members, I would like to convey my heartiest congrat-

ulations and thank you to all authors who had successfully completed their original
work piece, participated and hence contributed to the conference. The conference
organisers also would like to acknowledge and appreciate all internal and external
reviewers who had spent their precious time in reading and suggesting improve-
ments on the selected manuscripts. Selected manuscripts were broadly inclined to
the conference theme which was “Synergising Sustainable Technologies for Future
Generation”, and again these subjects are either directly or indirectly related to the
general chemical engineering subjects.
Several selected and edited papers will be published in the “ICGSCE 2014”
conference proceedings with approval from established publisher, such as Springer.
Other selected and edited paper will be published in the Malaysian Journal of
Analytical Sciences, and in the Materials Science and Engineering Journal, pub-
lished by Transtech publications. Finally, and again, on behalf of editors, I would
like to thanks the conference organisers for their trust and confidence in the editorial
committee. The acknowledgment is also due to the Research and Management
Institute (RMI), UiTM for their valuable advice on publication matters.
Md Amin Hashim
[email protected]
List of Reviewers
Abdul Aziz Ishak, Universiti Teknologi MARA

Amin Azdarpour, Universiti Teknologi MARA
Amizon Azizan, Universiti Teknologi MARA
Arina Sauki, Universiti Teknologi MARA
Assoc. Prof. Dr. Md Amin Hashim, Universiti Teknologi MARA
Assoc. Prof. Dr. Norazah Abd Rahman, Universiti Teknologi MARA
Assoc. Prof. Dr. Rahmah Mohamed, Universiti Teknologi MARA
Assoc. Prof. Dr. Ruzitah Mohd Salleh, Universiti Teknologi MARA
Assoc. Prof. Hasnora Jafri, Universiti Teknologi MARA
Assoc. Prof. Dr. Ayub Md Som, Universiti Teknologi MARA
Assoc. Prof. Zulkafli Hassan, Universiti Teknologi MARA
Che Mohd Som Said, Universiti Teknologi MARA
Dr. Abdul Hadi, Universiti Teknologi MARA
Dr. Ahmad Rafizan Mohamad Daud, Universiti Teknologi MARA
Dr. Alawi Sulaiman, Universiti Teknologi MARA
Dr. Atikah Kadri, Universiti Teknologi MARA
Dr. Azil Bahari Alias, Universiti Teknologi MARA
Dr. Fazlena Hamzah, Universiti Teknologi MARA
Dr. Hazlina Hussin, Universiti Teknologi MARA
Dr. Ibtisam Abdul Wahab, Universiti Teknologi MARA
Dr. Istikamah Subuki, Universiti Teknologi MARA
Dr. Jagannathan Krishnan, Universiti Teknologi MARA
Dr. Jefri Jaapar, Universiti Teknologi MARA
Dr. Junaidah Jai, Universiti Teknologi MARA
Dr. Kamariah Noor Ismail, Universiti Teknologi MARA
Dr. Neni Kartini Che Mohd Ramli, Universiti Teknologi MARA
Dr. Noor Fitrah Abu Bakar, Universiti Teknologi MARA
Dr. Norhuda Ismail, Universiti Teknologi MARA
Dr. Norliza Ibrahim, Universiti Teknologi MARA
Dr. Nornizar Anuar, Universiti Teknologi MARA
Dr. Norun Farihah, International Islamic University Malaysia
vii
[email protected]
viii List of Reviewers
Dr. Nurul Fadhilah Kamalul Aripin, Universiti Teknologi MARA

Dr. Rafiziana Kasmani, Universiti Teknologi Malaysia
Dr. Rahida Wati Sharudin, Universiti Teknologi MARA
Dr. Safari Zainal, Universiti Teknologi MARA
Dr. Sherif Abdul Bari Ali, Universiti Teknologi MARA
Dr. Syed Shatir Asghrar Syed Hassan, Universiti Teknologi MARA
Dr. Tan Huey Ling, Universiti Teknologi MARA
Dr. Zulkifli Abdul Rashid, Universiti Teknologi MARA
Hanafiah Zainal Abidin, Universiti Teknologi MARA
Lim Ying Pei, Universiti Teknologi MARA
Miradatul Najwa Muhd Rodhi, Universiti Teknologi MARA
Mohibah Musa, Universiti Teknologi MARA
Dr. Nik Raikhan Nik Him, Universiti Teknologi MAR1A
Noorhaliza Aziz, Universiti Teknologi MARA
Nor Faeqah Idrus, Universiti Teknologi MARA
Norhasyimi Rahmat, Universiti Teknologi MARA
Norrulhuda Mohd Taib, Universiti Teknologi MARA
Nur Fadzeelah Abu Kassim, Universiti Teknologi MARA
Nurhaslina Che Radzi, Universiti Teknologi MARA
Prof. Dr. Ku Halim Ku Hamid, Universiti Teknologi MARA
Prof. Ir. Dr. Jailani Salihon, Universiti Teknologi MARA
Rasyidah Alrozi, Universiti Teknologi MARA
Rozaimee, Universiti Teknologi MARA
Sakinah Mohd Alauddin, Universiti Teknologi MARA
Sharmeela Matali, Universiti Teknologi MARA
Siti Shawalliah Idris, Universiti Teknologi MARA
Zalizawati Abdullah, Universiti Teknologi MARA
[email protected]
Contents
Part I Green and Sustainable Technologies
Overview on Chemical-Based, Bio-based and Natural-Based

Surfactants in EOR Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Hazlina Husin, Muhammad Najmi Ibrahim, Zulkafli Hassan,
Norrulhuda Mohd Taib, Ku Halim Ku Hamid,
Nik Khairul Irfan Nik Ab Lah and Muhammad Shafiq Mat Shayuti
OPEFB Filler from Biomass in Superabsorbent Polymer Composite

for Agriculture Application: A Comparative Study . . . . . . . . . . . . . . . 11
Wan Siti Nadiah Wan Yaacob, Saidatul Shima Jamari
and Suriati Ghazali
Exploring Non-wood Plants as Alternative Pulps: From the Physical

and Chemical Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Angzzas Kassim, Ashuvila Aripin, Zainuri Hatta and Zawawi Daud
Separation of Pyrrole from Isododecane Using Imidazolium

and Pyridinium Based Ionic Liquid at 298.15 K: Experiment
and COSMO-RS Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Nurshakirin Hazim Chan, Hanee Farzana Hizaddin
and Ramalingam Anantharaj
The Development of Amine-Based Potential Kenaf Sorbent

for Carbon Dioxide (CO2) Capture. . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Nabilah Zaini and Khairul Sozana Nor Kamarudin
Potential Effect of Palm Oil Fuel Ash as Micro-Filler

of Polymer Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
A.K. Nur Hafizah, M.W. Hussin, M. Ismail, M.A.R. Bhutta, M. Azman,
P.J. Ramadhansyah, A. Nur Farhayu and A.S.L. Nor Hasanah
ix
[email protected]
x Contents
Liquid Membrane Formulation for Succinic Acid Extraction

from Simulated Aqueous Waste Solution . . . . . . . . . . . . . . . . . . . . . . 51
Norela Jusoh, Norasikin Othman and Nur Alina Nasruddin
Preservation of Natural Textile Colorant from Xylocarpus

Moluccensis by Encapsulating into PVA Alginate . . . . . . . . . . . . . . . . 61
Norul Azilah Abdul Rahman, Ruziyati Tajuddin and Siti Marsinah Tumin
Coagulant from Chemically Modified Fish Scale for Textile

Wastewater Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Mohibah Musa, Ku Halim Ku Hamid, Hanafiah Zainal Abidin
and Miradatul Najwa Muhd Rodhi
Part II Chemical and Biotechnology Engineering
An Investigation on N-Containing Organic Compounds

as Corrosion Inhibitors for Carbon Steel in Saline Environment . . . . . 81
Md Amin Hashim, Izni Mariah Ibrahim and Sxureha Yunus
Evaluation on the Effect of Inhibitor Ratio to the Wax Deposition

in Malaysian Crude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
N. Ridzuan, Z. Yaacob and F. Adam
Solvent Extraction of Castor Beans Oil: Experimental Optimization

via Response Surface Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Jibrin Mohammed Danlami, Agus Arsad
and Muhammad Abbas Ahmad Zaini
Immobilization of Nano-sized TiO2 on Glass Plate for the Removal

of Methyl Orange and Methylene Blue . . . . . . . . . . . . . . . . . . . . . . . . 105
Lim Ying Chin, Lim Ying Pei, Razzana binti Rosli
and Nur Hazira binti Mohd Atni
Mercury Removal in Simulated Flue Gas by Oil Palm EFB Activated

Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Noor Hidayu Abdul Rani, Nor Fadilah Mohamad, Sharmeela Matali
and Sharifah Aishah Syed A. Kadir
Production of Glucose, Galactose and Mannose from the Skins

of Durian and Mangosteen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Zamila Mohd Senu, Maryam Husin, Abd Rashid Li, Rusnah Samsuddin,
Mohd Radzi Ahmad, Nik Roslan Nik Abd Rashid
and Nur Zalikha Mohd Taza
[email protected]
Contents xi
Reaction Kinetics Study for Microwave Energy Pretreated

Jatropha Curcas L In-Situ Transesterification . . . . . . . . . . . . . . . . . . . 131
Sintayehu Mekuria Hailegiorgis, Shuhaimi Mahadzir
and Duvvuri Subbarao
Induction Time of L-Isoleucine Crystallization with the Presence

of Electric Field. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Nik Salwani Md Azmi, Nornizar Anuar, Noor Fitrah Abu Bakar
and Mohammad Akmalhakim Zakaria
Gas Permeation Performance of Poly(lactic acid) Asymmetric

Membrane for O2/N2 Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
F. Mohamed, H. Hasbullah, W.N.R. Jamian, A.R.A. Rani, M.F.K. Saman,
W.N.H. Salleh and R.R. Ali
Particles Mixing in a Fluidized Bed by Using Digital Image

Processing and Thief Probe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Syaidatul Akma Mohd Zuki, Norazah Abd Rahman, Ihsan Mohd Yassin,
Nurul Asyikin Md Zaki and Mohamad Shafiq Izzudin Suhaimi
Polymeric Composite Membrane for CO2/CH4 Separation. . . . . . . . . . 167

Munawar Zaman Shahruddin and Muhammad Irfan Safwan Jasni
Comparative Study Between Isolated Xanthomonas Campestris

from Rotten Cabbage and Xanthamonas Campestris Culture
by YDC Media as Substrate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Mohibah Musa, Ku Halim Ku Hamid and Miradatul Najwa Muhd Rodhi
Part III Energy and Renewable Energy
1-Pyrenebutyric Acid Functionalized Reduced Graphene Oxide

(1-Pb-Rgo) Energy Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Ellie Yi Lih Teo, Mashitah Mohd Yusoff and Kwok Feng Chong
Clarifying the Palm-Based Drilling Fluids Potentials . . . . . . . . . . . . . . 195

Hazlina Husin, Zulkafli Hassan, Azlinda Azizi,
Mohd Fiqri Arman Mansor, Rosenadira Nasarauddin
and Sharifah Nawirah Syed Ariffin
The Study of Temperature Profile and Syngas Flare in Co-gasification

of Biomass Feedstock in Throated Downdraft Gasifier . . . . . . . . . . . . 203
Muddasser Inayat, Shaharin A. Sulaiman, Aini Abd Jamil,
Fiseha M. Guangul and Samson M. Atnaw
[email protected]
xii Contents
Evaluation of Energy Cost Saving and Pollutants Emission Reduction

for Solar Water Heater Development in Malaysia . . . . . . . . . . . . . . . . 211
Mohammed J.K Bashir, Ong Li Jing, Sumathi Sethupathi
and Ng Choon Aun
Part IV Process Control, Modelling, Simulations, Data Mining
Introducing Pandanaceae Via IOS-Based Software Application . . . . . . 221

Hannis Fadzillah Mohsin, Ibtisam Abdul Wahab
and Galoh Rashidah Haron
Investigation of Heat Treated Electrodeposited CoNiFe

on Microstructure and Hardness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Nor Azrina Resali, Koay Mei Hyie, M.N. Berhan and C.M. Mardziah
Computational Fluid Dynamics Modeling of Mercury Emissions

in Pulverized Coal Combustion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Nor Fadilah Mohamad, Noor Hidayu Abdul Rani,
Sherif Abdulbari Ali and Sharifah Aishah Syed A. Kadir
3D Sustainability Analysis of Integrated Process Design

and Control for Production of Cyclohexanone . . . . . . . . . . . . . . . . . . 245
Siti Aminah Zakaria, Mohd Jufri Zakaria, Mohamad Rizza Othman
and Mohd Kamaruddin Abd Hamid
Three-Dimensional Sustainability Analysis for Integrated Process

Design and Control for Separation of Benzene
and Toluene Mixture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Mohamad Zulkhairi Nordin, Mohamad Dzulfadzli Jais,
Mohamad Rizza Othman and Mohd Kamaruddin Abd Hamid
Nucleation Kinetics of Carbamazepine-Saccharin

(CBZ-SAC) Co-crystal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Khairool Azizul Mohammad, Syarifah Abd Rahim
and Mohd Rushdi Abu Bakar
Part V Advanced Materials and Nanotechnologies
Activation of Titanium Dioxide Under Visible-Light by Metal

and Non-metal Doping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
Krishnan Jagannathan, Sikirman Arman and Nerissa Mohamad Elvana
[email protected]
Contents xiii
Reflectometric Optosensor for Visual Detection of Ammonia

Based on Silica Pellet Sensing Material. . . . . . . . . . . . . . . . . . . . . . . . 281
Nur Syarmim Mohamed Noor, Ling Ling Tan, Siti Nurhafizah Zainuddin,
Kwok Feng Chong, Lee Yook Heng and Saiful Nizam Tajuddin
Low Noise and Properties of Double Layer Concrete

Paving Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Euniza Jusli, Hasanan Md. Nor, Ramadhansyah Putra Jaya,
Zaiton Haron, Musli Nizam Yahya, M. Azman and Lee Kuan Lim
Synthesis of Modified Covalent Organic Framework-1 (COF-1)

and Its Characterizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Muhammad Falaq Muhammad Faisal, Ahmad Rafizan Mohamad Daud
and Kamariah Noor Ismail
One Pot Synthesis of γ-Valerolactone from D-Glucose

Over H3PW12O40/ZrO2 Catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Dorairaaj Sivasubramaniam and Nor Aishah Saidina Amin
Carbon Cryogel from Lignin-Furfural as Acid Catalyst

in Esterification of Oleic Acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Muzakkir Mohammad Zainol, Nor Aishah Saidina Amin, Mohd Asmadi
and Norzita Ngadi
Surface Structure Study of CeZrO2 Nanocatalyst Doped

with Different Transition Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
S. Hanim Md Nor, M. Nazri Abu Shah, Kamariah Noor Ismail
and Abdul Hadi
Electrical Properties of Ammonium Iodide Doped Cellulose

Acetate Based Polymer Electrolyte . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
N.S. Samsi, R.M. Ali, R. Zakaria, M.Z.A. Yahya and A.M.M. Ali
Synthesis, Characterization and Corrosion Inhibition Studies

of 2-Methylbenzoylthiourea Derivatives . . . . . . . . . . . . . . . . . . . . . . . 339
Noor Khadijah Mustafa Kamal, Adibatul Husna Fadzil
and Karimah Kassim
Performance Studies of Anode Microbial Fuel Cells Using

Geobacter Sulfurreducens as a Biocatalyst . . . . . . . . . . . . . . . . . . . . . . 347
Noor Fazliani Shoparwe, Muhammad Hekarl Uzir
and Suhairi Abdul Sata
[email protected]
xiv Contents
Effects of Chain Length on the Thermotropic and Lyotropic Phase

Behaviours of Maltosides by Small-angle X-ray Diffraction Study . . . . 355
Hairul A.A. Hamid, Rauzah Hashim, John M. Seddon
and Nicholas J. Brooks
Thermal Properties Comparison Between Alumina Filled

and Organic Nano-crystal Filled UPR/EPS Composite. . . . . . . . . . . . . 363
Syed Anas Bin Syed Mustafa, Rahmah Mohamed and Husni Bin Rustam
Part VI Safety, Policies and Regulations
Hierarchy of Controls Analysis for Equipment Failures Prevention . . . 371

Nor Afina Eidura Hussin, Anwar Johari, Kamarizan Kidam
and Haslenda Hashim
Analysis the Effect of Explosion Efficiency in the TNT Equivalent

Blast Explosion Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Zulkifli Abdul Rashid, Azil Bahari Alias, Ku Halim Ku Hamid,
M. Shahnor Bani and Mohanad El Harbawi
Disparities in Generated Noise Between Predictions

and Measurements from Construction Sites . . . . . . . . . . . . . . . . . . . . 391
Zaiton Haron, Nadirah Darus and Khairulzan Yahya
Flame Retardancy of Polymeric Building Material

with Recycled Expanded Polystyrene Filler . . . . . . . . . . . . . . . . . . . . . 403
Syed Anas Bin Syed Mustafa, Rahmah Mohamed, Wan Bahira
and K. Rasidan
Grey Model for Accident Prediction in Data-Scarce Environment . . . . 411

Ali Al-shanini, Arshad Ahmad and Faisal Khan
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
[email protected]
Part I
Green and Sustainable Technologies
[email protected]
Overview on Chemical-Based, Bio-based
and Natural-Based Surfactants in EOR
Applications
Hazlina Husin, Muhammad Najmi Ibrahim, Zulkafli Hassan,

Norrulhuda Mohd Taib, Ku Halim Ku Hamid,
Nik Khairul Irfan Nik Ab Lah and Muhammad Shafiq Mat Shayuti
Abstract Energy consumption in the world is increasing every year while

expanding technology trying to catch up with the demand of oil production. One of
the technologies which has been widely used to enhance oil recovery is surfactant
flooding. A variety of surfactants promote oil recovery are reviewed in this paper,
with particular emphasis on the ability: (i) to lower the interfacial tension (ii) to
adjust the pH and (iii) to stand the high salinity conditions of reservoirs. To date,
surfactants studied in enhanced oil recovery applications can be categorized as
chemical-based, bio-based and natural-based surfactants.
Keywords Surfactant Interfacial tension EOR
H. Husin (&) M.N. Ibrahim ZulkafliHassan N.M. Taib K.H. Ku Hamid

N.K.I. Nik Ab Lah M.S. Shayuti
Faculty of Chemical Engineering, Universiti Teknologi MARA, 40450 Shah Alam,
Selangor Darul Ehsan, Malaysia
e-mail: [email protected]
M.N. Ibrahim
ZulkafliHassan
N.M. Taib
K.H. Ku Hamid
N.K.I. Nik Ab Lah
M.S. Shayuti
e-mail: mshafi[email protected]
H. Husin
CoRe of Frontier Materials & Industry Applications, Universiti Teknologi MARA,
40450 Shah Alam, Selangor Darul Ehsan, Malaysia
© Springer Science+Business Media Singapore 2015 3

M.A. Hashim (ed.), ICGSCE 2014, DOI 10.1007/978-981-287-505-1_1
[email protected]
4 H. Husin et al.
Introduction
Oil recovery using primary recovery methods which used reservoir pressure to
bring the oil up to the surface is only *20 %. When secondary recovery methods
are used, the oil recovery would only be increased up to *40 % maximum and thus
more oil residual is still left behind. Enhanced oil recovery (EOR) is another
method known to maximize oil recovery up to *60 % of the original oil-in-placed
remains in the reservoirs (also known as tertiary recovery methods).
In EOR, surfactant flooding is classified under chemical flooding methods.
Surfactant flooding is when a surfactant is employed in order to lower the interfacial
tension (IFT) between oil and water so that the oil becomes easier to flow.
Surfactant also alters the rock wettability (from oil-wet to water-wet) and adjusts pH
and salinity. This review focuses on chemical-based, bio-based and natural-based
surfactants. These surfactants are unique from one another depending on compo-
sition and chemical properties.
Surfactants are compounds that have both hydrophilic groups (heads) and
hydrophobic groups (tails). It means that the surfactant have both water soluble
component and oil soluble component. Surfactant will disperse in water and adsorb
between oil and water. The hydrophilic head will expand into the water phase while
the hydrophobic tail will extend into the oil phase.
Chemical-Based Surfactants
Surfactants manufactured from chemical synthesis processes are called chemical-

based surfactant. For example, ethoxylated sulfonates and petroleum sulfonates.
Ethoxylated sulfonates have a strong hydrophilic group and have a very high
electrolyte tolerance [1]. However, ethoxylated sulfonates have a tendency to form
liquid crystals. These liquid crystals can be avoided by an addition of co-solvent or
elevating the temperature. In contrast, petroleum sulfonates have a lower tolerance
to salt with optimal salinity of 1–2 % [2].
Recently, ionic liquids are reported to shown the ability as chemical-based
surfactants. Researchers have described that ionic liquid is able to tolerate harsh
conditions of salinity. For example, 1-dodecyl-3-methylimidazolium chloride
([C12mim][Cl]) and octylmethylnaphthalene sulfonate surfactants. Besides both
surfactants can survive harsh conditions of salinity, the first surfactant can lower the
IFT up to 70 mN m−1 difference at *1000 ppm [3] while the second surfactant can
lower the IFT up to 10−6 mN m−1 at a very low concentration [4].
[email protected]
Overview on Chemical-Based, Bio-based … 5
Biosurfactant
Bio-surfactant is an EOR method employed under microbial enhance oil recovery

(MEOR) techniques. Bio-surfactant is produced from the excretion of microbial
cells that grown on hydrocarbon. It is complex bio-polymer and has the prospective
in moving heavy crude oil from the reservoir. According to Al-Sulaimani et al. [5],
bio-surfactant produced by Bacillus licheniformis was able to lower the IFT from
46.6 to 3.28 mN m−1 in incubation for 16 h. Bacillus licheniformis are genetically
stable and can cause significant IFT reduction in different mediums [6]. Some other
bio-surfactants are able to reduce the surface tension of water from 72 to
27 mN m−1 [7]. The effectiveness of bio-surfactant during MEOR techniques is
contributed by its ability to lower the IFT, can withstand temperature up to 75 °C
and pH values of up to 12 [8] and the microorganism that produced the
bio-surfactant can also improve the sweep efficiency and selectively plug high
permeability zones.
Natural-Based Surfactant
Recently, studies are inclined towards surfactants derived from natural resources,
which also are categorized as natural-based surfactants [9]. These natural-based
surfactants are biodegradable, low toxicity, cheap and renewable. Due to their
unique properties, these natural-based surfactants are preferable to be used in EOR
as compared to the chemical-based surfactants. Furthermore, the impact to the
environment can be reduced without lowering the oil production. Although natural-
based surfactants have never been produced commercially, large scale industrial
production has been considered [10].
Studies [11–14] shown that sugar-based surfactants are effective surfactant
because it can lower the IFT and have excellent adsorption at solid-liquid interface
[15]. Additionally, glucose-based and sucrose-based surfactants have a range of
beneficial physical and performance properties, including high levels of surfactancy
(surface and interfacial activity), rapid biodegradability, low toxicity, effective
emulsification properties and effective surface interactions [16]. In terms of
adsorption mechanism, sugar-based surfactant cannot be classified as only one
simple chain of surfactant. This is because there are a lot of combination of
hydrophilic group and hydrophobic alkyl chain. The hydrophilic group might
consist of various sugar units while the hydrophobic group might consist of one or
more alkyl chains [17]. Figures 1 and 2 show some sugar-based surfactants that
have two or more hydrophobic chain.
Saponin can be classified as amphipathic glycosides with a distinctive foaming
characteristic. It can be found in various plant species. Saponins are surface active
substances because they contain both hydrophilic regions (such as rhamnose,
xylose, arabinose, galactose, fucose and glucuronic acid) and hydrophobic regions
[email protected]
6 H. Husin et al.
Fig. 1 Linkages between the hydrophilic and the hydrophobic in sugar surfactants: a ether bond,
dodecyl-β-D-glucoside; b ester bond, β-methyl-6-O-dodecanoyl-D-glucoside; c amine bond, N-
dodecyl-glucamine [17]
Fig. 2 A glycolipid: 1,3-di-O-dodecyl-2-O-β-D-glucosylglycerol [17]
(such as quillaic acid and gypsogenic acid) on the same molecule [18]. Example of
saponin surfactant that has been extracted from the Quillaja saponaria is shown in
Fig. 3 [18, 19]. Yang et al. [18] conducted a comparable study between a saponin
surfactant (Q-Naturale®) and a synthetic surfactant (Tween 80). They reported that
the saponin surfactant can lower the IFT as effective as the synthetic surfactant (as
in Fig. 4). Another research on saponin which was conducted by Shahri et al. [20]
have shown that the IFT decreases as the saponin concentrations increase. In
summary, saponin is a suitable candidate to replace the chemical surfactant because
it can lower the IFT as low as the chemical surfactant can perform, biodegradable
and can be produced at low cost. It does not have any adverse effect to the envi-
ronment as compared to chemical surfactant. This is a very important characteristic
as the reservoir can be preserved from any pollution that might change the condition
of hydrocarbon while increasing the oil recovery [20].
Malic acid-based surfactant is another suitable candidate to act as natural-based
surfactant for EOR because they contain active functional groups. Malic acid is an
[email protected]
Fig. 3 Structural representations of Quillaja saponin. This structure is presented in [19]
Fig. 4 The influence of

saponin (Q-Naturale®) and
synthetic (Tween 80)
surfactants at different
concentration on the
interfacial tension [18]
organic acid with chemical formula of HO2CCH2CHOHCO2H. Methods to produce

malic acid are described in many studies [21–25]. Lassila and Slone [26] investi-
gated on malic acid diester surfactant in waterborne coating, inks, adhesive,
[email protected]
8 H. Husin et al.
fountain solution and agricultural applications. Malic acid diester surfactant was
found to lower the IFT in water based systems. The adsorption mechanism study
[27, 28] showed the stable conformational molecular structure of malic acid subject
to particle surface. A presence of two different types of hydrogen bonds in the
racemic malic acid and only one type in the pure enantiomeric form was reported.
Conclusion
Although the use of surfactant for EOR has been studied and tested for many years,
the specific mechanisms involved in oil recovery are not understood fully or known
specifically for a wide variety of reservoir conditions, crude oils, and surfactant-
type systems. If MEOR or natural-based surfactant is to become an industrially
accepted alternative EOR process in the industry, the oil recovery mechanisms such
as IFT, phase and volume diagrams, and adsorptionmechanism need to be clarified.
The natural-based surfactant offers cost-effective processes as compared to existing
chemical-based surfactants.
Acknowledgments The author gratefully acknowledge the financial support from the Ministry of
Education, Malaysia (600-RMI/FRGS 5/3 (93/2013)) and for the continuous research supports
provided by the Universiti Teknologi MARA (UiTM). The author wish to thank the referee for
making this a better paper.
References
1. Carmona I, Schecter RS, Wade WH, Weerasooriya U (1985) Ethoxylated oleyl sulfonates as
model compounds for enhanced oil recovery. Soc Petrol Eng J 25:351–357
2. Skauge A, Palmgren O (1989) Phase behaviour and solution properties of ethoxylated anionic
surfactants. In: SPE international symposium on oilfield chemistry, 1989 Copyright 1989,
Society of Petroleum Engineers, Houston, Texas
3. Hezave AZ, Dorostkar S, Ayatollahi S, Nabipour M, Hemmateenejad B (2013) Dynamic
interfacial tension behaviour between heavy crude oil and ionic liquid solution (1-dodecyl-3-
methylimidazolium chloride ([C12mim][Cl]) + distilled or saline water/heavy crude oil)) as a
new surfactant. J Mol Liq 187:83–89
4. Zhao Z, Li Z, Qiao W, Cheng L (2005) Dynamic interfacial behaviour between crude oil and
octylmethylnaphthalene sulfonate surfactant flooding systems. Colloids Surf A 259:71–80
5. Al-Sulaimani H, Al-Wahaibi Y, Al-Bahry S, Elshafie A, Al-Bemani A, Joshi S, Zargari S
(2011) Optimization and partial characterization of biosurfactants produced by Bacillus
Species and their potential for ex-situ enhanced oil recovery. Soc Petrol Eng J 16:672–682
6. Thomas CP, Duvall ML, Robertson EP, Barrett KB, Bala GA (1993) Surfactant-based EOR
mediated by naturally occurring microorganisms. SPE Reservoir Eng 8:285–291
7. Illias RM, Ooi SW, Idris AK, Rahman WAWA (1999) Production of biosurfactant and
biopolymer from Malaysian oil fields isolated microorganisms. In: SPE Asia Pacific improved
oil recovery conference, Society of Petroleum Engineers, Kuala Lumpur, Malaysia
8. Gabitto JF (2006) Combined microbial surfactant-polymer system for improved oil mobility
and conformance control. In: SPE annual technical conference and exhibition, Society of
Petroleum Engineers, San Antonio, Texas, USA
[email protected]
9. Deymeh H, Shadizadeh SR, Motafakkerfard R (2012) Experimental investigation of Seidlitzia

rosmarinus effect on oil–water interfacial tension: usable for chemical enhanced oil recovery.
Scientia Iranica 19:1661–1664
10. Piispanen P, Persson M, Claesson P, Norin T (2004) Surface properties of surfactants derived
from natural products. Part 1: syntheses and structure/property relationships—solubility and
emulsification. J Surfact Deterg 7:147–159
11. Jarek E, Wydro P, Warszyński P, Paluch M (2006) Surface properties of mixtures of surface-
active sugar derivatives with ionic surfactants: theoretical and experimental investigations.
J Colloid Interface Sci 293:194–202
12. Rico-Lattes I, Lattes A (1997) Synthesis of new sugar-based surfactants having biological
applications: key role of their self-association. Colloids Surf A 123–124:37–48
13. Shi J, Li QD, Xia XC (2006) Synthesis and properties of sugar-based gemini surfactants.
Tenside Surfactants Deterg 43:204–209
14. Söderberg I, Drummond CJ, Furlong DN, Godkin S, Matthews B (1995) Non-ionic sugar-
based surfactants: self assembly and air/water interfacial activity. Colloids Surf A 102:91–97
15. Zhang L, Somasundaran P (2006) Adsorption of mixtures of nonionic sugar-based surfactants
with other surfactants at solid/liquid interfaces: I. Adsorption of n-dodecyl-β-d-maltoside with
anionic sodium dodecyl sulfate on alumina. J Colloid Interface Sci 302:20–24
16. Rojas OJ, Stubenrauch C, Lucia LA, Habibi Y (2009) Interfacial properties of sugar-based
surfactants
17. Stubenrauch C (2001) Sugar surfactants—aggregation, interfacial, and adsorption phenomena.
Curr Opin Colloid Interface Sci 6:160–170
18. Yang Y, Leser ME, Sher AA, McClements DJ (2013) Formation and stability of emulsions
using a natural small molecule surfactant: Quillaja saponin (Q-Naturale®). Food
Hydrocolloids 30:589–596
19. Mitra S, Dungan SR (1997) Micellar properties of Quillaja Saponin. 1. Effects of temperature,
salt, and pH on solution properties. J Agric Food Chem 45:1587–1595
20. Shahri MP, Shadizadeh SR, Jamialahmadi M (2012) Applicability test of new surfactant
produced from Zizyphus Spina-Christi leaves for enhanced oil recovery in carbonate
reservoirs. J Jpn Petrol Inst 55:27–32
21. Miyama M, Nakayama K (1993) D-malic acid production from DL-malic acid by microbial
assimilation of L-malic acid: screening of microorganism. Biotechnol Tech 7:105–110
22. Chibata I, Tosa T, Yamamoto K (1978) Industrial production of L-malic acid by immobilized
microbial cells. In: Pye EK, Weetall H (eds) Enzyme engineering. Springer, US, pp 463–468
23. Ado Y, Kawamoto T, Masunaga I, Takayama K, Takasawa S, Kimura K (1982) Production of
1-malic acid with immobilized thermophilic bacterium, Thermus rubens Nov. sp. In: Chibata I,
Fukui S, Wingard L Jr (eds) Enzyme Engineering. Springer, US, pp 303–304
24. Gong CS, Cao N, Sun Y, Tsao GT (1996) Production of L-malic acid from fumaric acid by
resting cells of Brevibacterium sp. In: Wyman C, Davison B (eds) 17th symposium on
biotechnology for fuels and chemicals. Humana Press, New York, pp 481–487
25. West T (2011) Malic acid production from thin stillage by Aspergillus species. Biotechnol Lett
33:2463–2467
26. Lassila KR, Slone CS (1999) Malic acid diester surfactants. Air Products and Chemicals
27. Husin H, Leong YK, Liu J (2012) Molecular attributes of an effective steric agent: yield stress
of dispersions in the presence of pure enantiomeric and racemate malic acids. Adv Powder
Technol 23:459–464
28. Husin H, Leong YK, Liu J (2011) Conformational molecular structure-interaction force
correlation in influencing effective steric between alumina and malic acid. In: CHEMECA
2011: the 41th Australasian chemical engineering conference, Engineers Australia, Sydney,
Australia, p 109
[email protected]
OPEFB Filler from Biomass
in Superabsorbent Polymer Composite
for Agriculture Application:
A Comparative Study
Wan Siti Nadiah Wan Yaacob, Saidatul Shima Jamari

and Suriati Ghazali
Abstract An agriculture activity is one of the major sectors which contribute a

significant share to Malaysia economy. Therefore it is important to maintain the soil
fertility. Superabsorbent polymer composite (SAP’c) was produced using graft
polymerization technique with the addition of the oil palm empty fruit bunch
(OPEFB) filler to enhance the absorbance properties and decomposition process.
Effect of filler addition at different size to the SAP’c was investigated. Water
absorbency test was conducted to determine the absorbency and swelling properties
while Fourier Transform Infrared (FTIR) and Scanning Electron Microscopic
(SEM) analysis was conducted to investigate the functional groups and morphology
of the SAP’c. Water absorb highest for SAP’c with coarse OPEFB while fine
particle size showed a decrement in swelling behaviour. FTIR spectra shows that
–C ≡ C–H with C–H bend existed in all samples. SEM showed that the addition of
the OPEFB give a fibrous morphology and build a surface contact. In conclusion
the addition of biomass filler in the superabsorbent polymer (SAP) will increase the
water absorbency and swelling properties of the hydrogel.
Keywords Superabsorbent polymer Superabsorbent polymer composite Oil

palm empty fruit bunch
W.S.N.W. Yaacob S.S. Jamari (&) S. Ghazali

Faculty of Chemical and Natural Resources Engineering, Universiti Malaysia Pahang,
Lebuhraya Tun Razak, 26300 Kuantan, Pahang, Malaysia
W.S.N.W. Yaacob
S. Ghazali

[email protected]
12 W.S.N.W. Yaacob et al.
Introduction
Nowadays, agricultural activities are becoming a major growing sector and con-
tribute a lot of income to the Malaysia economy. In addition, this activity also
provides employment opportunities to the citizen. In order to ensure that agricul-
tural activities continues to contribute significantly, measures should be taken to
safeguard the sustainability of this sector, for example, to maintain the soil fertility.
Fertilizers made from chemicals are beneficial towards the plant however, offer
adverse effects to the neutrality of the soil pH. Therefore, this research was pro-
posed to synthesis SAP’c as a soil conditioning mechanism.
In 1960s, the United States Department of Agriculture has developed superab-
sorbent polymer. The early attempt for the commercialization of superabsorbent
polymer was in early 1970s as it is commercialized in the form of starch/acrylo-
nitrile/acrylamide based polymer (superslupers). At first, this material was focused
to be used in agriculture/horticulture sector. In 1982, superabsorbent polymer for
the diapers application were introduced to the Japanese market by Unicharm, while
in 1983 by KAO and Procter and Gamble introduced in 1985, [1, 2]. In general, the
definition of superabsorbent polymer is a material that can absorb huge amount of
water while retain its original shape without loss the water that had been absorb.
The example applications that used this type of materials were in hygiene product
such as diapers. The most common technique to synthesize this material was by
graft polymerization method. In graft polymerization method, usually acrylic acid
or methacrylic acid was used as monomers and sometimes acrylamide and meth-
acrylamide also can be used. The addition of acrylamide into acrylic acid can help
to increase the absorbency of superabsorbent polymer besides induce the graft
polymerization [3].
Currently, the world major concern is towards the environmental problems. In
relative to the problem, superabsorbent polymers composite, which produce
throughout the research, were synthesized by adding the OPEFB from biomass
waste as filler. The addition of the OPEFB will enhance the biodegradable ability
thus it will be an eco-friendly product. Employing a biomass waste as filler will
reduce the land pollutant cause by the waste besides reducing the cost needed to
decompose the waste and production cost of the product. In recent years, Malaysia
has become the second largest (after Indonesia) palm oil exporter thus contributes
up to 39 wt% of the total world palm oil production. Due to the demand of the palm
oil industries around the world, few countries such as Indonesia, Colombia, Ivory
Coast and Papua New Guinea had identified other good potential of the palm oil
resulting in increasing of the fresh fruit bunch (FFB) production that leads to the
increasing of waste streams [4].
As the OPEFB is one of the main waste biomass products in Malaysia,
Malaysian Palm Oil Board (MPOB) in 2005 has record that about 20–25 % residue
of OPEFB produced for every tonne FFB after it was steamed and fruitlets are
removed. Study done before on the ten years application of empty fruit bunches in
an oil palm plantation on soil chemical properties showed that the decomposition
[email protected]
OPEFB Filler from Biomass in Superabsorbent … 13
time of OPEFB was decreased when it is applied to soil. Within 7–15 weeks of
decomposition process, empty fruit bunch (EFB) has lost it dry matter up to 50 %
with the total decomposition of dry OPEFB in the soil only take less than
12 months. Other than that, the application of OPEFB in soil has increase the pH
value of the soil [5]. So that, the application of OPEFB as a biomass filler in
superabsorbent polymer will help in naturally for the polymer chain to decompose
itself in soil besides increasing the fertility of the soil as the OPEFB help to increase
the pH value of the acidic soil.
Materials and Experimental
Materials
OPEFB was chosen as biomass filler in SAP’c composites. For synthesis of SAP’c,
AA was used as a monomer, MBA as a crosslinker agent while APS was used as an
initiator.
Experimental
Preparation of OPEFB as a Filler
Fresh OPEFB were cut into a small pieces and dried in an oven for 5 days at 70 °C.
After that, high speed grinder was used to reduce the size of dried OPEFB. Grinded
OPEFB particles were than sieved to get a uniform particle size.
Synthesis of SAP’c
SAP’c was synthesized by using graft polymerization method. In an empty three

neck flask, acrylic acid solution was prepared. Then, sodium hydroxide (5 M) was
added to neutralize the acrylic acid solution and OPEFB was added as the filler.
Mixture was continuously stirred throughout the process and crosslinking agent
(MBA) was added under a nitrogen atmosphere. After 30 min of stirring, APS was
added into it and temperature slowly raised up to 70 °C. SAP’c gel will formed after
30 min of reactions. All the reactions were carried out using double boil technique.
The obtained gel was then washed using distilled water for several times and dried
in an oven at temperatures, 70 °C for 24 h. Dried SAP’c was then milled to small
particles (powder form). Similar step was repeated for pure SAP’c, without the
addition of OPEFB filler.
[email protected]
Water Absorbency Measurement
0.53 g of SAP’c was weighed and added into a tea-bag. The tea bag was then
immersed in a 1 L distilled water at room temperature. SAP’c will slowly swell
until the equilibrium is reached. After 24 h, swollen gel was allowed to drain and
reweighed to calculate its water absorbency by using the following equation.
gH O m2 m1
Water Absorbency; Qð 2 Þ ¼
gsample m1
Water absorbency is expressed in grams of water stored in gel by a gram of dried

gel. m2 and m1 represents the weight of gel after immersed in water and weight of
dried gel before immersed in water. The test was repeated for three times.
Results and Discussions
Water Absorbency Test
Water absorbency test was carried out for 24 h to observe and determine the
swelling capacity of the all SAP’c. Table 1 shows that the weight of all samples is
increased up to 200 times from the original weight. The application of the OPEFB
filler has increased the swelling capacity. However, the particle size gave effect to
the swelling behavior of the SAP’c. It showed that SAP’c with a coarse OPEFB can
absorbed more water with the water absorbency increment up to 0.94 % but SAP’c
with fine OPEFB decreased to 5.82 %,compared to the pure OPEFB. Refer to the
Sadeghi et al. studied on the swelling behavior of the SAP, dense crosslinking in
SAP’c can decrease the swelling behavior of the hydrogel. It is due to the higher
crosslinking density has decreased the space between the copolymer chains thus
resulted in the rigid structure that cannot be expanded and hold large amount of
water [6]. So in this case, fine particle size give a large surface area for the OPEFB
to make contact and crosslink with the copolymer thus increased the crosslinking
density of SAP’c.
Table 1 The weight of SAP’c before and after the water absorption test
Samples Pure SAP’c SAP’c with coarse OPEFB SAP’c with fine OPEFB
Initial weight (g) 0.53 0.53 0.53
Final weight (g) 115.55 116.62 108.85
[email protected]
FTIR Analysis
FTIR analysis has been carried out to determine the functional group that exists in
OPEFB and SAP’c. As shown in Fig. 1, there is a strong and broad OH-stretch with
H-bonded at the band 3420.87 cm−1. At the spectra with medium peak of
2920.43 cm−1, C–H with asymmetric and symmetric stretch (alkanes group) formed
while at the band of 1738.42, 1639.33, and 1055.21 cm−1, there is C double bond O
stretch, medium peak of N-H bend in amines-primary group and medium C–N
stretch in aliphatic amines group for OPEFB sample.
The comparison of FTIR spectra in Fig. 2 for the three samples of SAP’c before
immersed in water showed that at peak 3425.52 cm−1 for pure SAP’c, the peak shift
Fig. 1 FTIR spectra for OPEFB
Fig. 2 FTIR spectra for a SAP’c with coarse OPEFB size b SAP’c with fine OPEFB size c pure
SAP’c
[email protected]
to 3464.64 cm−1 for the SAP’c with the fine size of filler and for the coarse filler
size, the peak shift back to 3420.61 cm−1. This peak shows the existence of strong
and broad O-H stretch with H-bonded. These bonds indicate the presence of –OH
group in the samples. For all three samples, N–H and N = O bending at peaks
1637.38 and 615.00 cm−1 for pure SAP’c, 1639.19 and 621.78 cm−1 for SAP’c
with fine filler and for SAP’c with coarse size filler was at 1639.19 and
1384.55 cm−1 and also –C ≡ C–H with C–H bend exist for three samples at peaks
in range between 614.00 and 621.8 cm−1 indicates that the graft polymerization
occurs in all samples and AA was grafted to the OPEFB during the graft poly-
merization reaction occurs [7].
Surrface Morphology Analysis Using Scanning Electron Microscopy

(SEM)
From the SEM analysis, it can be seen that the addition of the OPEFB as a filler in
SAP’c able to increase the space capacity in the SAP’c as the parallel alignment of
the OPEFB help the water to absorb faster and thus occupied the gel space.
Compared Fig. 3c, e the addition of the OPEFB give a fibrous morphology and
build a surface crosslinking thus enhance the absorption ability of the SAP’c gel.
Fig. 3 Surface morphology of a coarse OPEFB b fine OPEFB c pure SAP’c d SAP’c with fine
OPEFB e SAP’c with coarse OPEFB
[email protected]
Conclusion
In conclusion, superabsorbent polymer composite can be synthesized by using graft

polymerization techniques. The application of well distributed OPEFB in SAP’c
will enhance and optimize the absorbency and swelling properties of SAP’c.
However, it is important to choose the suitable particles size to prevent SAP’c from
losing their good absorbency and water retention characteristics.
References
1. Nexant Inc. (2004) Superabsorbent polymers (SAP). Nexant Chem System. PERP Program 1–6
2. Kabiri K, Omidian H, Hashemi SA, Zohuriaan-Mehr MJ (2003) Synthesis of fast-swelling
superabsorbent hydrogels: effect of crosslinker type and concentration on porosity and
absorption rate. Eur Polymer J 39:1341–1348
3. Doane WM, Doane SW and Savich MH (2004) Superabsorbent polymers in agriculture
applications. Absorbent Technologies, Inc
4. Jamari SS, Howse JR (2012) The effect of the hydrothermal carbonization process on palm oil
empty fruit bunch. Biomass Bioenergy 47:82–90
5. Abu Bakar R, Darus S, Kulaseharan S, Jamaluddin N (2010) Effects on ten year application of
empty fruit bunches in oil palm plantation on soil chemical properties. Nutr Cycl Agroecosyst
89(3):341–349
6. Sadeghi M (2012) Synthesis of biocopolymer carrageenam-g-poly (AAm-co-IA)/montmoril-
onite superabsorbent hydrogel composite. Braz J Chem Eng 29(02):295–305
7. Zhang J, Wang Q, Wang A (2007) Synthesis and characterization of chitosan-g-poly (acrylic
acid)/attapulgite superabsorbent composites. Carbohydr Polym 68(2):367–374
[email protected]
Exploring Non-wood Plants as Alternative
Pulps: From the Physical and Chemical
Perspectives
Angzzas Kassim, Ashuvila Aripin, Zainuri Hatta and Zawawi Daud
Abstract The increasing demand in wood fibre consumptions especially in pulp

and paper making has pushed forward the search for alternative fibre resources.
Non-woods derived fibre could be good candidates due to its abundance avail-
ability. The objective of this study is to determine the potential of non-wood plants
as alternative fibres for pulp and paper-based industries based on its chemical and
physical properties. The chemical properties involved in this study (cellulose,
hemicellulose, lignin, hot water and 1 % NaOH solubilities and ash contents) were
determined according to relevant TAPPI test, Kurscher-Hoffner and Chlorite
methods. Meanwhile, the physical properties (fibre length and diameter) were
determined according to the Franklin method. In order to propose the suitability of
the studied non-wood plants as alternative fibre resources in pulp and paper-based
industries, the obtained results are compared to properties of published wood
resources. Results show that lignin content (5.67 %) and 1 % NaOH solubility
(19.64 %) of cogon grass are the lowest compared to cocoa pod husk and oil palm
leaf. These contents influenced the production of higher pulp yield. Although cogon
grass contains short fibres length than oil palm leaf, the paper product will have
higher strength due to the good inter-fibre bonding. This study conclude, based on
A. Kassim (&) A. Aripin

Faculty of Engineering Technology, Universiti Tun Hussein Onn Malaysia,
Batu Pahat, Malaysia
A. Aripin
Z. Hatta Z. Daud
Faculty of Civil and Environmental Engineering, Universiti Tun Hussein Onn Malaysia,
Batu Pahat, Malaysia
Z. Daud

[email protected]
20 A. Kassim et al.
the chemical and physical properties, the cogon grass, cocoa pod husk, and oil palm
leaf are suitable to be used as alternative fibre resources in pulp and paper-based
industries with various applications such papers, boxes and craft materials.
Keywords Cellulose Fibre Hemicellulose Lignin Paper
Introduction
Perennial non-wood plant (cogon grass) and agro-wastes (cocoa pod husk and oil
palm leaf) are considered problematic because they possess no economic impor-
tance for any specific purpose. The depleting natural resources, growing environ-
mental concerns, abundant and inexpensive of non-wood plant resources are the
major driving forces to utilize these annually renewable plants for various industrial
applications [1]. Generally, these plants have the compositions and properties that
are suitable to be used as composite, textile, pulp and paper-based manufacturing as
well as in producing fuel, enzymes, chemicals and animal feed [1, 2].
In Malaysia, the plantation of cocoa was around 20,645 ha with 7019 tons
production in 2010 [3]. Cocoa pod husk (CPH) represents between 70 to 75 % of
the whole cocoa fruit weight [4]. That is, an average of 5000 tons of cocoa pod husk
was generated in that year. Oil palm trees are also one of the most valuable crops in
Malaysia with approximately 4.85 million ha of plantation area in 2010 [5]. The
large plantation area has generated large biomass residues (more than 70 million
tons) as 1 ha of plantation area may produce between 50 and 70 tons of wastes [6].
The residues included oil palm leaves (OPL), empty fruit bunches, shell, kernel and
trunks [7]. Moreover, the cogon grass (CG) is also widely found in Malaysia as
scattered patches along highways and open area. This invasive grass is problematic
to plantation area and a bad plant as feedstock for animals due to the rough and
sharp leaves [8].
When left on the plantation floor, these wastes create great environmental
problems that constitute a barrier for easy movement and serves as habitat for pests
and diseases causing organisms on the farm [9]. Inevitably, these massive wastes
from CPH and OPL were dumped in the landfill for disposal.
To promote a sustainable and greener environment, the economic importance of
the cogon grass, cocoa pod husks and oil palm leaves is being put forward. The first
step is to evaluate the fundamental properties comprise within them. Therefore, the
objective of this study is to evaluate the chemical and physical properties of cocoa
pod husk, cogon grass and oil palms leaves and to make comparison of their
suitability with the published literature. The information generated from this study
will be utilised in exploring the potential of those perennial non woods in the
applications as pulp in paper-based industries.
[email protected]
Exploring Non-wood Plants as Alternative … 21
Experimental
Materials
For this study, samples were collected from Pusat Pembangunan Komoditi Parit
Botak, Jabatan Pertanian Malaysia, Batu Pahat, Johor. Prior to air-dried, CG and
OPL were cut into 2 to 5 cm length whereas CPH was chopped approximately to
1.5 × 1.5 cm2. The samples were thoroughly washed to eliminate sand and other
contaminants. For chemical analysis, the air-dried samples were ground to 0.40–
0.45 mm and stored in air tight container for further analyses.
Methods
Chemical properties, Chemical composition of CG, CPH and OPL were performed
according to Technical Association of the Pulp and Paper Industry (TAPPI) Test
Method. The samples were first submitted to soxhlet extraction for 6 h according to
method T 264 om-88. The evaluation of extractive substances was carried out in
different liquids according to experimental parameters: hot water solubility (T
207 cm-08) and 1 % sodium hydroxide solubility (T 212 om-07) and ash content (T
211 om-07). The amount of lignin, holocellulose, hemicelluloses and cellulose were
assessed by using the following respective standard methods: T 222 om-06,
Chlorite and Kurschner-Hoffner methods. All experiments were conducted in
triplicates.
Physical properties, To measure the fibre length of the materials, Fraklin method
(Han et al. 1999) was applied. In this method, the materials were immersed into
glacial acetic acid and hydrogen peroxide for a period of 24 h at 60 °C or until they
were defibred. The fibre length of materials were measured under profile projector
microscope (V-12, Japan) and the fibre diameters were measured by using Quata
200 Scanning Electron Microscopy.
Chemical properties, Table 1 shows the percentages of various components present

in CG, CPH and OPL. Results indicated significant difference (p < 0.05) between
different samples (of the same properties). The data shows that OPL (43.79 %)
exhibited the highest content in cellulose compared to CPH (35.44 %) and CG
(37.13 %). In pulp and paper-based industries, cellulose is an important constituent
in papermaking. A high content of cellulose produces a high pulp yield and a good
quality paper [10].
[email protected]
22 A. Kassim et al.
Table 1 Chemical properties of non-wood plant fibres

Chemical properties Non-wood plants
(% w/w) Cocoa pod husk (CPH) Cogon grass (CG) Oil palm leaf (OPL)
Cellulose 35.44 ± 0.33* 37.13 ± 0.43* 43.79 ± 1.51*#
Hemicellulose 37.00 ± 0.50 *#
27.36 ± 0.40 *
36.41 ± 1.73*
Lignin 14.71 ± 0.35 *
5.67 ± 0.26 *#
19.66 ± 2.41*
Hot water solubility 17.63 ± 0.23 *
3.83 ± 0.30 *#
18.43 ± 0.39*
1 % NaOH solubility 20.23 ± 0.59 *
19.64 ± 0.53 *#
41.95 ± 2.92*
Ash 12.26 ± 0.23 *
8.24 ± 0.28 *
5.73 ± 0.94*#
*
Significant difference (p < 0.05) of different plant fibres; the most desirable content
#
On the other hand, hemicellulose content of CPH (37 %) is the highest compared
to CG (27.36 %) and OPL (36.41 %) as shown in Table 1. High hemicellulose
content is desirable for pulp and paper-based industries because it is correlated with
high strength of the paper produced especially in tensile and burst indices [11, 12].
Therefore, paper products from CPH are expected to have high tensile and burst
strength.
Lignin contents are at satisfactory level if the value is less than 30 % [10].
Remarkably, the lignin content of CG was found to be 5.67 %, which is the lowest
compared to CPH (14.71 %) and OPL (19.66 %). Low lignin content is considered
advantageous in the pulp and paper-based industries as it utilize less chemicals and
energy during the pulping process [13, 14]. Low lignin will also generate whiter
paper and therefore requires less bleaching.
Hot water solubility in water solvent is the measure of the extractive components
of the plants [10]. The hot water solubility in CG (3.83 %) is the lowest compared
to CPH (17.63 %) and OPL (18.43 %) as shown in Table 1. Low hot water
solubility indicates less extractive materials in the plants and potential of generating
high pulp yield after the pulping process [10].
The 1 % NaOH solubility indicates the level of degradation of the material in
weak alkali. CG (19.64 %) contains the lowest content of 1 % NaOH solubility
compared to CPH (20.23 %) and OPL (41.95 %). Low solubility suggests less fibre
degradation during the pulping process and high pulp yield [15].
Finally, the ash content indicates the mineral component of lignocellulosic
materials. The ash content was found highest in CPH (12.26 %), followed by CG
(8.24 %) and OPL (5.73 %). High ash content will contribute to decrease in the
strength properties of the product and reducing pulp yield [16].
Physical properties, Fibre length is an important parameter in measuring the
quality of pulp to be used in paper-based industries [17], due to the strong rela-
tionship between fibre length and strength properties of the resultant pulp [18].
Table 2 shows the fibre length and diameter of CPH, CG and OPL with significance
difference (p < 0.05) between different plants.
There are three classifications of fibre length; long, medium and short. Results in
Table 2 indicates fibres obtained from CPH, CG and OPL are considered as short
[email protected]
Exploring Non-wood Plants as Alternative … 23
Table 2 Physical properties of non-wood plants

Physical properties Non-wood plants
Cocoa pod husk (CPH) Cogon grass (CG) Oil palm leaf (OPL)
Fibre length, mm 0.73 ± 0.09*# 0.94 ± 0.05* 1.13 ± 0.06*
Fibre diameter, µm 11.09 ± 2.40*
6.63 ± 0.40*
6.33 ± 0.06*#
Felting rate 65.83 141.78 178.52
*
Significant difference (p < 0.05) of different plant fibres; #the most desirable content
fibre (lengths between 0.2 and 1.2 mm). Fibre length of CPH (0.73 mm) is the
shortest compared to CG (0.94 mm) and OPL (1.13 mm). The shorter fibres length
contributes to excellent formation and increase inter-fibre bonding as well as
increases the strength properties of the produced paper [19]. In addition, short fibre
length will also give smoothness of the surface morphology of produced paper [20].
Fibre diameter is also important to determine the quality of fibres in pulp and
paper-based industries. The fibre diameter of OPL (6.33 µm) is slightly smaller than
CPH (6.63 µm), followed by CG (11.09 µm). A small diameter of fibre indicates
that less fibre flexibility.
The strength properties of paper-based products were found to be positively
correlated to the felting rate (fibre length/fibre diameter). Generally, the acceptable
value for felting rate is more than 33 [10]. The felting rates of all the materials are
acceptable as a good fibre; CPH (65.83); CG (141.78) and OPL (178.52).
Conclusion
In conclusion, the non-wood plants (cocoa pod husk, cogon grass and oil palm leaf)
could be viewed as potential resources as alternative fibres in pulp and paper-based
industries. Although the chemical and physical properties are different, these plants
are still appropriate to be utilised in the paper-based industries depending on the
suitable applications. For example, the oil palm leaf with a greater fibre length (i.e.
rougher paper surface) as compared to CPH and CG might be more suitable for
craft industries or carrier bags. The chemical properties indicate cogon grass has the
most favourable parameters (three out of six), indicating it might be the most
suitable fibre as pulp in paper industry. In addition, cogon grass has short fibre
length and considered to have good arrangement of fibre that increase the strength
property in the resultant paper. Based on the chemical and physical properties, all
proposed plants are suitable to be used as alternative fibre depending on its
application with cogon grass being the most suitable.
Acknowledgment This research was financially supported by the Fundamental Research Grant
Scheme (FRGS) and scholarship from Universiti Tun Hussein Onn Malaysia (UTHM). The
authors are thankful for this financial support.
[email protected]
24 A. Kassim et al.
References
1. Reddy N, Yang Y (2005) Biofibers from agricultural byproducts for industrial applications.
Trends Biotechnol 23(1):22–27
2. Abdul Khalil HPS, Siti Alwani M, Mohd Omar AK (2006) Chemical composition, anatomy,
lignin distribution and cell wall structure of Malaysian plants waste fibers. BioResource 1
(2):220–232
3. Jabatan Perangkaan Malaysia, JPM (2011) Perangkaan Ekonomi Malaysia, Siri Masa
Retrieved on 08 May 2012. www.statistics.gov.my
4. Cruz G, Pirilä M, Huuhtanen M, Carrión L, Alvarenga LE, Keiski RL (2012) Production of
activated carbon from cocoa (Theobroma cacao) pod husk. J Civil Environ Eng 2(2):1–6
5. Malaysia Palm Oil Board Statistics, MPOB (2011). Retrieved 10 Nov 2011. http://www.
mpob.gov.my
6. Shuit SH, Tan KT, Lee KT, Kamaruddin AH (2009) Oil palm biomass as a sustainable energy
source: a Malaysia case study. Energy 34(9):1225–1235
7. WanRosli WD, Law KN (2011) Oil palm fibers as papermaking material: potentials and
challenges. BioResources 6(1):901–917
8. Coile NC, Shilling DG (1993) Cogongrass, Imperata cylindrical (L.) Beauc: a good grass gone
bad! Botany Circular 28:1–3
9. Agele SO, Agbona AI (2008) Effects of cocoa pod husk amendment on soil and leaf chemical
composition and growth of cashew (Anacardium occidentale L).seeding in the nursery. Am-
Euraasian J Sustain Agric 2(3):219–224
10. Shakhes J, Marandi MAB, Zeinaly F, Saraian A, Saghafi T (2011) Tobacco residuals as
promising lignocellulosic materials for pulp and paper industry. Bioresources 6(4):4481–4493
11. Jacobs RS, Pan WL, Fuller WS, Mckean WT (1999) Genetic and environmental influences on
the chemical composition of Washington state wheat straw. In: 1999 pulping conference
proceedings of Atlanta, TAPPI Press. pp 839–847
12. Wathén R (2006) Studies on fiber strength and its effect on paper properties. Ph.D. Thesis,
Helsinki University of Technology
13. Al-Mefarrej HA, Abdel-Aal MA, Nasser RA (2013) Chemical evaluation of some
lignocellulosic residues for pulp and paper production. Am-Eurasia J Agric Environ Sci 13
(4):498–504
14. Marques, Rencoret J, Ana G, José CDR (2010) Evaluation of the chemical composition
of different non-woody plants fibers used for pulp and paper manufacturing. Open Agric
J 4:93–101
15. Tran AV (2006) Chemical analysis and pulping study of pineapple crown leaves. Ind Crops
Prod 24:66–74
16. Ai J, Tschirner U (2010) Fiber length and pulping characteristics of switchgrass, alfalfa stems,
hybrid poplar and willow biomasses. Bioresour Technol 101(1):215–221
17. Saijonkari-Pahkal K (2001) Non-wood plants as raw material for pulp and paper. Master’s
Thesis, University of Helsinki
18. Akpakpan AE, Akpabio UD, Ogunsile BO, Eduok UM (2011) Influence of cooking variables
on the soda and soda-ethanol pulping of Nypa Fruticans petioles. Aust J Basic Appl Sci 5
(12):1202–1208
19. Chandra M (1998) Use of nonwood plant fibres for pulp and paper industry in Asia: potential
in China. Master’s Thesis, State University
20. Heriksson G, Elisabet B, Helena L (2009) The trees. In: Monica Ek, Göran G, Heriksson
Gunnar (eds) Wood chemistry and wood biotechnology vol 1. Walter de Gruyter GmBH and
Co, United State of America, pp 13–44
[email protected]
Separation of Pyrrole from Isododecane
Using Imidazolium and Pyridinium Based
Ionic Liquid at 298.15 K: Experiment
and COSMO-RS Prediction
Nurshakirin Hazim Chan, Hanee Farzana Hizaddin

and Ramalingam Anantharaj
Abstract The removal of aromatic nitrogen compound from diesel oil is important
to produce cleaner fuel and reduce the environmental impact. Therefore, denitrifi-
cation of diesel oil is carried out by liquid-liquid extraction at ambient and moderate
condition. The aim of this work is to study liquid-liquid extraction (LLE) for the
mixture of ionic liquids (ILs); 1-ethyl-3-methylimidazolium ethylsulfate ([EMIM]
[EtSO4]) and 1-ethyl-3-methylpyridinium ethylsulfate ([EMPy][EtSO4], nitrogen
compounds; pyrrole (PYR) and model diesel compound; isododecane (ISOD). The
selectivity and solute distribution ratio values were calculated, four ternary dia-
grams were generated and denitrification efficiency was determined to evaluate the
effectiveness of the extraction. The Conductor-like Screening Model for Real
Solvents (COSMO-RS) model was used to predict the composition of the ternary
systems. The experimental values and COSMO-RS prediction were then compared
to determine the root-mean-square deviation (RMSD) for each system. The slope of
tie lines is positive for all ternary system and the average RMSD is 2.825 %. The
prediction gave RMSD of 1.12 % for [EMIM][EtSO4](1) + Pyrrole(2) + Isodecane
(3) system; 2.51 % for [EMPy][EtSO4](1) + Pyrrole(2) + Isodecane(3) system.

Keywords Ionic liquid Pyrrole Isododecane Solvent extraction COSMO-RS
prediction
N. Hazim Chan H.F. Hizaddin R. Anantharaj (&)

Department Chemical Engineering, Faculty of Engineering, University of Malaya,
50603 Kuala Lumpur, Malaysia
N. Hazim Chan
H.F. Hizaddin

[email protected]
26 N. Hazim Chan et al.
Introduction
Due to the increasingly strict environmental regulations on transportation fuel,

cleaner fuel has become very significant research subject. Therefore, denitrification
of diesel oil is the main interest of refineries worldwide nowadays. Diesel oil is a
complex mixture of saturated, unsaturated and aromatic hydrocarbon which has low
amount of nitrogen-based compounds. However, the nitrogen content will increase
with the boiling point of a particular oil fraction. Most of the nitrogen compound
found in diesel oil is present in aromatic compound which has a low reactivity and
high refractoriness compared to acidic and non-acidic sulfur compound. The
nitrogen-based compounds have to be removed from the diesel oil because they are
(a) main source of air pollution; (b) strong inhibitor for desulfurization process; (c)
emitting NOx during combustion which will be converted into NO2 that is the
contributor for troposphere ozone formation. The presence of nitrogen compounds
as low as <30 ppm can inhibits the efficiency of hydrodesulphurization process
through competitive adsorption, which make it more difficult to meet the low-
sulphur requirement. Thus, it is important to reduce the nitrogen content to achieve
the level of zero emission by reducing from >70 to <0.1 ppm of nitrogen compound
in diesel oil [1]. The nitrogen compound can be divided into non-basic compound
(5-membered ring) and basic compound (6-membered ring). The non-basic nitrogen
compounds are Pyrrole (C4H5N), Indole (C8H7N), Indoline (C8H9N), Carbazole
(C12H9N) and Benzocarbazole (C16H11N) while the basic nitrogen compounds are
Pyridine (C5H5N), Quinoline (C9H7N) and Benzoquinoline (C13H9N).
The aim of this work is to perform LLE by using ionic liquids to know their
selectivity toward nitrogen compound for the nitrogen removal in diesel oil at
298.15 K. Thus, the nitrogen compounds used in this work is pyrrole with isod-
odecane as the feed and two types of ionic liquids which are 1-ethyl-3-methyl-
imidazolium ethylsulfate ([EMIM][EtSO4]) and 1-ethyl-3-methylpyridinium
ethylsulfate ([EMPy][EtSO4]).
Theory
COSMO-RS Theory: COSMO-RS stand for Conductor-like Screening Model for

Real Solvents which is a unique method for a prior prediction of thermodynamics
phase behavior of liquid mixture. This method was developed to give a general
prediction without using any experimental data. The prediction is performed by
screening charge density of individual molecules to calculate their chemical
potential. In the quantum chemical calculation, the activity coefficient of molecules
can be calculated by dividing the molecules according to their unique area and
charge and get their ratio of charge segment and area segment [2]. The activity
coefficient for the component is the total of their segment activity coefficient. By
using the screening charge density of molecules, COSMO-RS need no functional
[email protected]
Separation of Pyrrole from Isododecane … 27
group parameters. The prediction and experimental values can be compared by

root-mean-square deviation (RMSD), which can be represented as:
" j 2 #12
Xa X b X 2
xik ^xikj
RMSD ð%Þ ¼ 100 ð1Þ
k¼1 i¼i j¼1
2mc
where a is the number of tie lines, b is the number of components; and 2 is the
number of phases. While xikj and ^xikj are the experimental and predicted values of
composition for component i at kth tie-line in phase j, respectively.
NMR Spectroscopy Theory: NMR spectroscopy is Nuclear Magnetic
Resonance spectroscopy which is one of the most information-rich techniques
because it exhibits a number of important characteristics for identification and
quantitative analysis. The basis of this technique is the resonance interaction
between high frequency field and nuclei of a compound placed in an external
magnetic field. The nuclei of isotope which is hydrogen that has a spin quantum
number of 1/2 will possess a magnetic moment by the spin of the nucleus [3].
Generally, it can be said that this analysis is for the measurement for the peak areas
of the hydrogen molecule (1H) of each component. The concentration of each
component can be calculated using the equation below [4]:
Hi
xi ¼ P3 ð2Þ
i¼1 Hi
where xi is the composition of component i; and Hi is the peak area of single

hydrogen atom in component i.
Methodology
Chemicals: Pyrrole was supplied by Sigma Aldrich, with a purity of 98 %. 1-ethyl-

3-methylimidazolium ethylsulfate ([EMIM][EtSO4]) with purity of ≥95 % was
supplied by Sigma Aldrich while 1-ethyl-3-methylpyridinium ethylsulfate([EMPy]
[EtSO4]) was supplied by Merck. Isododecane with 99 % purity was supplied by
Sigma Aldrich as well as deuterated chloroform (CDCl3) with purity >99.8 %
which was used in 1H NMR analysis of extract and raffinate phase.
Procedure: 10 pieces of 20 ml culture tubes were prepared. A culture tube was
put on the weighing balance and it was tare to zero. A 5 ml syringe was used to
transfer nitrogen compound into the culture tube for about 0.10 g of mass. The mass
of nitrogen compound was recorded and tare it to zero again to measure the mass of
next compound. A 10 ml syringe was used to transfer isododecane into the same
culture tube for about 1.90 g. The mass of isododecane in the culture tube was
recorded and the weighing balance was tare to zero again. A 10 ml syringe was
[email protected]
used to transfer ionic liquid for about 2.00 g into culture tube. The mass of ionic
liquid was recorded. The culture tube was then covered with a screw cap and
secured with parafilm tape to avoid loss due to evaporation. The screw cap was
labeled with the composition of nitrogen compound in the feed (5 %). The mixture
in the culture tube was mixed by shaking the tube. The same steps were repeated to
prepare for the other feed compositions (10–50 %). 10 culture tubes that were
prepared were put in a shaker port at which the speed and temperature of the shaker
were set at 130 rpm and 298.15 K, respectively. The culture tubes were kept in the
shaker port for about 6 h. After 6 h of mixing, let the mixture rest to settle for at
least 12 h. For NMR analysis, a syringe was used to transfer 3–4 drops of the top
layer (raffinate phase) into the NMR tube with 0.5 ml of CDCl3 solvent. The NMR
tube was then capped and sealed with parafilm tape and ready to proceeds with 1H
NMR analysis. The steps were repeated for bottom layer (extract phase) as well as
for other compositions.
Results and Discussion
LLE Studies
The ternary diagrams are presented in Figs. 1 and 2. It can be observed that in all
ternary systems, the slopes of tie-lines are all positive and they are increasing with
the solute (pyrrole) concentration. This indicates that the lower the solute con-
centration, the higher the solvent (IL) concentration required to extracts the solute.
From the diagrams, it can also be seen that as the tie-lines length increases, the
higher the immiscibility. In LLE studies, there two type of ternary system. Type I
system consist of a system with one pair of compound that partially miscible and
one immiscible region. Type II system is a system with two pair of compound that
partially miscible and one immiscible region. In this case, the tie-lines for all
systems are exhibiting both type of system, where the nitrogen compound and
hydrocarbon is partially miscible in ionic liquid while ionic liquid is immiscible in
hydrocarbon. Therefore, the raffinate phase is rich with hydrocarbon while ionic
liquid is rich in extract phase.
Solvent Selection Parameters
The parameters of solvent selection are solute distribution ratio (β) and selectivity
(S). It is significant to know whether the ILs used have high selectivity towards
nitrogen compound or not. Therefore, the ability of IL in extracting nitrogen
compound is measured by obtaining both of the parameters which can be defined by
following equations:
[email protected]
Pyrrole
Experiment
0.0 COSMO-RS model
1.0
0.1
0.9
0.2
0.8
0.3
0.7
0.4
0.6
0.5
0.5
0.6
0.4
0.7
0.3
0.8
0.2
0.9
0.1
1.0
0.0
Isododecane 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 [EMIM][EtSO4 ]
Fig. 1 Experimental tie-lines and COSMO-RS prediction for LLE of ternary diagram system
{[EMIM][EtSO4] (1) + Pyrrole (2) + Isododecane (3)}
Pyrrole
Experiment
0.0
1.0 COSMO-RS model
0.1
0.9
0.2
0.8
0.3
0.7
0.4
0.6
0.5
0.5
0.6
0.4
0.7
0.3
0.8
0.2
0.9
0.1
1.0
0.0
Isododecane 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 [EMPY][EtSO4]
Fig. 2 Experimental tie-lines and COSMO-RS prediction for LLE of ternary diagram system
{[EMPy][EtSO4] (1) + Pyrrole (2) + Isodecane (3)}
[email protected]
xII2
Solute distribution ratio; b ¼ ð3Þ
xI2
xII2 xI1
Selectivity; S ¼ ð4Þ
xI2 xII1
where x refers to the concentration or mole fraction; subscript 1 refers to the

hydrocarbon; subscript 2 refers to the ionic liquid; superscript I refers to the
hydrocarbon rich phase (raffinate layer); and superscript II refers to the ionic liquid
rich phase (extract layer). It can be seen that the selectivity at low concentration is
very high. As the concentration increases, the selectivity also increases until some
point where the concentration is high, it decreases. By comparing both of the ILs,
[EMPy][EtSO4] have higher selectivity compared to [EMIM][EtSO4]. On the other
hand, the solute distribution ratio for all the system is greater than unity (>1). This
indicates that the nitrogen compounds are successfully extracted by ILs. The higher
the concentration of nitrogen compound, the lower the quantity of IL required to
extracts the nitrogen compound. Since the nitrogen content in the commercial diesel
is very low, great amount of IL is required for the extraction process.
Denitrification Efficiency
The tendency of nitrogen removal can be represented by denitrification efficiency

which is defined in the following equation [5]:
xinitial xfinal
Denitrification efficiency ð%Þ ¼ 100 ð5Þ
xinitial
where x refers to the concentration of solute.

In this case, IL does not present in the raffinate phase. The denitrification effi-
ciency decreases with increasing of the solute concentration. Therefore, it can be
said that at the lower solute concentration the efficiency is very high. Some
researcher found that the efficiency can be increased by increasing IL: Oil mass
ratio [5]. This means that the amount of IL have to be increased in order to get
higher nitrogen removal efficiency. However, it is less viable to be operated. Thus,
multi-steps denitrification efficiency have to be considered and the ideal IL: Oil
mass ratio is 1 because the value beyond this is not economic for application in the
industry. Denitrification efficiency of 94.26 % for [EMIM][EtSO4](1) + Pyrrole
(2) + Isododecane(3) system and 95.06 % for [EMPy][EtSO4](1) + Pyrrole
(2) + Isododecane(3) system.
[email protected]
COSMO-RS Prediction
Figures 1 and 2 are illustrated to compare the experimental LLE data with COSMO-
RS prediction of the ternary systems. At the immiscible region, the shape and length
of immiscible region are slightly deviated, particularly in extract phase. The
deviation may due to interactions between pyrrole with IL in the mixture where the
hydrogen bonding plays a main role in the interaction. From Figs. 1 and 2, it can be
observed that the experimental value for pyrrole is nearly matched to the COSMO-
RS prediction at the raffinate phase. Pyrrole is 5-membered ring nitrogen compound
while pyridine is 6-membered ring nitrogen compound. Therefore, pyrrole shows
higher selectivity because it has higher electronegativity of heteroatom. The
nitrogen atom in pyrrole molecule contains lone-pair electrons which are not tied up
in the π–electron cloud of the ring [6]. The nitrogen atom has hydrogen atom
attached to it in which it can interact with the electronegative atom in IL. Thus,
experimental LLE data for pyrrole is nearly matched to the prediction at the raffi-
nate phase. The deviation of experimental LLE data from COSMO-RS prediction is
measured by root-mean-square deviation (RMSD). The RMSD is 1.12 % for
[EMIM][EtSO4](1) + Pyrrole(2) + Isododecane(3) system; 2.51 % for [EMPy]
[EtSO4](1) + Pyrrole(2) + Isododecane(3) system. The average RMSD for all ter-
nary system is 2.825 %.
Conclusion
In this work, the following ternary systems, which are [EMIM][EtSO4](1) + Pyrrole
(2) + Isododecane(3), [EMPy][EtSO4](1) + Pyrrole(2) + Isododecane(3) system
were studied. The experimental LLE data were generated in ternary diagram cor-
responding with COSMO-RS prediction. Generally, the tie-lines for all the ternary
system have positive slope. The selectivity of ionic liquid is higher at low solute
concentration. Ionic liquid [EMPy][EtSO4] have higher selectivity compared to
[EMIM][EtSO4]. The solute distribution ratio is greater than unity for all the sys-
tem. The denitrification efficiency for system containing pyrrole is higher by using
[EMPy][EtSO4] with denitrification efficiency of 94.26 and 95.06 % respectively.
Lastly, the experimental data were predicted with COSMO-RS model which give
an average RMSD of 2.825 % for all ternary system.
Acknowledgment The modeling and experimental work reported in this article was financially
supported by a research grant number RP006E-13SUS under University Malaya Research Grant
(UMRG) from University of Malaya, Malaysia.
[email protected]
References
1. Anantharaj R, Banerjee T (2010) COSMO-RS-based screening of ionic liquids as green

solvents in denitrification studies. Ind Eng Chem Res 49:8705–8725
2. Anantharaj R, Banerjee T (2013) Liquid-liquid equilibrium studies on the removal of thiophene
and pyridine from pentane using imidazolium-based ionic liquids. J Chem Eng Data 58:
829–837
3. Perry RH, Green DW, Maloney JO (1984) Perry’s chemical engineers’ handbook, 6th edn.
McGraw-Hill Book Company, New York, pp 4–17
4. Ravilla UK, Banerjee T (2012) Liquid liquid equilibria of imidazolium based ionic
liquid + pyridine + hydrocarbon at 298.15 K: experiments and correlations. Fluid Phase
Equilib 324:17–27
5. Wilfred CD, Chong FK, Man Zakaria, Azmi Bustam M, Ibrahim MMM, Chan ZP (2012)
Extraction of dibenzothiophene from dodecane using ionic liquids. Fuel Process Technol
93:85–89
6. Anantharaj R, Banerjee T (2011) Fast solvent screening for the simultaneous hydrodesulfu-
rization and hydrodenitrification of diesel oil using ionic liquids. J Chem Eng Data 56:
2720–2785
[email protected]
The Development of Amine-Based
Potential Kenaf Sorbent for Carbon
Dioxide (CO2) Capture
Nabilah Zaini and Khairul Sozana Nor Kamarudin
Abstract Adsorption process becomes a promising, cost-effective and efficient

separation technique for CO2 capture. Inspired by the commercialized technique for
capturing and sequestering CO2, the development of solid amine sorbents relies on
support material has become a great interest amongst researchers worldwide. Based
on this new invention, a research on kenaf as cost-potential agricultural source was
conducted to study the adsorptive performance towards carbon dioxide (CO2). The
research work was initiated by impregnating a series of amine groups such as MEA,
DEA and MDEA on kenaf via wet impregnation method. The CO2 adsorption study
was carried out by flowing 300 cm3/min of purified CO2 to the PSA column and
compressed up to 1.5 bar. Result obtained indicates that MEA has achieved the
highest amount of CO2 adsorbed amongst the other types of amines. Therefore,
MEA was selected for further study by varying the ratio of kenaf to MEA as 1:0.5,
1:0.7, 1:1, 1:2, 1:5, 1:7 and 1:10. It is shown that the ratio of kenaf to MEA of 1:1
have recorded the highest amount of CO2 capture capacity (2.07 mmol CO2/g
sorbent) as compared to 1:0.5, 1:0.7, 1:2, 1:5, 1:7 and 1:10 with 0.78, 1.71, 1.77,
1.20, 0.96 and 0.89 mmol CO2/g sorbent, respectively. This study inferred that the
introduction of MEA has enhanced the adsorptive capability of kenaf towards CO2.
Keywords Kenaf Amine Impregnation Adsorption CO2 Pressure swing

adsorption (PSA)
N. Zaini (&) K.S.N. Kamarudin

Faculty of Petroleum and Renewable Energy Engineering, Universiti Teknologi Malaysia,
UTM Skudai, 81310 Johor Bahru, Johor, Malaysia
K.S.N. Kamarudin

[email protected]
34 N. Zaini and K.S.N. Kamarudin
Introduction
For many years, there has been a growing concern about global warming and
climate change that prompted the development towards more-efficient and
improved processes for carbon dioxide (CO2) capture from various sources. The
greenhouse gases that mainly consist of CO2 have increased by 30 % over the past
200 years [1]. CO2 capture from large point sources was identified as a major option
to address the problem and known as CO2 capture and storage [2]. CO2 capture and
storage (CCS) technology was implemented via amine-based absorption, cryogenic
techniques, membranes diffusion, adsorption by using solid sorbent and oxygen
recovery from O2/CO2 recycle combustion [3]. Among of these options, amine-
based regenerative chemical absorption processes have been widely practiced for
capturing CO2. However, this technology exists several drawbacks such as low CO2
loading capacity, high energy consumption, high equipment corrosion rate, amine
degradation by SO2, NO2 and O2, large equipment size required, solvent leakage,
fouling of the process equipment, have limited amine concentration in aqueous
phase because of the viscosity and foaming issues and possible environmental
problems due to high volatility of amine [4–9]. Therefore, the cost-effective and
efficient process for CO2 capture should be replaced the existing technology in
order to mitigate the global warming issues.
The economic crisis has led the researchers to turn their interest and expertise
into adsorption field. Adsorption is a promising method since it is cost-effective and
efficient CO2 separation technique. It requires less regeneration energy than aque-
ous amine-based process, higher adsorption capacity and selectivity towards CO2,
low heat capacity than aqueous amine solvent and ease of handling [2]. Inspired by
the commercialized technique for capturing and sequestering CO2, the development
of solid amine sorbents had become a great interest amongst researchers worldwide.
The introduction of green material as a sorbent that originated from agricultural
sources makes them very competitive since it is cost-effective as compared to the
existing amine-based support. Therefore, a research on kenaf as agricultural source
was conducted to study the adsorptive performance towards CO2. To the best of the
authors’ knowledge, there are no literatures on the modification of kenaf by using
amine functional group for CO2 capture. Thus, this paper presents the modification
of kenaf by impregnating different types of alkanolamines onto kenaf surface in
order to determine the CO2 adsorption capacity.
Experimental
Material and Chemicals, In this study, kenaf core without outer periphery bark
with particle size larger than 6 mm was supplied by National Kenaf and Tobacco
Board (NKTB), Kelantan. The sample was then grinded into smaller particle size
(300 µm). The sample was stored in air-tight container to prevent moisture build up
[email protected]
The Development of Amine-Based Potential Kenaf Sorbent … 35
and fungi infections and was used for impregnation procedure without having
pre-treatment process. All chemicals used in this study were of highly concentrated
reagent grade; MEA, DEA and MDEA. Methanol (CH3OH, Merck) was selected as
solvent in amine impregnation procedure since it has lower boiling point (64.7 °C)
than water (100 °C). This initial adsorption study involves the use of purified CO2
(99.999 %, Mega Mount Industrial Gases Sdn Bhd) and purified N2 (99.999 %,
Mega Mount Industrial Gases Sdn Bhd) for regeneration process.
Amine-Impregnation Procedure, The synthesis of amine-impregnated kenaf

was carried out via conventional method known as wet impregnation method.
Initially, kenaf sample was wetted in methanol solution according to the ratio solid
to liquid of 1:20 (weight percent) and the mixture was maintained in contact under
mechanical agitation of 600 rpm for 20 min in two stages prior to be filtered and air-
dried overnight. Alcoholic amine solution was prepared by stirring MEA with
methanol for 20 min. Then, the dried kenaf sample was added into the alcoholic
amine solution for 15 min prior to be agitated for 5 h impregnation. Finally, the
product of MEA-impregnated kenaf was filtered and air-dried before being stored in
desiccators for further analysis. The similar procedure was repeated for DEA and
MDEA.
CO2 Adsorption and Regeneration Study, In this study, the CO2 adsorption
process was carried out by using a single-column pressure swing adsorption system
(PSA). The column used has a dimension of 15 cm height and 1 cm diameter. The
PSA operation system involves four phases; pressurization, adsorption, desorption
and purging. The adsorption process was conducted at 1.5 bar with a feed gas of
300 cm3/min for 5 min. Whilst, the regeneration process was implemented by
changing the pressure to 1 bar. The schematic diagram of single-column PSA
operation system is shown in Fig. 1. In this study, the adsorption process was
V2
GFC 2
PG
V3
GC
GFC 1 BED
PC
VENT
V4
N2 CO2 V1
Fig. 1 Schematic diagram of single-column PSA system
[email protected]
executed in three (3) consecutive cycles of operation. Gas chromatography (7820A)

from Agilent Technologies was used to measure the amount of gas leaving the PSA
system in each cyclic operation.
Result and Discussion
CO2 Adsorption Study, In this study, a series of amine groups containing

N-functional groups (MEA, DEA and MDEA) have been studied for the adsorption
of CO2. Initially, a sample containing 50 wt% amine-impregnated kenaf was
pressurized with purified CO2 at 1.5 bar in order to determine the highest types of
amine for capturing CO2. The results were presented in Fig. 2.
Figure 2 presents a comparison of CO2 capture capacity for MEA, DEA and
MDEA impregnated onto kenaf sorbent. It is demonstrated that kenaf that has been
impregnated by MEA has achieved the highest CO2 capture capacity (0.78 mmol
CO2/g) compared to DEA and MDEA with 0.68 and 0.66 mmol CO2/g, respec-
tively. Generally, the structures of alkanolamines containing at least one hydroxyl
group (–OH) and one amine group (–NH2) that reacts directly with CO2 to form
carbamate ions for primary and secondary amines and bicarbonate or carbonate for
tertiary amine. The reaction of primary and secondary amines with CO2 can be
represented in Eq. (1). Whilst, the reaction of tertiary amine with CO2 is indicated
in Eq. (2) [9].
2RR0 NH þ CO2 $ RR0 NCOO ðcarbamateÞ þ RR0 NHþ

2 ð1Þ
RR0 NCOO þ H2 O $ RR0 NH þ HCO

3 ðbicarbonateÞ ð2Þ
where R = C2H4OH. The ability to form carbamate ions result from a direct reaction
of primary and secondary amine with CO2 could produces faster CO2 capture
kinetics [10].
Fig. 2 CO2 capture capacity 1.0

CO 2 capture capacity (mmol/g)
for different types of

alkanolamines on kenaf at 0.78
0.8
1.5 bar 0.68 0.66
0.6
0.4
0.2
0.0
MEA DEA MDEA
[email protected]
Although the formation of carbamate ions proceed rapidly, the CO2 capture
capacity for MEA and DEA is reduced according to the stoichiometry requirement
of two amine molecules for each CO2 molecule reacted (Eq. 1). Since the adsorption
time is only 5 min, the tertiary amine did not transform into carbamates because of
slow reaction rate. Therefore, the adsorbent has low affinity towards CO2 which is
shown in Fig. 2. The impregnation of amine has introduced nitrogen functional
groups on kenaf that provides the additional sites for CO2 adsorption; hence improve
the capture capacity towards CO2. Moreover, this result could also be attributed to
the effect of molecular size since MEA has the lowest size of molecule (61.08 g/mol)
compared to DEA (105.14 g/mol) and MDEA (119.16 g/mol). The intrusion of
MEA could enhance the basicity of kenaf support through the numbers of amine
impregnated. In the other perspective, it can be said that the smaller size of amine
could increase the numbers of amine impregnated on kenaf support and vice versa.
This argument can be proved through the pH value of decanted amine solution
gained during experimental work. The pH values for each prepared sample were
recorded accordingly as tabulated in Table 1. It is clear that basicity of the sample
increases by ascending order of MDEA < DEA < MEA. The higher basicity of
prepared sample will increase the tendency as well as affinity towards acidic CO2,
thereby enhancing the capability of capturing CO2. To further study the adsorption
characteristics of CO2 on amine-impregnated kenaf, the amount of MEA was
increased up to 10 times of their weight and was presented in Fig. 3.
Table 1 Parameters for amine-impregnated kenaf sample

Sample pH values
Alcoholic amine solution Decanted alcoholic amine solution
50 wt% MEA 11.23 10.60
50 wt% DEA 10.62 10.24
50 wt% MDEA 10.21 10.18
2.4
CO2 capture capacity (mmol/g)
2.2 2.07
2.0 1.77
1.8 1.71
1.6
1.4 1.20
1.2
0.96
1.0 0.89
0.78
0.8 0.62
0.6
0.4
0.2
0.0
Raw 1 kenaf : 1 kenaf : 1 kenaf : 1 kenaf : 1 kenaf : 1 kenaf : 1 kenaf :
kenaf 0.5 MEA 0.7 MEA 1 MEA 2 MEA 5 MEA 7 MEA 10 MEA
Fig. 3 CO2 capture capacity for different MEA concentration on kenaf at 1.5 bar
[email protected]
From Fig. 3, it can be inferred that the performance of kenaf after MEA
modification increased significantly as compared to the raw kenaf. This is due to the
additional basic adsorption sites exist on the surface that was discussed previously.
Then, the MEA concentration was varied to 0.5, 0.7, 1, 2, 5, 7 and 10 times of
sorbent weight. Results gained show that the ratio of kenaf to MEA as 1:1 indicates
higher CO2 capture capacity (2.07 mmol CO2/g) compared to other ratios (1:0.5,
1:0.7, 1:2, 1:5, 1:7 and 1:10) with the capture capacity of 0.78, 1.71, 1.77, 1.20,
0.96 and 0.89 mmol CO2/g, respectively. It might be due to the effect of the number
of adsorption sites on sorbent. The increment of adsorption sites is directly pro-
portional to the MEA concentration up to 1:1. Further increased the MEA
concentration might block the pore channel of kenaf, less adsorption due to the
effect of steric hindrance. This finding reveals that the ratio of kenaf to MEA of 1:1
is the best concentration for MEA impregnation on kenaf sorbent. Further increase
of MEA concentration could reduce the CO2 capture capacity of MEA-modified
kenaf. It can be conveyed that the additional concentration of MEA could not give
any increment to the number of basic adsorption sites.
Regeneration Study, Regeneration study is a crucial and important step in order
to ensure the adsorbent can be used for the next cycle operation. In spite of having
high adsorption capacity towards adsorbate, the adsorbent must also display a stable
adsorption performance for cyclic operation. In this study, the regeneration process
was implemented by reducing the adsorption pressure of 1.5 bar to 1 bar for three
(3) consecutive cycles of operation. The regeneration performance for each sample
is shown in Fig. 4. It can be inferred that the adsorption capacity was almost
constant for three (3) consecutive cycles of adsorption and desorption. The results
indicate that the adsorption performance was stable under those conditions.
2.5
CO2 Capture capacity (mmol /g)
cycle 1
2.0 cycle 2
cycle 3
1.5
1.0
0.5
0.0
1 kenaf:0.5 1 kenaf:0.7 1 kenaf:1 1 kenaf:2 1 kenaf:5 1 kenaf:7 1 kenaf:10
MEA MEA MEA MEA MEA MEA MEA
Amine concentration (g/g)
Fig. 4 Regeneration process for different MEA concentration
[email protected]
Conclusion
This finding revealed that the functionalized of kenaf by MEA could improve the
adsorption of CO2 as compared to the raw kenaf. The optimum CO2 uptakes can be
achieved as high as 2.07 mmol CO2/g for a ratio of kenaf to MEA of 1:1 at 1.5 bar
and 298 K. The high adsorption of CO2 on MEA-modified kenaf can be ascribed by
the presence of nitrogen-functionalized group as well as the chemical adsorbate-
adsorbent interaction. However, further increase the amount of MEA loadings on
kenaf could reduce the CO2 capture capacity. This study also shows that MEA-
modified kenaf have a stable cyclic adsorption/desorption performance up to three
(3) cycles of operation. Indeed, kenaf with large pore size could be a good and
promising host for the development of novel CO2 adsorbent.
Acknowledgments The authors gratefully acknowledge the financial support given by Ministry
of Education Malaysia (MOE), Ministry of Science, Technology and Innovation (MOSTI) Vote
no. 4S047 and Universiti Teknologi Malaysia (UTM). An appreciation also dedicated to National
Kenaf and Tobacco Board (NKTB), Kelantan for providing the kenaf sample.
References
1. IPCC (2001) Chapters 1 and 2. Climate change: the scientific basis. Contribution of working
group to the third assessment report of the intergovernmental panel on climate change.
Cambridge University Press, Cambridge
2. Samanta A, Zhao A, Shimizu GKH, Sarkar P, Gupta R (2012) Post-combustion CO2 capture
using solid sorbents: a review. Ind Eng Chem Res 51:1438–1463
3. Gray ML, Soong Y, Champagne KJ, Pennline H, Baltrus JP, Steven RW Jr, Khatri R, Chuang
SSC, Filburn T (2005) Improved immobilized carbon dioxide capture sorbents. Fuel Process
Technol 86:1449–1455
4. Màmun S, Nilsen R, Svendsen HF, Juliussen O (2005) Solubility of carbon in 30 Mass %
monoethanolamine and 50 Mass % methyldiethanolamine solutions. J Chem Eng Data
50:630–634
5. Mandal BP, Bandyopadhyay SS (2005) Simultaneous absorption of carbon dioxide and
hydrogen sulphide into aqueous blends of 2–Amino–2–Methyl–1–propanol and
diethanolamine. Chem Eng Sci 60:6438–6451
6. Mandal BP, Bandyopadhyay SS (2006) Absorption of carbon dioxide into aqueous blends of
2–amino–2–methyl–1–propanol and monoethanolamine. Chem Eng Sci 61:5440–5447
7. Yue MB, Sun LB, Cao Y, Wang ZJ, Wang Y, Yu Q, Zhu JH (2008) Promoting the CO2
adsorption in the amine–containing SBA–15 by hydroxyl group. Microporous Mesoporous
Mater 114:74–81
8. Cao C, Zhang K, He C, Zhao Y, Guo Q (2011) Investigation into a gas–solid–solid three phase
fluidized–bed carbonator to capture CO2 from combustion flue gas. Chem Eng Sci 66:375–383
9. Yu CH, Huang CH, Tan CS (2012) A review of CO2 capture by absorption and adsorption.
Aerosol Air Qual Res 12:745–769
10. Filburn T, Helble JJ, Weiss RA (2005) Development of supported ethanolamines and modified
ethanolamines for CO2 capture. Ind Eng Chem Res 44:1542–1546
[email protected]
Potential Effect of Palm Oil Fuel Ash
as Micro-Filler of Polymer Concrete
A.K. Nur Hafizah, M.W. Hussin, M. Ismail, M.A.R. Bhutta,

M. Azman, P.J. Ramadhansyah, A. Nur Farhayu
and A.S.L. Nor Hasanah
Abstract This paper discusses the feasibility of using an agricultural waste called
palm oil fuel ash (POFA) as micro-filler in polymer concrete. As a cellulose waste
material, such potential of POFA is often neglected since it tends to uptake
excessive resin during the mixing process. Hence, the filler characterization was
involved in this study to investigate the potential of POFA become as micro-filler in
polymer concrete. The filler characteristics of unground and ground POFA
(UPOFA and GPOFA) were examined under microstructural examination, which
included particle size analyzer (PSA), morphology image, and termogravimetric
A.K. Nur Hafizah M.W. Hussin (&) A. Nur Farhayu A.S.L. Nor Hasanah
Institute of Smart Infrastructure and Innovative Construction (ISIIC), Faculty of Civil
Engineering, Universiti Teknologi Malaysia, 81310 Johor, Malaysia
A.K. Nur Hafizah
e-mail: fi[email protected]
A. Nur Farhayu
A.S.L. Nor Hasanah
M. Ismail
Departments of Structure and Materials, Faculty of Civil Engineering,
Universiti Teknologi Malaysia, 81310 Johor, Malaysia
M.A.R. Bhutta
Department of Civil Engineering, Faculty of Applied Science, University of British
Columbia, Vancouver, BC V5P 1X6, Canada
M. Azman P.J. Ramadhansyah
Department of Geotechnics and Transportation, Faculty of Civil Engineering,
Universiti Teknologi Malaysia, International Campus, 81310 Johor, Malaysia
P.J. Ramadhansyah

[email protected]
42 A.K. Nur Hafizah et al.
and differential thermal analyses (TGA and DTA). A total of ten mix designs of
polymer concrete were employed for flowability and compression test. Then, the
workability and strength of polymer concrete were investigated after incorporating
two types of filler with different filler content. Two major outcomes had been
achieved. Firstly, the filler had their physical features significantly changed after
physical surface modification. Secondly, finer filler had given superior workability
and compressive strength. In conclusion, the agricultural waste of POFA can be
incorporated as micro-filler in polymer concrete after being physically modified and
mixed with the appropriate mix design to obtain superior mechanical properties.
Keywords Agricultural waste Palm oil fuel ash Filler Polymer concrete
Introduction
Polymer concrete (PC) is produced solely using polymer resin as concrete binder,
dry inert granular of aggregate, and occasionally filler. Such PC mix does not
contain cement materials and water; its hardening takes the form of the polymer-
ization process when additives, catalysts, or accelerators are added. Generally, any
PC with filler tends to have denser packing of the solids concrete which enhances
the mechanical properties. The type of filling materials, i.e., natural, granulated, or
synthetic, affects the composite characteristics [1].
Currently, the incorporation of natural filler originating from calcium carbonate
and sand in PC production is still being debated. In spite of this, the incorporation
of green material with the waste in concrete that has PC enhancing characteristics
has recently come into the lime light as a more attractive alternative to researchers.
The characteristics of PC incorporated with filler are governed by several factors
such as the filler type, its size and shape, and the type as well as amount of binder
used [2]. The filler exists as small particles that fit into the gaps between larger
particles to enhance the mechanical properties of PC, and such has been proven by
several researchers. For instance, Atzeni et al. [3] have shown that the addition of
fly ash as filler in epoxy mortar can enhance their specimens’ mechanical properties
better than conventional quartz flour filler mainly because of the great fineness of fly
ash filler. Other than the aforementioned fly ash, palm oil fuel ash (POFA) is also
attractive alternative filler. POFA can be found in abundance as agricultural waste
that is often uncontrollably dumped in landfills. Nevertheless, not all waste ashes
have the potential to become PC filler. The wrong selection of filler material may
lead to worse PC quality. Not only that, it affects the PC’s workability and process
ability [4]. Most natural waste source from agricultural plant is cellulose [5, 6]
which has a structure that attracts liquid into PC. This can lead to excessive resin
[email protected]
Potential Effect of Palm Oil Fuel Ash … 43
consumption, which is not cost effective, and thus jeopardize the production of PC
even when filler is used. Until today, no study has so far been done on the
incorporation of POFA as PC filler because of the aforesaid potential setbacks.
The objective of this paper is to give an insight into the potential incorporation of
palm oil fuel ash (POFA) as micro-filler into PC, gauged through microstructure
and strength examination. The first outcomes of microstructure experiments on raw
and ground POFA (UPOFA and GPOFA) were compared with previous findings.
The workability and compressive strength of PC with appropriate mix design was
investigated to support the outcomes. The outcomes of this work will be beneficial
to researchers and engineers working in the field of cellulose filler in PC.
Material and Method
Binder, The polymer binders used in this study was the isophthalic unsaturated
polyester resin. Its general properties are shown in Table 1. Polyester resin has been
chosen in this study since it is widely used in the composites industry and is the
cheapest among thermoset resins. Additionally, the resin can be easily handled,
pigmented, filled and fiber reinforced in liquid form [7].
Aggregates and Filler, Crushed coarse aggregate and natural river fine aggregate
were used as the inert granular material. The aggregate was oven-dried at 100 ± 5 °C.
The moisture content of the inert granular material was limited below 0.1 %. The size
of coarse aggregate of between 10 and 12 mm was used in this study.
Micro-filler used in this study was the palm oil fuel ash (POFA) since it was
cheap and readily available. The POFA was collected from the southern part of
Peninsular Malaysia. It was oven-dried at 100 ± 5 °C and the moisture content was
kept constant below 0.1 % before being further divided into two portions, i.e.,
UPOFA and GPOFA. The UPOFA was obtained directly by passing the POFA
through the 300 µm sieve to separate their natural substances and homogenize the
particles size. Meanwhile, the GPOFA was obtained directly by passing the POFA
through the 45 µm sieve and then grinding them. After that, it was grounded using a
laboratory grinding mill with rod bar to obtain finer filler.
Microstructure Examination, Particle size distribution was also analyzed to
observe the changes in size before and after the particle grinding process. For both
GPOFA and UPOFA, this was done using the particle size analyzer that employed
the wetting method. The particle was dispersed using distilled water to avoid
agglomerate condition before the test was run to get convincingly accurate results.
Morphology test was conducted to examine the surface of GPOFA and UPOFA
under field emission scanning electron microscopy (FESEM) with 250 times
Table 1 Properties of isophthalic unsaturated polyester resin

Properties Isophthalic unsaturated polyester resin
Density (g/cm3) 1.1
Viscosity (Brookfield), 25 °C, 60 rpm (cps) 538
[email protected]
Table 2 Mix proportion of polymer concrete

Mix proportion
Polymer Polyester binder POFA filler Fine aggregate Coarse aggregate
concrete type (kg/m3) (kg/m3) (kg/m3) (kg/m3)
PC-GPOFA 132a 78 1240 750
PC-UPOFA 97 1190
116 1141
136 1091
155 1042
a
Polyester binder was mixed with 1 % MEKP and 0.5 % CoNp
magnifications. Before that, both GPOFA and UPOFA specimens were coated so
that significant morphology image could be captured.
Termogravimetric and differential thermal analysis (TGA and DTA) was
employed where automatic thermal analyzer was used to investigate the thermal
behaviour of POFA. The measurement was carried out between 30 and 1000 °C
with heating rate set at 10 °C/min in nitrogenic atmosphere. The finding was similar
since both GPOFA and UPOFA exhibited identical thermal behaviour due to both
POFA from similar origin.
Mix Proportion, Ten mix proportions of PC incorporating POFA as filler are
given in Table 2. Since resin is expensive, the low content of resins (binder) was
fixed at 12 % by weight of resin. Then, they were separately mixed with methyl
ethyl keton peroxide (MEKP) as catalysis (harderner) and cobalt naphthenate
(CoNp) as initiator. The unground and ground POFA micro-filler that occupied the
PC spaces varied from 8 to 16 % by weight of POFA. To achieve the strength of PC
inert granular material, fine and coarse aggregates were used. The coarse aggregate
content was fixed at 30 % by weight of coarse aggregate. The PC without any filler
was designated as the control specimen.
Specimen, The dry PC mix was prepared before the wet mix, and consisted of
inert granular aggregates and filler. The wet mix had additional polyester binder
mixed together with its initiator and promoter. Then, the wet mix was poured into
the dry mix and immediately mixed together using concrete mixer. Releasing agent
was applied onto the mould surface for easy demoulding. The fresh concrete was
thereafter poured into the mould and compacted with a vibrating table for 20 s. The
specimens were subjected to post-curing by oven-heating temperature at 70 °C for
4 h. The specimens were prepared strictly according to JIS A 1181 [8].
Flowability Test, Flowability test was conducted as per JIS R 5201 [9] after
mixing the PC to measure its workability, which was governed by the flowing
ability of actively-mobilized materials with respect to the flow spread. In this study,
the actively-mobilized materials were the polymer binder and filler, which were
used to fill in the gaps of inert materials. Both materials were blended at various
filler content and no compaction was applied during the test. The blended polymer
was poured into v-funnel slump cone (inside diameter of v-funnel con is 70 and
100 mm while the height is 50 mm) and immediately uplifted once the cone was
[email protected]
full of blended polymer. The flow spread diameter was measured five times in every
test after the flow spread ceased. This method was employed since PC is sticky and
has zero slumps. Similar test had been conducted to examine the flowing ability of
self-compacting mortar using v-funnel slump cone by Safiuddin et al. [10], which
proposed that the flow spread could be indicative of the workability of PC.
Strength Test, The cured specimens were tested for compressive strength
according to JIS A 1181 [8]. A total of 66 cubical specimens including control
specimens (without any filler) with 100 mm by 100 mm in cross-section were
prepared for the compression test. Compression test was carried out by using
compression machine with capacity of 200 kN. The appropriate mix design at
different filler content was investigated at this point.
Particle Size Distribution, The particle size distribution curve of UPOFA and
GPOFA are shown in Fig. 1. The distribution showed that both UPOFA and
GPOFA had well-graded particles present as both coarse and fine micro-particle,
but more than 90 % had passed through 300 and 45 µm sieve, respectively. The
mean particle size of GPOFA was 14.21 µm and that of UPOFA was 86.00 µm.
Smaller particle size was obtained by grinding. These were confirmed clearly in the
FESEM image as shown in Fig. 4a. The UPOFA was considered as having coarse
micro-particle since the original mean particle size in the study of Bakar et al. [11]
was not more than 18 µm; the UPOFA was comparatively coarser in this sense.
Particle Morphology, Fig. 2a, b shows the morphology of GPOFA and
UPOFA, respectively. The morphology results indicated that the UPOFA had a
combination of spherical and irregular shaped particles with a sizeable fraction and
porous cellular structures. Meanwhile, GPOFA had more uniform spherical shaped
particles as compared to UPOFA. The morphology of UPOFA was similar to that
found by Noorvand et al. [12] using SEM. The opening size of pores for UPOFA
ranged between 8.3 and 17 µm; this opening size may have affected the workability
Fig. 1 Particle size GPOFA UPOFA

distribution 100
80
Passing (%)
60
40
20
0
0.01 0.1 1 10 100 1000 10000
Diameter (µm)
[email protected]
Fig. 2 FESEM image of ground POFA (a) and unground POFA (b)
of fresh PC since it has higher tendency of absorbing liquid resin. This is in

agreement with Haidar et al. [13]. Other than that, it may have increased the number
of voids and capillaries in the structure, thus resulting in the reduction of densely
pack structure [12]. The existing UPOFA opening pore size also showed its original
particle features as cellulose based agricultural waste, which can be physically
modified by grinding to make it finer and reduce the particle pores (see Fig. 2a).
Additionally, a known fact, the fineness of particle give the highest surface area and
this was supported by Bakar et al. [11]. This also demonstrated the potential of
GPOFA as filler in PC which is higher than UPOFA.
Termogravimetric and differential thermal analysis (TGA and DTA), TGA
and DTA were obtained in concomitant in a single test as shown in Fig. 3. The
Figure has also depicted that the first DTA intensity peak of POFA occurs at the
boiling point of H2O, i.e., 100 °C. Major POFA weight loss was observed between
500 and 780 °C, which happened at the second DTA intensity peak. The significant
weight loss of POFA was interpreted to be due to the production of silica. In PC any
inherent moisture should be eliminated to enhance the bonding between materials,
102 0.001
100 TGA
DTG
Deriv. weight (mgs-1)
98 0.0005
Weight loss (%)
96
94
TGA
DTA
0
92
90
88 -0.0005
86
84 -0.001
0 200 400 600 800 1000
Temperature (oC)
Fig. 3 TGA and DTA analysis for POFA
[email protected]
Fig. 4 Flow spread diameter PC-GPOFA PC-UPOFA
Flow spread diameter (mm)

of blended polymer with 500
ground and unground POFA
400
300
200
100
0
0 8 10 12 14 16
Filler content (%)
and the moisture content in this regard is usually controlled below than 0.1 % at
100 ± 5 °C [14–16]. These conditions were kept in this study. The TGA and DTA
examinations are very important to gauge the POFA’s performance as PC filler at the
post-curing process and under high pressure processing [17, 18]. Additionally, the
results indicated that POFA is able to withstand thermal degradation up to 590 °C.
Workability, Fig. 4 presents the flow spread diameter of blended polymer
(polymer binder and POFA filler) as obtained in the flowability test. Higher flow
spread diameter indicates higher flowability and thus higher workability of PC and
vice versa. From the results, the overall flowability of various blended polymer had
gradually decreased when the filler content was increased. The decrease in flowability
of blended polymer reduced about 12 and 23 %, respectively for PC-GPOFA and PC-
UPOFA, respectively, as compared to neat polymer. Nevertheless, generally, the
blended polymer with GPOFA had shown superior flowability to that of blended
polymer with UPOFA. This showed that the grinding of POFA had significantly
improved the flowability of blended polymer. The worst flowabilty exhibited by
unground POFA (UPOFA) was predominantly caused by the open cellulose structure
of original POFA, as seen in Fig. 4b. Such structure has encouraged excessive uptake
of polyester resin and this has been observed by Bignozzi et al. [4] as well.
Compressive Strength, The compressive strength of different PCs with various
filler content is shown in Fig. 5. Generally, the filler had increased the specimens’
compressive strength approximately up to 57 and 18% for both PC-GPOFA and
Fig. 5 Compressive strength PC-GPOFA PC-UPOFA

of polymer concrete with 70
Compressive strength,σ c
ground and unground POFA 60

50
(MPa)
40
30
20
10
0
0 8 10 12 14 16
Filler content (%)
[email protected]
PC-UPOFA, respectively, as compared to neat PC. uced about 12 and 23 %,

respectively for PC-GPOFA and PC-UPOFA, respectively, as compared to neat
polymer. Moreover, the results demonstrated that an increment in ground POFA
added could give superior compressive strength for all POFA. Lower compressive
strength was obtained with unground POFA because of the earlier mentioned ori-
ginal cellulose-pores structure of POFA. This had caused a decrease in the concrete
mix’s workability. Results had also indicated that the cellulose-pores structure in
original POFA might have taken up the liquid resin; such is in agreement with the
findings of Bignozzi et al. [4], Haidar et al. [13]. Additionally, the compressive
strength had been influenced by the workability of the PC, which had increased
when GPOFA was added. However, the optimum filler content necessary for
Iso-GPOFA and Iso-UPOFA was 14 and 10 %, respectively. This demonstrated
that GPOFA had satisfactorily filled in the gaps between inert materials. In
conclusion, the agricultural waste of POFA has high potentiality to be used as
filling material in PC after being physically modified.
Conclusion
This paper had described the potential of agricultural POFA waste as filler for PC
with the following conclusions drawn:
1. The grinding process would give satisfactory POFA fineness.
2. The original cellulose-pores structure of POFA had significantly diminished
after grinding.
3. POFA is able to withstand thermal degradation up to 590 °C.
4. Ground POFA (GPOFA) had increased the flowability of blended polymer and
enhanced the workability and strength of PC.
5. The agricultural waste of POFA has the potential to be used as micro-filler in PC
after being subjected to physical surface modification.
Acknowledgment The authors wish to thank the Ministry of Education, Malaysia (MOE) and
Research Management Centre (RMC), Universiti Teknologi Malaysia (UTM) for sponsoring this
project under the research university grants (RUG) No. Q.J130000.7122.03H35 and No.
Q.J130000.2509.06H56. The authors are also thankful to the staff of Structures and Materials
Laboratory, Faculty of Civil Engineering for the facilities and support for experimental work.
References
1. Ateş E, Barnes S (2012) The effect of elevated temperature curing treatment on the
compression strength of composites with polyester resin matrix and quartz filler. Mater Des
34:435–443
[email protected]
2. Roger NR, Hancock M (2003) General principles guiding selection and use of particulate
materials. In: Rothon RN (ed) Particulate-filled polymer composites, 2nd edn. Rapra
Technology Limited, UK, pp 5–51
3. Atzeni C, Massidda L, Sanna U (1990) Mechanical properties of epoxy mortars with fly ash as
filler. Cement Concr Compos 12(1):3–8
4. Bignozzi M, Saccani A, Sandrolini F (2000) New polymer mortars containing polymeric
wastes. Part 1. Microstructure and mechanical properties. Compos A Appl Sci Manuf 31(2):
97–106
5. Raveendran K, Ganesh A, Khilar KC (1996) Pyrolysis characteristics of biomass and biomass
components. Sci Technol Fuel Energy 75(8):987–998
6. Kaddami H, Dufresne A, Khelifi B, Bendahou A, Taourirte M, Raihane M, Sami N (2006)
Short palm tree fibers—thermoset matrices composites. Compos A Appl Sci Manuf 37(9):
1413–1422
7. Yang H, Lee LJ (2001) Comparison of unsaturated polyester and vinylester resins in low
temperature polymerization. J Appl Polym Sci (January 2000) 1230–1242
8. Japanese Industrial Standard (2005) Test methods for polymer concrete. JIS A 1181, Japan
9. Japanese Industrial Standard (1997) Physical testing method for cement. JIS R 5201, Japan
10. Safiuddin M, West JS, Soudki KA (2011) Flowing ability of the mortars formulated from self-
compacting concretes incorporating rice husk ash. Constr Build Mater 25(2):973–978
11. Bakar BHA, Ramadhansyah PJ, Azmi MJM (2011) Effect of rice husk ash fineness on the
chemical and physical properties of concrete. Mag Concr Res 63(5):313–320
12. Noorvand H, Abdullah A, Ali A, Demirboga R, Farzadnia N (2013) Physical and chemical
characteristics of unground palm oil fuel ash cement mortars with nanosilica. Constr Build
Mater 48:1104–1113
13. Haidar M, Ghorbel E, Toutanji H (2011) Optimization of the formulation of micro-polymer
concretes. Constr Build Mater 25(4):1632–1644
14. Mun KJ, Choi NW (2008) Properties of poly methyl methacrylate mortars with unsaturated
polyester resin as a crosslinking agent. Constr Build Mater 22(10):2147–2152
15. Jo B, Park S, Kim C (2006) Mechanical properties of polyester polymer concrete using
recycled polyethylene terephthalate. Constr Build Mater 103:219–226
16. Jo B-W, Park S-K, Park J-C (2008) Mechanical properties of polymer concrete made with
recycled PET and recycled concrete aggregates. Constr Build Mater 22(12):2281–2291
17. Kubota H, Motors G (1975) Curing of highly reactive polyester resin under pressure: kinetic
studies by differential scanning calorimetry. J Appl Polym Sci 19:2279–2297
18. Waigaonkar S, Babu BJC, Rajput A (2011) Curing studies of unsaturated polyester resin used
in FRP products. Indian J Eng Mater Sci 18(February):31–39
[email protected]
Liquid Membrane Formulation
for Succinic Acid Extraction
from Simulated Aqueous Waste Solution
Norela Jusoh, Norasikin Othman and Nur Alina Nasruddin
Abstract Succinic acid is developing as one of the most competitive new bio-
based chemicals. However, the biological production of succinic acid results in
difficult product recovery and not economically competitive. Thus, the development
of economical method to extract and purify succinic acid is really essential. One of
the promising methods for succinic acid extraction is emulsion liquid membrane
(ELM). In this study, the ELM process will be develop to recover succinic acid
from simulated aqueous waste solution through screening of ELM components; the
carrier, diluent and stripping agent using liquid-liquid extraction method. Several
parameters such as the possibility of using palm oil as diluent, types of carrier and
their concentration, and stripping agents were examined. Outcomes indicate that
21.39 g/l of succinic acid and 0.41 g/l of acetic acid was extracted using Amberlite
LA2 in palm oil. Besides, 16.34 g/l succinic acid was recovered selectively over
acetic acid by Na2CO3 in stripping process. Therefore, it can be concluded that
Amberlite LA2, palm oil, and Na2CO3 has high possibility to be used in liquid
membrane formulation for succinic acid extraction.
Keywords Liquid membrane formulation Amberlite LA2 Succinic acid

Aqueous waste solution
N. Jusoh N. Othman (&) N.A. Nasruddin

Centre of Lipid Engineering and Applied Research (CLEAR), Department of Chemical
Engineering, Faculty of Chemical Engineering, Universiti Teknologi Malaysia,
81310 Johor Bahru, Johor, Malaysia
N. Jusoh
N.A. Nasruddin

[email protected]
52 N. Jusoh et al.
Introduction
Succinic acid is developing as one of the most competitive of the new platform
chemicals. The suitability of succinic acid as a platform chemical is based on the
ease of its biotechnological production and the huge variety of chemicals that can
be produced from it by chemical conversion. Some of the molecules that can be
derived from succinic acid are 1,4-butanediol, maleic anhydride, 2-pyrrolidinone,
and tetrahydrofuran. These molecules can be applied in a wide range of application,
such as in polymers [1], industrial solvents, and specialty chemicals [2]. Moreover,
succinic acid also can be applied in the food industry as flavoring agents [3] and as
the counterion in many pharmaceutical applications [4].
Nevertheless, the production of succinic acid through biological process is not
economically competitive because of low product concentration. In addition, the
existence of acetic acid as major by-product complicates the product recovery. In
order to increase the competitiveness of the biotechnological process over petro-
chemical process, the minimization of the production cost is very crucial. Various
methods have been proposed for succinic acid recovery, including crystallization
[5], liquid-liquid extraction [6], and membrane filtration [7], but these methods are
constrained by high energy or high chemical inputs.
One of the promising separation methods for succinic acid is emulsion liquid
membrane (ELM). ELM provides the advantage of combination of extraction and
stripping process [8]. This combination can remove the equilibrium limitation
between the aqueous and organic phases. Besides, with the use of appropriate
carrier for transport mechanism, specific molecular recognition can be achieved,
where succinic will be highly selective over acetic acid.
Therefore, in this study, an ELM process will be developed for selective sepa-
ration of succinic acid from aqueous waste solution by screening the ELM com-
ponents which are carrier, diluent and stripping agent. The screening process was
done by liquid-liquid extraction process, changing various experimental variables,
including different type of carrier, diluents, and stripping agents with different
concentration.
Materials and Methods
Materials, Amberlite LA2 as carrier was obtained from Merck Company.

Amberlite LA2 used, was a mixture of straight-chain secondary amine mixture
(M = 374 g/mol). Trioctylamine (TOA) (>93 % assay) and Tridodecylamine (TDA)
(>95 % assay) were purchased from Merck. Kerosene as diluent was purchased
from Sigma Aldrich. Palm oil, also as diluent used in this study, was ordinary
cooking oil (BURUH) acquired from a supermarket. Sodium Hydroxide (NaOH)
(98 % assay) was purchased from J. T. Baker, while Sodium Carbonate, Na2CO3
(99.5 % assay) was purchased from Merck. In addition, succinic acid (SA) (99.0 %
[email protected]
Liquid Membrane Formulation for Succinic Acid Extraction … 53
assay) and acetic acid (AA) (99.7 % assay) were purchased from Sigma Aldrich and
J. T. Baker, respectively. All these reagents and solutions were used directly as
received without further purification.
Experimental, The experiment on the carrier screening was carried out by
mixing an equal volume (10 ml) of organic solution (carrier in kerosene) with a
simulated aqueous waste solution (40 g/l succinic acid, 10 g/l acetic acid) at
320 rpm using a mechanical shaker (IKA-KS 130 Basic, Germany). The solution
was then carefully poured into a separating funnel for phase separation for about
15 min. Sample of aqueous phase at the bottom of the separating funnel was taken
for concentration measurement of succinic and acetic acid using High Performance
Liquid Chromatography (HPLC). The same procedures were repeated for other
parameters such as diluents and carrier concentrations.
After that, stripping process was carried out and it is the reverse extraction
step. Therefore, organic phase (succinic acid loaded phase) was taken from the
extraction process in section Materials and Methods. An equal volume of succinic
acid loaded organic phase and aqueous stripping solution were mixed together at
320 rpm using mechanical shaker for 24 h. The mixture was then carefully poured
in separating funnel and was left for about 15 min for the phase separation.
Aqueous phase at the bottom of the separating funnel was taken for succinic and
acetic acid concentration measurements using HPLC. The stripping procedures
were repeated for different stripping agent concentration. All experiments were
carried out at ambient temperature (±26 °C). The extraction and stripping perfor-
mance of succinic and acetic acid was evaluated using Eqs. (1) and (2).
½AiðaqÞ ½Af ðaqÞ

Amount extract ðg=lÞ ¼ ð1Þ
½AiðaqÞ
Amount strip ðg=lÞ ¼ ½AfsðaqÞ ð2Þ
where [A]i(aq) is the initial acid concentration in aqueous phase (g/l); [A]f(aq) is the
acid concentration in aqueous phase after extraction (g/l); [A]fS(aq) is the acid
concentration in aqueous phase after stripping (g/l).
Besides that, distribution ratio (D) and separation factor (α) for the extraction of
succinic and acetic acid were determined using the following equation;
½SAorg
DSA ¼ ð3Þ
½SAaq
½AAorg
DAA ¼ ð4Þ
½AAaq
DSA
SA ¼
aAA ð5Þ
DAA
[email protected]
54 N. Jusoh et al.
where DSA is the distribution ratio of succinic acid; DAA is the distribution ratio of
acetic acid; αSA/AA is the separation factor of succinic over acetic acid.
Effect of carrier on succinic acid extraction, The selection of suitable carrier is

very crucial for the ELM system to perform well and selectively form complex with
succinic acid. Generally, secondary or tertiary amine carriers are typical carriers of
carboxylic acid [9]. In this study, different types of amine carrier were used to extract
succinic acid, which is TOA, TDA, and Amberlite LA2 as shown in Fig. 1. The
current study clearly shows that Amberlite LA2 gives the best performance for
succinic acid extraction (19.98 g/l) compared to TOA (10.98 g/l) and TDA (3.19 g/l).
This is due to the Amberlite LA2 is one of the secondary amine, while TOA and
TDA are both tertiary amine. Secondary amine can easily form complex with suc-
cinic acid and form ammonium salt. The process of complex formation involved the
ion pair association between the alkylammonium cation and the acid radical. Since
Amberlite LA2 is a secondary amine, it can easily react with acid radical by losing
the H bonding. This finding in line with the results found by Ricker [10] who
reported that the extractability of secondary amine was higher than that of tertiary
amine.
Besides, the selectivity of succinic acid and acetic acid extraction was also
performed and tabulated in Table 1. The highest value of separation factor was
obtained using Amberlite LA2 as the carrier. This means that Amberlite LA2 not
only can extract high amount of succinic acid, but it also selectively forms complex
with succinic acid over acetic acid. Therefore, the use of Amberlite LA2 was
Fig. 1 Effect of carrier types for succinic acid extraction (diluent: kerosene; [succinic acid]: 40 g/l;
[acetic acid]: 10 g/l; [carrier]: 0.5 M; agitation speed: 320 rpm; extraction time: 24 h)
Table 1 Selectivity of Carrier DSA DAA αSA/AA

succinic from succinic and
acetic acid mixtures by TOA 0.54 0.25 2.15
various types of carrier TDA 0.11 0.01 8.48
Amberlite LA2 1.51 0.06 23.81
[email protected]
Fig. 2 Effect of diluent composition on extraction of succinic acid ([Amberlite LA2]: 0.5 M;
[succinic acid]: 40 g/l; [acetic acid]: 10 g/l; agitation speed: 320 rpm; extraction time: 24 h)
Table 2 Selectivity of Kerosene/palm oil % DSA DAA αSA/AA

succinic from succinic and
acetic acid mixtures by 100/0 1.41 0.01 134.73
various types of carrier 70/30 1.97 0.11 17.79
50/50 1.76 0.06 27.73
30/70 2.39 0.04 57.82
0/100 2.11 0.06 34.10
preferred for the ELM formulation to separate succinic and acetic acid and used in
the following section.
Effect of diluent on succinic acid extraction, In this study, kerosene and palm
oil and was used as diluents in different composition. The effect of these diluents on
separation of succinic and acetic acid were illustrated in Fig. 2. Besides, the
selectivity of succinic acid was shown in Table 2. The present study shows that
similar amount of succinic acid was extracted using different composition of ker-
osene and palm oil between 18–22 g/l of succinic acid. The results indicate that the
composition of kerosene and palm oil did not give much effect on succinic acid
extraction. Nevertheless, there was a really large different on the selectivity of
succinic acid as shown in Table 2, where kerosene gives the best selection of
succinic acid with 134.73 separation factor value. However, kerosene is not envi-
ronmentally friendly and the toxicity effect of kerosene will be harmful to human.
Hence, it is highly desirable to replace it with another material, such as palm oil.
According to Table 2, the combination of 30/70 kerosene to palm oil also leads to
high separation factor (57.82) followed by a comparatively separation factor using
pure palm oil (34.10). The 30/70 kerosene to palm oil provide larger separation
factor, but the combination can complicate the extraction process and kerosene is
still being used although in a small proportion. Therefore pure palm oil is selected
as possible diluent in this study. This is because the valuable property in palm oil
over kerosene seen as renewable, nontoxic and readily available. In addition, it
works well for the extraction of heavy metal [11] and phenol [12]. Hence, pure
palm oil will be considered as diluent for the next study.
Effect of stripping agent types, In the ELM system, driving force for the
succinic acid extraction is pH difference between the feed and receiving phase.
[email protected]
56 N. Jusoh et al.
Fig. 3 Effect of stripping agent types on succinic acid extraction ([stripping agent]: 1 M; agitation
speed: 320 rpm; stripping time: 24 h)
Therefore, alkaline solution such as NaOH and Na2CO3 were screened in this study
as stripping agent for succinic acid extraction. Figure 3 shows the effect of stripping
agent type of solute from succinic loaded organic phase. The results indicate there is
almost no acetic acid stripped in the process due to very small amount of acetic acid
in extracted into loaded organic phase. The amount of succinic acid stripped using
Na2CO3 is 16.34 g/l compared to NaOH which is only 7.59 g/l. This is because
Na2CO3 dissociates, and forms 2 mol of sodium ion in comparative to NaOH that
gives only 1 mol. Therefore, more succinic acid will be stripped using Na2CO3.
The result of this study seems to be consistent with other research, who reported
that Na2CO3 show better performance than NaOH in stripping process of succinic
acid [13].
Effect of Amberlite LA2 concentration on extraction of succinic acid, The
effect of concentration of Amberlite LA2 on the amount of succinic acid extracted is
shown in Fig. 4. The amount of succinic acid extracted was increased by increasing
the concentration of Amberlite LA2. This indicates that Amberlite LA2 plays an
important role in the extraction of succinic acid and it shows that the extraction of
succinic acid was very selective toward acetic acid. Besides that, this result shows
that smaller amount of Amberlite LA2 was required to extract succinic acid and
can be proved by examining the stoichiometric reaction of Amberlite LA2 and
succinic acid.
Fig. 4 Effect of Amberlite LA2 concentration on succinic acid extraction (diluent: palm oil;
[succinic acid]: 40 g/l; [acetic acid]: 10 g/l; agitation speed: 320 rpm; extraction time: 24 h)
[email protected]
Fig. 5 Stoichiometric plot of equilibrium extraction of succinic acid using Amberlite LA2
The reaction of Amberlite LA2 with undissociated succinic acid can be

described by the reaction:
mðH2 AÞðaqÞ þ nBðorgÞ ! ðH2 AÞm Bn ðorgÞ ð6Þ
where m and n is the number of moles,(H2A)(aq) is undissociated succinic acid in

aqueous phase, B(org) is Amberlite LA2 in organic phase, and (H2A)mBn(org) is
the complex formed.
Based on the extraction reaction and result from Fig. 4, the nature of extracted
solute was examined, by plotting a graph of log D versus log [Amberlite LA2] as
shown in Fig. 5. The slope obtained was 0.617, which implies that the Amberlite
LA2 form 1:1 complex with succinic acid and presented in Eq. 7.
ðH2 AÞðaqÞ þ BðorgÞ ! ðH2 AÞBðorgÞ ð7Þ
Effect of Na2CO3 concentration, Effect of concentration Na2CO3 on stripping

of succinic acid was shown in Fig. 6. According to the result, there is approximately
no acetic acid because it was stripped from organic loaded phased contained a very
low amount of acetic acid. On the other hand, the amount of succinic acid stripped
was increased when high concentrations of Na2CO3 were used. The result indicates
that the amount of succinic acid strip was dependent on concentration of Na2CO3.
The reaction of Na2CO3 and acid-amine complex is expressed by:
mðH2 AÞB þ nNaþ ! ðH2 AÞ2 Na þ B ð8Þ
Amount strip (g/l)

15
10
succinic
5 acetic
0 [Na2CO3 ]
0 0.5 1 1.5 2
Fig. 6 Effect of Na2CO3 concentration on succinic acid extraction (Agitation speed: 320 rpm;
time: 24 h)
[email protected]
58 N. Jusoh et al.
Fig. 7 Stoichiometric plot of equilibrium stripping of succinic acid by Na2CO3 stripping agent
To evaluate the nature of the stripped solute, a graph of log D versus log
[Na2CO3] was plotted as shown in Fig. 7. From the graph, slope obtained was 1.723
which is nearly to 2.0. Hence, molar ratio of succinic-amine complex to sodium ion
is found to be 1:2. Thus, it shows that one mole of complex reacts with 2 mol of
sodium ion in stripping reaction as shown in Eq. 9.
ðH2 AÞB þ 2Naþ ! ðH2 AÞ2 Na þ B ð9Þ
Conclusion
The finding of this study provides an approach for selective extraction of succinic
acid toward acetic acid. It can be concluded the best liquid membrane was for-
mulated using Amberlite LA2 as carrier, palm oil as diluent, and Na2CO3 as
stripping agent.
Acknowledgments The authors would like to acknowledge the Ministry of Higher Education
(MOHE), Universiti Teknologi Malaysia (RU Research Grant; GUP:Q.J130000.2544.04H47), and
Centre of Lipid Engineering and Applied Research (CLEAR) for financial support to make this
research possible.
References
1. Xu J, Guo BH (2010) Poly(butylene succinate) and its copolymers: research, development and
industrialization. Biotechnol J 5:1149–1163
2. Zeikus JG, Jain MK, Elankovan P (1999) Biotechnology of succinic acid production and
markets for derived industrial products. Appl Environ Microbiol 51:545–552
3. Tsukatani T, Matsumoto K (2000) Flow-injection fluorometric quantification of succinate in
food stuffs based on the use of an immobilized enzyme reactor. Anal Chim Acta 416:197–203
4. Thakker C, Martinez I, San KY, Bennet GN (2012) Succinate production in Escherichia coli.
Biotechnol J 7:213–224
5. Li Q, Wang D, Wu Y, Li WL, Zhang YJ, Xing JM, Su ZG (2010) One step recovery of
succinic acid from fermentation broths by crystallization. Sep Purif Technol 72:294–300
[email protected]
6. Kurzrock T, Weuster-Botz D (2011) New reactive extraction systems for separation of bio-
succinic acid. Bioproc Biosyst Eng 34:779–787
7. Wang C, Li Q, Tang H, Yan DJ, Zhou W, Xing JM, Wan YH (2012) Membrane fouling
mechanism in ultrafiltration of succinic acid fermentation broth. Bioresour Technol 116:
366–371
8. Othman N, Chan KH, Goto M, Mat H (2006) Emulsion liquid membrane extraction of silver
from photographic waste using CYANEX 302 as the mobile carrier. Solvent Extr Res Dev
13:191–202
9. Hong YK, Hong WH, Han DH (2001) Application of reactive extraction to recovery of
carboxylic acids. Biotechnol Bioprocess Eng 6:386–394
10. Ricker NL (1978) Recovery of carboxylic acids and related organic chemicals from
wastewater by solvent extraction. PhD. Dissertation, University of California, Berkeley
11. Sanna B, Rose FK (2011) A study of the heavy metal extraction process using emulsion liquid
membranes. Master’s thesis, Chalmers University of Technology, Sweden
12. Venkateswaran P, Palanivelu K (2006) Recovery of phenol from aqueous solution by
supported liquid membrane using vegetable oils as liquid membrane. J Hazard Mater 131:
146–152
13. Lee SC (2011) Extraction of succinic acid from simulated media by emulsion liquid
membranes. J Membr Sci 391:237–243
[email protected]
Preservation of Natural Textile Colorant
from Xylocarpus Moluccensis
by Encapsulating into PVA Alginate
Norul Azilah Abdul Rahman, Ruziyati Tajuddin

and Siti Marsinah Tumin
Abstract Natural dye/colorant extracts used for dyeing source are incapable to be
stored for a substantial period of time due to their unstable property in aqueous
medium. Therefore, this study explores a suitable method to preserve the extracts
for storage and produce natural textiles dyes with longer shelf life. PVA Alginate
was used as a pure/blank beads to encapsulate the colorant from Xylocarpus mo-
luccensis using ultrasonic cleaner approach at different ultrasonic volume and time.
Results show the highest encapsulation efficiency was achieved using ultrasonic
cleaner at optimum condition of high sonic volume in 80 min time. Fourier-
transform infrared spectrosopic (FTIR), scanning electron microscope (SEM) and
x-ray diffraction (XRD) were used to characterize the encapsulated colorant. The
characterization via these methods indicates the colorant was successfully encap-
sulated into the PVA Alginate beads. Thus, the encapsulation approach studied in
this work showed a promising process for preservation of natural textiles dyes.
Keywords Xylocarpus moluccensis PVA alginate Ultrasonic cleaner

Encapsulation
N.A. Abdul Rahman R. Tajuddin (&) S.M. Tumin

Faculty of Applied Sciences, Universiti Teknologi MARA (UiTM), 40450 Shah Alam,
Selangor, Malaysia
N.A. Abdul Rahman
S.M. Tumin

[email protected]
62 N.A. Abdul Rahman et al.
Introduction
There are many challenges in developing and preserving natural dye/colorant.

Natural textiles dyes extracted from plant materials are commonly in liquid aqueous
form which must be used immediately due to its low stability. It is suggested that
the liquid extracts should be converted to a solid form so that they can be preserved.
To date there is no reported research on any method to preserve natural textile dyes.
Nevertheless, preservation method based on encapsulation approach is a common
practice in food industry in order to entrap colorant or flavor within solid matrix as
for the purpose of converting some product into a more conveniently handled solid
form. β-cyclodextrin [1], Polyethyleneglycol [2], AIF (acetone insoluble fraction)
phospholipids [3], alginate [4] and PVA Alginate [1] are among the substances
commonly used as encapsulation agents.
PVA-alginate is widely used in enzyme activity [5] and bacteria growth
(microbiology) [6] studies as well as to remove hazardous waste (chemical, textile
and pharmaceutical industries) [7–9]. It has porous structure on the surface and
inner to entrap and immobilize compound. According to Dave and Madamwar [5],
PVA contributes to the strength and high crosslinking capacity whereas alginate
reduces the agglomeration and increases the surface properties. These claim
strongly support the rationale of choosing PVA-alginate as the encapsulation agent
in this study. PVA-alginate was exploited as the pure/blank beads to trap and
immobilize natural colorant extracted from the heartwood of Xylocarpus molucc-
ensis plant.
Experimental
Materials, Wood and bark wastes of Xylocarpus moluccensis plant were obtained
from craft small industries located in Pulau Carrey, Klang Selangor, Malaysia. The
sample was ground and stored at room temperature before use. Sodium hydroxide
and preparation beads, calcium chloride were purchased from Merck, Germany.
Boric acid was purchased from R & M Marketing, U.K. Poly(vinyl alcohol) and
Alginic acid sodium salt from brown algae (sodium alginate) were supplied by
Sigma-aldrich, UK.
Extraction Methods, Ultrasonic extractions were carried out as described pre-
viously in our reported study [10]. The optimum parameters were used.
Preparation of beads for encapsulation, Approximately 10 g of PVA and 1 g
of Sodium alginate were weighed and placed in a sample bottle. 100 mL of distilled
water was added and stirred using glass rod. The bottle was then placed into the
ultrasonic cleaning system (Powersonic 420, 100-420-230, AC 230 V, 50 Hz,
700 W, Hwashin Technology Seoul Korea) for 60 min at temperature of 0 °C and
medium sonic volume. The PVA Alginate mixture was then dropped into a mixed
solution of 100 mL boric acid (1 %) and calcium chloride (1 %) using peristaltic
[email protected]
Preservation of Natural Textile Colorant … 63
pump (Master Flex®4s®, USA) and silicone tubing to form beads. The PVA–
alginate beads were then stored at 4 °C for 24 h. For the encapsulation process, the
beads were dried using oven (Memmert, Germany) and added with colorant
extracts, followed by ultrasonic process. The ultrasonic condition was optimized by
varying the ultrasonic volume (low, medium and high) and encapsulation time (20,
30, 40, 60, and 80 min) while the temperature was read after the encapsulation
process. The encapsulated beads were then dried using oven. To determine
encapsulation efficiency (EE), the volume of colorant loaded into PVA Alginate
beads was calculated based on Eq. (1) below:
Volume of loaded colorant

%EE ¼ 100 % ð1Þ
Volume of initial colorant
Characterization Methods
Fourier-transform infrared (FTIR) spectroscopic analysis, The Fourier trans-

form IR spectra (FT-IR) of samples were recorded with a Perkin Elmer in the range
of 400 and 4000 cm−1 by averaging 16 scans number at 4 cm−1 resolution as KBr
pellets.
Scanning electron microscope (SEM), The surfaces of the samples were
imaged using a Carl Zeiss electron microscope (Supra 40VP, Germany) at 10 kV
with 3 k magnification. The samples were deposited on a copper stub and coated
with gold, using vapor deposition techniques.
X-ray diffraction (XRD), X-ray diffractograms of the samples were recorded
over a 2 theta range of 2–50° by X-ray diffractometer (Rikagu Ultima IV, Japan)
with a step angle of 0.02°/min. The voltage and current were utilized at 40 kV and
20 mA respectively.
Encapsulation efficiency, The encapsulation efficiency determined by the per-

centage of Xylocarpus moluccensis colorant adsorbed into PVA Alginate beads was
measured based on ultrasonic volume and encapsulation time as shown in a plot in
Fig. 1. As can be observed, the encapsulation efficiency increased with increasing
of ultrasonic volume and encapsulation time, and it reached maximum at 80 min
time with percentage adsorption of 98 %. Further increased in ultrasonic encap-
sulation time at high sonic volume, altered the structure of PVA Alginate beads
which caused a slight reduction of percentage adsorption. This indicates that the
equilibrium of PVA alginate and colorant solution was completely achieved within
80 min.
[email protected]
% Encapsulation efficiency
120
LSV
100 MSV
HSV
80
60
40
20
0
20 30 40 60 80 100
Time, minutes
Fig. 1 Percentage of encapsulation efficiency based on ultrasonic volume and encapsulation time.
LSV is lower sonic volume, MSV is medium sonic volume and HSV is high sonic volume
Characterization
Fourier-Transform Infrared Spectroscopic Analysis
FTIR spectra of PVA Alginate (pure/blank beads), XM (colorant extract from

Xylocarpus moluccensis) and HSV80 (encapsulated XM colorant at high sonic
volume in 80 min) are shown in Fig. 2 and the assignment of bands is presented in
Table 1. General characteristics of PVA Alginate were viewed at 3405.79 cm−1
(–OH stretching), 2927.53 cm−1 (–CH2 asymmetric stretching), 1627.85 cm−1 and
1429.13 cm−1 (C = O asymmetric stretching and C = O symmetric stretching
Fig. 2 FTIR spectra of PVA Alginate beads (pure bead/blank beads), XM (colorant extract of
Xylocarpus moluccensis) and HSV80 (encapsulated XM colorant at high sonic volume with
80 min encapsulation time)
[email protected]
Table 1 Assignment of FTIR bands for PVA alginate beads (pure bead/blank beads), XM (dye
extract Xylocarpus moluccensis) and HSV80 (encapsulated bead with XM at high sonic volume
with 80 min encapsulation time)
Samples OH CH2 C=O C=O CH3 C–O Aromatic
(cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1) (cm−1)
(a) PVA A 3405.79 2927.53 1627.85 1429.13 1331.58 1093.12
(b) XM 3405.79 2927.53 1627.85 1393.00 1093.12 1093.12–1393.00
(c) HSV80 3405.79 2927.53 1627.85 1429.13 1093.12 1093.12–1429.13
respectively), 1331.58 cm−1 (–CH3 symmetric bending) and 1093.12 cm−1 (C–O
bending). This result was similar to those reported by Ravindrachary et al. [11] and
De Queiroz [12].
For XM sample, the same characteristics were observed at 3405.79 cm−1 (–OH
stretching), 2927.53 cm−1 (–CH2 stretching) and 1627.85 cm−1 (C = O asymmetric
stretching), however the C = O symmetric stretching band was shifted to
1393.00 cm−1. A weak and broad absorption band attributed to benzene ring was
observed from 1093.12 to 1393.00 cm−1 and this overlapped with the absorption
band of C–O bending at 1093.12 cm−1. This result is in agreement with that
reported by Rosu et al. [13] for alizarin (9,10-anthraquinone) compound. They
suggested that the presence of absorption band at 1627/1629 cm−1 was due to the
conjugate carbonyl group vibrations (by hydrogen bonds with adjacent O–H
groups) that represented the characteristic absorption band of 9,10-anthraquinone
compound.
XM loaded into pure/blank beads (HSV80) resulted in a markedly increased
intensity of the absorption bands as shown at 3405.79 cm−1 (–OH stretching),
2927.153 cm−1(–CH2 stretching), 1093.12 cm−1 (C–O bending) and 1093.12–
1429.13 cm−1 (benzene ring), indicating an increase in the content of alizarin,
which might come from colorant extract of Xylocarpus moluccensis. The disap-
pearance of absorption band for –CH3 symmetric bending at 1331.58 cm−1 (pre-
viously shown in PVA-Alginate spectrum) also indicates the successful
encapsulation of XM colorant into the PVA Alginate.
Scanning Electron Microscope (SEM) Analysis
The morphology obtained for PVA Alginate and HSV80 samples were then imaged
by SEM as shown in Fig. 3. The aggregation and pores of PVA Alginate beads
were obviously observed in Fig. 3a, whereas in Fig. 3b, both aggregation and pores
of beads were almost invisible owing to complete adsorption of colorant after
encapsulation process. These aggregates seemed to be melting, combining with
each other and the outer surface was smooth due to the possibility of some XM
colorant content surrounding the particle surface.
[email protected]
(a) Encapsulation (b)

of dye extract
XM
At high sonic
volume in 80
minutes time
Fig. 3 SEM picture of a PVA Alginate (pure bead/blank beads), and b HSV80 (encapsulated XM
colorant at high sonic volume with 80 min encapsulation time). a PVA alginate (pure/blank beads),
b HSV80 (encapsulated colorant)
X-Ray Diffraction (XRD) Analysis
Figure 4 illustrates the X-ray diffractograms of PVA Alginate, XM and HSV80

samples. Peaks with significant intensity at 20° and a shoulder at 30° were present
in PVA Alginate diffractogram which were similar to a result reported by Udaseen
et al. [14]. The intensities of both peaks were increased after encapsulation process
as observed in HSV80 diffractogram. This implies that the remarkably higher
intensity of peak at 20° was the result of inclusion complexes between XM and the
PVA Alginate beads. In contrast, diffractogram of XM did not show any charac-
teristic peak for the colorant, indicating that the colorant extract was in an amor-
phous state.
Fig. 4 X-Ray diffractograms for PVA alginate (pure bead/blank beads), XM (colorant extract
Xylocarpus moluccensis) and HSV80 (encapsulated XM colorant at high sonic volume with
80 min encapsulation time)
[email protected]
Conclusion
In this study, colorant extract from Xylocarpus moluccensis was successfully

encapsulated into PVA Alginate using ultrasonic method. PVA Alginate was found
to be an efficient encapsulation matrix for colorant encapsulation. The encapsula-
tion efficiency of 98 % colorant adsorption was achieved at optimum condition of
high sonic volume in 80 min time. Characteristics of encapsulated colorant were
successfully determined using FTIR, SEM and XRD analysis which indicate the
potential achievement of this encapsulation process for preservation of natural
textile dyes.
Acknowledgments We wish to express our gratitude to the Faculty of Applied Sciences,

Universiti Teknologi MARA Shah Alam for the support in this research project
References
1. Norasiha H, Mimi Sakinah AM, Rohaida CM (2011) Characterization of β-cyclodextrin

complexes with natural dye. In: National conference on postgraduate research (NCON-PGR),
pp. 98–106
2. Santos DT, Albarelli JQ, Beppu MM, Meireles MAA (2011) Stabilization of anthocyanin
extract from jabuticaba skins by encapsulation using supercritical CO2 as solvent. Food Res
Int 1–8
3. El-Zawahry MM, El-Shami S, El-Mallah MH (2007) Optimizing a wool dyeing process with
reactive dye by liposome microencapsulation. Dyes Pigm 74:684–691
4. Sanjay S, Kumar A, Suneetha V, Mishra B, Gopinath R, Yadav S et al (2012) Synthesis and
activation of Immobilized beads by natural dye extracts. Int J Drug Dev Res 4:304–310
5. Dave R, Madamwar D (2006) Esterification in organic solvents by lipase immobilized in
polymer of PVA–alginate–boric acid. Process Biochem 41:951–955
6. Yujian W, Xiaojuan Y, Hongyu L, Wei T (2006) Immobilization of Acidithiobacillus
ferrooxidans with complex of PVA and sodium alginate. Polym Degrad Stab 91:2408–2414
7. Zhang Y, Kogelnig D, Morgenbesser C, Stojanovic A, Jirsa F, Lichtscheidl-Schultz I,
Krachler R, Li Y, Keppler BK (2011) Preparation and characterization of immobilized [A336]
[MTBA] in PVA–alginate gel beads as novel solid-phase extractants for an efficient recovery
of Hg(II) from aqueous solutions. J Hazard Mater 196:201–209
8. Idris A, Misran E, Hassan N, Abd Jalil A, Seng CE (2012) Modified PVA-alginate encapsulated
photocatalyst ferro photo gels for Cr(VI) reduction. J Hazard Mater 227–228:309–316
9. Kumar SS, Kumar MS, Siddavattam D, Karegoudar TB (2012) Generation of continuous packed
bed reactor with PVA–alginate blend immobilized Ochrobactrum sp. DGVK1 cells for effective
removal of N, N-dimethylformamide from industrial effluents. J Hazard Mater 199–200:58–63
10. Abdul Rahman NA, Tumin SM, Tajuddin R (2013) Optimization of ultrasonic extraction
method of natural dyes from Xylocarpus moluccensis. Int J Biosci Biochem Bioinf 3:53–55
11. Ravindrachary V, Ismayil SP Nayak, Dutta D, Pujari PK (2011) Free volume related
fluorescent behavior in electron beam irradiated chalcone doped PVA. Polym Degrad Stab
96:1676–1686
12. De Queiroz AAA, Passos ED, Alves SDB, Silva GS, Fernandez LRV, Higa OZ, Vítolo M
(2006) Alginate-poly(vinyl alcohol) core-shell microspheres for lipase immobilization.
J Apppl Polym Sci 102:1553–1560
[email protected]
13. Rosu M-C, Suciu R-C, Lazar MD, Bratu I (2013) The influence of alizarin and fluorescein on
the photoactivity of Ni, Pt and Ru-doped TiO2 layers. Mater Sci Eng 178:383–390
14. Udaseen S, Asthana S, Raveendran NT, Kumar K, Samal A, Pal K, Pramanik K, Ray SS
(2014) Optimization of process parameters for nozzle—free electrospinning of poly (vinyl
alcohol) and alginate blend nano-fibrous scaffolds. Int J Enhanced Res Sci Technol Eng 3
(1):405–411
[email protected]
Coagulant from Chemically Modified Fish
Scale for Textile Wastewater Treatment
Mohibah Musa, Ku Halim Ku Hamid, Hanafiah Zainal Abidin

and Miradatul Najwa Muhd Rodhi
Abstract Textile manufacturing is one of the largest industrial users of process

water. This production gives rise to the major pollutant in this industry where the
discharging residue from dyeing process contains organic and inorganic compounds
and some potentially toxic solvents. This effluent is the environmental challenge for
textile industry where the wastewater often rich in color, high pH, high Chemical
Oxygen Demand (COD) and Biological Oxygen Demand (BOD) also low biode-
gradability. Therefore, complete dye removing technology of users and environ-
mentally friendly is very necessary. In this research, biodegradable material for
coagulation-flocculation process tailored for textile wastewater treatment was
developed based on polymeric material from fish scales. This coagulant with a
particle size of 75 µm was tested for real wastewater from industry via jar tests to
determine the optimum operating condition such as the dosage of coagulant and pH.
The treated textile wastewater was found to be complying with the Parameter
Limits of Effluents of Standards A and B from Environmental Quality (Sewage and
Industrial Effluents) Regulations 1978, from the result, the optimum pH and dosage
is 5.0 mg/l of coagulant dosage at pH 5.0. COD reduction is about 98 % meanwhile
the reduction of colour achieves more than 90 %.
Keywords Fish scale Textile wastewater Polymeric material Coagulant

Biodegradable
M. Musa (&) K.H. Ku Hamid H. Zainal Abidin M.N. Muhd Rodhi

Selangor, Malaysia
K.H. Ku Hamid
H. Zainal Abidin
e-mail: hanafi[email protected]
M.N. Muhd Rodhi

[email protected]
70 M. Musa et al.
Introduction
The population of the world is increasing year by year and its effect considerable
amount of stress on the environment. Human needs water to survive and move on
with their activities and nowadays water will continue to become a major factor for
survival. Currently, water had been used by household approximately 50 % and the
rest had been used by industrial and agricultural activities. Textile manufacturing is
one of the largest industrial users of process water. Water is used extensively
throughout batik processing operations. Dyeing among the most water intensive
processes in batik production, the environmental challenge for the textile industry is
associated with liquid waste, which tends to dominate over air-emissions and solid
wastes in term of environmental effects [1].
The effluent from dyeing and finishing processes often rich in colour, containing
residual of dyes and chemical, and needs proper treatment before releasing into the
environment [2]. It’s also been characterized by high pH, high temperature, high
COD and low biodegradability [3].
The effluent from textile wet processing requires several treatment options before
it can be discharged. Conventional activated sludge is one of the technology that’s
been applied for the treatment. In this type of treatment, it has the capability to
remove large fractions of COD. Dye removal is most often achieved by sorption
processes, rather than by biodegradation. For effluents with COD over 5000 mg/l,
anaerobic treatment becomes more significant. A combined anaerobic and aerobic
treatment can also be effective for removal of azodyes [7]. However, this system
removes poorly most of the dyes used, and is clearly ineffective in decolourizing
textile effluent, even when mixed and treated together with sewage. Removal of
dyes is possible by activated carbon treatment. But, it is expensive and the level of
colour removal depends on the type of dye and complete colour removal is rarely
achieved [1, 4].
Ozone (O3) is one of the strongest oxidizing agents that is readily available. It is
used to reduce colour, eliminate organic waste, reduce odour and reduce total
organic in the water. Ozone is one of the strongest commercially available oxi-
dizers, making it popular for primary disinfection of potable water as well as for
colour and organic removal in wastewater applications. But, ozone cannot com-
pletely mineralize the organic dye to carbon dioxide and water even at high dose
usage [4]. This due to the decolonization rate decreasing with increasing initial dye
colour [5]. Up to date there is no economic and effective method to treat textile
wastewater. Therefore, a serious effort has to be carried out to solve this problem.
Colour from textile wastewater is hard to remove due to resistance of dyes to
biological degradation where they are not readily degraded under the aerobic
conditions prevailing in biological treatment plant [6]. The colour of textile effluent
is unacceptable under Malaysian Environmental Regulation besides the other
parameter such as COD, BOD, total iron, etc. [7]. Regulation for effluent and air
emissions, however, are becoming increasingly stringent and more rigorously
enforced, which is prompting the industry to develop alternative methods for the
[email protected]
Coagulant from Chemically Modified Fish Scale … 71
management of process wastes. Due to the factors stated above, balancing the
environmental protection and its economic viability and sustainable development is
the biggest challenges in the textile industry. In order to keep the economic growth,
new way of treatment is vital to preserve the environment. To face the current
challenges, textile effluents need an efficient treatment system. Coagulation is one
of the techniques used to remove the pollution base on suspended particles. This
method has been applied in removal of colour textile by using miscellaneous type
of coagulant.
Coagulation method using inorganic coagulants with the addition of polymer
flocculant gives almost 100 % of colour removal [4]. Removal of some reactive
dyes from synthetic wastewater by combining Al(III) coagulation /carbon adsorp-
tion process, achieved complete decolorisation and almost total elimination
(99.9 %) of reactive dyes from wastewater [8]. In this work, the use of fish scale as
the coagulant was studied. Amount of fish scale generated as waste allows an
abundant source of biomaterial. Fish scale found to be effectively used as an organic
wastewater coagulant [8].
The proteins which present in the organic fraction in fish scale seem to be the
major factor of adsorption ability. Other biomaterial such as Chitosan, had proven
its ability in dye removal where the charge neutralization in main mechanism for
dye coagulation with chitosan is at acidic pH [9]. For this work, fish scales were
dried and ground before being added as a coagulant in powder form. An experi-
mental study were carried out in a local textile wastewater to assess the removal
efficiency of a coagulation process. Up to date there is no serious effort to utilize
fish scale as a coagulant for wastewater treatment in Malaysia.
Experimental
Raw material preparation, The raw material (scales) were collected at the local
market. Firstly, this material were washed and thermally treated with distilled water
at temperature of 800 °C for 30 min. Then, this pre-treated material were dried a
convection oven at 80–90 °C for 18 h. Then, the scales were treated with 5 % HCL
at 40 °C for 6 h in order to produce organic-rich material and to make the scales
amber and very fragile. After 6 h, the solution will filter off and the scales will be
washed with distilled water until the pH of the washes becomes neutral. Then, it
will be dry in a convection oven at 80 °C for 18 h. The treated fish scale will be
ground into a powder form with particle size of 75 µm.
Coagulation-Flocculation Process, The process were carried out via Jar Test at
room temperature. For each test coagulant dose were varied from 1.0 to 5.0 g and
pH from 5 to 8 in order to determine the optimum operating conditions for colour
removal. Wastewater were transferred into the jar and stirred at 50 rpm for 5 min.
The pH of the wastewater were adjusted by adding 0.1 M HCl or 0.1 M NaOH. The
[email protected]
72 M. Musa et al.
sample will be stirred at 180 rpm for 10 min prior to additional of organic coagulant
from fish scale were added. The rapid mixing will be continued for another 2 min.
The speed of the mixing will be reduced to 50 rpm for 15 min before leaving the
sample for sedimentation process about 4 h.
Sample analysis, At each condition applied (different dosage and pH), COD,
BOD, colour, metallic ion level and colour Removal Efficiency (CRE) will be
determined. For metallic ion level, the test were run using ICP and amount of
metallic ion present in the sample before and after jar test treatment were measured.
The Colour Removal Efficiency (CRE) was calculated from the difference between
initial and final absorbance value. The observance was measured using UV-VIS
Spectrophotometer. The CRE will be calculated as follows:
ð1 A1Þ
CRE ¼ 100 %
A0
where A0 and A1 are the absorbance of the samples before and after jar test
treatment. The final result were compared to the Parameter Limits of Effluents of
Standards A and B from Environmental Quality (Sewage And Industrial Effluents)
Regulations 1978, DOE.
To investigate the reason of high efficiency and the textile wastewater treatment
mechanism of fish scale based coagulant, the variation of zeta potential hydrolysis
products as a function of pH was measured. In this series of experiments, the pH
value was varied by the addition of HCl or NaOH solution. It needs to be pointed
out that the pH values in zeta experiments were the same as the pH of coagulant
solution (0.5 g/L). The working solution used in this study was distilled water. The
data obtained are summarized in Fig. 1. Zeta potential is a controlling parameter of
charge neutralization of coagulant can usually used to interpret the trend of
Fig. 1 The variation of zeta

potential with respect to pH
for fish scale based coagulant
[email protected]
Table 1 Wastewater quality Properties Unit Value

for industrial dye wastewater
Chemical oxygen demand (COD) ppm 4700.00
Colour m−1 2120.00
Cadmium (Cd) ppm 0.31
Chromium (Cr) ppm 2.23
Copper (Cu) ppm 6.78
Iron (Fe) ppm 2.62
Mercury (Hg) ppm 3.33
Plumbum (Pb) ppm 16.13
coagulation efficiency. The zeta potential of fish scale based coagulant decreased
from 32.2 to 15.8 mV as the pH was increased from 4.5 to 7.5. The value of the zeta
potential is practically constant in a wide range of pH (from 4 to 8). The textile in
water has a negative charge on its surface, so the higher the positive charge on the
coagulant surface, the higher the latex concentrate removal efficiency.
Table 1 shows the wastewater quality of industrial dye wastewater, the sample
was collected from batik industries from Kuala Terengganu. The main parameter
that was measured is mentioned in the table below. From the study the COD value
of the wastewater is around 4700 ppm. Value for colour is around 2120 m−1. In
addition, heavy metal ion concentration been also measured in this study where Cd,
Cr, Cu, Fe, Hg and Pb concentration approximately 0.31, 2.23, 6.78, 2.26, 3.33 and
16.13 ppm respectively.
Figure 2 shows the result of COD removal by difference coagulant dosing i.e.
1.0, 2.0, 3.0, 4.0 and 5.0 g/ml. The lowest value of COD was found in sample with
a pH value around 5.0 ppm at coagulant dosing about 5.0 g/ml with concentration
of 130 ppm. The highest value for COD is found in sample at pH 8.0, the value
reaches until 4650 ppm compare to other.
4650
Fig. 2 Result for COD

4600
5000
4528
removal by difference 4500

4025
coagulant dosing
3804
3599
4000
concentration (ppm)
3243
3500
3002
3000 5.0
2500
2480
2340
6.0
2500
2056
1840
7.0
1670
2000
1560
8.0
1080
1500
1008
1000
480
260
500
130
0
1.00 2.00 3.00 4.00 5.00
dosing
[email protected]
74 M. Musa et al.
1890
Fig. 3 Result for colour
1870
1870
1850
2000
1800
1780
1770
removal by difference
1700
1680
1660
1800
coagulant dosing
1550
1600
1440
1360
1350
1330
1400
1210
1200
colour removal
1190
5.0
1200
970
6.0
1000
7.0
800
8.0
600
340
400
200
0
1.00 2.00 3.00 4.00 5.00
dosing (g/ml)
Figure 3 shows the result of colour removal by difference coagulant dosing. The
result show that the highest value of colour at 1890 m−1 was detected at doses of
coagulant around 2.0 g/ml the value is around. Meanwhile, the lowest value of
colour was detected at pH 5.0 and coagulant dosing about 5.0 g/ml, the value
reaches to 340 m−1.
Figures 4, 5, 6, 7 and 8 shows the value for removing heavy metal ions by
difference in coagulant dosing i.e. 1.0, 2.0, 3.0, 4.0 and 5.0 g/ml. Several common
heavy metal ions were measured to indicate the performance of the coagulant. For
the cadmium ion the lowest value is about 0.01 ppm the value was indicated by the
experiment using pH 5.0 at coagulant dosing about 4.0 and 5.0 g/ml. For chromium
ion the lowest value is 0.39 ppm. The value found from experiments using pH 5.0
and dosing value is around 5.0 g/ml. The pattern of lowest value also show for the
other remaining metal ion. The lowest value was shown at pH 5.0 and at doses of
coagulant around 5.0 g/ml.
Fig. 4 Result for removal of 16.00

13.93
heavy metal ion by 1.0 g/ml 14.00

of coagulant dosing
concentration (ppm)
12.00
10.00 5.0
6.0
8.00
6.75
6.60
6.56
7.0
5.89
6.00 8.0
4.67
3.29
3.26
4.00
2.80
2.60
2.63
2.59
2.34
2.37
2.23
2.23
2.22
2.13
2.08
1.10
2.00
0.30
0.31
0.31
0.22
0.00
Cd Cr Cu Fe Hg Pb
heavy metal
[email protected]
Fig. 5 Result for removal of
9.29
10.00
heavy metal ion by 2.0 g/ml 9.00
of coagulant dosing 8.00
concentration (ppm)
6.72
6.46
7.00
5.43
6.00 5.0
6.0
4.32
5.00
4.19
7.0
4.00
3.22
8.0
2.80
2.61
2.57
2.28
3.00
2.20
2.21
2.18
1.92
1.90
1.80
1.81
1.60
2.00
0.40
0.29
0.29
0.30
1.00
0.07
0.00
Cd Cr Cu Fe Hg Pb
heavy metal
7.10
heavy metal ion by 3.0 g/ml 7.00 6.10
6.03
of coagulant dosing
concentration (ppm)
6.00
4.56
5.00 5.0
3.79
6.0
4.00
3.26
2.96
7.0
2.58
2.58
3.00 2.32 8.0

2.24
2.22
2.19
2.10
1.85
1.80
1.59
2.00
0.49
0.43
1.00
0.30
0.29
0.18
0.17
0.05
0.00
Cd Cr Cu Fe Hg Pb
heavy metal

6.14
heavy metal ion by 4.0 g/ml

5.89
5.53
6.00
of coagulant dosing
concentration (ppm)
5.00
4.16
5.0
4.00
6.0
3.11
3.01
7.0
2.56
2.55
3.00
2.34
2.20
2.20
2.12
2.08
8.0
1.81
1.73
1.53
2.00
1.48
1.00
0.28
0.28
0.21
0.18
0.17
0.10
0.01
0.00
Cd Cr Cu Fe Hg Pb
heavy metal
[email protected]
76 M. Musa et al.
5.40
5.22
heavy metal ion by 5.0 g/ml
of coagulant dosing 5.00
concentration (ppm)
4.03
4.00
3.46
5.0
3.05
2.94
6.0
3.00
2.53
2.54
7.0
2.19
2.18
2.13
1.94
1.82
8.0
2.00
1.60
0.80
1.00
0.59
0.39
0.29
0.27
0.18
0.09
0.04
0.04
0.01
0.00
Cd Cr Cu Fe Hg Pb
heavy metal content
Conclusion
From the result, the optimum dosage and pH is 5.0 mg/l of coagulant dosage at pH
5.0 respectively. COD reduction is about 98 % meanwhile the reduction of color
achieve more than 90 %. The treated textile wastewater was found to be complying
with the Parameter Limits of Effluents of Standards B from Environmental Quality
(Sewage and Industrial Effluents) Regulations 2013.
Acknowledgments The authors would like to thank the Ministry of Education Malaysia for
financial support via Fundamental Research Grant Scheme (FRGS). The authors also grateful
acknowledge use of services and facilities of Chemical Engineering Faculty and Universiti
Teknologi MARA.
References
1. Ahmad AL, Harris WA, Syafiie Ooi BS (2002) Removal of dye from wastewater of textile
industry using membrane technology. Journal Teknologi 36(F): 31–44
2. Chakraborty S, Purkait MK, DasGupta S, De S, Basu JK (2003) Nanofiltration of textile plant
effluent for color removal and reduction in COD. Sep Purif Technol 31:141–151
3. Shyam SundarP, KarthikeyanN, Prabhu KH (2002) Wastewater and its treatment in textile
industry. Institute of Chemical Technology, University of Mumbai
4. Ahmad AL, Puasa SW, Abiding S (2005) Crossflow ultrafiltration for removing direct-15 dye
from wastewater of textile industry. School of Chemical Engineering, USM, Penang
5. Tang C, Chen V (2002) Nanofiltration of textile wastewater for water reuse. Desalination
143:11–20
6. Koyuncu I (2002) Reactive dye removal in dye/salt mixtures by nanofiltration membranes
containing vinylsulphone dyes: effect of feed concentration and cross flow velocity.
Desalination 143:243–253
[email protected]
7. Tan BH, Teng TT, Omar AKM (2000) Removal of dyes and industrial dye wastes by
magnesium chloride. Wat Res 34(2):597–601
8. Hood LF, Zall RR (1980) Recovery, utilization and treatment of seafood processing wastes. In:
Conell JJ (ed) Advances in fish science and technology. Fishing News Books, Ltd., Surrey,
England
9. Guibal E, Roussy J (2007) Coagulation and flocculation of dye-containing solutions using a
biopolymer (chitosan). React Funct Polym 67:33–42
[email protected]
Part II
Chemical and Biotechnology
Engineering
[email protected]
An Investigation on N-Containing Organic
Compounds as Corrosion Inhibitors
for Carbon Steel in Saline Environment
Md Amin Hashim, Izni Mariah Ibrahim

and Sxureha Yunus
Abstract The adsorption types of corrosion inhibitors are typically organic

compounds made-up of element comprising of N, O, P and S. In this study, the
relative inhibition effect of synthesized fatty amides mixture, pyridine and pyrrole as
corrosion inhibitors on carbon steels in saline water had been compared using linear
polarization resistance method (LPRM). It is still unknown which of these organic
compounds performed best than another. The experimental results had shown that
compounds containing elemental N performed as corrosion inhibitor in protecting
carbon steels when immersed in saline water. The synthesized fatty amides mixture
is found to be better than pyridine and pyrrole, indicating the presence of various
chain lengths of fatty amides molecules provide a better surface on carbon steels.

Keywords Organic corrosion inhibitor Electron lone pair Physisorption

Chemisorption Carbon steel Static saline environment
Introduction
Corrosion inhibitors are admixture of organic and inorganic chemicals with the
effect that delays the corrosion rate of carbon steels when added in small quantity in
an aqueous environment. The presence of stringent environmental regulations
prohibited inorganic corrosion inhibitors such as chromates, nitrites or other oxides
to be used for protection of carbon steels. Previous researchers suggested that they
M.A. Hashim (&) I.M. Ibrahim S. Yunus

Faculty of Chemical Engineering, Universiti Teknologi MARA,
40450 Shah Alam, Selangor, Malaysia
I.M. Ibrahim
S. Yunus

[email protected]
82 M.A. Hashim et al.
could be replaced by organic compounds such as carboxylic acid derived organics

[1]. Further, carboxylic acid derived organics are well established industrial posi-
tion, whilst their applications are widespread and increasing. The main reason for
using carboxylic acid derived organics is due to its availability and relatively
inexpensive. They are derived from non-edible animal and vegetables fats.
In this research, as-synthesized fatty amides, pyridine and pyrrole were used as
the working corrosion inhibitors. These organics performed with relatively high
efficiency as adsorption type of corrosion inhibitors, inhibiting carbon steels when
immersed in saline water at room temperature and under static environment [2].
When the concentrations of organic inhibitors are low, the adsorbed inhibitor will
be parallel alignment to the steel surfaces. When a further increase in the inhibitor
concentration was added, this causes the adsorbed inhibitor molecules to align in
parallel with each other and perpendicular to the steel surface. In this manner, the
steel surface becomes hydrophobic. Hence, the adsorbed inhibitors molecules can
act as waterproof barrier on the steel surface [3].
Experimental
Preparation of fatty amides. Mixtures of fatty amides were synthesized from

as-received palm fatty acid distillates (PFAD) with analytical grade isoproylamine,
added in excess, and at reaction temperature of 80 °C. The selected reaction path
followed closely as suggested by Cossy and Pale-Grosdemange [4]. A selected
amount (in ppm) of the as-synthesized fatty amides was used in the tests to
determine the efficiency of inhibitor.
Pyridine and pyrrole. As-received, analytical grades (AR) of the chemicals
were used in a prescribed concentration in 3.5 wt% of saline water at room tem-
perature and at static condition.
Test specimen and solution. Commercially available carbon steel was used as
test specimen, and prepared for polarization resistance measurements as described
by Hashim et al. [5]. The carbon steel was analyzed for chemical composition using
Inductively Coupled Plasma (ICP) with Optical Emission Spectrometer (OES) and
CHNS analyzers. The chemical composition is as shown in Table 1. The test
medium used was a 3.5 wt% sodium chloride solution.
A 3-electrode system was used to determine the Rp values of the steel specimen
immersed in the test solution in static condition and at ambient temperature of 25 °C.
The VoltaLab, Model PGP 201 potentiostat was used in the Linear Polarization
Table 1 Chemical Element wt% Element wt%

composition of a typical
carbon steel test specimen Carbon 0.04 Copper 0.059
Aluminum 0.23 Chromium 0.060
Manganese 0.74 Zinc 0.54
Nickel 0.056 Lead 0.02
Iron Balance
[email protected]
An Investigation of N-Containing Organic Compound … 83
(a)
100
80
60
%T
40
C-N bond in
20
pyridine
0
500 1000 1500 2000 2500 3000 3500
-1
cm
(b)
100
80
60
%T
40 C-N bond in
pyrrole
20
500 1000 1500 2000 2500 3000 3500 4000

-1
cm
(c)
95
85 N-H bond in
synthesized
75 fatty amides
%T
65 C=O bond
represents
fatty acid in
55 PFAD
45
3800 3300 2800 2300 1800 1300 800
cm-1
Fig. 1 Observed FTIR spectra of a pyrrole, b pyridine, and c synthesized fatty amides used as
corrosion inhibitors [2, 7]
Resistance Method (LPRM) mode. The inhibition effects of the other organics were
similarly tested with increasing volumetric quantities. Surface morphology of spec-
imen was examined using SEM before and after conducting polarization tests.
[email protected]
Steel specimen chemical composition. Table 1 shows chemical composition of a

typical carbon steel specimen.
Characterization of synthesized fatty amides, pyridine and pyrrole organ-
ics. Typical FTIR spectrums for these organics are as shown in Fig. 1.
Table 2 shows differences in the observed and literature FTIR spectrum for the
organics used as corrosion inhibitors.
From Table 2, it is observed that elemental N is commonly in existence in the
selected organic molecules which made-up the corrosion inhibitors organics.
Elemental nitrogen is known to have a pair of lone electron from the sp2 electron
configuration which promotes surface adsorption [8]. Surface adsorption, either
physisorption or chemisorptions enhances delay in the diffusion time of oxygen
molecules toward the metallic surfaces to enhance cathodic reaction.
Liner Polarization Resistance Method (LPRM) measurement, Rp values
obtained were measured by using the available potentiostat and galvanostat.
Through measured Rp values, the corrosion rate is determined to be inversely
proportional to the Rp of the test specimen [9]. Examples of the Rp-values obtained
from the measurements done are as shown in Fig. 2a–d.
Observations from Fig. 2a, b show large difference in the measures Rp-values
between fatty amides mixture added in various concentrations to the pyrrole and
pyridine organics added in the same concentration. It is also observed that the
inhibitive mechanism of the added fatty amides mixture is spontaneous in nature
and it is an effective inhibitor for first hour after addition of the organics.
Surface analysis, Fig. 3a shows surface morphology of the steel specimen
immersed in the saline solution, for a test period of 1 h, without any form of surface
protection. It is observed to be covered with surface hydroxides and for a specimen
which had been immersed in the saline solution for a prolonged period of time and
dried, the surface layer is observed to be covered with numerous, short length
surface cracks indicating corrosion activity had taken place.
However, Fig. 3b, c do not exhibit phenomenon observed in Fig. 3a. There were
some form surface protections shown in Fig. 3b, c. Figure 3c did not revealed any
forms of surface cracks indicating a better surface protection from the addition of
Table 2 Difference in FTIR wavelength: observed versus literature

Radical Wavelength (cm−1) Inferences
Observed Literature value
value [7]
C–H 838–699 680–860 Presence of C–H bond in the organic
N–H 3500–3180 3500–3300 Presence of N–H bond in the organics
C–N 1200–1025 1200–1029 Presence of C–N bond in the organics
C=O 1710–1558 1800–1678 Presence of C=O in the synthesized
inhibitor
[email protected]
An Investigation of N-Containing Organic Compound … 85
(a)
Polarization resistance (ohm.cm )
10000
2
8000 20 ppm fatty amide

50 ppm fatty amide
200 ppm fatty amide
4000 20 ppm pyridine
50 ppm pyridine
2000 100 ppm pyridine
200 ppm pyridine
0
10 20 30 40 50 60
Time (min)
(b)
10000
Polarization resistance (ohm.cm )
20 ppm fatty amide

2

100 ppm fatty amide
20 ppm pyrrole
4000 50 ppm pyrrole
100 ppm pyrrole
2000 200 ppm pyrrole
10 20 30 40 50 60
Time (min)
Fig. 2 Differences in the measured Rp-values in the test aqueous solution containing a fatty
amides mixture and pyridine and, b fatty amides mixture and pyrrole. All tests were conducted at
room temperature and in static condition
(a) (b) (c)
2µm 2µm 2µm
Fig. 3 Comparative steel specimen surface morphology by scanning electron microscopy.

a Surface morphology without inhibitor, b with addition of 20 ppm pyrrole, and c with addition of
200 ppm of fatty amides mixture [5, 6]
the fatty amides mixture. The surface protection offered by fatty amides mixture is
due to elemental nitrogen adsorption on the steel surface which is either physi-
sorption or chemicsorption in nature. Previous work by Hashim et al. [10] indicated
chemisorption is the type of surface adsorption preferred due to the presence of an
electron lone- pair on elemental nitrogen in fatty amide molecule.
[email protected]
Conclusion
The research demonstrated that pyrrole, pyridine and the synthesized fatty amides
mixture are corrosion inhibitors for protection of carbon steels when exposed to
saline water at room temperature and under static fluid environment. Under the test
conditions, as-synthesized fatty amides mixture out-performed other organics.
Acknowledgment The authors would like to express their thanks and gratitude to the Research
Management Institute (RMI), the Academic and International Affairs (HEA) and the Faculty of
Chemical Engineering of the Universiti Teknologi MARA for financial and other advisory sup-
ports in completing and presenting results of this research project.
References
1. Kern P, Landolt D (2001) Electrochim Acta 47:589–598

2. Ibrahim IM, Yunus S, Hashim MA (2013) Int J Sci Eng Res 4:1–12
3. Blinks BP, Fletcher PDI, Hicks JT, Horsup DI, Durnie WH (2005) Proceedings of NACE,
Corrosion Paper No.05307. NACE International, Houston, Texas
4. Cossy J, Pale-Grosdemange C (1989) Tetrahedron Lett 30:2771–2774
5. Hashim MA, Shamsudin S, Lim CG, Morat A, Toff MRM (2001) Investigation into inhibitive
effect of synthesized fatty amides on commercial-grade low-carbon steel, Presented at the 2nd
Asian Science and Technology Week Conference. Universiti Brunie Darussalam
6. Ibrahim IM, Jai J, Hashim MA (2014) Appl Mech Mater 575:206–209
7. Coates J (2000) In: Meyers RA (ed) Interpretation of infrared spectra, a practical approach.
Wiley, Chichester, pp 10815–10837
8. Al A, Hamzi H, Zarrok A, Zarrouk R, Salghi B, Hammouti S, Al-Deyab M, Bouachrine A,
Amine F (2013) Guenoun. Int J Electrochem Sci 8:2586–2605
9. Revie RW, Uhlig HH (2011) Uhlig’s corrosion handbook. Wiley, Chichester
10. Hashim MA, Salleh SNM, Toff MRM (2009) Mater Res Innov 13:291–294
[email protected]
Evaluation on the Effect of Inhibitor Ratio
to the Wax Deposition in Malaysian Crude
N. Ridzuan, Z. Yaacob and F. Adam
Abstract It is important to understand the fundamental nature of petroleum wax

formation since this problem is still a challenging issue in the crude oils industry.
The application of a wax inhibitor was introduced in this study to reduce the wax.
Poly (ethylene-co-vinyl acetate) (EVA) and cocoamide-diethanolmines (DEA) with
different ratios were used to evaluate their performance as an inhibitor using cold
finger equipment. A total of 10 ml inhibitor at different ratio of EVA and DEA was
injected into crude oil sample at constant cold finger and water bath temperature of
10 and 50 °C, respectively. The Percent Inhibition Efficiency (PIE) was increased as
the amount of EVA in DEA was increased. The use of pure EVA had successfully
showed the highest PIE value which is 23.08 % compare to other ratios. The
GC-FID analysis found that the C28H58 (n-octacosane) wax solid deposit was
reduced almost 90 % if pure EVA was used. The addition of DEA in the EVA
inhibitor does not give a significant effect to the prevention of wax solid formation.
It is highly recommended to use pure EVA as a single wax inhibitor in controlling
wax solid deposition.

Keywords Cold finger Wax deposition Ethylene-co-vinyl acetate (EVA)
Cocoamide-diethanolamine (DEA)
N. Ridzuan (&) Z. Yaacob F. Adam

Faculty of Chemical and Natural Resources Engineering, Universiti Malaysia Pahang,
Lebuhraya Tun Razak, 26300 Kuantan, Pahang, Malaysia
Z. Yaacob
F. Adam

[email protected]
88 N. Ridzuan et al.
Introduction
An increase in world crude oil demand has driven oil production industry to tap into
the deeper ocean to increase the offshore production. However, wax tends to deposit
from crude oil has become a very critical concern in the seabed due to the low
temperatures condition. This problem can cause a major problem in oil exploration
and transportations facilities [1–5]. Crude oil is a complex mixture of hydrocarbons,
consisting of paraffins, asphaltenes, resins, aromatics and naphthenes [6]. Wax
deposits are commonly contributed by asphaltenes and paraffins. For paraffins, the
deposit commonly originated from alkanes with high molecular weight range of
C18–C70. Paraffin can precipitate out from crude when the equilibrium condition is
changed resulting the loss of solubility of the paraffin in crude or when the tem-
perature of the crude drops below the cloud point temperature [7].
Many techniques including removal and prevention approach have been intro-
duced to minimize the wax problems such as chemical, mechanical and thermal
methods. Among them chemical method or also known as ‘paraffin inhibitors’ was
widely applied. Paraffin inhibitor can be divided into three types which are solvents,
dispersants and crystal modifiers. The effectiveness of certain paraffin inhibitor was
varied from one crude oil well to another well. It is required to find the best
inhibitor for the crude oil from specific well [8].
The paraffin inhibitors are commonly exist as polymeric compounds which
constituted of hydrocarbon chain and polar section. The hydrocarbon section will
connect the interaction between the additive and paraffin. While polar section
content in the inhibitor is responsible for the wax crystals morphology modification
which is necessary to inhibit the aggregation stage. It is strongly believed that both
segments can reduce or eliminate the build-up of wax. An effective wax inhibitors
can create a weaker deposit which subject to easy wax removal process if shears
force is applied the field flow. Several polymers have been evaluated as wax crystal
modifiers. EVA co-polymers is widely used due to the good efficiency as crude oil
flow modifiers and wax deposition inhibitor [5, 9]. EVA acts as Pour Point
Dispersant (PPD) where its capabilities of building into wax crystals through the
altering growth and surface characteristics of the crystal. Thereby, it is reduces the
tendency of the crystals to stick to metal surfaces such as pipe walls. In addition, it
has a the tendency to reduce viscosity and influence rheology behaviour by low-
ering down the pour point and wax appearance temperature [10]. Lucas and co
worker [9] reported that the different amount of vinyl acetate functional group
content in EVA will give different results in wax deposition efficiency.
On the other hand, DEA had been reported as one of good oil-in-water emulsifier
which can prevents the gelling problem in crude oil and can act as an inhibitor.
Although the individual mechanism of each PPD is well understood, but there is
very limited studies have been conducted on the combined effect of PPD with
emulsifier such as DEA in reducing the wax deposition rate. Hence, the main
[email protected]
Evaluation on the Effect of Inhibitor … 89
objective of this study is to evaluate the performance of the EVA:DEA ratio using
cold finger method by determining the best formulation of EVA:DEA ratio which
can demonstrate the best paraffin inhibition efficiency (PIE).
Experimental
Materials
Chemicals, Poly (ethylene-co-vinyl acetate) (vinyl acetate 25 wt.%, melt index

19 g/10 min), cocoamide-diethenolamine (DEA) (purity 99.5 %), cyclohexane
(purity 99.5 %), acetone (purity 99.5 %), heptane (purity 99.5 %) and petroleum
ether were obtained from Sigma Aldrich. Raw crude oil was kindly supplied by
Petronas Refinery from Kertih, Terengganu, Malaysia.
Preparation of EVA-solvent, VA-solvent solution (400 ppm) was prepared by
dissolving appropriate amount of EVA in cyclohexane under continuously stirring
at 300 rpm and 60 °C for one hour.
Measurement of Sample
Physical Properties
Table 1 showed the physical properties of the light crude oil sample used. Wax
appearance temperature (WAT) was determined using differential scanning calo-
rimetry (DSC). The viscosity was determined by Brookfield DV-III using spindle
type 31. Pour point measurement using cloud and pour point bath was supplied by
Koehler Company. The density has been was measured by pycnometer apparatus
Model Micromeritis AccuPyc II 1340. The wax content was determined by Gas
Chromatography with Flame Ionization Detector (FID). A capillary column ZB1,
30 m long and with 0.32 mm internal diameter and 0.25 μm film thickness, was
used. The samples were injected into the column using an auto-sampler at injector
temperature of 250 °C and a split/splitless ratio of 1:50. The temperature of the
detector was set up at 300 °C. Analysis was performed under a helium flow rate of
1.2 ml/min and a temperature program of 40–280 °C at 12 °C/min and held for
5 min at the highest temperature.
Table 1 Physical properties of the crude oil sample

Sample Viscosity mPa.s Pour Density (abs) SG °API WAT,
name at 40 °C point (°C) g/cm3 sample o
C
Light 5.18 3 0.8505 0.8505 34.87 33
[email protected]
Cold Finger Experimental Set up
The cold finger apparatus as shown in Fig. 1 was fabricated to evaluate the rate of
wax deposition in crude oils. This devices is suitable to correlate wax deposition
data to the temperature differences between the bulk oil and wall [4, 11, 12]. The
cold finger was placed within the center of stainless steel jar filled with 300 ml of
crude oil. The cold finger and jar were then placed inside a heating water bath to
control the crude oil temperature. An impeller was also attached at the middle of the
cold finger to make the field flow influences both the shear stress field and the rate
of heat transfer at the cold finger surface. Crude oils with and without wax inhibitor
were tested in the cold finger experiment. The experiment was carried out for two
hours to presume the aging effect upon the wax deposited onto the cold finger.
Evaluation of Paraffin-Inhibition Efficiency
Paraffin inhibition efficiency (PIE) was calculated using Eq. 1:
Wf Wt
PIEð%Þ ¼ 100 ð1Þ
Wf
where Wf is the reference amount of wax deposition without chemical treatment (g)
and Wf is the amount of paraffin deposition with chemical treatment (g) [12]. Five
different EVA:DEA ratios were studied.
Fig. 1 Schematic diagram of cold finger apparatus
[email protected]
Comparative Study on the Effects of Different

Wax Inhibitor Ratios
Five different ratios of EVA:DEA were examined based on the capability to

decrease the gross amount of paraffin deposit. Figure 2 shows the effect of inhibitor
ratio to the wax deposit and PIE at 0 rpm. The solid line represents the amount of
wax deposit in gram and the dotted line represents the PIE (%). The wax deposit or
crystal solid formed without the presence of the inhibitor (for control purpose) is
about 1.30 g. However, under the same running condition, the amount of wax
deposit reduced to 1.00 g if pure EVA was used. This reduction is equivalent to the
23.08 % in PIE value. The physical state of the deposited wax was observed
visually be a soft amorphous gel. The amount of wax deposit was slightly reduced
if the amount of EVA in DEA solution was increased as shown in Fig. 2. This
graphical trend describes the presence of more amount of EVA in the DEA solution
will reduce the formation of wax on the cold finger.
If the crude oils contain of pure DEA (10 ml of DEA), the PIE value was 6.15 %.
Nevertheless, DEA can helps to suppress the wax formation is some extent when
compared to the control experiment using pure crude oil without any inhibitor. It
might suggest that the activity coefficient of DEA is lower compared to EVA
inhibitor. The waxes solid formed was the highest at 100 % DEA concentration.
This result shows that pure DEA performance exhibit 50 % lesser than EVA alone.
The solid state of the deposited wax was found to be hard amorphous gel. The solid
state of the deposit at 5 ml:5 ml ratios of EVA:DEA was observed to be slightly
softer than the wax deposited from the pure DEA. The solid deposit was easily
Fig. 2 The effect of inhibitor ratio to the wax deposit and PIE at 0 rpm
[email protected]
removed from the cold finger compared to pure DEA. This concludes that the
presence of more EVA in DEA helps the solid structure softer with less amount of
solid wax.
The study reported that EVA will aggregate to the structure of the paraffin wax
by burying itself inside the wax crystals and absorbed itself on the surface of the
wax crystals. Therefore, it can coalesce with wax crystals and interfere with the
growth of crystals. Thus, the ability of crystals to interconnect and form networks
decreases. Therefore, more crystals with smaller size can be produced [5].
Figure 3 shows an example of a wax solid deposited which was analyzed by GC-
FID. This graph represents the solid formed during an experiment run at equal ratio
of EVA:DEA. The major compound n-octacosane (C28H58) was used to represent
wax solid because the highest composition amount presence in crude oil. Table 2
shows the details analysis of C28H58 percentage in wax solid and its percent
inhibition compared to the wax deposit (blank sample). This result gives an indi-
cation of the inhibitor performance. The amount of C28H58 wax solid decreased as
the amount of EVA in DEA was increased. The percentage of C28H58 in wax solid
(wt.%) also found a similar trend to total amount (weight) of wax solid as described
in Fig. 2. The inhibition percentage of C28H58 wax solid also increased significantly
when the 100 % of EVA used. It can be concluded that EVA performed a sig-
nificant chemical inhibitor compared to DEA in particular to inhibit the major
Malaysian crude compound n-paraffin C28H58 to deposit or precipitate.
Fig. 3 An Example of GC-FID chromatogram of cold finger wax solid deposit at equal EVA:
DEA ratio
[email protected]
Table 2 Percentage of C28H58 wax compound in total wax and it inhibition percentage
Sample Ratio of EVA to % of C28H58 in the total Percent inhibition of C28H58 in
DEA wax (wt.%) wax total (%)
Vial A 0:10 1.52 5.50
Vial B 2:8 1.56 3.42
Vial C 5:5 1.20 25.93
Vial D 8:2 0.24 85.41
Vial E 10:0 0.19 88.32
Vial F Without inhibitor 1.61 NA
Conclusion
The study found that EVA inhibits the formation of waxes solid more significantly
compared to DEA. The increasing of DEA amount in EVA does not significantly
reduce the wax solid formation because of the less activity coefficient. GC-FID
analyzed that the amount C28H58 wax reduced when 100 % of EVA was used as
inhibitor. This study was successfully evaluated the suitable inhibitor ratio to inhibit
the wax solid deposit for Malaysian crude oils. The details understanding on the
behavior of a chemical inhibitor to wax solid should be further studied to under-
stand the role of both chemical inhibitors at molecular level perspective.
Acknowledgment The authors express high appreciation to Universiti Malaysia Pahang and the
Ministry of Higher Education, Malaysia, through Research Acculturation Grant Scheme (RAGS-
RDU 121414) for having this research project successfully conducted.
References
1. Misra S, Baruah S, Singh K (1995) Paraffin problems in crude oil production and
transportation: a review. SPE Pro Facil 10:50–54
2. Jang YH, Blanco M, Creek J, Tang Y, Goddard WA (2007) Wax inhibition by comb-like
polymers: support of the incorporation-perturbation mechanism from molecular dynamics
simulations. J Phys Chem B 111:13173–13179
3. Lee HS (2008) Computational and rheological study of wax deposition and gelation in subsea
pipelines. University of Michigan, Michigan
4. dos Santos JDST, Fernandes AC, Giulietti M (2004) Study of the paraffin deposit formation
using the cold finger methodology for Brazilian crude oils. J Pet Sci Eng 45:47–60
5. Jafari Ansaroudi HR, Vafaie-Sefti M, Masoudi S, Behbahani TJ, Jafari H (2013) Study of the
morphology of wax crystals in the presence of ethylene-co-vinyl acetate copolymer. Pet Sci
Technol 31:643–651
6. Singh P, Venkatesan R, Fogler HS, Nagarajan N (2000) Formation and aging of incipient thin
film wax-oil gels. AIChE J 46:1059–1074
7. Abdallah EM, El-khatib N (2011) Modeling wax deposition in crude oil presented at National
Postgraduate Conference (NPC). Kuala Lumpur, 19–20 Sept 2011, pp 1–7
[email protected]
8. Lashkarbolooki M, Esmaeilzadeh F, Mowla D (2011) Mitigation of wax deposition by wax-

crystal modifier for Kermanshah crude oil. J Dispers Sci Technol 32:975–985
9. Lucas EF, de Machado ALC (1999) Poly (ethylene-co-vinyl acetate)(EVA) copolymers as
modifiers of oil wax. Pet Sci Technol 17:1029–1041
10. Pedersen KS, Rønningsen HP (2003) Influence of wax inhibitors on wax appearance
temperature, pour point, and viscosity of waxy crude oils. Energy Fuels 17:321–328
11. Jennings DW, Weispfennig K (2006) Effect of shear on the performance of paraffin inhibitors:
coldfinger investigation with Gulf of Mexico crude oils. Energy Fuels 20:2457–2464
12. Bello O, Fasesan S, Teodoriu C, Reinicke K (2006) An evaluation of the performance
of selected wax inhibitors on paraffin deposition of Nigerian crude oils. Pet Sci Technol
24:195–206
[email protected]
Solvent Extraction of Castor Beans Oil:
Experimental Optimization via Response
Surface Methodology
Jibrin Mohammed Danlami, Agus Arsad

and Muhammad Abbas Ahmad Zaini
Abstract Optimization of percentage of oil yield extracted from castor seed was
carried out using response surface methodology and Box-Behnken design. Castor oil
was extracted using solvent extraction with three different solvents; hexane, petro-
leum-ether and ethanol. The linear and quadratic effects of the two variables studied
were most significant in affecting the oil yield. The maximum yield of oil extracted
was 50.9, 49.0 and 59.5 % using hexane, petroleum ether and ethanol respectively as
solvents and the optimum conditions were at a time of 3 h and an average particle
size of 1 mm which were within the experimental domain. The model equations gave
a good fit with the coefficient of determination of 0.9216, 0.9587 and 0.9005 for
hexane, petroleum ether and ethanol, respectively, and hence the suitability of the
response surface methodology for the optimization of percentage oil yields.
Keywords Solvent extraction Response surface methodology Optimization

Castor Ricinoleic acid
Introduction
Castor oil is viscous, pale yellow and non-drying oil with a taste and in most times
used as purgative [1]. Castor oil can either be applied in its refined, hydrogenated
form or crude form. Typically, 28 % is refined, 65 % of it is processed, 12 % is
J. Mohammed Danlami M.A. Ahmad Zaini

Centre of Lipid Engineering and Applied Research (CLEAR),
Ibnu Sina Institute for Scientific and Industrial Research,
Universiti Teknologi Malaysia, 81310 Johor Bahru, Malaysia
M.A. Ahmad Zaini
A. Arsad (&)
Enhanced Polymer Research Group (ENPRO), Department of Polymer Engineering,
Faculty of Chemical Engineering, Universiti Teknologi Malaysia,
81310 Johor Bahru, Malaysia

[email protected]
96 J. Mohammed Danlami et al.
hydrogenated, 5 % is processed to manufacture other derivatives and 20 % is

dehydrated. Castor oil is extracted from castor beans by either aqueous traditional
technique, mechanical expression such as hydraulic pressing or screw pressing or
through solvent extraction [2, 3]. The preferred technique largely depends on some
socioeconomic factors and also the technology available within the area of use, [4]
mechanical pressing is often used because it ensures the extraction of non con-
taminated oil at a comparatively low cost [5, 6]. The production and quality of oil
extracted using mechanical expression depends on the pre-pressing process con-
ditions, such as the purity, dehulling, particle size and moisture content of the seed.
Since processing conditions modifies the chemical properties of vegetable oils,
appropriate combination of the processing conditions is needed to extract crude oil
with properties close to that of the refined oil. However, studies have shown that
moisture content, heating temperature and heating time had pronounced effects on
both the quality and the quantity of oil expressed from some oil seeds [4, 7–9].
Generally, extracted castor oil contains small amounts of free fatty acids,
phospholipids, mono- and diacylglycerols, protein degradation products, waxes,
hydrocarbons, moisture, dirt, and other contaminants, a refining process has to be
applied prior to derivatization of the oil. A typical refining process consists of the
following steps: degumming, neutralization, bleaching, deodorization, and some-
times winterization [10–12]. The degumming step is employed to reduce the
phospholipids and the metal content of the crude oil by combining it with water and
acid, such as citric or phosphoric acid [10–12]. The efficiency of solvent extraction
is affected by extraction time and particle size. To study the effect of these variables
on the output, the response surface methodology (RSM) is an effective technique
[13]. The principles and applications of RSM have been well described [14, 15].
The aim of the present research is to get a better understanding of the relation
between the variables (time and particle size) and determine the optimum condi-
tions for the extraction of castor oil. RSM is an efficient statistical technique for
modelling and analyzing the influence of several variables on the response of a
system and to optimize this response with minimum number of experiments. RSM
is a faster and more economical method for gathering research results than classic
one-variable-at-a-time or full-factors experimentation.
Materials
The castor bean seed was obtained from Ancient Greenfield PL, India. All
Chemicals used for seed oil extraction were hexane, petroleum ether and ethanol
and were purchased from Sigma Aldrich, Malaysia.
[email protected]
Solvent Extraction of Castor Beans Oil: Experimental Optimization … 97
Determination of Moisture Content of the Seeds
200 g of the cleaned and ground sample was weighed and dried in an oven at 90 °C
for 24 h and the weight was taken after every 2 h. The procedure was repeated until
a constant weight obtained. After each 1 h, the sample was removed from the oven
and placed in the desiccator for 30 min to cool. It was then removed and re-
weighed. The percentage moisture in the seed was calculated from the formula:

100 W1 W2
Moisture ¼ ð1Þ
W1
where W1 = Original weight of the sample before drying; W2 = Weight of the

sample after drying.
Soxhlet Extractor Operation and Percentage Oil

Determination
150 mL n-hexane solvent was measured and poured into a round bottom flask of
1000 mL. 10 g of the measured seed sample was inserted in a thimble placed in the
extractor. The round bottom flask was heated to 60 °C equivalent to the boiling
point of hexane. As the solvent boils, there was a rise of the solvent vapour into the
condenser to the top through a vertical tube. The condensate solvent dropped into
the thimble containing the castor seed sample. The extract, then fills the siphon tube
by seeping through the pores of the thimble and refluxed into the flask. The circle
continues for l, 2 and 3 h depending on the operating condition. The waste was then
dried, cooled in a desiccator and weighed to quantify the extracted oil. The entire
process was repeated in duplicate by placing another 10 g of the sample into the
thimble for the next batch of extraction. The ensuing mixture (miscella) containing
the oil at the end of the extraction was heated to recover the solvent from the oil by
using a rotary evaporator. The same process was repeated for ethanol and petroleum
ether as an extracting solvent and recovery.
Castor Oil Recovery
The % yield of castor oil was calculated using Eq. (2)
Y1 Y2
%Yield ¼ ð100Þ ð2Þ
Y2
where Y1 and Y2 are the weights of the beans after extraction, in grams.
[email protected]
Table 1 Experimental design for each of the independent variables used in developing the
experimental data for optimizing the solvent extraction of oil from castor seeds
Independent variables Symbol Coded levels
Coded Uncoded −1 0 1
Time, (min) X1 T 1 2 3
Particle size, (mm) X2 P 1 2.36 3.35
Statistical Analysis
Design of Experiment
To achieve the optimum oil yield by employing solvent extraction, the variables
considered (time and particle size) were optimized using RSM as summarized in
Table 1. RSM is a mathematical tool used for planning experiments, developing
polynomial models for predicting response, assessing the substantial effects of
factors and optimizing the required function [16].
Box–Behnken design was chosen with three factors to design the experiment.
The RSM was used to optimize the key variables influencing the extraction. The
RSM of this extraction experiment was conducted to achieve the optimum degree
extraction and optimum oil yield. In this statistical analysis, the relationship
between the coded and the actual variables can be expressed by Eq. (3).
Xi ¼ ðTi P0 ÞDTi ð3Þ
where Xi is the independent variable, Ti is the actual independent value; Po is the

independent real value at the centre point and ΔTi step change value. The oil yield
is taken as the dependent variables. Therefore, the proposed model for the response
is given by Eq. (4) [17, 18].
Yi ¼ b0 þ b1 X1 þ b2 X2 þ b11 X21 þ b22 X22 þ b12 X1 X2 ð4Þ
where Yi are independent variables, βo is the constant term, while β1, β2 are the
linear coefficient, β11, β22 are the squared terms and β12 is the interaction effect.
STATISTICA 6.0 was used for analysis of variance (ANOVA), and multiple
regression analyses of the data was obtained.
Experimental Data
The results of the experimental design (table not shown) shows the experimental
design and predicted responses for oil yield from castor bean seed.
[email protected]
Regression Model
The results of experimental design and predicted responses (table not shown) were
used to run ANOVA and Multiple Regression Analysis in STATISCA using
polynomial model Eq. (2). This allows the optimum oil yield and other responses to
be predicted. The coefficients of the model equation were used to predict the
optimum oil yield by multiple regression analysis using STATISCA 6.0 and are
shown as Eqs. (5–7).
Y1 ¼ 37:398 þ 4:948X1 2:341X2 þ 3:35X21 þ 1:667X22 ð5Þ
Y2 ¼ 37:873 þ 6:568X1 2:618X2 3:525X21 þ 1:992X22 ð6Þ
Y3 ¼ 51:297 þ 0:302X1 0:805X2 þ 5:775X21 1:142X22 ð7Þ
with Y1, Y2 and Y3 representing hexane, petroleum ether and ethanol solvent
extractions and the responses for the tested variables in coded terms; time of
extraction (X1) and average particle size (X2). The significance of each of the
variable is calculated as F-value of the model from the mean square. The F-value
for hexane, petroleum ether and ethanol are 14.16, 11.01 and 29.0 respectively,
with a very low probability value of <0.000001, <0.000001 and <0.00002
respectively. This shows a very high significant effect on the yield. It has been
reported that regression models with P values less than 0.05 indicates a statistical
significant. Analysis of variance (ANOVA) of the model indicates a good model
performance with an R2 value of 0.9216, 0.9587 and 0.9005 for hexane, petroleum
ether and ethanol respectively. An R2 of at least 0.80 is required for a model to be fit
[19]. More so, the probability values P > F and less than 0.05 (table not shown)
shows that the model terms are significant. Therefore, the percentage oil recovery
indicates that the model is most significant, as the calculated F-values representing
the three solvent of extraction are much greater than the table experimental value of
F(2,15,0.05) = 3.68.
Response Surfaces Analysis
The effect of both independent variables on the dependent variable can easily be
spotted using response plots generated from the STATISTICA software. The
combined effect of extraction time and average particle size are shown in Fig. 1a
using hexane as solvent of extraction. The optimum oil yield % was obtained at
extraction time of 3 h and particle size of 1 mm with the linear effect of extraction
time being dominant. Figure 1b also shows the effect of extraction time against
average particle size using petroleum ether as solvent of extraction. The illustration
shows the same pattern as that of hexane, with an extraction time being linear and
dominant. The time factor therefore, plays a significant role in determining the oil
[email protected]
Fig. 1 3D response surface plots for oil yield % using a hexane and b petroleum ether as an
extracting solvent versus extraction time and average particle size
Fig. 2 3D response surface plots for oil yield % using ethanol as an extracting solvent versus
extraction time and average particle size
yield. The effect of extraction time against average particle size is also shown in
Fig. 2 using ethanol as an extracting solvent, indicates that the interaction effect of
average particle size is linear and dominant over extraction time.
[email protected]
(a) Pareto Chart of Standardized Effects; Variable: Oil Yield (%)

2 3-level factors, 1 Blocks, 18 Runs; MS Pure Error=.3705556
DV: Oil Yield (%)
(2)Time (hr)(L) 22.85755
(1)Particle Size (mm)(L) -11.2865
Particle Size (mm)(Q) 10.31221
-.547586
Time (hr)(Q) -.547586
p=.05
Standardized Effect Estimate (Absolute Value)
(b) Pareto Chart of Standardized Effects; Variable: Oil Yield %

DV: Oil Yield %
(2)Extraction Time (hr)(L) 42.13735
Average Particle Size (mm)(Q) 19.23084
(1)Average Particle Size (mm)(L) -14.3884
Extraction Time (hr)(Q) -10.9088
p=.05
Fig. 3 Pareto Chart using a Hexane and b Petroleum ether as extracting solvent
Paretor Chart
The Pareto chart showed the most influence variable in the extraction process.
Figure 3a, b, shows the most influence variable was time indicating linear term with
values of 22.85755 and 42.137 using hexane and petroleum ether as the extracting
solvent. Followed by the average particle size in linear and quadratic term with
values of −11.2865 and −19.23084. However, ethanol shows a linear effect of
[email protected]
Pareto Chart of Standardized Effects; Variable: Oil Yield %

DV: Oil Yield %
(1)Average Particle Size (mm)(L) -8.01951
(2)Extraction Time (hr)(L) 5.932799
3.236321
Extraction Time (hr)(Q) 3.236321
3.058552
Average Particle Size (mm)(Q) 3.058552
p=.05
Fig. 4 Pareto chart using ethanol as an extracting solvent
average particle size with a value of −8.0195 and followed by the extraction time in
linear term with a value of 5.932799 as shown in Fig. 4. The positive sign indicates
that the dependent variable can be operated at the maximum level while the neg-
ative sign indicates that the dependent variable can be operated at a minimum level
to get the best response.
Optimization
The models obtained were used in determining the optimum conditions for pre-
diction of oil recovery at an average particle size and time of 1 mm and 2.53 h
respectively for ethanol (table not show). These conditions corresponded to 57.8 %
of predicted oil recovery. Confirmatory experiments were run with these optimum
values, average particle size and extraction time. The observed value of oil recovery
was 59.5 % indicating 1.7 % error between observed and predicted value. This error
is considered small as the observed value is within the 5 % significance level.
Conclusion
Response surface methodology (RSM) was successfully applied for optimizing oil
yield for solvent extraction. The linear term of extraction time significantly affected
the oil recovery while using hexane and petroleum ether as an extracting solvent. In
addition, the interactions between extraction time and average particle size had a
significant effect on the oil recovery. The parametric analysis was adopted to find
the optimum operating conditions of average particle size of 1 mm and an
extraction time of 2.53 h. The experimental oil recovery of 50.9, 49 and 59.5 %
[email protected]
were achieved using the optimum conditions, which is compatible with the pre-
dicted value of 51.6, 49.2 and 57.8 % of hexane, petroleum ether and ethanol
solvent respectively. Therefore, the model used in predicting the oil recovery gave a
good fit with the experimental variables and hence the suitability of the RSM
method for the optimal performance.
References
1. Othmer K (1979) Encyclopaedia of chemical technology. Wiley, New York

2. Oyinlola A, Ojo A, Adekoya LO (2004) Development of a laboratory model screw press for
peanut oil expression. J Food Eng 64:221–227
3. Takashi Y (2006) Development trend of oil expression and degumming. Oleoscience 6:133–138
4. Faborode MO, Owolarafe OK, Lasisi AA, Kasali SA, Oguntuase KS (2003) Assessment of
seed-oil extraction technology in some selected states of Nigeria. Technovation 23:545–553
5. Singh J, Bargale PC (2000) Development of a small capacity double stage compression screw
press for oil expression. J Food Eng 43:75–82
6. Bamgboye A, Adejumo A (2007) Development of a sunflower oil expeller. CIGR E-J IX,
Manuscript EE 06 015
7. Ibrahim A, Onwualu AP (2005) Technologies for extraction of oil from oil-bearing
agricultural products: a review. J Agric Eng Technol 13:58–70
8. Abidakun OA (2010) Mechanical expression and lubricity of Irvingia gabonensis (Dika nut) oil.
M.Sc. Thesis, Department of Mechanical Engineering, Obafemi Awolowo University, Ile Ife
9. Mwthiga G, Moriasi L (2007) A study of yield characteristics during mechanical oil extraction
of preheated and ground soybeans. J Appl Sci Res 3:1146–1151
10. Anderson D (2005) A primer on oils processing technology. In: Shahidi F (ed) Bailey’s
industrial oil and fat products. Edible Oil and Fat Products: Processing Technologies, Wiley,
Hoboken, pp. 1–56
11. Kemper TG (2005) Oil extraction. In: Shahidi F (ed) Bailey’s industrial oil and fat products.
Edible Oil and Fat Products: Processing Technologies, Wiley, Hoboken, pp. 57–98
12. Ali MF (2005) Edible oils, fats and waxes. In: Ali MF, El Ali BM, Speight JG (eds) Handbook
of industrial chemistry: organic chemicals. McGraw-Hill, New York, pp. 86–121
13. Bas D, Boyaci IH (2007) Modeling and optimization I: usability of response surface
methodology. J Food Eng 78:836–845
14. Giovanni M (1983) Response surface methodology and product optimization. Food Technol
37:41–45
15. Myers RH, Montgomery DC (1995) Response surface methodology: process and product
optimization using designed experiments. Wiley, New York
16. Zhang Q, Tang L, Zhang J, Mao Z, Jiang L (2011) Optimization of thermal-dilute sulphuric
acid pre-treatment for enhancement of methane production from cassava residues. Bioresour
Technol 102:3958–3960
17. Shehu MS, Abdul Manan Z, Wan Alwi SR (2012) Optimization of thermo-alkaline
disintegration of sewage sludge for enhanced biogas yield. Bioresour Technol 114:69–74
18. Cornell JA (1990) How to apply response surface methodology, revised edition. tammy griffin.
Book Crafters, United State of America
19. Guan X, Yao H (2008) Optimization of viscozyme L assisted extraction of oat bran protein
using respond surface methodology. Food Chem 106:345–350
[email protected]
Immobilization of Nano-sized TiO2
on Glass Plate for the Removal
of Methyl Orange and Methylene Blue
Lim Ying Chin, Lim Ying Pei, Razzana binti Rosli

and Nur Hazira binti Mohd Atni
Abstract Titanium dioxide (TiO2) is the most researched photocatalyst due to its
potential for dye removals in wastewater treatment. However, the TiO2 powder
suffers from agglomeration which hindered its process of industrialization. Thus,
this study was carried out to immobilize nano-sized TiO2 on glass plate via doctor
blade technique and the crystal structure and surface morphology of the samples
were accomplished using XRD and FESEM. The effects of operational parameters
such as effect of light and pH in the removal of methyl orange (MO) and methylene
blue (MB) using the as-prepared TiO2 films were investigated. MO achieved a
removal of 95 %, while only 47 % removal efficiency for MB after 120 min under
UV-light illumination. On the other hand, percentage removal for MO and MB are
recorded as 65 and 28 % respectively under LED light irradiation. The photocat-
alytic removal of MO and MB was found to be mainly due to the adsorption-
photocatalysis process, while the photolysis process is almost negligible. The
kinetic study indicates that the photodegradation rate of MO and MB is fitted to the
pseudo-first order rate model.
Keywords Adsorption Photodegradation Dyes Titanium dioxide
L.Y. Chin (&)

Faculty of Applied Sciences, Universiti Teknologi MARA,
L.Y. Pei R.b. Rosli N.H.b. Mohd Atni
R.b. Rosli
N.H.b. Mohd Atni

[email protected]
106 L.Y. Chin et al.
Introduction
Large quantities of dyes are extensively used in the textile industries particularly
those involved in the finishing processes. As such, the textile effluents contain large
amount of organic compound that is not easily to be treated using conventional
methods. Environmental pollution by organic dyes possesses severe ecological
problem as the organic dyes are toxic to microorganisms and have long degradation
time in the environment. Consequently, wastewater treatment methods such as
adsorption, reverse osmosis, flocculation and biological degradation have been used
to treat such effluents. Ultimately, advanced oxidation process (AOPs) for the
destruction of synthetic organic species that are highly potent and strongly oxi-
dizing radical allows the destruction of a wide range of organic chemical substrate
with no selectivity [1]. Among the AOPs, heterogeneous photocatalysis employing
TiO2 has proven as efficient method for degrading both aquatic and atmospheric
organic contaminants. TiO2 has emerged as an excellent photocatalyst due to its
non-toxicity, high corrosion stability, high photocatalytic efficiency and the
photogenerated holes are highly oxidizing [2].
TiO2 can be used either in suspension mode or immobilized it on a substrate as
thin film. Practically, immobilization method is more preferable over the suspension
mode as it saves unnecessary cost for catalyst separation and simplify the purifi-
cation process. In addition, TiO2 tends to agglomerate especially at high concen-
trations and this set a limitation towards continuous flow system [3]. Therefore, the
objective of this work is to immobilize the TiO2 suspension on the glass plate and to
evaluate the photocatalytic activity of the immobilized TiO2 samples in degradation
of methyl orange (MO) and methylene blue (MB). Further more, the effects of
different light source and initial dye solution pH were investigated.
Experimental
Materials
TiO2 powder was obtained from Riedel-de Haën. Polyvinyl alcohol and formal-
dehyde were purchased from Sigma Aldrich. Hydrochloric acid (HCl) obtained
from JT Baker and sodium hydroxide pellets (NaOH) purchased from Merck were
used to adjust the solution pH. All the materials were used as received without
further purification. Methyl orange purchased from Bendosen and methylene blue
obtained from Unilab were selected as the model pollutant. The MO and MB
solution was prepared through dissolving the MO and MB powder in ultra pure
water (Direct-Q3 Water Purification System).
[email protected]
Immobilization of Nano-sized TiO2 on Glass Plate … 107
Preparation of Immobilized TiO2
1.5 g of anatase TiO2 powder was added to 15 mL of ultra pure water and the
mixture was stirred for an hour to ensure homogeneity. In order to attach the
prepared TiO2 suspension to the substrate, the binder was prepared by mixing 5 g of
polyvinyl alcohol to 80 mL of formaldehyde in water bath at 70 °C under con-
tinuous stirring until it became viscous and sticky. Next, glass plate with a
dimension of 2.54 cm × 7.62 cm was weighed prior to coating. Then, the plates
were applied with a thin layer of binder followed by TiO2 suspension. The layered
of binder and TiO2 suspension was dried and the coated samples were weighed
again to determine the amount of TiO2 coated onto glass plate. Morphological
characterization of the sample was carried out using a Field Emission Scanning
Electron Microscope (FESEM, Carl Zeiss Supra 40 V). The structure of the sample
was identified using an X-ray Diffractometer (XRD, Shimadzu D6000) using Cu
Kα radiation (λ = 1.5406 Å).
Adhesion Strength Test
The strength test of the immobilized TiO2 was carried out to evaluate the adhesion
property of the binder as described in [4]. Briefly, the immobilized TiO2 plate was
immersed into a beaker containing pure water and was sonicated for 30 s using a
37 kHz ultrasonicator (Elmasonic S30/(H). The differences in weight of the glass
plate before and after every interval of 5 s of ultrasonication were measured to
obtain the percentage of TiO2 particles adhere on the glass plate.
Dyes Removal
A custom-made glass photoreactor measured by 17 cm × 12 cm × 12 cm was used

for all the photodegradation experiments. Five pieces of TiO2 plate were used as
photocatalyst and irradiated with an UV fluorescent lamp (Generic, 9 W) at a
distance 10 cm from the top of the dye solution. Oxygen was continuously supplied
into the solution using an aquarium pump (Regent calm, RC-002) with a flow rate
of 37 cm2/s. In typical experiment, 200 mL of MO and MB solution with initial
concentration of 10 ppm was used as model pollutant. To investigate the effect of
light source on the dyes removal, the UV light was replaced with LED lamp
(Aletko, 9 W). The initial pH of the MO and MB solution was pH 6.2 and 4.7,
respectively. It was adjusted by adding either 0.1 M HCl or 0.1 M NaOH to the
desired values in order to study the influence of pH value (pH range 3–9) on the dye
degradation.
[email protected]
5 mL of the samples were taken every 15 min for a period of 120 min. The
concentration of MO and MB was determined by measuring the absorbance at
λ = 465 nm and 650 nm for MO and MB, respectively using a DR 2800 HACH
spectrophotometer. The photocatalytic decolorization efficiency was calculated
using Eq. 1 and the kinetics was determined using the first-order kinetic equation as
shown in Eq. 2:
Dye removalð%Þ ¼ ðC0 CÞ=C0 100 % ð1Þ
ln C=C0 ¼ k1 t ð2Þ
where C is the concentration of substrate in solution at time t of illumination (ppm);

C0 is the concentration of the substrate in solution at time zero of illumination
(ppm); and k1 is the pseudo first-order rate constant (min−1). The rate constant, k1
was obtained from the slope of the straight-line plots of ln C/C0 versus time, t.
Control Sets of Experiment
In order to verify the photocatalytic degradation of the samples, two sets of control
experiments were carried out on MO and MB for 120 min. First set of experiment
was photolysis (direct illumination of UV lamp without TiO2). This is important to
test the dye degradability under light exposure. The second set was adsorption
(presence of TiO2 without light illumination) with the aim to examine whether
removal of dye occured via adsorption on TiO2 surface. The MO and MB solution
used for control experiments was acidic without adjusting the pH.
Characterization of TiO2
Adhesion strength, Fig. 1 shows the percentage of TiO2 remained at different

sonication times. As can be seen from Fig. 1, the amount of TiO2 decreased to about
67 % for the first 15 s and remained constant thereafter, suggesting good adherence
of TiO2 films on the glass plate. This result also implied that the mixture of
polyvinyl alcohol and formaldehyde offers good adhesion and acted as good
adhesive to strengthen the TiO2 coating on the glass plate. Similar results using
other inorganic binders such as sodium silicate and zinc phosphate were reported by
Park et al. [5, 6].
XRD analysis, Fig. 2 shows the XRD pattern of TiO2 sample. It is discernible
that the sample exhibits crystalline anatase at 2θ = 25.2°, 36.9°, 37.8°, 38.5°, 48.0°,
[email protected]
Fig. 1 Amount of TiO2 (%) 100
The amount of TiO2

remained on the glass plate at
different sonication time 80
remained (%)
60
40
20
0
0 5 10 15 20 25 30
Times
Fig. 2 XRD pattern of TiO2 2000

powder A
Intensity /counts
1500
1000
500 A
A AA
A A
0
20 25 30 35 40 45 50 55 60
2 theta/degree
53.9° and 55.0° corresponding to (101), (103), (004), (112), (200), (105) and (211)
plane, respectively (JCPDS No. 21-1272). The broad diffraction peaks are mainly
due to small tetragonal crystal size.
FESEM analysis, Fig. 3 shows the FESEM images for immobilized TiO2 onto
glass plate. At 5000× magnification, the glass plate was well covered with TiO2
without any cracks. 50,000× magnification view of the sample revealed that the
sample comprised of almost spherical-like structure with diameter approximately
95 ± 5 nm. TiO2 nanoparticles are more desirable as this could provide higher
surface area and more active site for photodegradation compare to that of bulk TiO2.
Photodegradation of MO and MB, Control tests, The colour removal of MO
and MB at different conditions is shown in Fig. 4. From Fig. 4, the photolysis rate for
both MO and MB dye under photolysis is almost negligible (9 %), indicating that
MO and MB is highly stable towards photolysis. However, the adsorption of MO
was found to be 67.3 %, while for MB is 18.4 %. MO molecules are known to be
anionic in nature and under acidic conditions, the positive charged TiO2 surface
being more available for the absorption of the MO molecules. For both MO and MB,
the highest degradation rate was obtained through photocatalysis, indicating that the
presence of UV light is necessary for the degradation of azo dye using TiO2.
Effect of different light source, Fig. 5a depicts the degradation of MO and MB
under UV light illumination. MO achieves a removal rate of 95 %, while MB only
47 % after 120 min of photodegradation. This result implied that with the presence
[email protected]
Fig. 3 FESEM images of immobilized TiO2 onto glass plate with a 50000x and b 5000x
magnification
(a) (b)
Colour Removal of MO
120 120
Colour Remobal of MB
100 100
80 80
Adsorption
(%)
60 60
(%)
Photolysis
Photocatalysis
40 40 Adsorption
Photolysis
20 20 Photocatalysis
0 0
0 50 100 150 0 50 100 150
Time (min) Time (min)
Fig. 4 Effect of different conditions on colour removal of a MO and b MB
(a) 1.2 (b) 1.2 MO

Methyl Orange
MB
1.0 Methylene Blue 1.0
0.8 0.8
C/Co
C/Co
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
0 100 200 0 100 200
Time (min) Time (min)
Fig. 5 Degradation of MO and MB as a function of irradiation time under illumination of a UV

lamp and b LED lamp
[email protected]
of UV light, the photodegradation of both MO and MB has further enhanced by

30 % compared to the adsorption process. Under UV light illumination, electrons
excite from a filled valence band to an empty conduction band, giving rise to
electron–hole pairs. The photo-generated holes and electrons can migrate to the
surface of TiO2 particles and serve as redox sources that react with adsorbed
reactants, leading to the formation of superoxide radical anions, hydrogen peroxide
and hydroxyl radicals involved in the oxidation of dye. The mechanism of the
photodegradation of dye by using TiO2 is proposed as follows [7]:
TiO2 þ hv ! TiO2 ðe ; hþ Þ

hþ þ OH ! OH;
e þ O2 ! O
2
dye þ O
2 or OH ! peroxylated or hydroxylated intermediate
! degraded products
Figure 5b depicts the degradation of MO and MB under LED light illumination.

After 120 min, MO achieves a removal efficiency of 65 %, while MB yields 28 %.
This showed that photodegradation of MO under LED illumination does not vary
from the adsorption process as the energy from the LED light is not sufficient to
generate the electron-hole pairs, However, LED does enhance the removal of MB
with approximately 10 % of rises. Similarly to the one under UV light illumination,
photodegradation of MB with LED illumination follows the first order kinetic.
Table 1 shows the value of the rate constant and half time for both the MO and MB
with different light source. As seen from the correction factors, R2, the result shows
that the photodegradation of MO and MB follows the first order kinetic.
Effect of different pH, The degradation of MO at different pH is shown in
Fig. 6a. The ability of TiO2 to degrade MO in alkaline medium was lower com-
pared to that in acidic medium. It was observed that the decolorization efficiency
increases with increase in pH, exhibiting maximum efficiency (95.4 %) at pH 6,
beyond which the rate of degradation decreased at pH 9 (72 %). The initial pH of
the dye solution did affect the degradation of MO and MB. Point of zero charge
(pHpzc) of TiO2 is 6.8. Therefore, the TiO2 surface becomes positively charged
with the species TiOH2+ at pH < pHpzc, negatively charged with the species TiO−
at pH > pHpzc and neutral with the species TiOH at pH = pHpzc [8]. Furthermore,
Table 1 Kinetics data for MO and MB under UV and LED light illumination
Light Dye First order rate constant, Half-life time, Correlation
source k (× 10−3 min−1) t1/2 (min) factor, R2
UV MO 26.3 26.3 0.95
MB 5.2 133.3 0.98
LED MO 7.7 90.0 0.82
MB 2.7 256.7 0.87
[email protected]
(a) Time (min) (b) Time (min)

0 20 40 60 80 100 120 0 20 40 60 80 100 120
0 0
R² = 0.8313 R² = 0.8851
-1 -0.5 R² = 0.9263
R² = 0.9537
ln C/Co
ln C/Co
R² = 0.5035
-2 -1
R² = 0.9753
R² = 0.7942
-3 -1.5 pH3.0 pH 5.2
pH 3.0 pH6.29 pH7.0 pH 9.0
pH9.0 Linear (pH 3.0) Linear (pH3.0) Linear (pH 5.2)
Linear (pH6.29) Linear (pH9.0) Linear (pH7.0) Linear (pH 9.0)
-4 -2
Fig. 6 Effect of pH on the degradation of a MO and b MB
protonation and deprotonation condition of the photocatalysts surface by taking

TiO2 as a model, can be shown as follows [9, 10]:
TiOH þ Hþ ! TiOHþ2 in acidic solution

TiOH þ OH ! TiO þ H2 O in alkaline solution
In acidic medium, the TiO2 surface is positively charged. Since MO is an anionic

dye, the electrostatic attraction between TiO2 and MO can be achieved in acidic
medium and thus results in higher percentage removal [11]. Oppositely, the surface
of TiO2 becomes negatively charges in alkaline medium. This causes a repulsive
force between the surface of TiO2 and MO, leading to the reduction of percentage
removal.
Contrast trend was observed for the removal of MB. The ability of TiO2 to
degrade MB in alkaline medium was higher compared to that in acidic medium.
The percentage of removal for MB was the highest at pH 7 (70.7 %) followed by
pH 9 with a degradation efficiency of 58.4 %. Since MB is a cationic dye, the
electrostatic attraction between TiO2 and MB can be achieved in alkaline medium
and thus results in higher percentage removal. As expected, the percentage removal
for MB at pH 3 is only 20.5 % ascribed to repulsive force between the surface of
TiO2 and MB.
Conclusion
Nano-sized TiO2 suspension has been successfully immobilized on the glass plates
for the removal of azo dyes. The photodegradation of both MO and MB was found
to be due to combination of photocatalysis and adsorption process, while the
photolysis was almost negligible. Under illumination of different light source, UV-
light irradiation has shown better degradation efficiency compared to LED for both
MO and MB dye. The photo decolorization rate was the highest in pH 6 for MO
[email protected]
and in pH 7 for MB with a maximum efficiency of 95.4 and 70.7 %, respectively.

The kinetic data shows that the photodegradation of MO and MB follows the
pseudo-first order reaction.
Acknowledgment The authors would like to acknowledge the financial support from the
Universiti Teknologi MARA through Research Intensive Faculty Fund (RIF with No. 600/RMI/
DANA 5/3/RIF 381/2012) and the Ministry of Education Malaysia via Fundamental Research
Grant Scheme (FRGS/1/2014/SG01/UiTM/02/1).
References
1. Andreozzi R, Caprio V, Insola A, Marotta R (1999) Advanced oxidation processes (AOP) for
water purification and recovery. Catal Today 53(1):51–59
2. Yu JC, Yu J, Zhao J (2002) Enhanced photocatalytic activity of mesoporous and ordinary
TiO2 thin films by sulfuric acid treatment. Appl Catal B Environ 36(1):31–43
3. Pal M, Serrano JG, Santiago P, Pal U (2007) Size-controlled synthesis of spherical TiO2
nanoparticles: morphology, crystallization, and phase transition. J Phys Chem C 111
(1):96–102
4. Nawi MA, Zain SM (2012) Enhancing the surface properties of the immobilized Degussa P-25
TiO2 for the efficient photocatalytic removal of methylene blue from aqueous solution. Appl
Surf Sci 258(16):6148–6157
5. Park S, Choi GR, Lee JC, Kim YC, Lee JH (2012) Effects of zinc phosphate binder on the
immobilization properties of photocatalytic ZnO nanopowders synthesized by a solution
combustion method. J Ceram Process Res 13:S170–S173
6. Park S, Lee JH, Yoo K, Park HJ, Kim HS, Lee JC (2008) Adhesion properties of inorganic
binders for the immobilization of photocatalytic ZnO and TiO2 nanopowders. J Phys Chem
Solids 69(5):1461–1463
7. Su Y, Yang Y, Zhang H, Xie Y, Wu Z, Jiang Y, Wang ZL (2013) Enhanced photodegradation
of methyl orange with TiO2 nanoparticles using a triboelectric nanogenerator. Nanotechnology
24(29):295401
8. Chiou CH, Wu CY, Juang RS (2008) Photocatalytic degradation of phenol and m-nitrophenol
using irradiated TiO2 in aqueous solutions. Sep Purif Technol 62:559–564
9. Habib MA, Ismail IMI, Mahmood AJ, Ullah MR (2012) Photocatalytic decolorization of
brilliant golden yellow in TiO2 and ZnO suspensions. J Saudi Chem Soc 16:423–429
10. Liao J, Lin S, Zhang L, Pan N, Cao X, Li J (2012) Photocatalytic degradation of methyl
orange using a TiO2/Ti mesh electrode with 3D nanotube arrays. ACS Appl Mater Interfaces 4
(1):171–177
11. Yang H, Zhang K, Shi R, Li X, Dong X, Yu Y (2006) Sol–gel synthesis of TiO2 nanoparticles
and photocatalytic degradation of methyl orange in aqueous TiO2 suspensions. J Alloys
Compd 413(1):302–306
[email protected]
Mercury Removal in Simulated Flue Gas
by Oil Palm EFB Activated Carbon
Noor Hidayu Abdul Rani, Nor Fadilah Mohamad, Sharmeela Matali

and Sharifah Aishah Syed A. Kadir
Abstract This study investigated the effects of simulated flue gas temperature and
mercury inlet concentration on the mercury removal efficiency using activated
carbon produced from oil palm empty fruit bunch (EFB) which is prepared using
physical (steam) activation method. The elemental, proximate, porous structure and
particle size were performed for the oil palm EFB activated carbon. At low tem-
perature, performance of activated carbon showed a good potential for adsorption of
mercury. However, as temperature increase from 90 to 180 °C, the efficiency of
mercury was found to be decreased from 92 to 64 %. When the inlet mercury
concentration was increased, the mercury removal efficiency is increased from 55 to
98 %. It was observed that mercury removal was favoured at lower temperature,
such as 30 and 90 °C and higher mercury concentration, such as 0.192 ppm for
virgin oil palm EFB activated carbon.

Keywords Mercury Adsorption process Activated carbon Oil palm empty

fruit bunch Simulated flue gas
N.H. Abdul Rani (&) N.F. Mohamad S. Matali S.A. Syed A. Kadir
40450 Shah Alam, Malaysia
N.F. Mohamad
S. Matali
S.A. Syed A. Kadir

[email protected]
116 N.H. Abdul Rani et al.
Introduction
Mercury pollution is a global crisis encountered by society today, because it is a

toxic and hazardous element that can severely threaten human health and the
environment. The emissions of mercury in atmospheric come from flue gas, which
from anthropogenic sources. Flue gas consists of three forms which are oxidized
form (Hg2+), particulate form (Hgp) and elemental form (Hg0). The elemental
mercury (Hg0) is the predominant form of atmospheric Hg, which has a long
residence time in the atmosphere from 0.5 to 2 years and can be dispersed and
deposited in remote places even 1000 km away from the sources [1, 2].
With the exception of mercury mining itself, the mercury emissions arise from
mercury that is present as an ‘impurity’ in the fuel or raw material used. The main
‘by-product’ emissions are from sectors that involve combustion of coal or oil,
production of pig iron and steel, production of non-ferrous metals, and cement
production. So, due to the growth of these industries, concerns on the mercury
emission should not be ignored as the level of mercury in the air will potentially
increase. This concern has led to the development of technologies and researches
for reduction of mercury in flue gas especially from the largest global single-known
source of anthropogenic i.e. coal-fired power plant. However, existing pollution
abatement technologies such as flue gas desulfurization and baghouses are inef-
fective for the capture of Hg0 because of its low melting point, high equilibrium
vapor pressure and low solubility in water. Therefore, most Hg0 is released into the
atmosphere. Recent researches have proved that application of activated carbon can
be effective adsorbent for the removal of vapor phase mercury in a variety of
conditions [3, 4].
However, using commercial activated carbon as adsorbent for the removal Hg0
will be increasing the cost of operating process because the adsorbent costs are
relatively high. Therefore, many researchers have focused on alternative low cost
adsorbents generated from agricultural wastes [5]. Extensive previous studies have
shown that agricultural wastes such as corn cob, coconut shell, grain sorghum, olive
stone, walnut shell, palm shell and date palm seed have been found to be suitable
precursors owing to their high carbon and low ash contents [6]. Plus, they have
unique properties, including large surface area, a high degree of surface reactivity,
universal adsorption effect, and favorable pore size [7].
In this study, oil palm EFB is choosing to be utilized as activated carbon since it
is abundantly available in Malaysia and has very low market value. Indeed, con-
version of oil palm EFB to activated carbon would directly solve part of the
environmental problem. So, this paper aims to investigate the performance of Hg0
adsorption on oil palm EFB activated carbon surface. In order to accomplish the
targets, mercury adsorption tests have been conducted in a fixed bed stainless steel
reactor under simulated flue gas and activated carbon was prepared under physical
(steam) activation.
[email protected]
Mercury Removal in Simulated Flue Gas … 117
Experimental
Materials. Oil palm EFB was provided by Meru Palm Oil Mill Sdn Bhd in Klang,
Selangor. The bio-char of oil palm EFB was produced using pyrolysis at temper-
ature 500 °C.
Preparation of activated carbon. The activated carbon was prepared by
physical (steam) activation process of the bio-char using nitrogen as the carrier in a
tubular reactor. The superheated steam was injected at the flow rate 120 mL/h with
the activation temperature at 764.5 °C and activation time of 77 min. This optimum
operating condition was obtained for steam activation of bio-char using response
surface methodology (RSM).
Characterization of activated carbon. Proximate analysis was performed using
thermogravimetric analyzer, TGA (Mettler Toledo, United States) according to the
ASTM D7582-10 and the results are shown in terms of moisture, volatile content,
fixed carbon and ash contents. Ultimate analysis was carried out using CHNS-O
Analyzer, (FlashEA, 1112 Series, United States) to determine carbon, hydrogen,
nitrogen and oxygen (by difference) of the samples. BET (Brunauer, Emmett, and
Teller) surface area and pore structure of the precursor and activated carbon was
determined by automated gas adsorption Analyzer, AUTOSORB-1 (Quanta
Chrome Instruments, USA).
Mercury adsorption process. The mercury removal efficiency of the virgin oil
palm EFB activated carbon from gaseous phase was carried out in a fixed bed
reactor. The schematic diagram of the experimental setup was shown in Fig. 1. The
reactor was stainless steel reactor with 20 mm inner diameter and 20 cm long. A
screen was fitted in the reactor to support the activated carbon. The concentration of
mercury in the process gas was controlled by varying water bath temperature. The
permeation device (VICI Metronics) was used to provide a vapor phase Hg0 source
Gas mixing
chamber
FLUE GAS INLET
L
1L/min Types of AC
Stainless Steel - AC EFB
Reactor -Modified AC
Rotameter Rotameter Rotameter
Temperature
ACTIVATED
Controller
CARBON
F
Rotameter 500mg
90-180 oC
F
Water Bath
Hg Permeation
device
Impingers solution
Vent
Impingers solution FLUE GAS OUTLET
Vent
Mercury Sampling
CO2 O2 N2 and Detecting
Ice Bath Device
12% 6% Balance
Ice Bath
Fig. 1 Experimental setup for mercury adsorption process
[email protected]
with nitrogen (200 ml/min) as carrier gas. Premixed gas mixtures of carbon dioxide
(12 %), oxygen (6 %) and nitrogen (balance) was passed through the tube and
mixed together at gas mixing chamber for a certain time to ensure perfectly mixing
is achieved. The total flow rate of flue gas is set to 1.0 L/min. After mixing, the
simulated flue gas was flowing directly to stainless steel reactor and kept inside the
reactor at selected temperatures, i.e. 30, 90, 140 and 180 °C for 2 h in order to
determine the efficiency of mercury adsorption in flue gas. For virgin EFB activated
carbon, 500 mg sample was used in the reactor. Impinger solution (4 W/V%
KMnO4, 10 V/V% H2SO4) was placed downstream of reactor for total mercury
analysis. Atomic absorption spectrometry, AAS (Perkin Elmer, Lambda 25 UV
Spectrophotometer) was used measure the concentration of mercury.
Characterization of activated carbon. The elemental and proximate analysis were

performed in Table 1 for raw oil palm EFB and activated carbon. It showed that
after activation, the carbon content of activated carbon increased from 48.48 to
68.32 wt%. This is because at high temperature, volatile content was released [8].
With high carbon content, there is high possibility the amount of active site is
increased which is helped for mercury adsorption [9]. For proximate analysis, EFB
has high fixed carbon and volatile contents. As a result, EFB is a good and viable
options precursor for the production of activated carbon and with low ash content,
surface area of activated carbon is increased and gave the higher adsorption effi-
ciency [11].
The BET surface area, micropore area, total pore volume, and micropore volume
of the activated carbon are shown in Table 2. It can be seen that the BET of
activated carbon is higher compared to its raw material. This is due to at high
temperature, fragile structure was generated, resulting in increased pore size and
BET surface area [10]. The result also shows a high proportion of micropore
volume (about 90 % of total pore volume) which means that oil palm EFB activated
carbon indicates a higher volume of wide micropores and presence of small mes-
opores. With high proportion of micropores, high mercury adsorption capacity
because micropores are responsible for mercury adsorption while mesopores serves
as transport routes [3].
Table 1 Proximate and ultimate analysis of oil palm EFB and activated carbon
Sample Proximate analysis (wt%) Ultimate analysis (wt%)
Moisture Volatile Fixed carbon Ash C H N Oa
Oil palm EFB 2.44 73.63 18.67 5.26 48.48 7.14 0.64 43.74
AC-EFB 7.53 15.23 67.66 9.58 68.32 3.12 2.12 26.44
a
Oxygen by difference
[email protected]
Table 2 Surface area and pore structure analysis of oil palm EFB
Sample BET surface Total pore Micropore Average pore Micropore
area (m2/g) volume volume diameter (Å) percentage (%)
(cm3/g) (cm3/g)
Oil Palm EFB 2.005 0.013 5.545 × 10-4 265.400 4.27
AC-EFB 720.0 0.341 0.304 18.890 89.15
Mercury removal,%
100
80
60
40
20
0
0 20 40 60 80 100 120
Time,min
30°C 90°C 140°C 180°C
Fig. 2 Effect of flue gas temperature on mercury removal efficiency of AC-EFB (adsorbent
amount = 500 mg; C0 = 0.07–0.09 mg/L)
Effect of simulated flue gas temperature. Figure 2 shows the effect of on the
Hg0 removal efficiency of the oil palm EFB activated carbon. As the temperature is
increased from 30, 90, 140 and 180 °C, mercury adsorption efficiency is decreased
from 98, 92, 70 and 62 %. It is indicated a physisorption mechanism occurred due
to van de Waals forces between the molecules Hg0 and the oil palm EFB activated
carbon which only favored at low temperature [4]. The result obtained is in similar
to previous study made by De et al. [4], Shen et al. [11], and Zeng et al. [12] which
proposed that physisorption was responsible for mercury removal by unimpreg-
nated activated carbons [4, 11, 12]. Activated carbon is a porous matter which has a
high BET surface area (720 m2/g), so it has extremely excellent adsorption capacity.
In this process, Hg0 molecules easily penetrated through a bed of highly porous
activated carbon and allowed a large mercury capacity to enter. Then, Hg0 mole-
cules diffused into pores of activated carbon and are adsorbed onto the surface of
pores activated carbon by weak forces of intermolecular cohesion, resulting higher
percentage mercury removal [13].
Effect of initial mercury concentration. Figure 3 shows the effect of inlet
mercury concentration on the Hg0 of the oil palm EFB activated carbon with varying
water bath temperatures (30, 40 and 50 °C) in order to get the different value of inlet
mercury concentration. According to previous research, the adsorption capacity of
mercury increased as the inlet mercury concentration increased [4, 14, 15]. In this
study, with an increase in inlet mercury concentration from 0.05 mg/L (30 °C) to
0.07 mg/L (40 °C), mercury removal efficiency at 90 °C increased from 55 to 92 %
and continue rising to 98 % when the water bath temperature is increased to 50 °C.
[email protected]
Fig. 3 Effects of inlet

100
mercury concentration on
Mercury Removal,%
mercury removal efficiency of 80
AC-EFB (adsorbent
amount = 500 mg; 60
TREACTOR = 90 °C)
40
20
0
0 20 40 60 80 100 120
Time,min
Co=0.05 mg/L Co=0.072 mg/L Co=0.192 mg/L
This may be caused by the fact that when concentration of inlet mercury in feed gas is
higher, the atoms of Hg have more chances to interact and be attached to the carbon
surface and finally increased the efficiency of adsorption [4]. According to Ping et al.
[16], mercury molecules become more active when inlet concentration mercury
increase because with high inlet concentration mercury, the gas phase driving force
for mass transfer will be increased and make the reaction rate increases and finally
leading to an attraction between the surface of activated carbon [16].
Conclusion
This work showed that oil palm empty fruit bunch (EFB) can be used as the good
raw material with steam as the activating agent for the production of activated
carbon. From the results, with a high temperature of flue gas indicated lower
efficiency of mercury removal. This is due to the breaking of weak Van der Waals
bond during physisorption process. When the inlet mercury concentration was
increased, the mercury removal efficiency was increased. This may be caused by the
fact that when concentration of inlet mercury in feed gas is higher, the mercury
atoms have more chances to interact and be attached to the carbon surface. As a
conclusion, for the unimpregnated activated carbon, the physical properties of
activated carbon such as surface area, pore volume and pore size play an important
role in mercury adsorption.
Acknowledgment This work was financially supported by the Research Management Institution
(RMI) through Excellence Fund (600-RMI/DANA 5/3/RIF (311/2012)) and UiTM Shah Alam.
The authors also thanks to Prof. Masaki Takaoka for his contribution in this research.
[email protected]
References
1. Li P, Feng XB, Qiu GL, Shang LH, Li ZG (2009) Mercury pollution in Asia: A review of the
contaminated sites. J Hazard Mater 168:591–601
2. Hutchison A, Atwood D, Santilliann-Jiminez QE (2008) The removal of mercury from water
by open chain ligands containing multiple sulfurs. J Hazard Mater 156:458–465
3. Yao Y, Velpari V, Economy J (2014) Design of sulfur treated activated carbon fibers for gas
phase elemental mercury removal. Fuel 116:560–565
4. De M, Azargohar R, Dalai AK, Shewchuk SR (2013) Mercury removal by bio-char based
modified activated carbons. Fuel 103:570–578
5. Tan ZQ, Xiang J, Su S, Zeng HC, Zhou CS, Sun LS, Hu S, Qiu JR (2012) Enhanced capture
of elemental mercury by bamboo-based sorbents. J Hazard Mater 239:160–166
6. Tongpoothorn W, Sriuttha M, Homchan P, Chanthai S, Ruangviriyachai C (2011) Preparation
of activated carbon derived from Jatropha curcas fruit shell by simple thermo-chemical
activation and characterization of their physico-chemical properties. Chem Eng Res Des
89:335–340
7. Muyibi SA, Kamaldin N, Gombak J (2008) Production Of activated carbon from oil palm
empty fruit bunches for removal of zinc. In: Presented at the twelfth international water
technology conference, IWTC12. Alexandria, Egypt
8. Arami-Niya A, Wan Daud WMA, Mjalli FS, Abnisa F, Shafeeyan MS (2012) Production of
microporous palm shell based activated carbon for methane adsorption: Modeling and
optimization using response surface methodology. Chem Eng Res Des 90:776–784
9. Hidayu AR, Mohamad NF, Matali S, Sharifah ASAK (2013) Characterization of activated
carbon prepared from oil palm empty fruit bunch using BET and FT-IR techniques. Proc Eng
68:379–384
10. Hidayu AR, Mohamad NF, Matali S, Sharifah ASAK (2014) Preparation and characterization
of activated carbon made from oil palm empty fruit bunch. Key Eng Mater 594-595:44–48
11. Shen Z, Ma J, Mei Z, Zhang J (2010) Metal chlorides loaded on activated carbon to capture
elemental mercury. J Environ Sci 22:1814–1819
12. Zeng H, Jin F, Guo J (2004) Removal of elemental mercury from coal combustion flue gas by
chlorideimpregnated activated carbon. Fuel 83:143–146
13. US Environmental Protection Agency (2008) In: Adsorption, control of gaseous emissions
(Chapter 4)
14. Takaoka M, Takeda N, Fujiwara T (2002) Control of mercury emissions from a municipal
solid waste incinerator in Japan. J Air Waste Manag Assoc 52:931–940
15. Carey TR, Hargrove OW, Richardson CF, Chang R, Meserole FB (1998) Factors Affecting
Mercury Control in Utility Flue Gas Using Activated Carbon. J Air Waste Manag Assoc
48:1166-1174
16. Ping F, Chao-ping C, Zi jun T (2012) Experimental study on the oxidative absorption of Hg0
by KMnO4 solution. Chem Eng J 198–199:95-102
[email protected]
Production of Glucose, Galactose
and Mannose from the Skins of Durian
and Mangosteen
Zamila Mohd Senu, Maryam Husin, Abd Rashid Li,

Rusnah Samsuddin, Mohd Radzi Ahmad, Nik Roslan Nik Abd Rashid
and Nur Zalikha Mohd Taza
Abstract In this study, production of glucose, galactose and mannose from the
skins of tropical fruits, durian and mangosteen was carried out. Those sugars were
obtained by acid hydrolysis of micro-crystalline cellulose and polysaccharide rich-
fractions of the samples. The micro-crystalline cellulose (MCC) rich-fractions were
purified using nitric acid-acetic extraction while polysaccharide rich-fractions were
obtained using aqueous-acetone precipitation method. The hydrolysates consist of
glucose, galactose, mannose and other sugar residues were analyzed its total sugar/
carbohydrate contents and monosaccharide composition using phenol-sulfuric acid
and pre-column derivatization HPLC techniques, respectively. The results indicated
that yields of glucose that obtained from the MCC of durian skins is 14.23 % of the
dried skin of durian compared to 8.52 % of glucose from the MCC of mangosteen
skins. The yields of glucose, galactose and mannose from the polysaccharides of the
durian and mangosteen skins are very low, which are 0.98, 0.70 and 0.16 % of the
Z.M. Senu (&) M. Husin R. Samsuddin N.R. Nik Abd Rashid

Faculty of Applied Sciences, Universiti Teknologi MARA, 40450 Shah Alam,
Selangor, Malaysia
M. Husin
R. Samsuddin
N.R. Nik Abd Rashid
A.R. Li M.R. Ahmad N.Z. Mohd Taza
Phytochemistry Program, Natural Products Division, Forest Research
Institute Malaysia (FRIM), 52109 Kepong, Selangor, Malaysia
M.R. Ahmad
N.Z. Mohd Taza

[email protected]
124 Z.M. Senu et al.
dried skin of durian, respectively. While 0.23 % glucose, 0.18 % galactose and
0.17 % mannose were determined from the polysaccharide fraction of mangosteen
skins. Results proposed that durian and mangosteen skins are feasible for the
production of bioethanol from plant wastes.
Keywords Fruit skins Acid hydrolysis Hydrolysates Polysaccharides

Micro-crystalline cellulose
Introduction
Waste generation has increased steadily around the world for a number of reasons.
Tonnes of waste are being thrown away and buried in the landfill every day.
Agriculture produces several million tonnes of solid waste annually which are
disposed through various methods. In view of this, investigations have been
focused on the reuse of agri-food waste that contains numerous organic matters into
food ingredients, biofuel and other value-added applications [1]. These organic
wastes are converted from low value matters into valuable products.
Durian and mangosteen are the most popular and important seasonal tropical
fruits throughout Southeast Asia. The popularity of these fruits is continuously
increasing due to its delicious taste, good smell, nutrition and medicinal values.
Nowadays, both durian and mangosteen have been accepted by people around the
world and were commercialized into wide ranges of food products such as jam, ice-
cream, durian crepe, syrup and candy. In 2011, Malaysia produced 363,421 and
21,555 metric tonnes of durian and mangosteen with the plantation areas of 76,399
and 4,268 ha respectively [2]. During its season, waste from both fruits such as
skins and low quality fruits becomes abundant. The waste if not disposed properly
will be harmful to the environment. Disposal or burning the waste products usually
was undertaken by farmers and consumers which subsequently caused air, water
and soil pollutions.
Cellulosic materials are the main components of plant material composed of
lignin, hemicellulose and cellulose. Both hemicellulose and cellulose are comprised
of long chains of sugar molecules and are enclosed rigidly by lignin which contains
no sugar and can be extracted by acid hydrolysis [3–6]. Recently, researchers have
been focused on cellulose hydrolysis from the agricultural waste such as rice husk
[7], sisal pulp [8], orange mesocarp [9], oil palm biomass [3], rice straw, wheat
straw and corn stalks [10]. Polysaccharides cellulose obtained from the cellulosic
biomass can produce cellulosic ethanol (or bioethanol) [8]. The cellulosic ethanol is
produced from the natural yeast fermentation on soluble sugars obtained from
polysaccharide cellulose via acid or enzymatic hydrolysis.
This cellulosic ethanol could be used as one of the alternatives to petroleum-based
fuel. Ethanol that was blended with gasoline (85 % ethanol–15 % gasoline) has
caused the increment in oxygen content, causing better hydrocarbon oxidation and
diminishing greenhouse gasses [11]. Nowadays, with the price of petroleum
[email protected]
Production of Glucose, Galactose and Mannose from the Skins … 125
worldwide almost doubled during the last ten years and the use of petroleum also
contributed towards the accumulation of greenhouse gases in the atmosphere [12],
the introduction of cellulosic ethanol as petroleum-based fuel is an added advantage.
The purpose of this study is to extract micro-crystalline cellulose (MCC) and
polysaccharide rich-fractions from dried skins of durian and mangosteen. Then both
fractions were treated with diluted acid hydrolysis to obtain glucose, galactose and
mannose. The total contents of those sugars were determined using pre-column
derivatization HPLC method combined with phenol-sulfuric acid method. This
study has focused on those sugars because natural yeast can only ferment glucose,
galactose and mannose into bioethanol [13].
Experimental
Mangosteen and durian skins were obtained from the fruit market in Shah Alam,
Selangor. The fresh skins obtained were washed and cut into smaller pieces then
dried at 105 °C for 24 h. The dried skins were finely ground before storing at room
temperature for further analysis. Methanol, acetic acid, nitric acid, hydrochloric acid
and sulfuric acid, used for hydrolysis were purchased from Merck, Malaysia. For
HPLC analysis, glucose, galactose, arabinose, mannose and other monosaccharides
(Sigma-Aldrich) were used as standards.
Isolation of carbohydrates. The fine powdered skins were subjected to repeated
extractions with methanol followed by aqueous extraction using Soxhlet apparatus.
The water extract was collected and then treated with acetone before freeze dried to
obtain polysaccharide rich-fraction. 80 % of acetic acid (20 ml/g biomass) and 65 %
of nitric acid (2 ml/g biomass) were added to the dried skin residue and then heated
for 1 h at 90–100 °C. The reaction mixture was filtered and neutralized with
distilled water. Dried cellulose was hydrolyzed with 2.5 N of hydrochloric acid in
the ratio 1:20 (cellulose:liquor) at 90–100 °C for 1 h. The mixture was allowed to
cool at room temperature, then filtered and washed with distilled water before being
dried at room temperature. The mass of MCC rich-fraction was recorded upon
drying. Concentrated sulfuric acid was used to hydrolyze the carbohydrates prior to
heating at 120 °C for 120 min. The hydrolyzed product was neutralized by using
0.5 M sodium hydroxide until the pH was 7.0. The salt was removed and the sugar
was extracted by using rotary evaporator.
Sample analysis. The analysis of monomeric sugars was carried out on a Series
LC-10A HPLC system (Shimadzu, Kyoto, Japan) equipped with an aminopropyl-
silyl column and refractive index detector (RID-10A). The separation was done
using acetonitrile/water, 7:3 (v/v) as a mobile phase at 80 °C of column tempera-
ture. The sugar extract was hydrolyzed with phenol solution and sulfuric acid prior
to heating and homogenized at 30 °C for 30 min in an incubating shaker. The
mixture was then read under UV absorbance 493 nm using UV-Vis spectropho-
tometer in triplicate for each sample. External calibration standards were performed
for the quantification of sugars.
[email protected]
The MCC and polysaccharide rich-fractions composition of dried durian and

mangosteen skins were shown in Table 1. Durian skins contained approximately
13.93 % of polysaccharide and 25.69 % of MCC. A similar result was reported on
orange mesocarp that was treated with 2.5 N of hydrochloric acid where the yield of
MCC was 25.30 % [9]. Meanwhile, mangosteen skins contained 15.09 % of
polysaccharide and the MCC content in mangosteen skins was 11.59 % less than
those in durian skins.
Typical HPLC chromatograms for monomeric sugar in MCC hydrolysate of
durian skins and mangosteen skins were shown in Fig. 1. Glucose and rhamnose
were two primarily sugars that contained in MCC hyrolysate of the durian and
mangosteen skins. Results for actual percentage of glucose obtained from pre-
column derivatization HPLC and phenol-sulfuric acid method for MCC of both
fruit skins were shown in Table 2. From 25.69 % of the MCC rich-fraction com-
position of durian skins, 98.60 % of glucose and 1.40 % of rhamnose were obtained
Table 1 MCC and Sample (%) Durian skins Mangosteen skins

polysaccharide rich-fractions
composition of dried skins of MCC 25.69 14.10
durian and mangosteen Polysaccharide 13.93 15.09
Fig. 1 HPLC chromatograms for monomeric sugar in MCC rich-fraction hydrolysate of a durian
skins; b mangosteen skins
Table 2 Actual percentage of sugar compound in MCC of durian skins and mangosteen skins
Sample (%) MCC of durian skins MCC of mangosteen skins
Glucose percentagea 98.60 97.72
Rhamnose percentagea 1.40 2.28
Total sugar contentb 61.30 56.69
Actual percentage of glucose 60.44 55.40
a
Result obtained from the pre-column derivatization HPLC method
b
Result obtained from phenol-sulfuric acid method
[email protected]
and the actual percentage of glucose was 60.44 % of the total sugar content
(61.30 %). It was found that the glucose yield in this study was higher than those
reported by Unhasirikul (50.71 g/L) on sulfuric acid hydrolysis of durian skins with
the total sugar was 53.18 g/L [14]. Mangosteen skins contained 14.10 % of MCC
where 97.72 % of it was glucose and the other 2.28 % was rhamnose. The actual
percentage of glucose in the MCC of mangosteen skins was 55.40 % of the total
sugar content. The results showed that glucose was the main composition in the
MCC of both durian and mangosteen skins. As glucose was mainly produced by
acid hydrolysis of cellulose [14, 15], galactose, xylose and mannose were not
detected in both hydrolysates.
For hydrolysis of polysaccharides of durian skins and mangosteen skins, glu-
cose, galactose, rhamnose and mannose were detected by pre-column derivatization
HPLC techniques as shown in Fig. 2. The other major compound detected was
galacturonic acid. From the HPLC analysis, glucose, galactose, rhamnose and
mannose percentage in polysaccharide of durian skins obtained were 23.60, 16.00,
4.20 and 3.90 % respectively where the actual percentage of glucose, galactose and
mannose were 7.28, 4.94 and 1.20 % of the total sugar content respectively.
However, the actual percentage of glucose, galactose and mannose in polysac-
charide of mangosteen skins obtained were slightly less than those in durian skins
which are 1.51, 1.20 and 1.17 % respectively of total sugar content as shown in
Table 3.
Yield of glucose, galactose and mannose in the MCC and polysaccharide
hydrolysates of durian skins and mangosteen skins were tabulated in Table 4. The
highest glucose yield was reported from the MCC of durian skins which was
14.23 % of weight of dried sample. Similar to results for polysaccharide of both
fruits skins, the yield of fermentable sugars were high in durian skins compared to
those in mangosteen skins. The yield of glucose, galactose and mannose in the
polysaccharide hydrolysate of durian skins were 7.28, 4.94 and 1.20 % respectively.
The monomeric sugars yield from both MCC and polysaccharide of mangosteen
skins were less than the yield from durian skins where the glucose from MCC was
8.52 % while yield from the polysaccharide for glucose, galactose and mannose
were 1.51, 1.20 and 1.17 % of weight of dried sample respectively.
Fig. 2 HPLC chromatograms for monomeric sugar in polysaccharide rich-fraction hydrolysate of

a durian skins; b mangosteen skins
[email protected]
Table 3 Actual percentage of sugar compound in polysaccharide of durian skins and mangosteen
skins
Sample (%) Polysaccharide Polysaccharide of
of durian skins mangosteen skins
Galacturonic acid percentagea 52.30 44.70
Glucose percentagea 23.60 21.00
Galactose percentagea 16.00 16.70
Rhamnose percentagea 4.20 1.30
Mannose percentagea 3.90 16.20
Total sugar contentb 30.86 7.21
Actual percentage of glucose 7.28 1.51
Actual percentage of galactose 4.94 1.20
Actual percentage of mannose 1.20 1.17
a
Result obtained from the pre-column derivatization HPLC method
b
Result obtained from phenol-sulfuric acid method
Table 4 Yield of glucose, galactose and mannose in the MCC and polysaccharide hydrolysates of
durian skins and mangosteen skins
Sample (%) Durian skins Mangosteen skins
MCC Polysaccharide MCC Polysaccharide
Glucose 14.23 7.28 8.52 1.51
Galactose – 4.94 – 1.20
Mannose – 1.20 – 1.17
The fermentable monosaccharide sugars were obtained from both MCC and
polysaccharide rich-fractions via diluted acid hydrolysis treatment. When the yield
of monomeric sugars from these two rich-fractions are summed, the yield of fer-
mentable glucose, galactose and mannose from the durian skins were 21.51, 4.94
and 1.20 % respectively while the yield from the mangosteen skins were 10.03,
1.20 and 1.17 % respectively as shown in Fig. 3. From another study [16], the
25
Durian skins
20 Mangosteen skins
Yield (%)
15
10
0
Glucose Galactose Mannose
Fig. 3 Total yield of glucose, galactose and mannose in the MCC and polysaccharide hydrolysate
of durian skins and mangosteen skins
[email protected]
maximum glucose yield from two-stage diluted sulphuric acid-catalyzed hydrolysis

of bark-rich sawmill was 13.60 % by combining the yield from first and second
stage hydrolysis while the yield of galactose and mannose was 10.50 % of the
original dry feedstock. The glucose yield was higher in this study might be
attributed to effective extraction of rich-fraction of MCC by nitric acid-acetic
extraction and polysaccharide rich-fraction by aqueous-acetone precipitation
method. Glucose was the main product obtained from hydrolysis of the durian skins
[14] as well as the mangosteen skins. Those sugars were very important in the
production of bio-ethanol from plant wastes. Besides glucose which was widely
known, galactose and mannose could also be converted into bioethanol with the
presence of microorganism.
Conclusion
Durian skins and mangosteen skins were successfully fractionated into MCC and
polysaccharide rich-fractions by nitric acid-acetic extraction and aqueous-acetone
precipitation method, respectively. Both of MCC and polysaccharide rich-fractions
were effectively hydrolyzed to produce fermentable sugars. It was noted that those
fractions contained fermentable sugars. As an approximation, sugars could be
considered as originating not only from MCC but also from the polysaccharide of
the fruit skins of durian and mangosteen. Acid hydrolysis of MCC and polysac-
charides rich-fraction of durian skins and mangosteen skins produced mainly glu-
cose followed by galactose and mannose. Based on the phenol-sulphuric acid and
pre-column derivatization HPLC techniques, the maximum yield of glucose,
galactose and mannose from the durian skins were 21.51, 4.94 and 1.20 %
respectively while the yield from the mangosteen skins were 10.03, 1.20 and
1.17 % respectively. Durian skins and mangosteen skins showed a high potential to
be converted to monomeric sugars and those sugars could be further utilized for the
production of cellulosic ethanol via natural yeast fermentation.
Acknowledgment This study was financially supported by the Research Excellence Fund—
Research Intensive Faculty (600-RMI/DANA 5/3/RIF (567/2012)) of Universiti Teknologi
MARA. A special thanks to support staff of Natural Product Division of Forest Research Institute
Malaysia for laboratory assistance.
References
1. Makris DP, Boskou G, Andrikopoulos NK (2007) Polyphenolic content and in vitro

antioxidant characteristics of wine industry and other agri-food solid waste extracts. J Food
Compos Anal 20(2):125–132
2. DOA (2011) Fruit crop statistics, 1st edn. Peninsular Malaysia
[email protected]
3. Mohamad MH, Eichhorn SJ, Hassan A, Jawaid M (2013) Isolation and characterization of
microcrystalline cellulose from oil palm biomass residue. Carbohydr Polym 93(2):628–634
4. Das K, Ray D, Bandyopadhyay NR, Sengupta S (2010) Study of the properties of
microcrystalline cellulose particles from different renewable resources by XRD, FTIR,
Nanoindentation, TGA and SEM. J Polym Environ 18(3):355–363
5. Terinte N, Ibbett R, Schuster KC (2011) Overview on native cellulose and microcrystalline
cellulose I structure studied by X-Ray diffraction (WAXD): comparison between measurement
techniques. Lenzinger Berichte 89:118–131
6. Oliveira RL, Silva Barud H, Assunção RMN, Silva Meireles C, Carvalho GO, Filho GR,
Ribeiro SJL (2011) Synthesis and characterization of microcrystalline cellulose produced from
bacterial cellulose. J Therm Anal Calorim 106(3):703–709
7. Johar N, Ahmad I, Dufresne A (2012) Extraction, preparation and characterization of cellulose
fibres and nanocrystals from rice husk. Ind Crops Prod 37(1):93–99
8. Paula MP, Lacerda TM, Zambon MD, Frollini E (2012) Adding value to the Brazilian sisal:
acid hydrolysis of its pulp seeking production of sugars and materials. Cellulose 19(3):975–992
9. Ejikeme PM (2007) Investigation of the physicochemical properties of microcrystalline
cellulose from agricultural wastes I: orange mesocarp. Cellulose 15(1):141–147
10. Nuruddin M, Chowdhury A, Haque SA, Rahman M, Farhad SF, Sarwar Jahan M et al (2011)
Extraction and characterization of cellulose microfibrils from agricultural wastes in an
integrated biorefinery initiative. Cellul Chem Technol 45:347–354
11. Hill J, Nelson E, Tilman D, Polasky S, Tiffany D (2006) Environmental, economic, and
energetic costs and benefits of biodiesel and ethanol biofuels. Proc Natl Acad Sci USA 103
(30):11206–11210
12. Hossain ABMS, Fazliny AR (2010) Creation of alternative energy by bio-ethanol production
from pineapple waste and the usage of its properties for engine. Afr J Microbiol Res 4
(9):813–819
13. Wallheimer B (2010) New yeast can ferment more sugar, make more cellulosic ethanol. http://
www.purdue.edu/newsroom/research/2010/100607MosierYeast.html. Accessed 17 Dec 13
14. Unhasirikul M, Narkrugsa W, Naranong N (2013) Sugar production from durian (Durio
zibethinus Murray) peel by acid hydrolysis. Afr J Biotechnol 12(33):5244–5251
15. Jahan MS, Saeed A, He Z, Ni Y (2010) Jute as raw material for the preparation of
microcrystalline cellulose. Cellulose 18(2):451–459
16. Kim KH, Tucker M, Nguyen Q (2005) Conversion of bark-rich biomass mixture into
fermentable sugar by two-stage dilute acid-catalyzed hydrolysis. Bioresour Technol 96
(11):1249–1255
[email protected]
Reaction Kinetics Study for Microwave
Energy Pretreated Jatropha Curcas
L In-Situ Transesterification
Sintayehu Mekuria Hailegiorgis, Shuhaimi Mahadzir

and Duvvuri Subbarao
Abstract In the present work, in-situ methanolysis of microwave irradiation heat

pre-treated Jatropha curcas l oil in the presence of alkaline benzyltrimethylam-
monium hydroxide (BTMAOH) as a phase transfer catalysis (PTC) was investi-
gated. Combined use of microwave heat pretreatment of Jatropha curcas l seed
particles and alkaline BTMAOH as a PTC drastically reduced the reaction time
from 240 to 30 min. The yield of fatty acid methyl ester (FAME) was also increased
from 49.7 to 93.5 %. Reaction kinetics was investigated to study the extent of
reaction rate. Reaction kinetic study demonstrated that the order of the reaction is a
first order reaction at all reaction temperature under investigation. Activation energy
of the reaction was found to be 21,641 J/mole. Combined use of microwave heat
treatment of Jatropha curcas l seed particles and PTC is promising to enhance the
reaction rate of in-situ methanolysis.

Keywords Jatropha curcas l Phase transfer catalysis Microwave irradiation

Benzyltrimethylammonium hydroxide Fatty acid methyl esters
Introduction
The rising concerns of energy demand and the dependence on fossil fuels along
with its environmental impact have aggravated the search for renewable alternative
fuels. Biodiesel is one of the most promising such options to substitute petroleum
based diesel [1]. Biodiesel production is based on transesterification of vegetable
oils and fats through the addition of lower alcohols such as methanol and a catalyst,
giving glycerol as a co-product. The main feedstocks for the production of biodiesel
were edible oil sources such as rapeseeds, sunflower seeds, soy seeds and palm oil
S.M. Hailegiorgis (&) S. Mahadzir D. Subbarao

Chemical Engineering Department, Universiti Teknologi PETRONAS, 31750 Bandar Seri
Iskandar, Tronoh, Perak, Malaysia

[email protected]
132 S.M. Hailegiorgis et al.
seeds [1, 2]. As the demand for vegetable oils of food has increased at alarming rate,
it is difficult to justify diversion of edible oils to biodiesel production which
adversely affect food as well as biodiesel industry [3]. Hence, the contribution of
non-edible oils sources such as Jatropha curcas l for biodiesel production is
drawing increasing attention.
Vegetable oils processing such as degumming, deacidification, dewaxing, dep-
hosphorization, dehydration, etc contributes to the high cost of biodiesel processing
(70 %). It is desirable to reduce or eliminate those oil processing steps to make
biodiesel cost competitive enough to make it attractive. Harrington and D’Arcy-
Evans [4] developed a method of biodiesel production that cuts out the expensive
intermediary steps with a process known as in-situ transesterification. In-situ
transesterification is a biodiesel production method that utilizes the original agri-
cultural component as the source of triglycerides for direct transesterification that
eliminates the costly hexane extraction process and works with any lipid-bearing
material. Investigations on various oil seeds [5, 6] supported the observations of
Harrington and D’Arcy-Evans. Transesterification is a very slow reaction as veg-
etable oil and methanol are sparingly soluble. In order to enhance the contact
between two phases, several techniques such as co-solvent addition, ultrasonication,
higher temperature and pressure or super critical methanol, super critical reactive
extraction, microwave irradiation and phase transfer catalysis [7] were investigated.
Microwave energy is more recently used to increase the reaction rate of conven-
tional transesterification reaction though microwave heating of oil-alcohol reaction
mixture can cause risk of handing high volatile alcohol under microwave irradiation
particularly at commercial scale of biodiesel processing. However, microwave heat
pretreatment of seed particles prior to in-situ transesterification reaction can make
oil molecules more reactive. In this study, the effect of microwave heat pretreatment
of Jatropha curcas l seed particles prior to in-situ transesterification of the seed with
alkaline PTC was investigated. The reaction kinetics was studied to determine the
rate of reaction and activation energy.
Experimental
Materials and chemicals. Jatropha curcas l seeds produced in Malaysia was

purchased from Agro Innaz Resources, Malaysia, a local Jatropha curcas l seeds
and oil trading company. The seeds were stored in an oven adjusted at 30 °C to
avoid any moisture contamination. Chemicals used were methanol (CH3OH, purity
≥99.7 %), n-hexane (CH3(CH2)4CH3, purity ≥99 %), NaOH (purity ≥99 %), KOH
(purity ≥85 %), iso-propanol (CH3CH2OHCH3, purity 99.8 %), benzyltrimethy-
lammonium hydroxide (C10H17NO) as phase transfer catalysts were procured from
Sigma Aldrich in Malaysia. Reagent grade acetic acid, diethyl ether (C4H18O) and
other pro-analysis chemicals were purchased from R & M chemicals, Malaysia. Gas
chromatographic analytical grade biodiesel standard chemical kits were purchased
from Sigma Aldrich, Malaysia.
[email protected]
Reaction Kinetics Study for Microwave Energy … 133
In-situ transesterification reaction experimental approach. A two necks

round bottom flask reactor equipped with a reflux condenser (to prevent loss of
alcohol), a magnetic stirrer and a thermometer was used. The experiments were
conducted and the results were analysed according to the methods presented in our
previous work elsewhere [7]. For reaction kinetics study, in-situ transesterification
reaction with alkaline methanol using BTMAOH as a PTC was investigated at
optimal conditions in a batch reactor at temperatures of 30, 40, 50, and 60 °C. The
reaction was terminated at the end of specific reaction time (30, 60, 90, 120, 150,
180 and 210 min) and the product oil layer was recovered by washing with water.
The product oil layer was analyzed using gas chromatography (GC) technique as
per ASTM D 65 84-00 testing procedure [8]. Molar concentration of triglycerides,
diglycerides, monoglycerides and FAME were estimated from the chromatograms
using the internal reference standards with reaction time.
In-situ transesterification experiment. The yield of FAME produced using both

microwave irradiation pretreated and untreated seed particles were plotted as a
function of reaction time as shown in Fig. 1a. For a reaction catalyzed by only
NaOH as alkaline catalyst, pretreatment of Jatropha curcas l seed particles with
microwave irradiation has increased FAME yield from 49.7 to 84.3 % w/w while
reducing the reaction time from 240 to 120 min as compared to microwave
untreated Jatropha curcas l seed particles reaction. Observing the positive effect of
microwave pretreatment of Jatropha curcas l seeds, further investigation of the
effect of microwave pretreatment of Jatropha curcas l seed particles on the in-situ
methanolysis reaction rate in the presence of alkaline BTMAOH as a PTC was
investigated and results are presented in the Fig. 1b.
Figure 1b exhibited the yield of FAME increased drastically when in-situ
transesterification reactions were conducted with microwave pretreated seed parti-
cles as compared to microwave untreated seeds. Figure 1b demonstrates during in-
(a) (b)
100 100
Methyl esters yield
Methyl esters yield
90 90
80 80
70
(% w/w)
(%w/w)
60 70
50 60 BTMAOH+NAOH+MWH
40 50
30 40 BTMAOH+NaOH
20 NaOH + MWHT
10 NaOH 30
0 20
0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 210
Reaction time (min) Reaction time (min)
Fig. 1 Effect of microwave irradiation pretreatment of Jatropha curcas l particles on alkaline in-
situ methanolysis. a Without BTMAOH as a PTC and b With BTMAOH as a PTC
[email protected]
situ methanolysis assisted by alkaline BTMAOH as a PTC 93.5 % w/w maximum

FAME yield was achieved in 30 min as compared to 89.8 % FAME yield observed
in 90 min for microwave untreated seeds of the same reaction condition. From this
result, it can be deduced that microwave treatment of seed particles has significant
effect to increase the rate of reaction and reduce the reaction time from 90 min to
about 30 min while increasing the yield of FAME.
Empirical Reaction Kinetics
Jatropha curcas l oil is mainly a mixture of triglycerides of C16 to C18 fatty acids
with an average molecular weight of 878 g/mol. Fatty acid triglycerides can be
further transesterified with lighter alcohols in presence of a catalyst (alkaline/acidic/
enzymatic materials) to produce biodiesel and glycerol. The reaction can be slow as
the alcohols and oils are not very soluble. Soap may also be formed by the
undesirable fatty oils reaction with alkalis. In spite of the heterogeneity due to
limited solubility among the different phases, estimation of an effective empirical
rate constant assuming the system to be a pseudo homogeneous phase can provide
useful information for the reactor design calculations. The expected overall reac-
tions are as shown in Eq. (1)
NaOH; BTMAOH
Triglycerides þ 3 Alcohol !
3 Biodiesel þ Glycerol ð1Þ
One mole of triglycerides requires 3 mol of alcohol to produce 3 mol of biodiesel

and one mole of glycerol. The undesirable saponification reaction which can pro-
duce soap needs to be suppressed to improve the economics. The alkaline transe-
sterification reaction is mildly exothermic and reversible in nature though the
reverse reaction is very slow. Use of excess alcohol can drive the reaction to
completion. Dependence of the reaction kinetics on concentration of reactants and
temperature through activation energy needs to be experimentally measured and
correlated by empirical of the reaction rate Eq. (2).
d[TG]
¼ k ½TGn ð2Þ
dt
where [TG] is triglycerides concentration, ƙ apparent reaction rate constant and t is

reaction time, n order of the reaction.
The apparent rate constant k can depend on concentration of catalyst, ratio of
alcohol to oil, temperature and level of mixing [9]. Molar concentration of tri-
glycerides, diglycerides, monoglycerides and FAME estimated from the chro-
matograms using the internal reference standards with reaction time are shown in
Fig. 2.
[email protected]
0.035
0.03
Concentration
0.025
(mol/ml)
TG
0.02
DG
0.015
MG
0.01
FAME
0.005
0
0 30 60 90 120 150 180 210
Reaction time (min)
Fig. 2 Reaction profiles of TG, DG, MG and FAME and FAEE as a function of reaction time
Table 1 The reaction rate equation for triglycerides conversion at different reaction temperature
Temp, K Rate equation, −r[TG], (mol/ml.min) 1/T R. order, n Rate constant, k
0:8587
303 rTG ¼ 0:04328½TG 0.0033 1 0.04328
313 rTG ¼ 0:04856½TG0:9704 0.0032 1 0.04856
1:1768
323 rTG ¼ 0:05498½TG 0.0031 1 0.05498
333 rTG ¼ 0:06479½TG1:264 0.0030 1 0.06479
It can be seen that diglycerides and monoglycerides are intermediates which are
formed and get converted to biodiesel simultaneously. The overall reaction rate of
triglycerides conversion was estimated by differential analysis [10] to obtain the
order of reaction and rate constant as per Eq. (2) for triglycerides conversion during
in-situ methanolysis at various temperatures and the corresponding reaction rate
equations are shown in Table 1.
The result shows that the order of the reaction is a first order for the range of
operating conditions investigated. With the first order reaction, rate constants were
re-evaluated at each temperature for further kinetics analysis and presented in
Table 1. Reaction rate constant is a temperature dependent term that can be rep-
resented by Arrhenius’ law as shown in Eq. (3);
Ea
k ¼ kA e RT ð3Þ
where Ea is activation energy of the reaction, kA is frequency factor and R is

universal gas constant.
Arrhenius plots are presented in Fig. 3. The activation energy of the reaction was
found to be 21,641 J/mole. A similar investigation reported by Marjanović et al.
[11] demonstrated that the activation energy of base-catalyzed sunflower oil
transesterification reaction was in the range of 8300–35,100 J/mole. However,
Doell et al. [12] reported in their work the activation energy of trans-methylation of
soybean oil as 63,000 J/mol. Present values are relatively less and this could be due
to diffusion effects inherent in in-situ transesterification reactions. The empirical
[email protected]
-3
0.0029
-3.2 0.003 0.0031 0.0032 0.0033 0.0034
-3.4
-3.6
lnk
-3.8
-4
-4.2 y = -2603.x + 4.430
R² = 0.941
-4.4
1/T
Fig. 3 Arrhenius’ plots of lnk versus 1/T
rate equation obtained using experimental results can be rewritten in terms of

reaction temperature and conversion as a function of reaction time as;
rTG ¼ 84½TGo eð8:314T Þ ð1 XTG Þ

21641
ð4Þ
Validation of the Rate Equations
The reactor used in the present work is a batch reactor; the performance report for a
batch reaction can be evaluated based on resident time of the reaction using Eq. (5).
ZXTG
dXTG
t ¼ TG0 ð5Þ
rTG
0
For a first order equation
XTG ¼ 1 ekt ð6Þ
Then, the rate constant, k for the present work is
k ¼ 84 eð8:314T Þ
21641
ð7Þ
The results of triglycerides conversion, XTG obtained using Eq. (6) and substi-
tuting the values of reaction rate constant k of Eq. (7) for in-situ methanolysis as a
function of reaction time, t at different reaction temperatures compare well with
experimental observations as shown in Fig. 4.
[email protected]
Conversion, XTG (%) 100
Conversion, XTG, (%)

(a) 100
80 (b)
80
60
60
40 40
20 20
0 0
0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 210

100
(c) 100
80 (d)
80
60
60
40 40
20 20
0 0
0 30 60 90 120 150 180 210 0 30 60 90 120 150 180 210
Fig. 4 Comparisons of experimentally achieved conversion of triglycerides (filled square) with

batch reactor performance Eq. (6) (filled diamond) at different reaction temperatures. a at 30 °C,
b at 40 °C, c at 50 °C, d at 60 °C
Conclusion
In-situ methanolysis of Jatropha curcas l seed was investigated. To increase the

extractability of oil and enhance the reaction rate, the seed was pretreated with
microwave irradiation. Benzyltrimethylammonium hydroxide (BTMAOH) was
used as a phase transfer catalyst to enhance the rate of reaction. Fast conversion of
triglyceride to fatty acid methyl esters was observed when the reaction was assisted
by BTMAOH as a PTC for a reaction conducted using seeds pretreated with
microwave energy. The reaction kinetics was investigated to study the extent of
reaction rate. The reaction kinetic study indicated activation energy of the reaction
is 21,641 J/mole. Microwave heat treatment of Jatropha curcas l seed and use of
BTMAOH as a PTC is promising to enhance the reaction rate of in-situ
methanolysis.
Acknowledgment The authors would like to thank Universiti Teknologi PETRONAS, Malaysia
for financial support, laboratory facilities and encouragement to carry out this work.
[email protected]
References
1. Leung Dennis YC, Xuan Wu, Leung MKH (2010) A review on biodiesel production using
catalyzed transesterification. Appl Energy 87:1083–1095
2. Lin L, Cunshan Z, Vittayapadung S, Xiangqian S, Mingdong D (2011) Opportunities and
challenges for biodiesel fuel. Appl Energy 88:1020–1031
3. Hailegiorgis SM, Mahadzir M, Subbarao D (2011) Enhanced in-situ etanolysis of Jatropha
curcas l in the presence of phase transfer catalyst. Renew Energy 36:2502–2507
4. Harrington KJ, D’Arcy-Evans C (1985) A comparison of conventional and in-situ methods of
transesterification of seed oil from a series of sunflower cultivars. J Am Oil Chem Soc
62:1009–1013
5. Haas MJ, Scott KM, Marmer WN, Foglia TA (2004) In-situ alkaline transesterification: an
effective method for the production of fatty acid esters from vegetable oils. J Am Oil Chem
Soc 81:83–89
6. Jianli Zeng, Xiaodong W, Zhao Bing, Sun Jingcan, Wang Yuchun (2009) Rapid in-situ
transesterification of sunflower oil. Ind Eng Chem Res 48:850–856
7. Hailegiorgis SM, Mahadzir M, Subbarao D (2013) Parametric study and optimization of in-
situ transesterification of Jatropha curcas l assisted by benzyltrimethylammonium hydroxide
as a phase transfer catalyst via response surface methodology. Biomass Bioenergy 49:63–73
8. American Standard Test Method (ASTM) (2002) Test methods for determination of free and
total glycerin in B-100 biodiesel methyl esters. ASTM Designation D6584-08
9. Hailegiorgis SM (2013) Enhanced in-situ transesterification of Jatropha curcas l in the
presence of alkaline phase transfer catalyst. PhD thesis, Universiti Teknologi PETRONAS,
Malaysia
10. Levenspiel O (1999) Chemical reaction engineering, 3rd edn. Wiley, New York
11. Marjanović AV, Stamenković OS, Todorović ZB, Lazić ML, Veljković VB (2010) Kinetics of
the base-catalyzed sunflower oil ethanolysis. Fuel 89:665–671
12. Doell R, Konar SK, Boocock DGB (2008) Kinetic parameters of a homogeneous
transmethylation of soybean oil. J Am Oil Chem Soc 85:271–276
[email protected]
Induction Time of L-Isoleucine
Crystallization with the Presence
of Electric Field
Nik Salwani Md Azmi, Nornizar Anuar, Noor Fitrah Abu Bakar

and Mohammad Akmalhakim Zakaria
Abstract Nucleation signifies the start of phase transition where nuclei of crystalline
forms in the solution prior to crystal growth. There is an increasing demand in
investigating reliable method to measure the nucleation rate in order to control the
process an it has been reported that electric field has been successfully used to control
the nucleation consequently the crystallization process. Hence, it becomes the
objective of this paper to examine the effect of electric field on the induction time
crystallization of L-isoleucine. The experiment was conducted in a 200 mL jacketed
reactor where constant stirring was provided by retreat curve impeller and the tem-
perature was regulated using two refrigerated baths. The induction time in this work
was found to be lower than system without the application of electric field due to the
hydrophobicity of L-isoleucine molecule and competition to form hydrogen bond
with water when electrolysis occurs. The interfacial tension calculated shows no
significant change with concentration while nucleation rate increases as the super-
saturation increases. The critical radius calculated also proves to be smaller and
decreases when supersaturation increases. The result also suggests the presence of
new crystalline phase for the crystals produced at 48 g/L and supersaturation ratio of
1.08 and 1.14.
Keywords Interfacial tension Nucleation rate Critical radius Polymorphism
N.S. Md Azmi N. Anuar (&) N.F. Abu Bakar M.A. Zakaria

Universiti Teknologi MARA Malaysia, 40450 Shah Alam, Selangor, Malaysia
N.S. Md Azmi
N.F. Abu Bakar
e-mail: drnoorfi[email protected]
M.A. Zakaria

[email protected]
140 N.S. Md Azmi et al.
Introduction
L-Isoleucine is an amino acid with many applications in various sectors, such as

pharmaceutical, food and fine chemical industries. Along with other types of amino
acid, it is a constituent of infusions and special dietary products [1]. L-isoleucine is
classified as an essential amino acid, in which the body cannot produce and has to
be acquired through dietary supplementation. L-isoleucine is known as the most
hydrophobic amino acid and exists as zwitterions in solution [2], making it an
interesting material to be studied. L-isoleucine helps in maintaining optimal level of
blood sugar and energy [3], and improves the nutrition of hepatic disease patient
[4]. It exists in crystalline form meaning that it can be produced through crystal-
lization process.
Crystallization refers to the formation of solid crystal from a homogenous
solution [5, 6]. It is vital in the pharmaceutical industry as it functions as a sepa-
ration process and frequent acts as final step in manufacture of active pharma-
ceutical ingredients (APIs) [7]. Supersaturation is the driving force of crystallization
process. A supersaturated solution is metastable, meaning that crystal does not
necessarily formed in the solution as a certain period of time is needed before the
appearance of the first nucleus can takes place. The metastability of a solution is
inversely proportional to the supersaturation, which means that the higher the
supersaturation, the faster the appearance of a crystal from a clear solution [5],
hence the faster the nucleation process occurs. Crystal nucleation is an important
step in crystallization process as it denotes the start of phase transition where nuclei
of crystalline forms in the solution before continually growing out to form larger
crystal [8]. The demand for developing a reliable method to measure the nucleation
rates is growing as it provides validation to nucleation theories, and control over the
crystal product quality in industrial crystallization process can be attained [9]. It is a
well known fact that a stable crystal grows in the metastable zone region, which is
kinetically inactive [10]. For the nucleation to occur in the metastable region,
external forces needs to be introduced and electric field has been successfully used
to induce the nucleation process as reported by previous researchers [11–13].
Hence, it becomes the main focus of this paper to investigate the effect of electric
field on the crystallization of L-isoleucine through induction time experiment.
Theory
Referring to the basis theory of the classic homogenous nucleation for the super-
saturated solution, the rate of nucleation, J can be expressed in the form of
Arrhenius reaction velocity [14]:
[email protected]
Induction Time of L-Isoleucine Crystallization … 141

DG
J ¼ Aexp ð1Þ
kT
Since the Gibbs free energy, DG is the summation between the volume free energy,
DGV (=4/3πrDGv) and the surface free energy, DGS (= 4πrγ), and taken into account
the Gibbs-Thomson relation, Eq. (1) becomes Eq. (2):
!
16pc3 v2
J ¼ Aexp ð2Þ
3k3 T 3 ðlnSÞ2
It is assumed that the induction time, tind to be inversely proportional to nucleation

rate, J, (tind α 1/J) and by applying logarithm to both side of the Eq. (2), it can be
rewritten as Eq. (3):
1 16pc3 v2
ln ¼ ln A ð3Þ
tind 3k 3 T 3 ðlnSÞ2
where tind is the induction time, γ is the interfacial tension, S is the supersaturation,
k is the Boltzmann constant, T is the temperature, A is the pre-exponential factor,
and v = M/(ρNA) is the molecular volume in which M is the molecular weight, ρ is
the density and NA is the Avogadro’s number with value of 6.023 × 1023 no./mol.
Meanwhile, the number of molecules in a critical radius, N* can be determined
using Eq. (4) which is derived and rearrange from [15, 16]:
N ¼ 4pr 3 =3v ð4Þ
where r* is the critical radius size and can be calculated using Eq. (5) [14]:
r ¼ 2cv=kTlnS ð5Þ
Experimental
Material, L-isoleucine with molecular formula of C6H13O2N and purity of higher

than 99 % was chosen as the study material. It was purchased from Merck and
distilled water was used to make up the solution. Carbon electrodes were used to
transfer the current to the solution. Pre-treatment of electrodes was conducted by
immersing the electrodes in acetone to remove any organic impurities and rinsing
them with distilled water, prior to induction time experiment.
Induction time crystallization, L-isoleucine was dissolved in a jacketed reactor
containing 200 mL distilled water where constant stirring was produced by retreat
curve impeller. The solution was heated to temperature 15 °C higher than the
saturation temperature for 60 min. 5 V electric current was supplied to the solution
[email protected]
via electrodes and the solution was cooled to the saturation temperature and
maintained at that temperature for 60 min. Then the solution was drastically cooled
to the desired temperature and was kept constant until the first crystal formed and
the induction time was recorded.
Characterization of product crystal, The product crystal was characterized
using X-ray powder diffraction (XRPD) with Cu Kα radiation at 40 mA, 40 kV, rate
of 0.01°/s and 2θ between 3° and 40°.
Induction time is defined as the elapse time between the creation of supersaturation
and the formation of solid phase [8, 17, 18]. Ghader et al. [17] stated that induction
time is dependent on temperature, interfacial tension, external factors (i.e impurities
and seed), and supersaturation, although the later is the most influential factors in
determining the induction time [14]. Figure 1 shows the relationship between the
induction times with supersaturation. The result clearly shows that the relationship
for each of the concentrations are made up of two distinguishable straight line with
different slopes based on the system nucleating at low and high supersaturations as
reported in Anuar et al. [19]. From experiment, it was observed that shorter
induction time was achieved when the supersaturation increases. This occurrence is
probably due to higher supersaturation leads to higher coagulation of clusters and
smaller critical size of nuclei hence successfully reduces the induction time [14, 17].
It is also clear that the induction time reduces when the solution concentration
increases. At higher solution concentration, more supply of fresh solution con-
taining particles is available in the solution, leading to the prolonged lifetime of
1/T 3 ln(S)2
0.0E+00 5.0E-06 1.0E-05 1.5E-05 2.0E-05
-4
42g/L
44g/L
-5 46g/L
48g/L
-6 42g/L(Anuar et al.,2009)
-7
-8
High Low
Supersaturation Supersaturation
-9
Fig. 1 The relation between the induction time and supersaturation of L-isoleucine at 5 V electric
field, showing nucleation occurs at high supersaturation and low supersaturation system
[email protected]
nuclei, which then enhances the crystal growth. The data was compared with the
data obtained from Anuar et al. [19] for isothermal crystallization of L-isoleucine
without the application of electric field which was also plotted in Fig. 1. It shows
that the induction time for the solution system with the presence of electric field is
lower compared to the induction time without electric field. This indicates that the
electric field has the capability to increase the onset of crystallization. This is in
agreement with a study by Koizumi et al. [12], where the nucleation rate for crystal
formation indeed increases with the application of electric field. The application of
electric field causes the electrolysis process to occur. Water exists as H3O+ and OH−
ions in solution due to self-ionisation. Since H3O+ and OH− ions have higher
affinity to be discharged compared to the L-isoleucine molecule, H3O+ and OH−
ions are easily reduced and oxidised at the cathode and anode, forming H2 and O2
gas, respectively. L-isoleucine on the other hand, is a hydrophobic molecule due to
the presence of the hydrophobic alkyl group [2] and it tends to aggregates close to
each other in to reduce the contact with water. When H3O+ and OH− ions are
discharged, this creates a competition for L-isoleucine to form hydrogen bond with
water near the vicinity of the electrodes, hence it is easier for L-isoleucine aggre-
gation to nucleate. Once the nucleation process starts, it will grow rapidly to form
crystal.
The nucleation phenomena and induction time data is very useful in determining
the interfacial tension of solid-liquid. The interfacial tension, γ was estimated from
the slope of Eq. (3) (16πγ3v2/3k3) and as depicted in Fig. 1. The slope was determined
from the graph in Fig. 1 and by applying the appropriate value into the slope equation,
the interfacial tension, γ was determined and shown in Fig. 2a. It can be seen in
Fig. 2a that there is no significant change in the interfacial tension of L-isoleucine
across the concentration, however the value of the high supersaturation system
indeed is higher compared to the low supersaturation system. The average interfacial
tension values calculated in this work for low and high supersaturation is 0.61 and
Interfacial surface tension, (J/m2)
Nucleation rate, Jx10 -3 (no./m.s)
(a) (b)
1.6 9.0
48g/L
1.4 8.0
High 46g/L
1.2 7.0
Supersaturation 44g/L
1.0 6.0
5.0 42g/L
0.8
Low 4.0
0.6 Supersaturation 3.0
0.4 2.0
0.2 1.0
0.0 0.0
40 42 44 46 48 50 1.05 1.10 1.15 1.20
Concentration (g/L) Supersaturation, S
Fig. 2 The changes of a interfacial surface tension at high and low supersaturation across the
concentration, and b nucleation rate at high supersaturation, at 5 V electric field
[email protected]
1.44 J/m2 respectively. It is stated in [20] that large value of interfacial tension
indicates that a large time is needed for the nucleation to occur. This seems to be the
opposite of what has been happening in this work since at higher supersaturation, the
induction time proves to be decreasing as depicted in Fig. 1. Meanwhile, the
nucleation rate calculated (Fig. 2b) shows an increasing trend as the supersaturation
increases at high supersaturation system and increases as the concentration increases.
High nucleation rates means that the formation of nucleation in the solution is faster
and these nuclei generally have relatively lower probability to grow into large crystal
compared to the lower number of nuclei formed under similar conditions [15, 21]. At
higher level of supersaturation, aggregation of molecules to build up new cluster
reaching critical size is faster than to lose it as the molecules in the solution
has become energetically favorable once the radius reach or exceed the critical
radius [22].
Figure 3a, b shows the dependency of the critical radius size of nuclei and
number of molecules in a critical radius size on supersaturation. From Fig. 3a, b, the
graph clearly shows that at low supersaturation the critical radius increases when
the concentration increases from 42 to 48 g/L, but at high supersaturation the
critical radius size decreases as the concentration increases. The values for the
critical radius size at 48 g/L for high and low supersaturation obtained in this work
ranges between 10.13 – 13.74Å and 8.23 – 10.80Å respectively while for 42 g/L, it
ranges between 11.38 – 15.43Å and 6.59 – 8.64Å for high and low supersaturation.
The values of critical radius size were compared with [19] and it is proven that the
critical radius is smaller due to the application of electric field. In classical nucle-
ation theory for crystals formation, aggregates of molecules needs to form a nucleus
with size exceeding the critical radius, r* which in turn reduces the total free energy
and hence the nucleus increases in size to form a crystal [14]. Theoretically, the
application of electric field further reduces the free energy of formation as described
(a) (b)
14 60
No. ofmolecules, N* (no./m. s)
42g/L 42g/L
13 44g/L 44g/L
50
Critical Radius, r* (Å)
46g/L 46g/L
12 48g/L
48g/L
High S 40
11 Low S High S
Low S
10 30
9
20
8
10
7
6 0
1.00 1.05 1.10 1.15 1.20 1.00 1.05 1.10 1.15 1.20
Supersaturation, S Supersaturation, S
Fig. 3 Supersaturation and concentration effect on a critical radius size, r*, and b number of
molecules in a critical radius, N*, at high supersaturation (High S) and low supersaturation (Low S)
[email protected]
A
New
peak
B
A A
A A A B New
A A
B peak
A
Intensity
A BA B B A
A A A
S=1.18 A
A B
A B
A A B B
B B
S=1.10
Form B
Form A
8 13 18 23 28 33 38
2 (°)
Fig. 4 Presence of peaks of Form A and Form B as well as the additional new peaks for crystal
produced in 48 g/L solution concentration at 1.08 and 1.14 supersaturation
in Saban et al. [23] making it easier for the nuclei to nucleate. Figure 3a also shows
that at high and low supersaturation, the critical radius size decreases when the
supersaturation increases where the same observation is reported in [15, 21].
Meanwhile, the number of molecules in a critical radius depends on the critical
radius value (Fig. 3b).
Polymorphism is a phenomenon where material is able to form more than one
form of crystal structure [2]. L-isoleucine exists in two form which is Form A and
Form B as reported in [19]. In this work, the crystals grown in high supersaturation
at 1.08 and 1.14 were characterized and compared with the existing data obtained
from [19]. Not all crystal can be collected especially at low supersaturation due to
the difficulties in growing the crystal. Figure 4 shows the XRPD profile for crystal
produced in a 48 g/L solution concentration at different supersaturation. The nec-
essary data is plotted and the peaks are shown in the figure. The data in this work
shows the crystals collected consists of mixtures of Form A, Form B and possibly
new crystalline phase. The results shows new peaks are present at approximately 2θ
of 26.18° at supersaturation of 1.08, and at 26.24°, 30°, 36.64°, 37.12°, 39.36° and
39.88° at supersaturation of 1.14. The presence of the new peak could mean that
new crystalline phase is formed in the solution due to the presence of electric field.
Conclusion
Electric field has been proven to have significant impact in the crystallization of
L-isoleucine. In this study, it can be concluded that electric field is capable to
increase the onset of crystallization, hence the crystallization process can be
achieved faster even at higher solution temperature. Electric field also is capable to
[email protected]
induce the nucleation rate of the solution. However, the XRPD result shows that the
crystal formed in the solution are in a mixture of L-isoleucine polymorphis form of
Form A, Form B and new crystalline phase, and this pose a problem to industrial
crystallization.
Acknowledgment The authors would like to express their gratitude to Universiti Teknologi
MARA and the Ministry of Education, Malaysia for funding of this work. This work has been
financially supported under the grant no. 600-RMI/FRGS 5/3(92/2013) and 600-RMI/ERGS 5/3
(29/2012).
References
1. Eggeling L, Morbach S, Sahm H (1997) The fruits of molecular physiology: engineering the l-
isoleucine biosynthesis pathway in Corynebacterium glutamicum. J Biotechnol 56:167–182
2. Florence AT, Attwood D (2006) Physiochemical principles of pharmacy, 4th edn.
Pharmaceutical Press, London
3. Kohlmeier M (2003) Isoleucine. In: Kohlmeier M (ed) Nutrient metabolism. Academic Press,
London, pp 377–383
4. Xie XL, Wei M, Yunoki T, Kakehashi A, Yamano S, Kato M, Wanibuchi H (2012) Long-term
treatment with L-isoleucine or L-leucine in AIN-93G diet has promoting effects on rat bladder
carcinogenesis. Food Chem Toxicol 50:3934–3940
5. Pessôa Filho PA, Medeiros Hirata GA, Watanabe ÉO, Miranda ÉA (2011) 2.46—precipitation
and crystallization. In: Moo-Young M (ed) Comprehensive biotechnology, 2nd edn. Academic
Press, Burlington, pp 651–663
6. O’Ciardha CT, Frawley PJ, Mitchell NA (2011) Estimation of the nucleation kinetics for the
anti-solvent crystallisation of paracetamol in methanol/water solutions. J Cryst Growth 328
(2011):50–57
7. Chen J, Sarma B, Evans JMB, Myerson AS (2011) Pharmaceutical crystallization. Cryst
Growth Des 11:887–895
8. Kulkarni SA, Kadam SS, Meekes H, Stankiewicz AI, ter Horst JH (2013) Crystal nucleation
kinetics from induction times and metastable zone widths. Cryst Growth Des 13:2435–2440
9. Jiang S, ter Horst JH (2010) Crystal nucleation rates from probability distributions of
induction times. Cryst Growth Des 11:256–261
10. Hammadi Z, Veesler S (2009) New approaches on crystallization under electric fields. Prog
Biophys Mol Biol 101:38–44
11. Koizumi H, Fujiwara K, Uda S (2009) Control of nucleation rate for tetragonal hen-egg white
lysozyme crystals by application of an electric field with variable frequencies. Cryst Growth
Des 2420–2424
12. Koizumi H, Tomita Y, Uda S, Fujiwara K, Nozawa J (2012) Nucleation rate enhancement of
porcine insulin by application of an external AC electric field. J Cryst Growth 352:155–157
13. Azmi NSM, Anuar N, Abu Bakar NF, Shukri UR, Mokhlis MH (2014) Solubility study of
electrocrystallized L-isoleucine in aqueous phase. J Appl Sci 14:1403–1408
14. Mullin JW (2001) Crystallization, 4th edn. Elsevier Butterworth-Heinemann, Oxford
15. Abdel-Aal EA, El-Shazly AN, El-Shahat MF (2013) Crystal nucleation of nano crystallite
strontium malonate without and with additives. Cryst Growth Des 13:4395–4401
16. Boomadevi S, Dhanasekaran R, Ramasamy P (2002) Investigations on nucleation and growth
kinetics of urea crystals from methanol. Cryst Res Technol 37:159–168
17. Ghader S, Manteghian M, Kokabi M, Mamoory RS (2007) Induction time of reaction
crystallization of silver nanoparticles. Chem Eng Technol 30:1129–1133
[email protected]
18. Flaten EM, Seiersten M, Andreassen J-P (2010) Induction time studies of calcium carbonate in
ethylene glycol and water. Chem Eng Res Des 88:1659–1668
19. Anuar N, Daud WRW, Roberts KJ, Kamarudin SK, Tasirin SM (2009) An examination of the
solution chemistry, nucleation kinetics, crystal morphology, and polymorphic behavior of
aqueous phase batch crystallized L-isoleucine at the 250 mL scale size. Cryst Growth Des
9:2853–2862
20. Wantha L, Flood AE (2012) Nucleation kinetics of the γ-polymorph of dl-methionine. Chem
Eng Technol 35:1024–1030
21. El-Shall H, Jeon J-H, Abdel-Aal EA, Khan S, Gower L, Rabinovich Y (2004) A study of
primary nucleation of calcium oxalate monohydrate: I-effect of supersaturation. Cryst Res
Technol 39:214–221
22. McPherson A, Kuznetsov YG (2014) Mechanisms, kinetics, impurities and defects:
consequences in macromolecular crystallization. Acta Crystallographica Section F 70:384–403
23. Saban KV, Thomas J, Varughese PA, Varghese G (2002) Thermodynamics of crystal
nucleation in an external electric field. Cryst Res Technol 37:1189–1199
[email protected]
Gas Permeation Performance of Poly
(lactic acid) Asymmetric Membrane
for O2/N2 Separation
F. Mohamed, H. Hasbullah, W.N.R. Jamian, A.R.A. Rani,

M.F.K. Saman, W.N.H. Salleh and R.R. Ali
Abstract Membrane gas separation has been one of the most favorable techniques
used in industry due to its efficiency and low operating cost. One of the main
problems in using membrane in gas separation is the non-biodegradable nature of
the membrane material. In this study, PLA membrane was fabricated for oxygen/
nitrogen separation system. The dope solution was prepared by using dichloro-
methane as a solvent with different concentration of 15, 17.5, 20 and 25 wt%. The
membrane was cast at a fixed casting speed and solvent evaporation time of
0.0413 ms−1 and 60 s, respectively by using pneumatically-controlled casting
system via dry/wet phase inversion technique. The prepared membrane morphol-
ogy, mechanical strength and gas permeation performance was characterized using
scanning electron microscope (SEM), tensile testing machine and soap bubble
meter, respectively. Based on SEM result, the increase in dope solution concen-
tration was proven to decrease the membrane porosity and increase its compactness
that leads to the improvement of membrane gas permeation performance and
mechanical strength. However, the further increase of concentration led to mem-
brane brittleness with low elasticity causing the membrane become hard to handle
for further testing.
Keywords Asymmetric membrane Biodegradable polymer Gas separation

Dope solution concentration
F. Mohamed H. Hasbullah (&) W.N.R. Jamian W.N.H. Salleh

Advanced Membrane Technology Research Centre (AMTEC), FPREE,
Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia
F. Mohamed H. Hasbullah W.N.R. Jamian A.R.A. Rani M.F.K. Saman W.N.H. Salleh
Faculty of Petroleum & Renewable Energy Engineering (FPREE),
Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia
R.R. Ali
Faculty Chemical Engineering, Universiti Teknologi Malaysia, 81310
Johor Bahru, Johor, Malaysia

[email protected]
150 F. Mohamed et al.
Introduction
Gas separation involves number of techniques to separate gases, either to give

multiple products or to purify a single product. Amine scrubbing technology had
been established over 60 years ago for removal of hydrogen sulphide and CO2 from
gas streams in the oil and chemical industries. Commercially, it is the most well
establish techniques available for CO2 capture. However, this technique is mainly
being practiced in gas streams which are chemically reducing, the opposite of the
oxidizing environment of a flue gas stream [1]. Another method is the cryogenic gas
separation, where separation occurs by cooling and condensation processes and
commercially used for streams that already have high concentrations, typically
more 90 % but not for more dilute streams [1].
Membrane gas separation is another method in industry that becoming more
popular these days. A gas separation membrane operates by allowing one com-
ponent in a gas stream to pass through faster than the other component. In order to
achieve a high product purity and yield (by selectivity) and a high throughput (by
permeability) in specific application with substantial market potential, permeability
and selectivity balance of membrane with respect to specific compositions of the
gases, vapors, or liquids to be separated require an extremely high degree of
optimization [2]. Several membranes with different characteristics may be required
to separate high-purity gas, while solvent assisted membranes are being developed
to combine the best features of membranes and solvent scrubbing. Much devel-
opment is required before membranes could be used on a large scale for capture in
power stations.
Among the current concerns regarding of membrane separation are its biode-
gradability where the convenient polymer membrane gives a major drawback, which
is difficult to degrade or dispose when it reached the limit of usage. A polymer based
on a C–C backbone tends to resist degradation, whereas heteroatom-containing
polymer backbones might possess biodegradability tendency. Therefore, a biode-
gradable membrane is introduced in this study. This research is conducted by using a
commercially available and stable biodegradable polymer that is poly(lactic acid)
(PLA). Biodegradable polymers such as PLA are used in medical application such as
controlled drug releasing agent, plain membranes for guided tissue regeneration and
surgical fixation to avoid a second operation to remove them [3].
In the preparation of biodegradable thin layer asymmetric membrane made from
poly(lactic acid) biopolymer, the selection of solvent, its temperature and concen-
tration, casting support and shear rate are among the important factors that influence
the membrane structure, tensile strength, stability, and gas separation performance.
Thus, for this study, the main objective is to investigate the effect of concentration
of PLA on the morphology, gas permeation, thermal and mechanical properties of
the biodegradable membrane.
[email protected]
Gas Permeation Performance of Poly(Lactic Acid) … 151
Experimental
Materials, Poly(lactic acid) (PLA) was used as the membrane material provided by
Shanghai Guanghe Biotech Co. Ltd. Dichloromethane (DCM) was used as solvent
for poly(lactic acid) polymer and supplied by MERCK Schuchardt and tap water
was used as coagulant medium.
Dope preparation, Prior to the dope solution preparation, the polymer was dried
overnight in a vacuum oven at about 60 °C. Firstly, DCM was poured into a bottle
and stirred. Then, PLA polymer was slowly added to avoid agglomeration and the
mixing was continued for about 24 h to ensure a homogenous solution was pre-
pared. The dope solutions were prepared with different concentration of 15, 17.5,
20, 25 wt% PLA. Then, the solution was degassed using an ultrasonic bath to
remove trapped microbubbles prior to casting process.
Preparations of asymmetric flat sheet membrane, Asymmetric flat sheet
membranes were prepared according to the dry/wet phase inversion process. The
polymer solution was cast on a clean glass plate using a unique pneumatically
controlled membrane casting system. The casting knife consists of a steel blade,
which rested onto two runners, arranged to form a precise gap between the blade
and glass plate. The solvent evaporation time and casting speed for the pneumat-
ically controlled casting system was fixed at 60 s and 0.0413 ms−1 respectively for
producing 13 in. length membranes. After the additional freestanding evaporation
step, the membranes were immersed in the water bath at room temperature (*28 °
C) (wet phase inversion) and remained there for about 24 h. Finally, the produced
membranes were air dried under the atmospheric conditions for another 24 h before
storing them in sealed plastic bags prior to testing.
Scanning Electron Microscopy (SEM), Scanning electron microscope was
used to observe the morphology and structure of prepared membrane. Prior to
positioning on a metal holder for gold coating using sputter coating operated under
vacuum, these samples were cryogenically prepared. The SEM micrographs of both
surface and cross-section of the membranes were taken.
Tensile test, The mechanical property of membranes can be obtained by mea-
suring its tensile strength. Universal Testing Machine (UTM, Lloyd Instruments,
UK) was employed to test the tensile strength of the membrane sample using ASTM
standard for sample preparation. Samples were measured at a rate of 2 mm/min strain.
Gas permeation test, Gas permeation of pure oxygen (O2) and nitrogen (N2) were
measured using soap bubble meter at 28 °C and pressures of 6 7, 8, 9 and 10 bar.
Membrane morphology, The morphology and structure of four different weight

percentage of PLA 15, 17.5, 20, and 25 wt% were observed using scanning electron
microscopy. The PLA asymmetric membrane surface and cross sectional structure
were shown in Figs. 1 and 2 respectively. Form Fig. 1a–c, as the concentration
[email protected]
increased from 15 to 20 wt%, the porosity of the membrane surface decreased. The
decrease in membrane surface porosity might be due to the presence of more PLA
molecules that cause the membrane intermolecular forces to increase resulting the
membrane surface to be packed and well aligned. Besides, Hachisuka (1996)
suggested that asymmetric membranes prepared using a low polymer concentration
tends to have finger voids or very porous structures [4]. However, as the concen-
tration was further increased from 20 to 25 wt%, the membrane surface showed the
appearance of pinholes that may due to the increase in dope solution viscosity that
cause by the presence of too much PLA molecules. The increase in dope solution
viscosity may lead the difficulties for the solvent to leave from the membrane
surface and non-solvent to penetrate from the bottom of the membrane. At first, the
penetration of non-solvent leaves no traces on the membrane surface, however,
when the membrane start to harden, the non-solvent penetration through the
membrane may leaves traces on membrane surface in the form of pinhole.
The membrane cross sectional structure that shown in Fig. 2a–d revealed that
membrane porosity decrease as the increase in PLA concentration. The increase in
PLA concentration may cause the bonding between PLA molecules to increase. The
increase in membrane intermolecular forces may cause the molecule to align
themselves to a greater molecular orientation during membrane casting [5]. The
significant increase of the dope solution viscosity cause by the increase in PLA
concentration may creates the difficulty of non-solvent penetration into the mem-
brane during the wet phase inversion process [6], thus reducing the membrane
Fig. 1 Micrographs of the PLA membrane surface with difference concentration of a 15 %;

b 17.5 %; c 20 %; and d 25 %
[email protected]
Fig. 2 Micrographs of PLA membrane cross sectional structure with difference concentration of
a 15 %; b 17.5 %; c 20 %; and d 25 %
porosity. Another possibility that creates the difficulty of non-solvent penetration is

the agglomeration of the PLA polymer during membrane casting.
Membrane tensile strength, The polymer concentration was varied at 15, 17.5.
20 and 25 wt% to determine the effect of concentration on membrane tensile
strength. Figure 3 shows the relationships between PLA concentration on membrane
tensile strength and young’s modulus. As shown in the Fig. 3, the increase in PLA
concentration from 15 to 17.5 wt% had increase the membrane tensile strength.
These increases may due to the membrane compactness cause by the greater
membrane molecular alignment [7]. As the concentration increase, it is expected to
increase the membrane tensile strength. However, as the PLA concentration further
increase to 20 and 25 wt%, the membrane tensile strength decrease. The decrease in
membrane tensile strength may due to the improper membrane formation resulted by
the polymer agglomeration causing the membrane to be brittle.
Membrane gas permeation performance, Increasing the concentration during
dope preparation is expected to increase the skin layer thickness with tighter sup-
port structure, thus, decreasing the pressure-normalized flux but improving the
overall gas separation performance in term of gas pair selectivity [8]. Figure 4
shows the gas separation performance of 15 wt% PLA asymmetric membrane. PLA
membrane with concentration of 17.5 wt% is expected to perform better in gas
permeation test than 15 wt%. Although the 17.5 wt% PLA membrane has tensile
strength higher compared to membrane with other concentration, the elasticity of
the membrane is lower thus the membrane is highly brittle and make it currently not
[email protected]
Fig. 3 Effect of
concentration on mechanical
properties of PLA membrane
Fig. 4 Gas separation

performance for PLA at
concentration 15 %
readily to be further tested in gas separation process. As mentioned above, the

membrane prepared with 20 and 25 wt% PLA was not properly formed, and it is
proven to possess low tensile strength as well as severely low elasticity. Hence,
these membranes were also difficult to handle and could not withstand the pressure
used in the gas permeation test. Thus, the only sample was able to be tested was the
membrane with concentration of 15 wt%. The permeance for all gas tested are
fluctuated with varied pressure in the range of 6–10 bar. The gas permeance for all
gases increase as the increase in pressure might be due to the increasing in the
[email protected]
driving force for all gasses to pass through the membrane. At pressure 6 bar, the O2
gas permeance in higher than N2 gas may be owed to the O2 gas characteristic of
smaller kinetic diameter than N2 gas as accordance to Graham’s Law that make O2
travel pass through the membrane 3–4 times faster than N2. As the pressure is
further increased, the N2 gas permeance is higher than O2 gas. This happens may be
due to the prominent membrane surface defects that cause the gases to permeate
faster through the membrane. As the pressure increase, the permeance of all gases
increase, however, the O2/N2 gasses selectivity decrease. The decrease in selectivity
is as accordance to the trade of between permeability and selectivity as discussed by
Robeson et al. where the increase in gases permeability will decrease the selectivity
of the membrane [9]. The separation factors were not commercially desirable but
the potential of PLA as a membrane for gas separation cannot be denied.
Conclusion
The development of innovative biopolymers has been underway for a number of

years and continue to be an area of interest for the modern scientists. Biodegradable
material used for membrane gas separation is the new research interest concerning
the issue of disposing the damaged non-biodegradable membranes into the envi-
ronment. Thus, finding an environmental friendly material suitable for membrane
production is crucial such as poly(lactic acid) that is very popular in the food
packaging and medical application industry. The effect of PLA concentration on
morphology, mechanical properties, and gas separation performance has been
investigated in this study. The results shows that the prepared PLA membrane
consists of skin layer on top and porous support layer on the bottom. The increase
in PLA concentration might increase the bonding between membrane molecules
thus reducing the membrane porosity, however the further increase on PLA con-
centration causing the dope solution viscosity to increase thus increasing the
membrane surface defect. Based on the tensile test, the optimum PLA concentration
for the sample to act as a gas separation membrane is at 15 wt% along with the
optimum tensile strength. The tensile strength of the membrane is expected to
increase with the increase in PLA concentration, however higher PLA concentra-
tion have cause the membrane to be more brittle with a reduction of membrane
elasticity and currently not readily to be tested for gas permeation performance. The
separation factor of PLA membrane with concentration of 15 wt% produced was
not commercially desirable; however it shows a huge potential as a membrane
material for gas separation with added benefit of being biodegradable. The mem-
brane is recommended for further testing by coating the membrane using silicon
rubber coated material (PDMS).
Acknowledgment We gratefully acknowledge the financial support from the Universiti

Teknologi Malaysia and Ministry of Education (Grant No: 04H70 and 4L116).
[email protected]
References
1. Gupta M, Coyle I, Thambimuthu K (2003) CO2 capture technologies and opportunities in

Canada: strawman document for CO2 capture and storage (CC & S) technology
roadmap. CANMET Energy Technology Centre, Natural Resources Canada
2. Ulbricht M (2006) Advanced functional polymer membranes. Polymer (Guildf) 47:2217–2262
3. Wei XW, Guo G, Gong CY, Gou ML, Qian ZY (2011) Biodegradable polymers: research and
applications. In: A handbook of applied biopolymer technology: synthesis, degradation and
applications, no. 12, Chengdu, China, pp. 365–387
4. Hachisuka H (1996) New type asymmetric membranes having almost defect free hyper-thin
skin layer and sponge-like porous matrix. J Memb Sci 116:265–272
5. Hasbullah H, Be N, Ibrahim N (2013) Mindel S-1000 based asymmetric membranes for O2/N2
separation: effect of polymer concentration. Chem Eng Trans 32:2035–2040
6. Chung TS, Shieh JJ, Lau WWY, Srinivasan MP, Paul DR (1999) Fabrication of multi-layer
composite hollow fiber membranes for gas separation. J Memb Sci 152:211–225
7. Ghasem N, Al-Marzouqi M, Duidar A (2012) Effect of PVDF concentration on the morphology
and performance of hollow fiber membrane employed as gas–liquid membrane contactor for
CO2 absorption. Sep Purif Technol 98:174–185
8. Pesek SC, Koros WJ (1993) Aqueous quenched asymmetric polysulfone membranes prepared
by dry/wet phase separation. J Memb Sci 81:71–88
9. Robeson LML (1991) Correlation of separation factor versus permeability for polymeric
membranes. J Memb Sci 62:165–185
[email protected]
Particles Mixing in a Fluidized Bed
by Using Digital Image Processing
and Thief Probe
Syaidatul Akma Mohd Zuki, Norazah Abd Rahman, Ihsan Mohd

Yassin, Nurul Asyikin Md Zaki
and Mohamad Shafiq Izzudin Suhaimi
Abstract Particle size plays a major role in segregation phenomena in mixing

process. In pharmaceutical industry, it is essential to have a proper analysis to make
sure that the good product is well mixed and to prevent segregation. In this study,
Digital Image processing has been used as an alternative method to thief probe, to
study the effect of particle size on mixing and segregation process. The experiment
started with execution of image acquisition process of mixing process. The images
taken were further analyzed using Image Processing Tool in MATLAB software.
The experiments were done with two different arrangement of particle in the bed
using air velocity of 0.94 ms−1. The mixture was considered to be well mixed when
the intensity of red of colour histogram is approximately constant. For the layered
position, the good mixing time for set 1, 2, 3 and 4 were achieved at 62, 84, 86 and
74 s respectively. Then, for random position, homogeneous mixture were achieved
at 62, 76, 82 and 72 s for set 1, 2, 3 and 4 respectively. This finding has been
quantified by using Lacey Mixing Index where the value obtained was nearly to 1.0
which showed uniform mixing.
S.A. Mohd Zuki N. Abd Rahman (&) N.A. Md Zaki M.S.I. Suhaimi
N. Abd Rahman
CoRe of Frontier Materials and Industry Applications, Universiti Teknologi MARA,
N.A. Md Zaki
CoRe of Green Technology and Sustainable Development, Universiti Teknologi MARA,
I. Mohd Yassin
Faculty of Electrical Engineering, Universiti Teknologi MARA, 40450 Shah Alam,
I. Mohd Yassin
CoRe of Advanced Computing and Communication, Universiti Teknologi MARA,

[email protected]
158 S.A. Mohd Zuki et al.
Keywords Digital image processing

Spherical polystyrene Fluidization

Lacey mixing index Segregation Intensity
Introduction
Particle mixing is a process of combining two or more particles to form a mixture.

Typically, the particles involved in the process have different characteristic or
properties; namely density, moisture ratio, size, porosity and shape. The difference
in their properties lead to a condition called segregation, where the mixture
potentially not to be mixed properly or unmixed [1]. Other than that, if a well-mixed
mixture continuously undergoes the mixing process, they have a high tendency to
segregate unless the process is stopped when the quality of mixing has been
achieved.
Some applications of mixing require simple combination of particles, while
others demand a high degree of homogeneity. The segregation process causes
problem in some industries such as in pharmaceutical industry, where mixing is
considered as a critical process. This is because mixing process is used in its
formulation where different chemical substances such as active drug and inactive
additives are mixed to produce a final product. For an orally taken drug, segregation
of the mixture will produce poor distribution of the drug which has been incor-
porating into a tablet or a capsule [2]. Therefore, the mixing process has to be
properly monitored to ensure a high degree of homogeneous mixture and prevent
segregation process, thus achieving high product quality. The mixing process has to
be stopped as soon as the mixing efficiency is achieved to optimize cost and time, as
well as to prevent segregation process.
Over the years, the study of mixing is gradually evolving due to its importance in
many industries. Two methods are generally available to identify mixing efficiency;
namely invasive and non-invasive. Invasive method is a traditional manual pro-
cedure that involves continuous monitoring of the mixing process [3]. Weaknesses
of the invasive method are disruption of the mixing process due to regular sampling
intervals, and long mixing times. Because of these weaknesses, non-invasive
methods such as Discrete Element Model (DEM) simulation [4, 5], Magnetic
Resonance Imaging (MRI) [6–8], Positron Emission Particle Tracking (PEPT) [9,
10] and Digital Image Processing (DIP) [11–13] have been used. Recently, DIP has
shown great potential to be studied further because of its advantage of economy,
simplicity and non-invasiveness [14].
In this study, the two methods have been used where thief probe and a web
camera have been used to represent an invasive and non-invasive method respec-
tively. The main objective is to show that simple image processing technique like
Multivariate Analysis RGB images produced from low cost web camera can be
used to assess mixing performance for spherical polystyrene with similar sizes but
different colours.
[email protected]
Particles Mixing in a Fluidized Bed … 159
The material that has been used in the study was two different sizes of free flowing
spherical polystyrene with same shape as well as particle density. Spherical poly-
styrene particles were sieved using mechanical sieve to obtain the particle sizes of
4.05 and 5.50 mm where both were coloured by red artificial dye. Fluidized bed
with column diameter of 20 cm and height of 21 cm made from perspex was used in
this study. Air was flown with velocity of 0.94 ms−1 from the bottom of the column.
The gas distributor with pore size of 63 µm was installed in the bottom of the
column to distribute the air during the fluidization process. 20 grams of white and
red polystyrene sphere were used with weight ratio of 50:50. It was observed that
the most suitable height for the placement of the web camera was at 25 cm from the
bottom.
The polystyrene spheres had been flown upward by two different position. The
red sphere polystyrenes were laid out at the bottom of the bed, followed by white
polystyrene spheres spread over the top. For the layered position, the red poly-
styrene spheres were laid out evenly before the white polystyrene spheres were
placed on top. Then, for the random position, the white polystyrene spheres were
placed at the top of red polystyrene spheres without laying them out evenly.
Figure 1 shows the illustration of the two positions for the placement of the
polystyrene spheres. Four sets of experiment were conducted for each position and
Table 1 shows the details of the arrangement used in the experiment.
10g white sphere

(a) (b)
10g white sphere
10g red 10g red sphere
sphere polystyrene
polystyrene
Fig. 1 Illustration of a layered position b random position
Table 1 Sets of experiment Set of Above (white) Below (red)

experiment (mm) (mm)
1 4.05 4.05
2 4.05 5.50
3 5.50 4.05
4 5.50 5.50
[email protected]
Non-invasive Method
For image acquisition system, a c920 Logitech web camera was used to capture the
mixing process. The mixing image was captured every 2 s and saved by executing a
MATLAB script to capture the images. The resolution of the image was 1920 x
1080 pixels with 24-bit color depth, and red, green and blue (RGB) representation.
The acquired images undergo several image processing steps in MATLAB, namely
cropping and histogram analysis. Maximum histogram points (describing the
highest number of intensity occurrence) for red colors were recorded across time.
The color histogram represents the color changes of the mixing image. The step of
analysis is shown in Fig. 2.
This study focuses on the red, green and blue color histograms. The color
histogram represented the color value at x-axis which has value from 0 to 255 and
the frequency of the color value. The color value indicates the degree of intensity of
a particular color. A high pixel value (near 255) indicates that the color intensity is
high, and vice versa. For this study, the three primary colors are red, green and blue.
But, as the polystyrene sphere was dyed with red, the study focused on the red color
histogram for further analysis. From the red color histogram, the maximum point is
recorded which represented the color value and the maximum intensity of that color
value. As mentioned by Mohd Zuki et al. [15] the mixture is homogeneous when
the distribution of the histogram is constant. It can be applied on maximum point of
the histogram. In order to ease the mixing analysis by using image processing,
maximum point of the histogram has been used to obtain the mixing time.
Invasive Method
For invasive method, thief probe was used to obtain the sample. Due to limitation of
the equipment, the time taken for sampling are from 1, 2 and 3 min at air velocity of
0.94 ms−1. After the time has elapsed, thief probe has was to obtain a small amount
of sample at several locations in the fluidized bed. This method is carried out to
support the finding from non-invasive method. In this study, Lacey mixing index
[16] has been used to quantify the degree of mixing or segregation.
S20 S2
M¼ ð1Þ
S20 S2R
1200
1000
800
600
400
cropped analyzed 200

0
0 50 100 150 200 250 300
Fig. 2 Step of image analysis
[email protected]
where S2 is the actual variance, S20 and S2R , respectively, represent the variances of
the completely segregated and well mixed states, given by
S20 ¼ p q ð2Þ
pq
S2R ¼ ð3Þ
N
where p and q are the volume fractions of white and red particles in a mixture and N
is the equivalent number of particles in the sample.
Figures 3 and 4 shows the graphs of maximum intensity of red value versus time of
mixing process for 4 sets of experiment in layered and random positions respec-
tively. Those four graphs started with high maximum frequency of red value
(colour value of 165–195) in range of 704–807 Hz at mixing time of 0 s. This is
because at this time, particles were not fluidized and the fluidized bed acted as a
fixed bed. The mixture was completely unmixed or completely segregated. The
image at this point showed the white polystyrene spheres at the top of the fluidized
bed. The maximum points of the red, green and blue histogram are approximately
the same. In theory, white color represented the same value of those three primary
500 800
Intensity, Hz
700
600
500
450 400
300
200
0 20 40 60 80 100 120
Intensty, Hz
400 Time, s
350
300
250
50 60 70 80 90 100 110
Time, s
set 1 set 2 set 3 set 4
Fig. 3 Intensity versus time for layered position
[email protected]
800
500
Intensity, Hz
600
400
450 200
0
0 20 40 60 80 100 120
Intensity, Hz
Time,s
400
350
300
250
60 70 80 90 100 110 120
Time, s
set 1 set 2 set 3 set 4
Fig. 4 Intensity versus time for random position
colors. If the images have high value of any one of the three colors, it will project
that color in the image.
As the mixing time was increased, red polystyrene particles started to move to
the top part of the bed and the maximum frequency of red value decreased. The
movement of red particles from the bottom layer is influenced by the wake of air
bubbles which induces mixing process and at the same time, white particles which
were at the top moved down due to drag force which induces segregation. The air
flowed around the spherical particles and made them moved and this movement is
anticipated to be the factor to mix the particles. The images that have been taken
during this process have two colors, which were white and red. Based on the range
of color value of red for white color, the maximum intensity in that range have been
plotted in Figs. 3 and 4. The change in the maximum frequency of red is caused by
the change of white and red distribution in the mixing image. The image that has
more white polystyrene spheres will result in high value of maximum intensity for
red value of 165–195 and vice versa. At one point, the graph begins to show a
constant trend. It indicated that the image have the similar intensity of the red value.
Based on the observation, it was concluded that the mixture had started to be
homogenous because the white and red polystyrene spheres were spread equally.
Therefore, the intensity of the red value in the image taken, were approximately the
same. The constant value had been taken as the time when the particle mixes well
[15]. At this state, a dynamic equilibrium state is achieved where the mixing
process balances the segregation process after several rearrangements of the par-
ticles [16]. The results of the experiments have been summarized in Fig. 5.
[email protected]
Fig. 5 Mixing time 90
80
70
Time, s
60
50
set set set set
1 2 3 4
layerred
position 62 84 86 74
random
position 62 76 82 72
The mixing times for two different position of material placement for 4 sets of
experiment were shown in Fig. 5. The mixing times for this experiment were in the
range of 62–86 s. As shown in Fig. 5, mixing time for set 2 and 3 are slightly higher
than set 1 and 2. The different in particle size for white and red polystyrene in set 2
and 3 give more time for the mixture to be well mixed. As mentioned by Hogg [17],
the difference in particle size give a main problem for mixing process. This factor
will lead to segregation phenomena as the process continues. The small particles
have the capability to fill the void age between the bigger particles in the expanded
fluidized bed. This filling process was assisted by gravitational force as the small
particle move downward during the process. The process is known as percolation.
Layered positioning gave higher mixing times compared to random positioning
for each set of experiment. The mixing and segregation process also can be affected
by the interaction between particle and the drag force. When the sphere polysty-
renes were filled in the fluidized bed as the random position, the particles are
concentrated in the middle of the bed like heap formation. There are no forces
applied during the filling process. The particles flowed downward freely and
formed the heap. At the middle of the heap which is a dense region, the drag force
of the particles are less than the gravitational force because there is no space for air
to flow upward [16]. The air bubbles take some time to flow upward after rear-
ranging the particles. However, the interaction between particles in random position
is less as compared to layered position since no force is applied to the particles. So,
longer time is needed to fluidize the particles for layered position as the force take
its effect during the filling process. The particle is flattened before the mixing
process start. Furthermore, for layered position, the dense regions are not only in the
center of the bed but spread out all over the area. So, the time taken for air bubble to
move upward is longer.
Figure 6 shows the effect of time on the mixing index for the two arrangements
when the superficial gas velocity is 0.94 ms−1. Due to limitation of the equipment,
the times taken to do the sampling were above 1 min. According to results obtained
from image processing, uniform mixture has already been achieved at the selected
time. Hence, mixing index showed value close to 1.
[email protected]
Fig. 6 Lacey mixing index Layered
Lavey mixing index

1
position set
0.995
2
0.99
Layered
0.985 position set
0.98 3
0.975 Random
0 1 2 3 position set
Time, minute 2
Conclusion
The graphs of maximum intensity of red value versus mixing time were plotted for
four sets of experiment for two different position of the placement of polystyrene
spheres; which were layered and random position. It was observed that the graph
starts to become constant at one point, where it indicates the time for well mixed
mixture. For the layered position, the well mixing time for set 1, 2, 3 and 4 are 62,
84, 86 and 74 s respectively. Then, for random position, the well mixing time is 62,
76, 82 and 72 s for set 1, 2, 3 and 4 respectively. This finding has been quantified by
using Lacey Mixing Index where the value obtained was near to 1.0 which showed
uniform mixing at selected time of operation. The experiment also can be further
studied to solve the segregation problem in mixing process. The well mixed mixture
is expected to segregate if the mixing process is continued further for a certain
length of time.
Acknowledgement The authors would like to thank the Research Management Institute of
Universiti Teknologi MARA for Excellence Fund (600-RMI/DANA5/3/RIF(90/2012)) for the
financial support, Faculty of Chemical Engineering and Universiti Teknologi MARA.
References
1. Susana L, Canu P, Santomaso AC (20100) Development and characterization of a new thief

sampling device for cohesive powders. Int J Pharm 416:260–267
2. Allen T (1981) Powder technology theory. Chapman & Hall, London
3. Kaye BH (1997) Powder mixing. Chapman & Hall, London
4. Chaikittisilp W, Taenumtrakul T, Boonsuwan P, Tanthapanichakoon W, Charinpanitkul T
(2006) Analysis of solid particle mixing in inclined fluidized beds using DEM simulation.
Chem Eng J 122:21–29
5. Di Renzo A, Di Maio FP, Girimonte R, Formisani B (2008) DEM simulation of the mixing
equilibrium in fluidized beds of two solids differing in density. Powder Technol 184:214–223
6. Richardson JC, Bowtell RW, Mäder K, Melia CD (2005) Pharmaceutical applications of
magnetic resonance imaging (MRI). Adv Drug Deliv Rev 57:1191–1209
7. Hardy EH, Hoferer J, Kasper G (2007) The mixing state of fine powders measured by
magnetic resonance imaging. Powder Technol 177:12–22
[email protected]
8. Müller CR, Holland DJ, Sederman AJ, Mantle MD, Gladden LF, Davidson JF (2008)
Magnetic resonance imaging of fluidized beds. Powder Technol 183:53–62
9. Rafiee M, Bakalisa S, Fryer PJ, Ingram A (2011) Study of laminar mixing in kenics static
mixer by using positron emission particle tracking (PEPT). Procedia Food Sci 1:678–684
10. Stellema CS, Vlek J, Mudde RF, de Goeij JJM, van den Bleek CM (1998) Development of an
improved positron emission particle tracking system. Nucl Instrum Methods Phys Res Sect A
404:334–348
11. Berthiaux H, Mosorov V, Tomczak L, Gatumel C, Demeyre JF (2006) Principal component
analysis for characterising homogeneity in powder mixing using image processing techniques.
Chem Eng Process 45:397–403
12. Le Coënt AL, Rivoire A, Briançon S, Lieto J (2005) An original image processing technique
for obtaining the mixing time: the box-counting with erosions method. Powder Technol
152:62–71
13. Muerza S, Berthiaux H, Massol-Chaudeur S, Thomas G (2002) A dynamic study of static
mixing using on-line image analysis. Powder Technol 128:195–204
14. Abdul Rahman N, Mohd Zuki SA, Mohd Yassin I (2012) A review of image processing
technique in particle mixing analysis. In: Proceeding of 2012 IEEE 8th international
colloquium on signal processing and its applications
15. Mohd Zuki SA, Abdul Rahman N, Mohd Yassin I (2013) Particle mixing analysis using
digital image processing technique. J Appl Sci 14:1392–1396
16. Feng YQ, Xu BH, Zhang SJ, Yu AB (2004) Discrete particle simulation of gas fluidization of
particle mixures. AIChE J 50:1713–1728
17. Hogg R (2009) Mixing and segregation in powders: evaluation. Mech Processes Powder Part J
27:3–17
[email protected]
Polymeric Composite Membrane
for CO2/CH4 Separation
Munawar Zaman Shahruddin

and Muhammad Irfan Safwan Jasni
Abstract Gas separation using polymeric membrane is associated with numerous

benefits compared to other separation technologies. This research was conducted to
study the separation performance of polymeric composite membranes and to
characterize the membranes in terms of mechanical strength and thermal stability.
For the purpose of this study, several types of membranes were fabricated by using
N-methyl pyrrolidone (79–82 wt%), polysulfone (18 wt%), bentonite clay (0–3 wt
%) and agarwood waste powder (0–3 wt%). Membranes were fabricated by solu-
tion blending and phase inversion methods. The separation performance of the
prepared membrane was identified through gas permeation experiment. In addition,
thermal stability of the membranes was analyzed by thermogravimetric analysis
(TGA). Furthermore, mechanical strength of the prepared membranes was deter-
mined by conducting tensile test. Based on result, the addition of bentonite clay into
polysulfone solution enhance the membrane performance in term of gas perme-
ability, strength and thermal stability. In the case of agarwood however, systematic
investigations conducted shows that membrane with 3 wt% agarwood waste
powder has a lower thermal stability and selectivity. Meanwhile, the performance of
the agarwood composite membrane such as mechanical properties and permeability
is better compared to Psf membrane.

Keywords Polymer composite membrane Polysulfone Agarwood Bentonite
Gas permeation
M.Z. Shahruddin (&) M.I.S. Jasni

Selangor, Malaysia
M.I.S. Jasni

[email protected]
168 M.Z. Shahruddin and M.I.S. Jasni
Introduction
Reducing carbon dioxide (CO2) emissions to the atmosphere has become a primary
concern worldwide in the aspect of environmental issue. CO2 emissions come from
a various sources and the source that is known to contribute the most shares in the
total emissions of CO2 is the power and industrial sector. The CO2 emissions to the
atmosphere can be reduced through capture of CO2 from the chemical and power
plant flue gas [1]. Besides that, natural gas originates from underground reservoir
and it composed mainly methane (CH4) as well as light and heavier hydrocarbons
and contaminating compounds of H2S, N2, He, CO2, Hg and etc. [2]. In natural gas,
the most common contaminant is CO2. The natural gas is said to have pipeline
quality standard as a consumer fuel if the natural gas has high calorific value and
low impurities. These requirements can be achieved by eliminating the impurities
from the natural gas. Another reason for removing the impurities from natural gas is
to prevent equipment and pipeline corrosion problem [2]. The captured CO2 then
can be directed into the gas fields and aquifers located in the subsurface or can also
be utilized as raw material to produce industrially important chemical [1, 3]. Based
on the issues highlighted above, the need for technology to capture CO2 is critical.
There are various technologies that was developed to capture carbon dioxide. These
technologies include absorption, adsorption and cryogenic distillation [4]. There are
several disadvantages associated with these technologies. As for cryogenic distil-
lation, the process demand huge amount of energy since condensation of gases is
involved. Besides that, the application of this technology to streams with low CO2
concentration can greatly decrease the overall plant efficiency. The limitation of
adsorption process is also associated with economical issue since large amount of
energy is required to regenerate the absorbent once equilibrium is achieved. The
adsorbent is regenerated by either thermal energy under thermal swing method or
by mechanical energy in the case of pressure swing method [2].
Currently, the most widely used carbon dioxide capture technology is absorption
column by means of an organic liquid. This technology removes the CO2 by
scrubbing the flue gas in several reactors in which the CO2 is bound by amines.
There are several disadvantages associated with this technology. Firstly, this
technology is not cost effective since it consumed large amount of energy. Also, this
technology causes pollution due to amine loss to the environment [1]. The
implementation of membrane to serve as a separation medium to selectively remove
CO2 from gas mixture could resolve the economic and the environment issue
imposed by the previously highlighted technologies.
Although separation technology using membrane is associated with numerous
benefits compared to other separation technologies, designing a membrane is always
a complicated and challenging process [5]. The membrane must be designed so that it
has long term stability in the specified process environment while retaining excellent
separation performance over its service period. A membrane is said to have long term
stability if it has high thermal and chemical stability. On the other hand, separation
performance of membrane is proportional to its selectivity and permeability [6]. At
[email protected]
Polymeric Composite Membrane for CO2/CH4 Separation 169
present, the approach taken to produce gas separation membrane with the
aforementioned properties is dispersing fillers in polymer matrix. Membrane fabri-
cated by this approach is known as polymer composite membrane [7].
Experimental
Materials, The polymer specified in this study is polysulfone (Psf) with average
molecular weight of 35,000 Da. In this study, N-methyl Pyrrolidone (NMP) with
99.5 % purity was used as a solvent. These two chemicals were supplied by Sigma-
Aldrich Co, USA. The Psf supplied was in pallet form. The additives involved in
casting the membranes include bentonite clay and agarwood waste. The bentonite
clay was also supplied by Sigma-Aldrich Co, USA. The bentonite clay supplied
was in powder form. Meanwhile, the agarwood waste used was supplied by
Environmental Lab of Faculty of Chemical Engineering, Universiti Teknonologi
MARA. Distilled water was used as non-solvent.
Membrane preparation, In this study, 3 types of membrane were fabricated by
using (1) 18 % polysulfone, (2) 18 % polysulfone with 3 % bentonite, and (3) 18 %
polysulfone with 3 % agarwood waste powder. All these membranes was prepared
by using NMP as the solvent. Polysulfone solution was prepared by dissolving
18 wt% of polysulfone in NMP under continuous stirring for 24 h at 30 °C until the
polymer dissolved. Next, Psf film of approximately 0.1 mm thick was casted using a
casting knife on a clean, dry, level glass plate and after 1 min evaporating time,
membrane was immersed in a distilled water for 24 h. After that, the membrane was
dried in open air for about 6 h to remove the remained solvent and distilled water.
Finally the dried membrane was cut into circular shape with 5.5 cm diameter. As for
the composite membrane, the preparation was started by addition of bentonite clay
and agarwood separately to the polymer solution. First additives material was
dispersed in NMP under continuous stirring for about 3. Then polymer was grad-
ually added up to 18 wt% concentration and the stirring was continued for 24 h to
achieve a homogeneous solution. After that, membranes are cast in the same way as
for Psf membranes [8].
Gas permeation experiment, Separation performance of membrane will be
tested in a gas permeation rig. Gas Permeation of pure gases was carried out by a
constant pressure method at ambient temperature and pressure of 4 × 105 Pa [8].
This study will utilized flat type module. A membrane (diameter of 5.5 cm) will be
placed between the feed gas chamber and the permeate chamber, with porous metal
disk as support and rubber O-rings as the seal. Pressure and flowrate will be
recorded and controlled by flow controller and flow indicator respectively. The
Permeation rate will be measured by a bubble flow meter [9]. Each test was
repeated 3 times and the average value presented. The gas permeability, P(mol.m/
(m2.s.Pa)) was calculated by using the equation below;
[email protected]
L
P¼ Q ð1Þ
Aðp2 p1 Þ
where L is film thickness (m), A is membrane area available for transport (m2), p2 is
feed absolute pressure (Pa), p1 is downstream absolute pressure (Pa) and Q is
permeate volumetric flow rate (mol/s). All experiments were conducted at atmo-
spheric downstream pressure (permeate). In gas separation with membrane, selec-
tivity is defined as the ratio of the individual gas permeabilities. Based on single gas
permeabilities of species “A” and “B”, ideal selectivity, aA=B can be expressed as;
PA
aA=B ¼ ð2Þ
PB
Gas permeation experiment, For the purpose of investigating the thermal

degradation, the prepared membranes was analyzed by thermogravimetric analysis
instrument. The sample was heated from 25 to 1000 °C at rate of 10 °C/min under
nitrogen atmosphere, with a nitrogen flow rate of 20 ml/min [10].
Mechanical strength evaluation, Mechanical strength of the membranes is
evaluated based on tensile strength and elongation at break. These two mechanical
strength parameters were determined by subjecting the membranes to a tensile test
to failure in a tensile strength measuring machine with a crosshead speed of
10 mm min−1 at room temperature [11].
Gas permeation result, The effect of dispersing bentonite and agarwood waste
powder in the polymer matrix on CH4 and CO2 permeation of the composite
membrane was investigated by performing the gas permeation experiment at
ambient temperature and pressure of 4 × 105 Pa. The result of the gas permeation
experiment for both Psf and composite membrane are shown in Fig. 1 and Table 1.
By referring to Table 1, it can be observed that the permeability of CH4 for
composite membrane with 3 wt% bentonite was greater compared to Psf membrane
which correspond to an increase of 63.9 %. Besides that, 3 wt% loadings of ben-
tonite to Psf matrix also improved the performance of the membrane in term of CO2
permeability by 56.8 %. The permeability of gases through the polymer is related to
its main properties such gas solubility, gas molecular size and also the mobility of
the polymer chain. Glassy polymer such Psf has a rigid structure. In addition, the
chain in the polymer is mobile. Based on these two characteristic, the diffusivity
ability of the gases in the polymer will cause the gases to permeate in the polymer.
Filler has the ability to restrict the chain packing of glassy polymer. This causes the
free volume in the polymer to increase. Besides that, the total free volume in the
polymer will also increase due to voids that form at the polymer-filler interface or
[email protected]
Fig. 1 The comparison of 90.00
Permeability ( ×10 -14 mol.m/m2.s.Pa))

gas permeability for pure Psf CH4
membrane and composite 80.00
CO2
membrane 70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
Psf 3 wt% bentonite 3 wt% agarwood
Table 1 Permeabilities and Membrane type Permeability Selectivity

ideal selectivity of pure Psf (×10−14 mol.m/
and composite membrane m2.s.Pa)
CH4 CO2 CO2/CH4
Psf 10.73 8.87 0.83
3 wt% Bentonite 29.74 20.54 0.69
3 wt% Agarwood waste 78.56 52.37 0.67
powder
between filler in filler aggregates. A higher diffusion and solubility coefficient is

achieve when the total free volume increases and this in turn result in a higher value
of gas permeability in composite membrane compared to pure polymer [8].
On the other hand, an increase in CH4 and CO2 permeability was recorded in the
case of composite membrane with 3 wt% agarwood waste powder which correspond
to an increase of 86.3 % in CH4 permeability and 83 % in CO2 Permeability. In
addition, the separation performance of the Psf membrane and the composite
membrane was also evaluated based on selectivity of CH4 and CO2. The result of
CH4/CO2 selectivity of Psf and composite membrane are illustrated in Table 1. The
CH4/CO2 selectivity of Psf membrane is 0.83. This value is much lower with 3 wt%
bentonite. This is probably due to the agglomeration of the fillers in the polymer
matrix causing voids to form at the interface between polymer and filler particles [8].
On the other hand, the CH4/CO2 selectivity achieved by composite membrane
with 3 wt% agarwood is also less than that of Psf membrane. The filler dispersed in
polymer matrix can either be porous or nonporous. The effect of these two fillers on
composite membrane is not identical and can be associated to their pore size and
structure. Agarwood is a porous filler. Based on Hg-porosimetry results for pure
wood powder from other studies, the pore radii range from 10 to 2511 nm and this
confirm that wood powder is a very porous material [12]. The separation by
composite membrane with porous filler occur through molecular sieving
[email protected]
mechanism. In this mechanism, gas molecule is separated by their size and shape.
As a result, composite membrane with porous filler usually have high permeability
and selectivity which is beyond the Robeson upper bound because porous filler has
concise apertures [13].
Thus, the presence of highly selectively porous filler in polymer matrix will result
in a composite membrane with a better permselectivity than that of the pure polymer
membrane since the porous filler in the composite membrane selectively allow the
desire component to pass through its pores. However, the result of this study is not
consistent with the aforementioned theory since only the permeability of the com-
posite membrane with 3 wt% agarwood is higher than that of Psf membrane instead
of both permeability and selectivity. The possible explanation for this finding is the
formation of non-selective voids that result from lack of adhesion at the agarwood-
polymer interface. The presence of these voids increase the permeability and cause
the apparent selectivity of the composite membrane to reduce. Therefore the agar-
wood become ineffective since gas molecule pass through the less resistance inter-
facial voids instead of passing through the pore of the agarwood [13].
Thermal stability test, In order to identify the effect of bentonite and agarwood
waste powder on thermal stability of Psf composite membrane, TGA was carried
out. The thermal stabilities as a function of weight (%) were evaluated via TGA
with heating rate of 10 °C/min and the result are shown in Fig. 2. One decompo-
sition step can be observed for Psf membrane and 3 wt% bentonite composite
membrane. Meanwhile, the decomposition of 3 wt% agarwood composite mem-
brane occurred in 3 steps. The first step is associated with the dehydration of
moister that was trapped inside the agarwood. The second step illustrates the
decomposition of hemicellulose and lignin that exist in the agarwood waste pow-
ders that are not in intimate contact with the polymer matrix [12]. This is because,
agarwood waste powder is a lignocellulosic material with a thermal stability lower
than that of pure polymers [14]. The third step on the other hand represents
decomposition of the composite [12].
The present of bentonite in the polymer matrix result in a membrane with better
performance in term of thermal stability since the total mass loss of the composite
membrane at 1000 °C was less than the Psf membrane. The Psf membrane
decomposed to a of mass of 30.8 % from 100 % at 1000 °C. In other word, Psf
TGA curve for Pure Psf membrane and composite

membrane
100
Mass loss, %
50
0
0 200 400 600 800 1000
Temperature, °C
Fig. 2 TGA curve of Psf membrane and composite membrane
[email protected]
membrane loss 69.2 % of its total mass at 1000 °C. Meanwhile when 3 wt% of
bentonite was added to Psf matrix, the resulted membrane loss only 54 % of its total
mass at 1000 °C. This result was observed probably because the presents of ben-
tonite in the polymer matrix act as barrier that obstruct the diffusion of oxygen
molecule into the composite membrane. Other than that, bentonite dispersed in the
Psf matrix result in a membrane with tortuous in structure which cause volatili-
zation to hold up [15].
On the other hand, a decrease in thermal stability is observed for composite
membrane with 3 wt% of agarwood waste powder since its mass loss at 1000 °C
was 7.1 % higher compare to the Psf membrane. This may result from the
decomposition of hemicellulose and lignin that exist in the agarwood waste pow-
ders that are not in intimate contact with the polymer matrix that occur before the
rest of the composite decompose [12]. When this occurs, some free radicals are
created which in turn accelerate the failure of the polymer at higher temperature
[14]. Agarwood waste powders are not in intimate contact with the polymer matrix
due to poor compatibility of the agarwood with the polymer. This is because
polymer are hydrophobic, while agarwood is hydrophilic. As a result, the polymer
and agarwood will not adhere to each other effectively [16].
Mechanical properties, The tensile strength and elongation at break for Psf
membrane and composite membrane was compared to determine the effect of
adding filler additive to Psf membrane on the mechanical strength of the Psf
membrane. The measurements done to determine the mechanical strength param-
eters show that the Psf membrane had a tensile strength of 7 MPa, and elongation at
break of 10.1 %. Generally, the addition of bentonite to pure Psf membrane is
expected to result in membrane with a lower elongation at break since the mem-
brane will become more brittle [17]. Ironically, the addition of bentonite had result a
better membrane in term of elongations at break which corresponded to an increase
of 78 % in a membrane with 3 wt% of bentonite as shown in Fig. 3.
12.0 40
11.0
35
10.0
Tensile strength, MPa
Elongation at break,%
30
8.0 7.7 25
7.0
6.0 20
18.00
15.16 15
4.0
10.10 10
Tensile strength-Bentonite, Mpa
2.0 Tensile strength-Agarwood, Mpa
Elongation at break-Bentonite, % 5
Elongation at break-Agarwood, %
0.0 0
0.5 1.5 2.5 3.5
wt%
Fig. 3 Effect of bentonite on tensile strength and elongation at break for Psf membrane and
composite membrane
[email protected]
Such result was obtained because there is a possibility that the clay platelets
rearrange in the direction of the deformation enable more deformations [18].
Besides that, the membrane with 3 wt% of bentonite can take a higher amount of
tensile stress before it break compared to Psf membrane, correspond to a value of
7.7 MPa in a membrane with 3 wt% of bentonite. The increase in tensile strength is
associated with the increase in the surface area of interaction between clay and
polymer matrix due to the insertion of the polymer chains between the silicate
layers [18]. The addition of agarwood waste powder also enhances the mechanical
properties of Psf membrane as shows in Fig. 3. The tensile strength of the Psf
membrane increased to 4 MPa with the presence of 3 wt% agarwood waste powder
in the polymer matrix. The agarwood waste powder may be regarded as acting
efficiently if it carries a relatively high proportion of the externally applied load.
This can result in higher strength. On the other hand, elongation at break of
membrane with 3 wt% of agarwood waste powder is 15 %. This shows that the
presences of agarwood waste powder increase the ability of the membrane to
deform by enhancing the mobility of the polymer chains. As a result, it is easier for
the segment of the membrane to slip past each other [12].
Conclusion
This research reveals that the addition of 3 wt% bentonite to polymer solution will
enhance the performance of membrane for gas separation in terms of gas perme-
ability, mechanical properties and thermal stability. However the 3 wt% bentonite
composite membrane achieved a lower CO2/CH4 Selectivity compared to Psf
membrane. Besides that, this study shows that a composite membrane with 3 wt%
agarwood waste powder has a lower thermal stability compared to Psf membrane.
Other than that, a higher gas permeability compared to pure Psf membrane was
observed for 3 wt% agarwood composite membrane although CO2/CH4 Selectivity
of 3 wt% agarwood composite membrane is lower than that of Psf membrane.
Nevertheless, the addition of 3 wt% agarwood waste powder improved the
mechanical properties of Psf membrane since a greater value of tensile stress and
elongation at break was recorded compared to pure Psf membrane. Based on the
performance and the properties of the membrane, Psf-bentonite composite mem-
brane is preferable for the separation of CH4/CO2.
Acknowledgment I take this opportunity to record my sincere thanks to Research Management

Institute (RMI), Universti Teknologi Mara for the cooperation, support and help throughout the
completion of this research via Research Intensive Faculty 600-RMI/DANA 5/3/RIF (310/2012).
Besides that, I would like to acknowledge the cooperation and sponsorship of Universiti Sains
Malaysia and Universiti Teknologi Petronas.
[email protected]
References
1. Nguyen QT, Sublet J, Langevin D, Chappey C, Marais S, Valleton JM, Poncin‐Epaillard F

(2010) CO2 Permeation with pebax®‐based membranes for global warming reduction. Membr
Gas Sep 9:255–277
2. Shimekit B, Mukhtar H (2012) Natural gas purification technologies—major advances for
CO2 separation and future directions. InTech Open Access Publisher
3. Aresta M, Tommasi I (1997) Carbon dioxide utilisation in the chemical industry. Energy
Convers Manage 38:S373–S378 (Supplement(0))
4. Moulijn JA, Makkee M, van Diepen AE (2013) Chemical process technology. Wiley,
Hoboken
5. Liu Ye, Wang Rong, Chung Tai-Shung (2001) Chemical cross-linking modification of
polyimide membranes for gas separation. J Membr Sci 189(2):231–239
6. Mulder M (1996) Basic principles of membrane technology, 2nd edn. Kluwer Academic Pub,
Dordrecht
7. Cong Hailin, Radosz Maciej, Towler Brian Francis, Shen Youqing (2007) Polymer—
inorganic nanocomposite membranes for gas separation. Sep Purif Technol 55(3):281–291
8. Momeni SM, Pakizeh M (2013) Preparation, characterization and gas permeation study of PSf/
MgO nanocomposite membrane. Braz J Chem Eng 30(3):589–597
9. Deng L, Kim TJ, Sandru M, Hägg MB (2009) PVA/PVAm blend FSC membrane for natural
gas sweetening. In: Paper presented at the proceedings of the 1st annual gas processing
symposium. Elsevier, Amsterdam
10. Zulhairun AK, Ismail AF, Matsuura T, Abdullah MS, Mustafa A (2013) Asymmetric mixed
matrix membrane incorporating organically modified clay particle for gas separation. Chem
Eng J 241:495–503
11. Anadão P, Priscila Sato, Fumie L, Wiebeck Hélio, Valenzuela-Díaz F Rolando (2010)
Montmorillonite as a component of polysulfone nanocomposite membranes. Appl Clay Sci 48
(1–2):127–132
12. Salemane MG, Luyt AS (2006) Thermal and mechanical properties of polypropylene—wood
powder composites. J Appl Polym Sci 100(5):4173–4180
13. Aroon MA, Ismail AF, Matsuura T, Montazer-Rahmati MM (2010) Performance studies of
mixed matrix membranes for gas separation: a review. Sep Purif Technol 75(3):229
14. Nemati M, Khademieslam H, Talaiepour M, Bazyar B, Khakifirooz A (2013) Studying the
effect of nanoclay on thermal properties of wood flour/recycled polystyrene composite. J Basic
Appl Sci Res 3(5):789–794
15. Lu H, Hu Y, Li M, Chen Z, Fan W (2006) Structure characteristics and thermal properties of
silane-grafted-polyethylene/clay nanocomposite prepared by reactive extrusion. Compos Sci
Technol 66(15):3035–3039
16. Özmen N, Çetin NS, Mengeloğlu F, Birinci E, Karakuş K (2012) Effect of wood acetylation
with vinyl acetate and acetic anhydride on the properties of wood-plastic composites.
BioResources 8(1):753–767
17. Nigam V, Setua DK, Mathur GN, Kar KK (2004) Epoxy-montmorillonite clay
nanocomposites: synthesis and characterization. J Appl Polym Sci 93(5):2201–2210
18. Anadao P, Montes RR, Larocca NM, Pessan LA (2013) Influence of the clay content and the
polysulfone molar mass on nanocomposite membrane properties. Appl Surf Sci 275:110–120
[email protected]
Comparative Study Between Isolated
Xanthomonas Campestris from Rotten
Cabbage and Xanthamonas Campestris
Culture by YDC Media as Substrate
Mohibah Musa, Ku Halim Ku Hamid and Miradatul Najwa Muhd

Rodhi
Abstract Xanthomonas campestris is a gram-negative bacterium which causes

black rot disease in varieties of brassicas which include cabbage, cauliflower,
broccoli, turnip and radish. The cabbage samples were collected locally from
Cameron Highlands, Pahang Darul Makmur. The bacteria were locally isolated and
cultured to produce pure species of the bacteria which later several analysis and a
series of fermentation was done in order to obtain Xanthan gum. Fermentation
procedure conducted through shake flask fermentation. Five different culture and
media combination were fermented to compare the final results of each substrate.
On specific time intervals e.g.: every 2 h sampling, several analysis were conducted
to determine bacteria cell mass and glucose concentration of the fermentation
samples. After the completion of the fermentation, Xanthan gums were recovered
through precipitation by isopropanol.
Keywords Fermentation Grinded rotten cabbage Xanthomonas campestris

Xanthan gum YDC
Introduction
In East Africa, the black rot disease attacking cabbages is still a nightmare to the
farmers. The plant pathogen has increasingly become harder to manage. On main
reason is due to the cultivation of barely prone to disease seeds to the plant path-
ogen [1]. Despite of the damage that the bacteria are capable of, it also is an answer
M. Musa (&) K.H. Ku Hamid M.N. Muhd Rodhi

Faculty of Chemical Engineering, Universiti Teknologi MARA, Selangor, Malaysia
K.H. Ku Hamid
M.N. Muhd Rodhi

[email protected]
178 M. Musa et al.
to the useful xanthan gum that is widely used. Usually, the bacteria are cultured in a
batch fermenter under respective conditions. The xanthan gum is recovered by
precipitation and it is produced by the fermentation of Xanthomonas Campestris,
the Gram negative plant-pathogenic bacterium. [2].
The fermentation of microbes can be done either in shake flask or in bioreactor.
In addition, these complex exopolysaccharides are containing D-glucose, D-man-
nosyl, and D-glucuronyl acid residues [3]. This is the answer for it that remarkably
found to be suitable for many temperature, pH and ionic strength [4]. These
properties are causing the gum to be more stabilized. Moreover, xanthan gum’s
ordered confirmation is very good for the stability. This stability also can be
enhanced with presence of salt to optimize the xanthan gum functions [5]. In
addition, they are so versatile till that they are being produced for many applica-
tions. Often, the xanthan gum are added for the food application, personal care
application, industrial applications, animal feed, and as well as in pharmaceuticals.
Xanthan gum is everywhere to be used daily.
Xanthan gum is not only acts as a stabilizer in most salads dressings and sauces,
but also in dairy products colloidal and solid materials where they are found to be
suitable to be stabilized by xanthan gum. Also, xanthan gum also found to be
reasonable to be used in frozen foods and beverages. In fact, it is possible to
undergo syneresis for the frozen foods. Syneresis is the process of extraction liquid
from respective gel with xanthan gum. In beverages, xanthan gum is good at
suspending the fruit pulps in order to give better taste and flavours. Apart of that,
xanthan gum gives desired texture in many ice-creams by enhances the binding of
free water for the product stability. Meanwhile, xanthan gum can improve the starch
cohesions in bakery that benefits for extending the products shelf-life where the
moisture preservation. It also prevents gas entrainment, where tiny bubbles always
form in cake making [6]. On the hands, it is very common for xanthan gum in
cosmetics and pharmaceuticals application. As for instance, xanthan gum plays an
important role as the thickener and stabilizer in the making of personal care product
like creams. As a result the xanthan gum gives gentle and soft feel as due to shear
thinning flow behaviour of xanthan gum. The shear thinning flow behaviour is also
significant for toothpaste which acts as the binder that improves the uniform dis-
persion on teeth and better rinsing during brushing of teeth. Next, xanthan gum is
also good in drilling fluids as thickener which are used in large quantities. The
fluids capable to bring out the solid cuts by drilling bit [7].
Nevertheless, xanthan gum production has its own condition to be achieved
where different sources and a few other limitations will affect the production itself.
It is wide opportunity for local people as if it is found that the sources are readily
available in order to produce local product as it is profitable to the nation economy.
Hence the development of locally produced xanthan gum is not impossible. The
research upon xanthan gum is initiate by the lab scale is necessary especially for the
preliminary research where the many basic properties of xanthan gum production
can be obtained [8].
[email protected]
Comparative Study Between Isolated Xanthomonas Campestris … 179
Experimental
Materials, Growth media, malt extract, yeast extract, peptone, glucose and all
apparatus used was provided by Universiti Teknologi MARA (UiTM) Shah Alam,
Malaysia. The research has been conducted in Bioprocess Laboratory, UiTM Shah
Alam and the equipment such as incubator shaker and spectrophotometer were
provided by UiTM Shah Alam. Rotten cabbage samples for the research were
collected locally from Cameron Highland, Pahang Darul Makmur.
Preparation of Growth Media, A media need to be prepared in order to grow
the bacteria. Yeast-Malt (YM) Agar in the form of solid and Yeast-Malt (YM)
Broth in the form of liquid was used for the growth. The composition of both media
as follows in Table 1.
Isolation of the Xanthomonas Campestris, For the first series, the black rot spot
were aseptically cut in small pieces and put on top of prepared YM agar. The
sample were labelled and incubated for 48 h at 28 °C. After the incubation com-
pleted for 48 h, the colonies formed on the sample were observed which later
inoculated into YM agar and YM broth. Both samples inoculated in YM agar and
YM broth were then incubated once again for 48 h at 28 °C. For the second series,
small pieces of rotten cabbage were dipped in YM Broth which then later incubated
for 48 h at 28 °C. The samples were not inoculated instantly instead, 100 ml of 24 h
YM Broth cultures were pippetted onto YM Agar before spread the culture thor-
oughly. The samples then taken for incubation for 48 h at 28 °C. After 48 h, the
culture inoculated into YM Agar and YM Broth.
Preparation of Fermentation Media, YDC media prepared consist of 10 g of
yeast extract, 5 g glucose and 5 g calcium carbonate. 20 g rotten cabbage was
grinded and mixed with 100 L of distilled water for fermentation substrate.
Innoculum Development, Samples from 48 h YM agar plate were inoculated
into 3 mL YM broth Media. The broth then will be incubated at 28 °C for 48 h.
100 μL of the previous sample cultures were pippetted into universal bottle con-
taining 10 mL of YM broth media. The culture incubated in the incubator shaker for
48 h at 28 °C. For further bioreactor study, 40 mL of prepared culture in the
universal bottle will be transferred into 400 mL flask of production media which
then will be incubated in the incubator shaker at 28 °C, agitation speed of 200 rpm
for 48 h.
Table 1 Composition of YM broth YM agar

growth media
Yeast 3 g/L 3 g/L
Malt 3 g/L 3 g/L
Peptone 5 g/L 5 g/L
Glucose 10 g/L 10 g/L
Distilled water 1L 1L
Agar 20 g/L
[email protected]
180 M. Musa et al.
Table 2 Substrate for Sample Substrate

fermentation process
1 YDC Media + Collection Culture
2 Grinded Rotten Cabbage
3 Grinded Rotten Cabbage + Collection Culture
4 Isolated Culture + YDC Media
5 Isolated Culture + Grinded Rotten Cabbage
Fermentation, Prior the fermentation process, fives type substrates for fer-
mentation of xanthan gum were prepared for further fermentation. The combination
of media and culture to form certain substrate can be refer on Table 2.
The effect of bacteria cell dry mass over time of fermentation, In comparison
with theoretical growth of bacteria versus time, growth trends of all 5 samples
(Fig. 1) satisfied with theoretical one although there are slight differences in term of
the exponential phase of the bacterial growth. It takes longer time for samples to
reach its maximum growth. The bacteria activities are at its peak between 32th and
36th h. After 36th h the bacteria activity start to decline due to limiting factor such
as nutrients provided in the fermentation broth. Samples 3 show the best results
among others.
The composition of the samples which consist of grinded rotten cabbage and
collection of xanthamonas campestris culture may provide rough hypothesis why it
is the best among others. Based on the trends, the usage of grinded rotten cabbage
in sample 2, 3 and 5 shows that grinded rotten cabbage may be the main contributor
to Xanthamonas Campestris cell dry mass as all three samples leave both sample
1 and 4 behind. This may not suggest that YDC media is not suitable for the growth
Fig. 1 Profile of cell dry

mass at different substrate
formula
[email protected]
Comparative Study Between Isolated Xanthomonas Campestris … 181
Fig. 2 Profile of xanthan

gum produced at different
substrate formula
of the bacteria, this is because the grinded rotten cabbage may consist of vast
amount of the bacteria during the initial process of fermentation compare to sub-
strate bases with YDC media.
The effect of glucose concentration level over time of fermentation, Glucose
analyses are important to determine the activities of bacteria in fermentation broth.
Theoretically, as the activity of the bacteria increase, the glucose concentration of
substrate decreases. Figure 2 shows the trend of glucose content of the substrates.
The trends meet the theory stated as the time passed the glucose content concen-
tration declines. The declination of glucose content satisfied with the increasing of
cell dry mass of the bacteria.
The Effect of Xanthan Gum Produced over Time of Fermentation, Based on
the Fig. 3, all 5 samples shows great increasing amount of xanthan gum by the time
passed up until it reaches its maximum amount of xanthan gum produced which is
in between 32 and 36 h. As the time passed on 36th h, amount of xanthan gum start
Fig. 3 Profile of glucose

concentration at different
substrate formula
[email protected]
182 M. Musa et al.
to decline dramatically. This is due to the decreasing amount of cultured bacteria to

precede the production of xanthan gum. Sample 1 show the best productivity
among others while sample 2 the worst among others. The trend shows both sample
1 and 4 which bases on YDC media provide better condition for xanthan gum
production. YDC media provide better condition for the fermentation process
despite having low amount of bacteria as mentioned before in Fig. 2.
Conclusion
Cell dry mass of the bacteria will increase as the fermentation proceeds up until its
peak condition which is called exponential phase of bacteria growth. The bacteria
reach its maximum activity at 32nd to 36th h and then will decline dramatically due
tu nutrient deficiency. Glucose content of the substrate trends over fermentation
period provide perfect example of nutrient used up by the bacteria. Xanthan gum
production increasing as the time passed up until 32nd to 36th h and the production
will decreases due to insufficient amount of bacteria for further fermentation. YDC
media bases substrate proved to be better condition for the xanthan gum production.
Acknowledgment The authors would like to thank the Ministry of Education Malaysia for
financial support via Fundamental Research Grant Scheme (FRGS). The authors also grateful
acknowledge use of services and facilities of Chemical Engineering Faculty and Universiti
Teknologi MARA.
References
1. Massomoa S, Mabagalaa R, Swaib I (2003) Evaluation of varietal resistance in cabbage against

the black rot pathogen, Xanthomonas campestris pv. campestris in Tanzania. Plant Pathogens
315–324
2. Moosavi- Nasab M, Pashangeh S, Rafsanjani M (2010) Effect of fermentation time on xanthan
gum production from sugar beet molasses. World Academy of Science, Engineering and
Technology pp 1244–1247
3. Ielpi L, Becker A, Katzen F, Puhler A (1998) Xanthan gum biosynthesis and application. Life
Science 145–152
4. Ciullo PA, Andersson M (2003) Xanthan gum, a clearly better stabilizer. SÖFW J Sonderdruck
5. Katzbauer B (1997) Properties and applications of xanthan gum. Sci Biol 59:82–96
6. Averre CW (2000, December 14) Plant pathology extension, College of Agricultural and Life
Sciences, North Carolina State University. Retrieved December 20, 2012, from Black Rot of
Cabbage and Related Crops: http://www.ces.ncsu.edu/depts/pp/notes/oldnotes/vg16.htm
7. Hiett K, Seal B (2009) Use of repetitive element palindromic PCR (rep-PCR) for the
epidemiologic discrimination of foodborne pathogens. Molecular epidemiology of microor-
ganisms, methods in molecular biology pp 49–58
8. Jensen B, Vicente J, Manan H (2010) Occurrence and diversity of Xanthomonas Campestris pv.
campestris in vegetable brassica fields in Nepal. Plant Disease, University of Warwick
institutional repository pp 1–34
[email protected]
Part III
Energy and Renewable Energy
[email protected]
1-Pyrenebutyric Acid Functionalized
Reduced Graphene Oxide (1-Pb-Rgo)
Energy Storage
Ellie Yi Lih Teo, Mashitah Mohd Yusoff and Kwok Feng Chong
Abstract Supercapacitors are a class of energy storage device which has high
energy density and high power density. As a material with unique 2D structure as
well as outstanding physical properties such as high electrical conductivity and
large surface area, graphene demonstrates great potential to be the electrode
material for supercapacitors. Despite graphene showing theoretical surface area as
high as 2630 m2/g, results acquired showed that not all the surface area were
utilized. This could be due to the tendency of the graphene layers to restack. In this
work, 1-pyrenebutyric acid (1-PB) was anchored to graphene with the pyrenyl
group via π-π stacking to prevent the restacking of graphene layers. The successful
functionalization of 1-PB on the hydrophobic surface of rGO was characterized
with UV-Vis Spectroscopy and Fourier Transformed Infrared Spectroscopy (FTIR).
The electrochemical performance of 1-PB-rGO was studied through cyclic vol-
tammetry (CV), galvanostatic charge-discharge (CD) and electrochemical imped-
ance spectroscopy (EIS). Using 6 M KOH as the electrolyte, we obtained an
enhanced specific capacitance for 1-PB-rGO. These findings indicates that the non-
covalent functionalization of 1-PB on rGO enhances the capacitive storage ability
and it show potential as an electrode material in the energy storage application.
Keywords Reduced graphene oxide Non-covalent functionalization Energy

storage
E.Y.L. Teo (&) K.F. Chong

Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang,
Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia
K.F. Chong
M.M. Yusoff
Research and Innovation Department, Universiti Malaysia Pahang,
Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia

[email protected]
186 E.Y.L. Teo et al.
Introduction
In the modern era of technology where portable electronics are a current trend
together with the depletion of fossil fuel have led to the search for a dependable and
durable energy storage solution. Among the various solutions which are being
explored, supercapacitor have gained the attention of the research community. The
supercapacitor is a class of energy storage device which has a high energy density
as compared to conventional capacitors and a higher power density compared to
batteries.
Since its discovery by Andre Geim and Konstantin Novoselov, graphene has
become a sensational material due to its exceptional properties. Graphene is a single
layer of sp2 bonded carbon atoms arranged in honeycomb crystal lattice. It possess
exceptional properties such as very high specific surface area (*2630 m2/g) and
high electrical conductivity which fits the criteria as the electrode material of the
supercapacitor [1]. This results in graphene to be highly explored in the application
of energy storage.
However graphene tends to agglomerate as a result of the Van der Waals
interaction. In order to prevent this, functionalization of graphene is performed by
the covalent or non-covalent modification technique in order to improve their
solubility, self-assembly properties and in applications [2]. The non-covalent
functionalization if preferred as it does not disrupt the conjugation of the graphene
sheet and improves its stability [3]. A few works have been reported on the non-
covalent functionalization of graphene with pyrene derivatives. Xu et al. [2]
functionalized graphene with 1-pyrenebutyric acid through a simultaneous func-
tionalization and reduction method. Supercritical carbon dioxide was used in the
research of Li et al [4] to assist in the functionalization process. While there are a
few literatures which reported on the different methods of functionalization, only
Ghosh et al. (2012) and An et al. (2010) reported on the application of pyrene
derivatives functionalized graphene as energy storage device. They studied on the
effect of 1-pyrenecarboxylic acid functionalization of graphene on its capacitive
storage ability and reported specific capacitance value of 200 and 120 F/g
respectively [1, 5].
In this work, 1-pyrenebutyric acid, a pyrene derivative which is reported to
interact with graphite with the pyrene group via π stacking is used to functionalize
reduced graphene oxide (rGO). 1-PB was anchored to the hydrophobic surface of
reduced graphene oxide (rGO) through a simple sonication process at room tem-
perature. Reduced graphene oxide was successfully functionalized with 1-pyre-
nebutyric acid (1-PB) with this method. The application of 1-pyrenebutyric
functionalized reduced graphene oxide (1-PB-rGO) is studied electrochemically
with cyclic voltammetry (CV), galvanostatic charge-discharge (CD) and electro-
chemical impedance spectroscopy (EIS).
[email protected]
1-Pyrenebutyric Acid Functionalized Reduced … 187
Experimental
Chemicals
All chemicals used in this experiment were purchased from Sigma-aldrich unless
otherwise stated and used as received.
Graphene Oxide Preparation
Briefly, 4.0 g of graphite powder was added into a mixture of concentrated sulfuric
acid, H2SO4 (30 mL), potassium persulfate, K2S2O8 (6.0 g) and phosphorus
pentoxide, P2O5 (6.0 g). The mixture was heated at 80 °C and kept stirred for 6 h
using oil bath. After 6 h, the solution was cooled and diluted with 2 L of deionized
water. The solution was filtered using Whatman filter paper and the residue was
dried naturally. The pre-treated graphite was added into concentrated H2SO4
(300 mL), followed by the addition of potassium permanganate, KMnO4 (35 g)
gradually with stirring. The mixture was stirred at 35 °C for 4 h and diluted with 2 L
of water by keeping the temperature under 50 °C. Hydrogen peroxide, H2O2
(100 mL) was then added to the mixture drop by drop. The solution was then
filtered and washed with 1:10 hydrochloric acid, HCl aqueous solution (2 L) fol-
lowed by deionized water (2 L). The resulting solid was filtered and dried at room
temperature.
Reduction of Graphene Oxide
60 mg graphite oxide was dispersed and sonicated in water (50 mL) for 4 h to
exfoliate it. The pH of the solution was then adjusted to pH 10. Hydrazine mono-
hydrate (1.2 mL) was added to the solution and it was refluxed for 24 h at 80 °C.
Functionalization of 1-Pyrenebutyric Acid (1-PB) on Reduced

Graphene Oxide (RGO)
10 mg of 1-PB and 10 mg of rGO were dispersed in 5 mL of ethanol, sonicated for

1 h stirred mechanically overnight. The resulting solution was filtered, washed with
deionized water and dried in oven at 80 °C for 3 h.
[email protected]
Characterization Methods
rGO, 1-PB and 1-PB-rGO were dispersed in deionized water in concentration of

0.1 mg/mL for UV-Vis measurement. The solutions were then scanned from 200 to
400 nm using Thermo Scientific Genesys 10S UV-Vis. Fourier Transform Infrared
Spectroscopy (FTIR) measurements were performed by pressing the materials with
potassium bromide (KBr) to form pellets. The pellets were scanned in the frequency
range of 400–4000 cm−1 using Perkin Elmer Spectrum 100.
Electrochemical Studies
The working electrodes were fabricated by mixing the material with carbon black
and polyvinylidene fluoride (PVDF) in the ratio of 80:15:5. The mixture was then
dispersed in N-methylpyrrolidone (NMP) and stirred overnight. The slurry was
thrush coated onto nickel foam and dried in the oven at 80 °C for 12 h. To assemble
the coin cell, two identical pieces of the fabricated electrode were sandwiched with
a microglass fiber separator (25 µm, Filter Fioroni) and placed inside the coin cell
(CR2032, MTI) with 6 M potassium hydroxide (KOH) as the electrolyte.
Cyclic voltammetry was performed at a potential window of −1–1 V at various
scan rates ranging from 100 to 5 mV/s. Galvanostatic charge-discharge measure-
ments were performed at different current densities for potential window of 0–1 V.
Electrochemical impedance spectroscopy (EIS) was carried out at open circuit
potential (OCP) for the frequency range of 500 kHz to 10 mHz. The experiments
were conducted at room temperature using PARSTAT 2273.
Characterization of RGO and 1-PB-RGO
Figure 1 shows the UV-Vis spectra for rGO, 1-PB and 1-PB-rGO. The spectra for
rGO exhibits an absorbance band at 266 nm which is the signature peak for rGO
that has a maximum at 270 nm. The presence of this peak is due to the electronic
π − π* transition of the C=C bond. It indicates that the electronic conjugation within
the rGO sheets were restored upon reduction with hydrazine. Meanwhile, the 1-PB-
rGO spectra exhibits three absorbance bands. An obvious band at 266 nm which
represents the presence of rGO and a shoulder peak at 242 nm and a weak peak at
343 nm. The shoulder peak and weak peak are contributed by 1-PB [4].
To further confirm the successful functionalization of 1-PB on rGO, FTIR
measurements were also conducted. Characteristic peaks for GO are observed in
[email protected]
266 nm
242 nm
Absorbance, a.u
346 nm
1-PB-rGO
1-PB
rGO
200 250 300 350 400

Wavelength, nm
Fig. 1 UV-Vis spectra of rGO, 1-AP and 1-AP-rGO
60
2918
2843
58
3442
1053
1386
1638
56
54
Transmission, %
1-PB-rGO
52
50
3418
rGO
1621
48
1119
46
44
42 GO
1055
1727
40
1623
3408
38
36
34
4000 3500 3000 2500 2000 1500 1000 500
Wavelength, nm
Fig. 2 FTIR spectra of GO, rGO and 1-AP-rGO
FTIR spectrum in Fig. 2 such as the peak due to broad and intense peak contributed
by O–H groups at 3408 cm−1, C=O group at 1727 cm−1 and C–O stretching at
1055 cm−1. The intense peak contributed by O–H group have diminished signifi-
cantly as a result of deoxygenation after reduction [6]. For the FTIR spectra of 1-
PB-rGO, the signature peaks for rGO at 3442 and 1638 cm−1 is observed which is
due to O–H group and C=O group respectively. A peak at 1053 cm−1 due to C–O
strectching is detected too. The peaks contributed by the C–H stretching of alkane
at 2918 and 2843 cm−1 and C–H rocking of alkane group at 1385 cm−1 are also
observed. These peaks are most likely due to the functional groups that are present
in 1-PB.
[email protected]
Electrochemical Studies
The electrochemical properties of rGO and 1-PB-rGO electrodes were studied using
CV, CD and EIS. The specific capacitance for the materials were calculated from
the discharge slope using the the following formula:
Csp ðF=gÞ ¼ 4 C/m: ð1Þ
C ¼ I=ðdV=dtÞ: ð2Þ
where C is the measured capacitance from the two-electrode system and m is the
total mass of active materials present in both electrodes. The multiplier 4 is used to
adjust the capacitance of the cell and the mass combined from the two electrodes to
the capacitance and mass of a single electrode. dV/dt is calculated from the slope of
the discharge curve [7].
Figure 3 shows the cyclic voltammogram of rGO and 1-PB-rGO at scan rate of
5 mV/s in 6 M potassium hydroxide (KOH) as the electrolyte. The cyclic vol-
tammogram for rGO yielded an rectangular curve which suggest the idea capacitive
behaviour of the rGO [8]. At the same scan rate, the CV curve for 1-PB-rGO is also
rectangular in shape with no obvious redox peak. However, the larger area under
the curve indicates that the 1-PB-rGO composite possess a better capacitive
behaviour.
The galvanostatic charge-discharge study was conducted to observe the sus-
tainability of the capacitive behaviour of the samples studied. Figure 4 shows the
galvanostatic charge-discharge curve for rGO and 1-PB-rGO at current density of
0.1 A/g. For both rGO and 1-PB-rGO samples, it can be observed that the charge-
discharge curve is triangular in shape which suggest that the charge discharge
process is highly symmetrical, leading to ideal capacitive behaviour [9]. The spe-
cific capacitance for rGO and 1-PB-rGO were calculated from Eqs. 1 and 2 to be
0.4
0.3
Current density, A/g
0.2
0.1
0.0
-0.1
-0.2
-0.3 1-PB-rGO
rGO
-0.4
-1.0 -0.5 0.0 0.5 1.0

Potential, V
Fig. 3 Cyclic voltammetry (CV) curves for rGO and 1-PB-rGO at scan rate of 5 mV/s
[email protected]
1.0
0.8
Potential, V
0.6
0.4
0.2
rGO 1-PB-rGO
0.0
0 200 400 600 800
Time, s
Fig. 4 Galvanostatic charge-discharge curve at current density of 0.1 A/g for rGO and 1-PB-rGO
respectively
150
Specific capacitance, F/g
1-PB-rGO
140
130
120
110
100
rGO
90
0.2 0.4 0.6 0.8 1.0

Current density, A/g
Fig. 5 Plot of specific capacitance of rGO and 1-PB-rGO electrode as a function of discharge
current density
105 and 150 F/g (as shown in Fig. 5) respectively. The large enhancement of
specific capacitance (approximately 45 %) observed is attributed to the intercalation
of pyrene group between the rGO layers which prevented agglomeration of rGO. It
is also attributed to the pseudocapacitance contributed by the –COOH functional
group which is attached to the surface of the rGO (Fig. 6).
The electrochemical impedance spectroscopy (EIS) is a powerful tool used in the
study of the ion transport at the interface and to probe the resistance limitations
present in the supercapacitor fabricated. The intercept of the x-axis of the Nyquist
plot at the high frequency end represents the solution resistance (Rs) as illustrated in
Fig. 6 [10]. It can be observed from the inset illustration that the Rs is at *0.6 ohm
for rGO and *0.5 ohm for 1-PB-rGO. This indicates that the functionalization of
rGO with 1-PB improves the Rs of rGO.
This is usually followed by a semi-circular arch which is a representation of the
interfacial charge transfer resistance (Rct) and the interfacial capacitance (Cs) which
[email protected]
200
1.0
1-PB-rGO
0.8
150 rGO
0.6
Z", ohm
Z", ohm
100 0.4
0.2
50
0.0
0.5 1.0 1.5 2.0 2.5 3.0
Z', ohm
0
0 20 40 60 80 100 120 140 160 180 200
Z', ohm
Fig. 6 Nyquist plot for 1-PB-rGO and rGO. Inset shows spectra at high frequency end
are connected at the electrode/electrolyte interface parallely. The Nyquist plots

shows that Rct for 1-PB-rGO is slightly larger than rGO which indicates that some
of the pores could be blocked by the 1-PB particles [11]. The last part is the vertical
line observed after the semi-circular arch. The slope for rGO and 1-PB-rGO is
recorded to be *83° and *82° which shows that the supercapacitor functions
closely to an ideal capacitor [10].
Conclusion
This work involves the non-covalent functionalization technique of rGO with 1-

pyrenebutyric acid which is simple, cost effective and eco-friendly as it requires only
simple sonication and no usage of strong oxidants or corrosive acids. The specific
capacitance calculated to be 150 F/g which is an enhancement of 45 % compared to
rGO. The enhancement is most probably due to the larger exposed surface area of the
material as well as due to the attachment of –COOH functional group on the surface
area of rGO. Therefore, 1-PB-rGO composite is an efficient choice of material to be
use in the fabrication of electrode for energy storage application.
Acknowledgment KF Chong and co-workers would like to thank the funding of RDU 121212
and RDU 121213.
References
1. Ghosh S, An X, Shah R, Rawat D, Dave B, Kar S, Talapatra S (2012) Effect of

1-pyrenecarboxylic acid functionalization of graphene on its capacitive energy storage. J Phys
Chem C 116:20688–20693
2. Xu Y, Bai H, Lu G, Li S, Shi G (2008) Flexible graphene films via the filtration of water
soluble non-covalent functionalized graphene sheets. J Am Chem Soc 130:5856–5857
[email protected]
3. Luo L, Zhang Z, Ding Y, Deng D, Zhu X, Wang Z (2013) Label free electrochemical
impedance genosensor based on 1-aminopyrene/graphene hybrids. Nanoscale 5:5833–5840
4. Li L, Zheng X, Wang J, Sun Q, Wun Q (2013) Solvent exfoliated and functionalized graphene
with assistance of supercritical carbon dioxide. ACS Sustain Chem Eng 1:144–151
5. An X, Simmons T, Shah R, Wolfe C, Lewis KM, Washington M, Nayak SK, Talapatra S, Kar
S (2010) Stable aqueous dispersions of noncovalently functionalized graphene from graphite
and their multifunctional high-performance applications. Nano Lett 10:4295–4301
6. Choi EY, Han TH, Hong J, Kim JE, Lee SH, Kim HW, Kim SO (2010) Non-covalent
functionalization of graphene with end functional polymers. J Mater Chem 20:1907–1912
7. Stoller MD, Ruoff RS (2013) Best practice methods for determining an electrode materials
performance for ultracapacitor. Energy Environ Sci 3:1294–1301
8. Zhu Y, Murali S, Stoller MD, Velamakanni A, Piner RD, Ruoff RS (2010) Microwave assisted
exfoliation and reduction of graphite oxide for ultracapacitors. Carbon 48:2106–2122
9. Mishra AK, Ramaprabhu S (2011) Functionalized graphene-based nanocomposites for
supercapacitor application. J Phys Chem C 115:14006–14013
10. Hastak RS, Sivaraman P, Potphode DD, Shashidhara K, Samui AB (2012) All solid
supercapacitor based on activated carbon and poly(2,5-benzimidazole) for high temperature
application. Electrochim Acta 59:296–303
11. Su Y, Zhitomirsky I (2013) Eletrophoretic assembly of organic molecules and composites for
electrochemical supercapacitors. J Colloid Interface Sci 392:247–255
[email protected]
Clarifying the Palm-Based Drilling Fluids
Potentials
Hazlina Husin, Zulkafli Hassan, Azlinda Azizi,

Mohd Fiqri Arman Mansor, Rosenadira Nasarauddin
and Sharifah Nawirah Syed Ariffin
Abstract With the goal of increasing awareness of toxicity level and environ-
mental impact of drilling fluid, and for providing continuous findings for a better
drilling fluid formulation, an experimental work based on a rheological evaluation
of drilling fluid was carried out. The use of palm-based and palm oil-based as an
alternative base component in drilling fluid formulation have been explored in
terms of sand content, pH, viscosity, yield point and gel strength. Standard labo-
ratory formulation including distilled water, sodium hydroxide, bentonite, barite
and 1,2-propylene glycol were used in this study. Tests conducted on palm-based
drilling fluid with properties of 17.0 ppg and 80/20 oil-water ratio revealed that
palm methyl ester C14 and palm kernel oil have sand content of less than 0.5 % vol.
with excellent thermal resistance on high temperature for palm kernel oil based.
Keywords Palm oil Palm methyl ester Palm kernel oil Drilling fluid
Rheological test
H. Husin (&) Z. Hassan A. Azizi M.F.A. Mansor R. Nasarauddin S.N. Syed Ariffin
Z. Hassan
A. Azizi
M.F.A. Mansor
R. Nasarauddin
S.N. Syed Ariffin
H. Husin
CoRe of Frontier Materials and Industry Applications, Universiti Teknologi MARA,

[email protected]
196 H. Husin et al.
Introduction
Drilling fluid or drilling mud is defined as any fluid used in a drilling operation.
Drilling fluid plays a major role in drilling operation where it is pumped or cir-
culated from the surface, down the drill string, through the bit and back to the
surface [1]. The fluid is used to lift drill cuttings to the surface, control subsurface
pressure, lubricate drill string, clean bottom-hole, aid formation evaluation process
and protect formation. Unfortunately, the costs of drilling fluid constitute a sig-
nificant fraction of the overall costs of drilling operation. In many cases, the cost
ascribed to the drilling fluid is the costs associated with waste disposal and solids
management. Thus, it is just as important to minimize costs associated with these
two aspects as it is to ensure that the drilling fluid fulfils its primary functions.
Among the three types of drilling fluid; the oil-based, the water-based and the
air/foam; oil-based drilling fluid has been the drilling fluid of choice for complex
conditions such as high temperature, high angle, extended-reach well and hydrat-
able shales. Also, oil-based drilling fluid produces low waste. However, a common
base component used in the oil-based drilling fluid formulation, the diesel oil, is
toxic and creates environmental problems when they are discharged [2]. According
to the Effluent Limitation Guidelines (ELGs) [3, 4], a discharging law by
Environmental Protection Agency (EPA), any waste produced from oil-based
drilling fluid cannot be directly discharged on-site. ELGs prohibit releases of free
oil, as detected by the static sheen test, from drilling fluid and drill cuttings
discharges.
There have been continuous studies on the replacement of diesel oil by palm
resources [5]. Study found that palm-oil has the lowest production cost from other
major oils [6]. The used of palm-oil is also advocated because it is assumed to
dramatically reduce CO2 emissions [7] and may satisfy demanding technical criteria
and strict environmental standards for offshore work [8]. This study aims to clarify
the potentials of a few compounds from palm-oil resources at different tempera-
tures, used as the base component in drilling fluid formulation.
Experimental
In this study, palm resources such as refined palm-oil, palm kernel oil and palm
methyl esters were employed as the base fluid into standard formulation compo-
nents of drilling fluid [9] as in Table 1 made of 17.0 ppg and 80/20 oil-water ratio.
Palm methyl esters are made through palm oil transesterification process. The
conversion of palm oil to palm methyl ester involves the use of methanol as a raw
material and a basic catalyst. Conversion of palm oil compounds into smaller
molecules of palm methyl esters allows them to behave similar to diesel fuel.
Procedures for rheology investigation was based on API recommended standard
(API-RP 13B-1 2009) [10]. The mixture was continuously stirred until a
[email protected]
Clarifying the Palm-Based Drilling Fluids Potentials 197
Table 1 Standard Component Quantity (g)

formulation of drilling
fluids [9] Distilled water 350
Calcium hydroxide 1.0
Bentonite 6.0
1,2-propylene glycol 0.7
Barite 130
homogeneous system obtained. The rheological properties of the mixture were

determined.
Sand Test, 200 mesh sand screen was employed to trap the sand and other
course particles in the drilling fluids. The volume of sand including the spaces
between the grains is expressed as a percentage of the volume of the drilling fluid.
The sand content of drilling fluid is the percentage (volume fraction) of particles of
diameter larger than 74 μm.
Rheology Test, The FANN (Model 35A) viscometer was used in order to get
viscosity, gel strength and yield point measurement. The FANN (Model 35A)
viscometer is a coxial cylindrical rotational type, used to determine single or
multiple-point viscosities. It has fixed speeds; 3 (GEL), 6, 100, 200, 300 and 600
RPM that are switch-selectable with the RPM knob. A freshly mixed sample was
placed in the cup, the upper housing of the viscometer was tilted back, the cup was
located under the sleeve (the pins on the bottom of the cup fit into holes in the base
plate), and the upper housing was lowered to its manual position. Then, the knurled
knob was turned between the rear support posts to raise or lower the rotor sleeve
until it was immersed in the sample of the scribed line. After that, the sample was
stirred for about 5 s at 600 RPM and then the RPM desired was selected for the
best. The dial reading was waited to stabilize and then the dial reading and RPM
were recorded.
Resistivity Test, 5 mL of the mixture was taken to undergo resistivity test using
the FANN (Model 653 Resistivity Meter) Analog Resistivity Meter.
There are two types of oil that can be produced from palm resources. These are the
palm oil which can be produced from the fibrous mesocarp of palm fruit and the
lauric oil from the palm kernel. During the milling process, the palm fruit bunches
are sterilised and stripped from the fruitlets. They are then digested and pressed to
extract the palm oil. The nuts are separated from fibre in the press cake and cracked
to obtain palm kernels which are crushed to obtain palm kernel oil and palm kernel
cake. Fractionation process of palm oil and palm kernel oil produces the liquid
stearin fraction and a solid stearin component. The fatty acid compositions of palm
oil products are presented in Table 2. The results showed that palm oil has a
[email protected]
198 H. Husin et al.
Table 2 Fatty acid Fatty Weight percentage

composition of palm oil acid
products, soy oil and coconut
Palm oil Palm kernel oil Coconut oil Soy oil
oil [11]
C6:0 0.3 0.2
C8:0 4.4 8
C10:0 3.7 7
C12:0 0.2 48.3 48.2
C14:0 1.1 15.6 18
C16:0 44 7.8 8.5 6.5
C18:0 4.5 2 2.3 4.2
C18:1 39.2 15.1 5.7 28
C18:2 10.1 2.7 2.1 52.6
Iodine 53.3 17.8 9.5 133
value
balanced ratio of saturated and unsaturated fatty acids while palm kernel oil has
mainly saturated fatty acids which are broadly similar to the composition of coconut
oil. When compared to soy oil, palm oil contains higher amount of saturated fatty
acids. This means that it will be more stable and less prone to oxidation at high
temperatures.
Table 3 shows the result for a sand content test on unconsolidated formation of
palm-based drilling fluids using different types of base fluid. In order to achieve
desired drilling fluid viscosity and drilling fluid weight, the sand content in the
drilling fluid formulation should be controlled to be less than 0.5 % vol [12]. This is
because the presence of high solid particles increases the drilling fluid viscosity and
drilling fluid weight from desired values hence affects the drilling efficiency, which
in turn increases maintenance costs. Furthermore, high sand content can lower the
rate of penetration which in turn increases horsepower required to circulate the
drilling fluid. According to Table 3, palm methyl esters C12 and C16 showed
intolerable range of sand content. In contrast, palm methyl ester C14 and palm
kernel oil showed a more reasonable range. Ismail and Khor studied palm methyl
ester C13 and C15 and found that the preliminary result of these palm oil esters too
has a potential to be used as based drilling fluid [2].
The study continues on the investigation of 2 sets of palm kernel oil-based
formulation at different temperature; 25.6 °C (labelled as PKO-1) and 100.7 °C
Table 3 Percentage of sand Base component Sand content (% vol)

content in drilling fluids with
different base component Palm kernel oil 0.50–0.80
Refined palm oil 3.50–5.00
Palm methyl ester, C12 1.00–6.00
[email protected]
(labelled as PKO-2). Investigation at high temperature aims to observe the proba-

bility of a thermal degradation process in the physicochemical structure of the fluid
occurred at elevated temperature via rheological test. Furthermore, the measurement
of rheological properties (such as viscosity) in drilling fluid is important in the
development of rheological models at high temperature because these fluids have a
very complex physicochemical structure which enables a non-Newtonian behaviour
to be exhibited. Initially, the formulation of PKO-1 at room temperature was made
for pH 9.79. Previous study has reported that the optimum range of drilling fluid pH
is between 8.0 and 10.5 [13]. When temperature was increased, the pH of the palm
kernel oil-based drilling fluid was found to be acidic at 100.7 °C (PKO-2). This
indicates that at high temperature condition, palm kernel oil-based drilling fluid will
require more alkaline components, such as soda ash or caustic soda, to elevate the
pH. Alternatively, a chemical inhibitor with/without scavenging chemical may be
needed. Commonly, the pH of drilling fluid is within an alkaline region so as to
minimize the corrosion effect and corrosion issues on drilling equipment.
Rheology test is the basis for all analyses of wellbore hydraulics [14].
Rheological properties exert a considerable influence upon hydraulic horsepower.
These include measurement of yield point, viscosity and gel strength (10′ and 10′′).
Measured using a viscometer [10], the results are vital in ensuring the performance
of drilling fluid controls. The viscosity of PKO-2 sample showed lower reading
than PKO-1. The plastic viscosity and apparent viscosity for PKO-2 is 13 cP and 14
cP respectively. This means that temperature has significant impact on the vis-
cosities of palm kernel oil-based drilling fluid. Comparing the viscosities at these
two temperatures, we found that gelation point was not observed at elevated tem-
perature, in particular at 100.7 °C. When compared to previous studies viscosities
of the investigated palm kernel oil-based drilling fluid is similar to that other
reported viscosities for drilling fluid, within a range of 14–18 cP [15], and this
reported value is also close to other commercial mineral oil-based drilling fluid
[16]. This shows that the palm kernel oil compound used as the base component in
the drilling fluid is compatible to the commercially systems in the sense of
molecular stability at elevated temperature condition.
Yield point is defined as the initial resistance to flow. Drilling fluid with high
yield point give better performance in carrying drill cuttings as compared to low
yield point of the same drilling fluid [17]. Palm kernel oil-based drilling fluids
(PKO-1 and PKO-2) demonstrated similar yield point value as 10.0 ppg palm
methyl ester-based drilling fluids [2]. Yield point of drilling fluid can be increased
by adding suitable gelling agent. Gel strength measured in 10 s–10 min period
represents the gelling effect of drilling fluid in static condition. The gel building up
property is vital in order to suspend the drill cuttings when drilling operations are
temporarily ceased. Increment of the gel strengths (as shown in Table 4) with time
confirms the thixotropic property of the palm kernel oil-based drilling fluid. The
PKO systems show an acceptable gelling effect. Gel strength for commercial
mineral oil-based [16] and palm methyl ester-based (C13 and C15) [2] drilling fluid
was reported to be 9/14 and 9/12 respectively. Gelling effect can be improved by
[email protected]
200 H. Husin et al.
Table 4 pH, rheology and resistivity tests

Sample pH Plastic Apparent Yield point Gel strength- Gel strength- Resistivity
viscosity (cP) viscosity (cP) (lb/100 ft2) 10 s (lb/100 ft2) 10 min (Ohm)
(lb/100 ft2)
PKO-1 9.79 59 62.5 7 9 12 12
PKO-2 5.69 13 14 2 9 16 12
increasing water content which behaves as pseudo particles and enhances suspen-
sion capability of the system [2].
Resistivity is a characteristic electrical property of a material and is equal to the
electrical resistance of a 1 m3 of the material to passage of a 1 A electric current
perpendicular to two parallel faces [18]. Solid materials exhibit intermediate elec-
trical properties depending on their physical and chemical properties. Control of the
resistivity of a drilling fluid is desirable in order to get better evaluation of formation
characteristics from an electric log. Resistivity of drilling fluid will be higher if
water disperses well in oil phase. It means that the mixture will produce a good
emulsion. In contrast, when water disperses hardly in oil phase the resistivity of
drilling fluid will be lower since the mixture formed bad emulsion. The stability of
emulsion of synthetic-base drilling fluid is reflected in the value of electrical sta-
bility. Since the 2 samples (PKO-1 and PKO-2) have the same formulation, there
was no change observed. The resistivity of the drilling fluids observed was 12Ω for
both low and high temperatures. This means that temperature has no effect on the
emulsion property of the palm kernel oil-based system.
Conclusion
Up to date, drilling fluid manufacturers are moving towards a diverse range of

formulation for a variety field condition in achieving their goal of efficient drilling
operation. This includes the formulation from palm resources, in particular palm
kernel oil. This study concluded that palm-based drilling fluid has a potential to be
used as based drilling fluid. Besides palm kernel oil-based drilling fluid is robust at
high temperature, usage of palm kernel oil will also allow manufacturers to develop
value-added drilling fluid formulations that meet the strongly increasing market
demand for environmental-friendly and sustainable sourcing of raw materials.
Acknowledgement The authors would like to thank Research Management Institute (RMI),
Universiti Teknologi MARA for the Excellent Fund (600-RMI/ST/DANA 5/3/Dst (57/2009)) for
funding the research project. Special grateful to the Ministry of Education, Malaysia (600-RMI/
ERGS 5/3 (30/2013)) for providing financial support. The authors wish to thank the referee for
making this a better paper.
[email protected]
References
1. Growcock F, Harvey T (2005) Drilling fluids processing handbook. Elsevier Inc., Oxford
2. Ismail AR, Khor SF (2001) Feasibility study of palm oil esters as based fluid in drilling
operation. In: Regional conference for young scientists, Skudai, Malaysia, pp 23–28
3. Burke CJ, Veil JA (1995) Potential environmental benefits from regulatory consideration of
synthetic drilling muds. In: Environmental assessment division. Argonne National Laboratory,
Illinois
4. Veil JA, Burke CJ, Moses DO (1995) Synthetic drilling fluids—a pollution prevention
opportunity for the oil and gas industry. In: Annual conference and exposition of the water
environment federation, Argonne National Laboratory, Illinois, Miami Beach, Florida
5. Salleh MK, von Tapavicza S (2005) Palm oil derived esters—an environmentally safe drilling
fluid. Oil Palm Ind Econ J (5)
6. Corley RHV (2009) How much palm oil do we need? Environ Sci Policy 1:134–139
7. Reijnders L, Huijbregts MAJ (2008) Palm oil and the emission of carbon-based greenhouse
gases. J Clean Prod 16:477–482
8. Tapavicza SV (2005) Special report: vegetable esters make drilling fluids more
environmentally friendly. In: Cognis Deutschland GmbH & Co. KG, Düsseldorf, Germany,
2005
9. Al-Bagoury M, Steele CD (2012) A new, alternative weight material for drilling fluids. In:
IADC/SPE drilling conference and exhibition, San Diego, California, USA
10. American Petroleum Institute (2009) Recommended practice for field testing water-based
drilling fluids. In: ANSI/API recommended practice RP 13B-1, 4th edn. American Petroleum
Institute
11. Ahmad S (2000) Non-food uses of palm oil and palm kernel oil. MPOPC Palm Oil Inf Ser.
MPOB, Kuala Lumpur, p 24
12. American Petroleum Institute (1996) Recommended practice for drilling fluid processing
systems evaluation. In: API recommended practice 13C, 2nd edn. American Petroleum
Institute
13. Amani M, Al-Jubouri M, Shadravan A (2012) Comparative study of using oil-based mud
versus water-based mud in HPHT fields. Adv Pet Explor Dev 4:18–27
14. Azar JJ, Samuel GR (2007) Drilling engineering. PennWell Corporation, Oklahoma
15. Azizi A, Ibrahim MSN, Hamid KHK, Sauki A, Ghazali NA, Mohd TAT (2013) Agarwood
waste as a new fluid loss control agent in water-based drilling fluid. Int J Sci Eng 5(2):101–105
16. Ismail AR (2001) Managing the environmental friendly drilling fluids in petroleum industries.
In: The 2nd international conference on disaster management, Surabaya, Indonesia
17. Amani M, Al-Jubouri M, Shadravan A (2012) Comparative study of using oil-based mud
versus water-based mud in HPHT fields. Adv Pet Explor Dev 4(2):18–27
18. A.S.S. Committee (2005) Drilling fluid processing handbook. Gulf Professional Publishing,
Oxford
[email protected]
The Study of Temperature Profile
and Syngas Flare in Co-gasification
of Biomass Feedstock in Throated
Downdraft Gasifier
Muddasser Inayat, Shaharin A. Sulaiman, Aini Abd Jamil,

Fiseha M. Guangul and Samson M. Atnaw
Abstract Biomass gasification is a common technology, which converted solid

biomass into gaseous fuel at high temperature reactions in the presence of gasifi-
cation agent. In this paper, co-gasification of lignocellulosic biomass materials with
oil palm fronds (OPF) in a downdraft gasifier is presented. The biomass feedstocks
considered were sugar cane bagasse (SCB) and wood (acacia mangium). Only one
material was co-gasified with OPF at a time, with blending ratios of 80:20, 50:50
and 20:80. The resulting temperature profiles in the reactor and the syngas flare
duration were recorded. It was found that the blend of 80:20 wood and OPF gave
the best result as it produced the longest steady flare duration (49.5 min). On the
other hand, a significant bridging problem was observed in the co-gasification OPF
and SCB, and thus implying the need for process improvement.
Keywords Co-gasification Downdraft gasifier Blending ratio Temperature

profile
M. Inayat (&) S.A. Sulaiman A. Abd Jamil F.M. Guangul

Department of Mechanical Engineering, Universiti Teknologi PETRONAS, 32610 Bandar
Seri Iskandar, Perak Darul Ridzuan, Malaysia
S.A. Sulaiman
A. Abd Jamil
F.M. Guangul
e-mail: fi[email protected]
S.M. Atnaw
Faculty of Technology, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 23600 Kuantan,
Pahang, Malaysia

[email protected]
204 M. Inayat et al.
Introduction
Fossil fuel reserves are depleting due to growth in population and human activities.
According to a forecast by the International Energy Agency (IEA), the demand for
energy by 2050 is expected to grow by three folds as compared to the current
consumption [1]. As a result, there is need to generate power from alternative
sources such as biomass [2]. Malaysia is bestowed with a huge quantity of biomass
resources due to its location near the equator with high precipitation rate and long
sunny days throughout the year. This makes biomass the most prominent source of
energy among other renewable sources in Malaysia. Forestry wood, oil palm
industry, and rubber cultivation are the main sources of the biomass waste stream in
Malaysia [3]. Although utilization of biomass as sources of renewable energy have
a many benefits, it has also several issues such as interruption and discontinuity in
feedstock supply [4]. The supply of some biomass materials can be affected by
natural disasters such as flood and heavy rain. These aforementioned issues could
be addressed through co-gasification.
The performance of gasification and quality of syngas depends upon numerous
factors such as gasifying medium (air, steam or O2), heating rate, temperature and
equivalence ratio. In addition, the characteristics of the feedstock, i.e., the elemental
composition, heating value, fixed carbon, volatile matters, moisture content and ash
content would determine the performance of gasification [5]. In gasification,
the temperature is a key factor that directly affects syngas composition [6]. High
gasification temperature favors increase in gas yield due to a high decomposition of
cellulose and hemicellulose. Steam reforming and water-gas reaction favor higher
H2 concentration and lower CH4 concentration in the syngas for gasification tem-
perature of between 750 °C and 800 °C. At a higher temperature between 850 °C
and 900 °C, boudouard reaction favors higher CO concentration in the syngas [7].
The aim of this paper is to study the performance of co-gasifying different
biomass feedstock, namely wood, oil palm fronds and sugar cane bagasse. The
study involved measurements of the temperature profile in the reactor and nature of
the flare resulted from the syngas produced.
Characterization of Feedstock
Oil palm fronds (OPF), wood and sugar cane bagasse (SCB) were selected as
feedstock in the study. The oil palm fronds were collected from the FELCRA palm
oil plantation in Bota Kanan, approximately 220 km to the Northwest of Kuala
Lumpur. Wood (acacia mangium) was collected from the landscape area within the
campus of Universiti Teknologi PETRONAS Malaysia. SCB was collected from
small scale sugar cane juicer hawker nearby the campus area.
The bagasse was chopped to the length of 25–40 mm, while, OPF and wood
were cut into small blocks with a maximum size of 25 mm. Feedstock was dried in
[email protected]
The Study of Temperature Profile and Syngas Flare … 205
Table 1 Characterization results of wood, OPF and SCB

Biomass Ultimate analysis (wt%) Proximate analysis HHV kJ/kg
(wt% on dry basis)
C H Oa N S MC VM FC Ash
Wood 43.54 3.59 51.70 1.00 0.16 4.25 88.07 10.77 1.16 17,526
OPF 42.57 5.71 51.02 0.42 0.29 6.15 80.55 16.43 3.02 17,045
SCB 42.02 4.01 51.95 1.28 0.74 20.67 65.30 11.35 2.69 17,486
a
By difference basis. MC Moisture content; VM Volatile matters. FC Fixed carbon. HHV Higher
heating values
an oven at 105 °C for 24 h. The maximum moisture content was 12 % on wet basis.
Ground 250 μm sieved feedstock samples were used for characterization tests.
A Leco CHNS-932 analyzer was used to determine carbon, hydrogen, nitrogen and
sulfur contents of the feedstock according to ASTM D3176-89 standard [8]. The
proximate analysis, for determination of moisture content, volatile matter, fixed
carbon and ash, was performed according to ASTM E1131-98 standard test method
by using STA 6000 TGA analyzer [9]. For measurements of higher heating value
(HHV), a Leco AC-350 bomb calorimeter was used. The calorific value was
measured through measurement of the heat release after combustion of the samples.
Characterization results of wood, OPF and SCB are presented in Table 1.
Experimental Set-Up
The experimental set-up is shown in Fig. 1. For temperature measurements, ther-

mocouples T1–T7 were mounted as shown in Fig. 1, and were recorded every 30 s
by the data logger and computer. A pre-mixed blend of feedstock was fed into the
gasifier through feeding holes. Ignition was started through the ignition hole by
using scrap paper. After steady combustion was achieved, all lids were closed and
air was supplied by blower at 400 L/min. The air flow rate was regulated and
measured by using a valve and a rotameter. The syngas was ignited at the flare
point. Char and ash were collected from the ash box and on the grate after the
process was ended and were then weighed.
The duration of the process to fully consume 12.8 kg blend of OPF (80 %) and wood
(20 %) was 90 min. The temperature variation with time for different zones inside the
gasifier is shown in Fig. 2. The flare was obtained only after 34 min from the start
time of the experiment, and the combustion zone temperature was recorded to be
504 °C. Continuous and stable flare was obtained for about 44 min. The maximum
[email protected]
Portable Thermocouple
Pc
Data Logger
Feeding
Hole
Thermocouple
Flare
T1
Downdraft
Gasifier
T2
T3
Rotameter Air inlet O150 mm
2000 mm
T4
Blower
T5
T7
T6 Cyclone
Ignition Hole Particulate and
Ash Port
O300 mm Condensate
Collecctor
2000 mm
Fig. 1 Experimental set-up
Fig. 2 Temperature profile for 80 % OPF and 20 % wood with time
temperature recorded at the flare point was 719 °C. A bright yellowish flare was
obtained at the start of gasification due to the excess of hydrocarbon contents present
in the syngas. However, the color of flare changed to light yellowish near the end of
gasification. It was noted that, during the last 10 min of gasification, the temperature
of the combustion zone increased to a peak value of 1112 °C. The significant increase
in combustion zone temperature near the end of the operation is attributed to the
[email protected]
Fig. 3 Temperature profile for co-gasification of 50:50 OPF and wood blend with time
batch fed nature of the gasification process, whereby most part of the volatile matter
components will be consumed during the steady operation duration, leaving com-
bustion to be the only dominant operation for the last few minutes of the operation,
causing the increase in temperature. The average temperatures in the combustion and
reduction zones were 766 and 545 °C, respectively, during 34–78 min (marked by
region between vertical lines) of gasification. The average temperature of syngas at
the outlet was 323 °C. Similar to that reported by Guangul et al. [6] bridging was
observed at start of experiment, probably due to the fibrous nature of OPF feedstock
and this was overcome by stirring the feedstock inside the gasifier as soon as a drop
in temperature in the reactor was detected.
The temperature variation for different zones with time for co-gasification of OPF
(50 %) and wood (50 %) is shown in Fig. 3. The experiment took 79 min, which was
shorter than that, for the mixture of 80 % OPF and 20 % wood. The syngas could
ignite after 27 min from the start of the experiment and combustion zone temperature
was reached 609 °C. The resulting flare was stable for 39 min. The maximum
temperature of the flare at top was 744 °C. The flare was bright yellowish at the start
of the process, changed to light yellow-bluish towards the end of the experiment.
The average temperatures of the combustion and reduction zones were 732 and
469 °C, respectively during 27–78 min of gasification. The maximum combustion
zone temperature was 1096 °C and sustained for 10.5 min of the experiment. The
significant increase in combustion zone temperature near the end of the operation
was similar to the Fig. 2. The average temperature of syngas outlet was 272 °C.
Different that the experiment of OPF (80 %) and wood (20 %), no bridging problem
was observed during the co-gasification of 50 % OPF and 50 % wood mixture.
The temperature variation for different zones of the gasifier with time for co-
gasification of 20:80 OPF and wood is shown in Fig. 4. The experiment took
94 min, which was longer than that all blends. The flare could ignite after 28 min of
experiment start and sustained for 49.5 min, average combustion zone temperature
at instant was recorded as 741 °C. The maximum temperature recorded at the top of
flare point was 719 °C. The colour of flare was bright yellow at the start of
gasification, however, it changed to light yellowish near the end of gasification. The
maximum combustion zone temperature was 1078 °C and sustained for 10 min near
[email protected]
Fig. 4 Temperature profile for co-gasification of 20:80 OPF and wood blend with time
Fig. 5 Temperature profile for co-gasification of 80:20 OPF and SCB blend with time
the end of the experiment it was similar to previous experiments. The average
temperature of combustion and reduction zones was 606 and 398 °C, respectively,
during the 28–77.5 min of the experiment. The average temperature of the syngas at
the outlet point was 315 °C. Throughout the experiment, there was no bridging
problem observed.
Co-gasification of OPF and SCB for all ratios was found to be difficult in
sustaining stable temperature profile and flare due to recurrence of bridging prob-
lem. It was observed that, as a ratio of SCB increases in blend bridging interval
increased and gasification time decreased. It was due to SCB have a low density
(2.16 g/cm3) light weight and higher specific surface area (4716 cm2/g) causes
bridging. The blend of 80:20 OPF and SCB (Fig. 5) encountered bridging twice.
For blends of 50:50 (Fig. 6); and 20:80 OPF and SCB (Fig. 7) encountered bridging
three times throughout the experiment. Therefore, to avert the problem, opened the
gasifier and, as a result, an abrupt decrease in the combustion zone and other zones
temperature were experienced. This bridging problem was due to light weight and
fibrous nature of the SCB. Hence, SCB in the form used in this work would not be
suitable for downdraft gasifier. In the co-gasification of 80:20 OPF and SCB,
an unstable flare was found only for 4.5 min. Whereas, in co-gasification of 50:50
(Fig. 6); and 20:80 OPF and SCB (Fig. 7) flare was not obtained at all.
[email protected]
Conclusion
OPF was co-gasified with wood and SCB in downdraft gasifier in order to inves-
tigate the effects of blending ratios on gasifier temperature profile, syngas flame
elapse time and co-gasification experiment time. All blends of OPF and wood
experiments temperature profile was found stable and average stable combustion
zone temperature was 766, 732 and 606 °C for 20:80, 50:50 and 80:20 OPF and
wood blends, respectively. On the contrary, all blends of OPF and SCB encoun-
tered the bridging problem and unstable temperature profile. Maximum 49.5 min
flare was recorded for 20:80 OPF and wood blend followed by 44 and 39 min for
80:20 and 50:50 OPF and wood blends, while, only 4.5 min flare was obtained for
80:20 OPF and SCB blend, while, 50:50 and 20:80 OPF and SCB blend no flare at
all. High wood ratio in blends increased the process time and flares ignition time
and sustained longer, while, high SCB ratio in blends increased the interval of
bridging and decreased process time.
Acknowledgment The authors acknowledge a grant-in-aid for research from “Exploratory

Research Grant Scheme (ERGS) No. 0153AB-I13.” Malaysia.
[email protected]
References
1. International Energy Agency (2013). Bioenergy. Available www.iea.org/topics/bioenergy

2. Saidur R, Abdelaziz EA, Demirbas A, Hossain MS, Mekhilef S (2011) A review on biomass as
a fuel for boilers. Renew Sustain Energy Rev 15:2262–2289
3. Shafie SM, Mahlia TMI, Masjuki HH, Ahmad-Yazid A (2012) A review on electricity
generation based on biomass residue in Malaysia. Renew Sustain Energy Rev 16:5879–5889
4. Ong HC, Mahlia TMI, Masjuki HH (2011) A review on energy scenario and sustainable energy
in Malaysia. Renew Sustain Energy Rev 15:639–647
5. Skoulou V, Zabaniotou A, Stavropoulos G, Sakelaropoulos G (2008) Syngas production from
olive tree cuttings and olive kernels in a downdraft fixed-bed gasifier. Int J Hydrogen Energy
33:1185–1194
6. Guangul FM, Sulaiman SA, Ramli A (2012) Gasifier selection, design and gasification of oil
palm fronds with preheated and unheated gasifying air. Bioresour Technol 126:224–232
7. Kumar A, Jones DD, Hanna MA (2009) Thermochemical biomass gasification: a review of the
current status of the technology. Energies 2:556–581
8. ASTM (2002) ASTM D3176-89: standard practice for ultimate analysis of coal and coke. In:
Annual book of ASTM standards. ASTM, West Conshohocken, PA
9. ASTM (2004) Annual book of ASTM standards, section five: petroleum products, lubricants,
and fossil fuels, gaseous fuels, coal and coke, vol 05.06. ASTM, West Conshohocken, PA
[email protected]
Evaluation of Energy Cost Saving
and Pollutants Emission Reduction
for Solar Water Heater Development
in Malaysia
Mohammed J. K Bashir, Ong Li Jing, Sumathi Sethupathi

and Ng Choon Aun
Abstract A potential solution to curb the environmental problems is the utilization

of renewable energy, including solar energy. Malaysia has abundant solar energy
resources throughout the year, promises the highest beneficial in using solar energy
for technology development such as solar water heater (SWH). However, peoples in
Malaysia are not aware of SWH’s advantages due to the high initial installation cost
of SWH. Thus, this study presents financial evaluation of SWH in Malaysia which
takes into account of the cost saving and pollution prevention by SWH instead of
electrical water heater (EWT). The benefit of SWH was analysed based on annual
effective solar radiation. From the analysis, the utilization of a SWH can save up to
RM 708.3 of energy cost per year and avoid 1583 kg CO2, 12 kg SO2 and 5 kg NOx
emissions per year.
Keywords Solar energy Emissions reduction Cost saving Malaysia
Introduction
The usage of electricity in Malaysia has extremely increased in last few years. This
is because of the technological development, population growth, economic and
industrial activities [1]. To meet the increased demand for electricity, the electricity
M.J.K. Bashir (&) O.L. Jing S. Sethupathi N.C. Aun

Faculty of Engineering and Green Technology, Universiti Tunku Abdul Rahman Jalan
Universiti, Bandar Barat, 31900 Kampar, Perak, Malaysia
O.L. Jing
S. Sethupathi
N.C. Aun

[email protected]
212 M.J.K. Bashir et al.
generation in Malaysia increases almost 100 % in 10 years from 2000 to 2010.

Indeed, Malaysia is highly depending on fossil fuels to produce electricity [2]. In
2010, fossil fuels, such as crude oil, coal and natural gas, generate almost 93 % of
electricity in Malaysia [1, 3, 4].
The usage of fossil fuels releases various gases such as greenhouse gases
(GHGs), carbon monoxide (CO), nitrogen oxide (NOx), sulphur dioxide (SO2) and
particulate matter (PM). These gases can cause harmful effects to the environment
such as smog, ozone depletion, acid rain and global warming [1, 5, 6]. A good
possible solution to limit this environmental problem is the utilization of renewable
energy. Currently, many countries considered solar, wind, hydro, bio-energy and
other renewable energy to create a clean energy future [6, 7]. The benefits derive
from the utilization of renewable energy fall into three categories: conservation of
conventional energy, pollution mitigation and generation of working opportunities
[6]. Solar energy is costless energy source, clean and naturally accessible source
[8]. There are several solar energy technologies include solar water heater, solar air
conditioner and solar drying system [9].
Malaysia has great potential for solar energy resources in comparison with other
countries due to its location near the equator. The average daily solar radiation is
18 MJ/m2, which is equivalent to 5 kWh/m2 and the average sunshine duration
about 12 h [8, 10, 11].
Saidur et al. [1] reported that about 985,704 households in Malaysia will own
EWH by 2013. The usage of EWH can increase the electricity bill monthly and
indirectly cause the emission of GHGs and air pollutants to atmosphere as it is
operated by electricity which is produced from the combustion of fossil fuels in the
power plant [1]. SWH performance is highly dependent on the climatic condition.
The climatic condition in Malaysia is favourable for the development of SWHs in
household and industrial applications. However, many families in Malaysia are still
using EWH to generate hot water. This is due to easy installation and inexpensive
installation cost of EWH compared to SWH, and on the other hand less public
awareness and understanding of potential benefits. As a result, SWH does not
attract attention from the general public yet [11].
Thus, the purpose of this study is to analyse the applicability of SWH in
Malaysia, by implementing the financial evaluation of SWH which takes into
account of cost saving and gas emission reduction by using SWH. Indeed, for a
SWH, if the daily solar radiation on today exceeds the amount of solar radiation
needed to heatwater, the excess solar radiation cannot be stored for later use,
therefore it is necessary to consider the solar radiation that used effectively to heat
water when analysing the true cost benefit of SWH development.
[email protected]
Evaluation of Energy Cost Saving and Pollutants Emission Reduction … 213
Methodology
Study area, The regional and seasonal solar radiation variation in Malaysia has to
be considered in order to analysis the significant effect of solar radiation variation
on the applicability of SWH in Malaysia. The variability of data can be measured
by the following equation:
r
CV ¼ ð1Þ
l
where CV = coefficient of variation; σ = standard deviation of data; µ = mean of

data.
To determine regional solar radiation variation, Kuala Lumpur as a study area
was selected for this study and the daily solar radiation variation data throughout
the year was obtained from National Aeronautics and Space Administration
(NASA) climate internet database [12].
Effective Solar Radiation (ER), The daily solar radiation data extracted was
compared with the minimum solar radiation required by SWH. The minimum solar
radiation required was calculated based on the temperature of tap water and the
design specifications of SWH by Eq. 2 [13]:
V q C DT
SRmin ¼ ð2Þ
cSWH A
where SRmin = minimum solar radiation required, kWh/m2; V = volume of SWH

storage tank, L; ρ = density of water, kg/L; C = specific heat capacity of water, MJ/
kg °C; DT = temperature difference between the desired water temperature and tap
water, °C; γSWH = collector efficiency of SWH; A = collector surface area, m2.
Table 1 lists the SWH specification assumptions considered in this study.
If the daily solar radiation is higher than the minimum solar radiation required,
the excessive solar radiation cannot be stored for later use. Therefore, the ER used
by SWH on that day will be equal to SRmin. In contrast, if the daily solar radiation is
lower than the minimum solar radiation required, the ER used by SWH will be
equal to the solar radiation on that day. The annual ER was then calculated by
following equation [13]:
For DRd [ SRmin ; ERd ¼ Smin ð3Þ
Table 1 Specification of Parameter Unit Value

SWH [13]
Tank volume L 250
Collector surface area m2 4
Collector efficiency % 50
Desired hot water temperature °C 45
[email protected]
Table 2 Pollutant emissions based on various energy sources [4, 14]

Energy Percentage in electricity generation mix on Emission factor (kg/kWh)
sources 2012 (%) CO2 SO2 NOx
Coal 38.9 1.18 0.0139 0.0052
Crude oil 1.0 0.85 0.0164 0.0025
Gas 52.7 0.53 0.0005 0.0009
Hydro 7.3 0.00 0.0000 0.0000
others 0.2 0.00 0.0000 0.0000
For DRd [ SRmin ; ERd ¼ DRd

X
Y ð4Þ
AR ¼ ERd
d¼1
where DRd = solar radiation on day d, kWh/m2; ERd = effective solar radiation on
day d, kWh/m2; AR = annual effective solar radiation, kWh/m2.
Table 2 shows the emission factors of different energy sources and the per-
centage of energy source in electricity generation mix on 2012. The emissions of
pollutant from the combustion of fossil fuels for electricity generation are dependent
on the characteristics of the energy source. Different source has different contents of
carbon, sulphur and nitrogen or their compounds [1].
Benefit Analysis of SWH, The benefit of SWH was calculated based on the
solar radiation in Malaysia which includes cost savings by replacing energy and
pollution mitigation.
The cost saving by replacing energy can be estimated by the amount of energy
generated by SWH under annual ER and the price of replaced energy. The replaced
energy is electricity with unit price of Ringgit Malaysia (RM) 0.334/kWh [15]. The
cost saving by replacing energy was calculated by Eqs. 5 and 6:
BE ¼ AR A cSWH ð5Þ
CBE ¼ BE P ð6Þ
where BE = total energy reduction, kWh; CBE = cost saving by replacing energy,
RM; P = unit price of replaced energy, RM/kWh.
The cost saving from pollution mitigation can be estimated by reduction of
GHGs and air pollutants emitted quantities once using SWH and the price of
treating those pollutants. Pan et al. [13] reported that the prices for treating 1000 kg
of CO2, NOx and SOx are approximately RM 80, RM 2698.5 and RM 2624.2,
respectively. The cost saving by pollution mitigation was calculated by Eqs. 7–8:
[email protected]
BPp;f ¼ BE EFp;f PFp;f ð7Þ

X
CBPf ¼ p
ðBPp;f TEp Þ ð8Þ
where EFpf = emission factor of pollutant p by energy source f, kg/kWh;

PEf = percentage of energy source f in mix; BPpf = amount of GHGs and air
pollutant p that reduces (e.g. CO2, NOx and SOx) when SWH avoids using elec-
tricity produced by energy source f, kg; CBPf = cost avoided to treat GHGs and air
pollutants emissions when SWH avoids using electricity produced by energy source
f, RM; TEp = unit price for treating pollutant p, RM/kg. Both EFpf and PEp values
are presented in Table 2.
The daily solar radiation data of study area was collected from National Aeronautics
and Space Administration (NASA) climate internet database [20]. The variability of
data was measured by coefficient of variation (CV), as in Eq. 1. If CV is low (<1),
the data have low variability and vice versa. From the data analysis, the annual solar
radiations at the study area have low CV (0.18). Thus, the daily solar radiations
during the year have low variability which indicates that the applicability of SWH
in Malaysia is not affected by seasonal solar radiation variation.
Benefit analysis, The benefit of using a SWH was analysed based on amount of
energy that generated under solar radiation includes cost savings by replacing
energy and pollution mitigation. The annual ER of the five study areas is roughly
60 % of its annual solar radiation, which means that only 60 % of the annual solar
radiation is effectively used by SWH to produce hot water. This is due to Malaysia,
which being hot and humid, has high tap water temperature throughout the year and
resulted in low SRmin. Besides, the average daily solar radiation in Malaysia is
5.0 kWh/m2, which is higher than SRmin. This indicated that most of the daily solar
radiation in Malaysia exceeds SRmin, as shown in Fig. 1.
Cost saving by replacing energy, Cost saving by replacing energy was
determined via Eq. 6 [13], based on the amount of energy generated under annual
ER once SWH replace electricity with solar energy to produce hot water. Based on
Fig. 1 Daily solar radiation

of study areas
[email protected]
Fig. 2 Carbon dioxide reduction in KL region
Fig. 3 Sulphur dioxide and nitrogen oxide reduction in KL region
Fig. 4 Cost avoided for reducing GHGs and air pollutants emission in KL region
calculation, a typical SWH can save an average RM 708.3 of energy cost per year,
which equivalent to RM 59 per month.
Cost saving by pollution mitigation, Cost saving by pollution mitigation, based
on the reduction in amount of GHGs and air pollutants when using SWH and the
price of treating those GHGs and air pollutants was estimated based on Eq. 8 and
Table 2.
Figure 2 and 3 show the estimated pollutants emissions reduction from various
energy sources (e.g., coal, gas, crude oil) in KL region, once the SWH employed
instead of EWH. Considering the estimated cost for treating 1 metric ton of CO2,
NOx and SO2 as approximately of RM 80, RM 2698.5 and RM 2624.2, respec-
tively, the cost avoided for reducing GHGs and pollutants emissions were esti-
mated. Thus, the avoided electricity generation from coal, oil and gas can save RM
120, RM 2.5 and RM 51.5 of treating cost per year as shown in Fig. 4. Although
[email protected]
CO2 has much lower unit treating cost than those of other air pollutants, however
for the three main energy sources, the cost avoided for reducing CO2 was the
highest. This is because the reduction in CO2 emission is the highest among all air
pollutants when replacing electricity with solar energy by the utilization of SWH.
Conclusion
In this study, financial evaluation of SWH in Malaysia was carried out which take
into account the SWH’s benefits. By applying SWH, the use of electricity for
heating water can reduce, consequently lower the electricity bill and reduce the
GHGs and pollutant emissions, produced by electricity generation of energy saving.
It was found that the installation of a SWH can save an average 2120.70 kWh of
electricity which equivalent to RM 708.3 of energy cost and avoid an average
1583 kg CO2, 12 kg SO2 and 5 kg NOx emissions. It can be concluded that SWH is
feasible to be used in Malaysia. However, To promote the widespread uses of SWH
in Malaysia, the attentions from both government and public are required such as
introduce policies by government that encourage people to use SWH. For example,
government should provide subsidies to SWH’s owner. Beside, a media seminar
can also conduct to increase the awareness and understanding of public towards the
benefits of SWH.
References
1. Saidur R, Masjuki HH, Jamaluddin MY, Ahmed S (2007) Energy and associated greenhouse
gas emissions from household appliances in Malaysia. Energy Policy 35:1648–1657
2. Shafie SM, Mahlia TMI, Masjuki HH, Andriyana A (2011) Current energy usage and
sustainable energy in Malaysia: a review. Renew Sustain Energy Rev 15:4370–4377
3. Ali R, Daut I, Taib S (2012) A review on existing and future energy sources for electrical
power generation in Malaysia. Renew Sustain Energy 16:4047–4055
4. EPU (Economic Planning Unit) (2013) The Malaysian economy in figures at http://www.epu.
gov.my/documents/10124/72ac36d7-fe5a-489b-a34c-a2cb2be073a6
5. Kaldellis JK, El-Samani K, Koronakis P (2005) Feasibility analysis of domestic solar water
heating systems in Greece. Renew Energy 30:659–682
6. Kalogirou SA (2004) Environmental benefits of domestic solar energy systems. Energy
Convers Manag 45:3075–3092
7. Al-Badi AH, Albadi MH (2012) Domestic solar water heating system in Oman: Current status
and future prospects. Renew Sustain Energy Rev 16:5727–5731
8. Mekhilef S, Safari A, Mustaffa WES, Saidur R, Omar R, Younis MAA (2012) Solar energy in
Malaysia: current state and prospects. Renew Sustain Energy Rev 16:386–396
9. Ibrahim A, Jin GL, Daghigh R, Salleh MH, Othman MY, Ruslan MH (2009) Hybrid
photovoltaic thermal (PV/T) air and water based solar collectors suitable for building
integrated applications. Am J Environ Sci 5(5):618–624
10. Chua SC, Oh TH (2012) Solar energy outlook in Malaysia. Renew Sustain Energy Rev
16:564–574
[email protected]
11. Ong HC, Mahlia TMI, Masjuki HH (2011) A review on energy scenario and sustainable
energy in Malaysia. Renew Sustain Energy Rev 15:639–647
12. NASA (National Aeronautics and Space Administration) (2013) NASA prediction of
worldwide energy resource (POWER) near real-time daily global radiation and meteorology.
http://power.larc.nasa.gov/cgibin/cgiwrap/solar/timeseries.cgi
13. Pan TC, Kao JJ, Wong CP (2012) Effective solar radiation based benefit and cost analyses for
solar water heater development in Taiwan. Renew Sustain Energy Rev 16:1874–1882
14. Mahlia TMI (2002) Emissions from electricity generation in Malaysia. Renew Energy 27:293–
300
15. Tenaga Nasional Berhad (2013) Pricing and Tariff. http://www.tnb.com.my/residential/
pricing-and-tariff.html
[email protected]
Part IV
Process Control, Modelling, Simulations,
Data Mining
[email protected]
Introducing Pandanaceae Via IOS-Based
Software Application
Hannis Fadzillah Mohsin, Ibtisam Abdul Wahab

and Galoh Rashidah Haron
Abstract Pandanaceae is an application, based on the iPhone Operating System

(iOS) that introduces the screwpines or pokok pandan, within a hand click. It
describes both medicinal and traditional uses, chemical constituents, scientific
references and geodata of this Asia Pacific plant species. For the purpose of the
application prototype, the user is introduced to eight selected plants within the
Pandanus genus of the Pandanaceae plant family. Fascinating pictorial views of
the whole Pandanus plant and the major plant parts are not to be missed. From the
users’ perspective, they can select the image of one specific Pandanus species and
the scientific information on the biological and pharmacological properties of that
particular plant will be displayed. The researchers’ years of experiences on sam-
pling the Pandanus and the laboratorial work can be shared via this software,
interactively. The molecular structures of the chemical ingredients of the Pandanus
extracts are also shown. Consequently, Pandanus research profiling can be con-
cluded; which touches on the scientists who are conducting experiments on this
plant and their area of studies. Owing to the lack of information regarding
Pandanaceae in a mobile application format, the concept is presently materialized.
Keywords Iphone Pandanaceae
H.F. Mohsin (&) I. Abdul Wahab

Faculty of Pharmacy, UiTM, 42300 Puncak Alam, Selangor Darul Ehsan, Malaysia
I. Abdul Wahab
G.R. Haron
MIMOS Berhad, Technology Park Malaysia, 57000 Kuala Lumpur, Malaysia

[email protected]
222 H.F. Mohsin et al.
Introduction
Pandanaceae is a plant family, which consists of more than six hundreds species. It
is also selected as a product label, given to a software application, which is designed
based on the iPhone Operating System (iOS). This application firstly introduces a
limited number of the most common pandan species (also known as the screwpines)
to the natural product lovers, within a hand click. It also describes both medicinal
and traditional uses, chemical constituents, scientific references and geodata (geo-
graphical locations) of this Asia Pacific plant species. The user of this application is
anyone who has the interest in learning such local plant. From a curiosity element, a
user can join the group of researchers and share the knowledge. Without having a
book, a digital library concerning pandan’s facts and images are reachable.
Comparable applications on other plant herbs could also be witnessed [1–3], and
now, local researchers’ years of experiences on sampling the Pandanus and
experimental work in the laboratory [4] can be shared via this application. The
usability of this prototype is currently described.
Experimental
The Pandanaceae application is developed based on the iOS, using Xcode software
(version 5.0), on a MacBook Pro, a Macintosh portable computer. For the purpose
of the application prototype, the user is introduced to eight selected plants within
the same Pandanus genus of the Pandanaceae plant family (Table 1). The selection
criteria are based on the most studied samples by local researcher [5, 6]. The
Pandanus plants are available locally; it can be definitely witnessed in the botanical
gardens, university campus and also planted as a landscaper along the roads. They
are ranging from a miniature species to a larger one, which can grow up to fifteen
meters in height. It is noteworthy that certain Pandanus are synonymous. For
example, Pandanus amaryllifolius is also botanically recognized as Pandanus
odorus [7].
Table 1 The selected No. Pandanus sample Local name

Pandanus plant for the
prototype 1 Pandanus pygmaeus Yellow pandan
2 Pandanus sanderi Unknown
3 Pandanus tectorius Unknown
4 Pandanus amaryllifolius Fragrant pandan
5 Pandanus dubius Mengkuang
6 Pandanus conoideus Red pandan fruit
7 Pandanus leram Nicobar breadfruit
8 Pandanus odoratissimus Unknown
9 Unidentified Pandanus Unknown
[email protected]
Introducing Pandanaceae Via IOS-Based Software Application 223
Fig. 1 The photograph of a Pandanus sample could be magnified, once the upmost left tile, is
selected (left). The ninth tiles are dedicated for a collection of unknown Pandanus (right)
Fascinating pictorial views of the whole Pandanus plant and the major plant parts
are successfully included (refer to Fig. 1). The screenshots display the characteristic
four-fold leaves of this plant. In addition, the stilt roots are also the major botanical
features for this species. Some of them display variegated leaves of white, yellow
and green stripes. Furthermore, some photographs of unknown wild Pandanus are
also included in this application (Fig. 1), in a hope to cultivate world web dis-
cussions for the purpose of the plant identification. From the users’ perspective,
they can select the image of one specific Pandanus species and the information on
the biological and pharmacological properties of that particular plant will be dis-
played. The user will have a flow of information of the species, from the root up to
the leaves of the species, interactively.
The molecular structures of the chemical ingredients of the Pandanus extracts
are also shown (refer to Fig. 2). Miscellaneous information that can be reached
includes the scientific evidences such as journals and other publications of the
biopharmaceutical properties of Pandanus. Consequently, Pandanus research
profiling can be concluded; which touches on the scientists who are conducting
experiments on this plant and their area of studies. A wide range of investigation is
involved, that includes Pandanus’ conservation proposals, taxonomic research,
ethnobotany and social studies, phytochemical and chemosynthetic work, myco-
logical and zoological observations of Pandanus.
[email protected]
Fig. 2 The icons are arranged accordingly for a specific Pandanus, which begin with its
photographic magnification, geodata, chemical constituents, microscopic observations and
scientific publications related to that sample (left). Example of a molecular structure was shown
(right)
Fig. 3 The location of a Pandanus sample could be identified (left). The home button will guide
the users back to the first window (right)
[email protected]
Introducing Pandanaceae Via IOS-Based Software Application 225
As shown in Fig. 3, the user could identify the location of a Pandanus sample,
once the icon for the map is clicked, thus introducing the national botanical garden
to the users. A window is also equipped with a home button, to guide users back to
the display of the eight Pandanus tiles, after looking and differentiating with the
unknown or wild Pandanus. In future, the user could also be able to turn for a
landscape view of this application. The notes to each Pandanus species could be
added, too. The user could also upload own captured Pandanus figures and interact
with the researchers, in order to exchange information of discovered species. Other
potential options for users would include the language change (either national
language or English), without the internet access.
Conclusion
This application is designed for both iPhone and iPad. It is anticipated that this
prototype should be easy to use and could offer the learnability of natural products
to human. Built-in quizzes could also be proposed in order for users to quickly learn
and distinguish between different Pandanus species. Later, to accommodate
Android smart phone based users, this effort will be extended to Android operating
system. Android users will also experience this pandan walk-about in their devices.
Owing to the lack of information regarding Pandanaceae in a mobile application
format, the concept is presently materialized. Continuous efforts could be planned,
in order to digitalize the information, related to another hundreds of Pandanus
recorded species. Thus, a comprehensive and fuller version of this application is
recommended and physically, achievable. The prime motivation factor on pushing
such technologies is aimed at positioning Malaysian academic and research insti-
tutions that contribute to the Apple store.
Acknowledgment This work is supported by Ministry of Education (Research Acculturation

Grant Scheme) and Universiti Teknologi MARA (Principal Investigator Support Initiative Grant).
References
1. Alphablind Studio, Wild Berries & Herbs—Nature Mobile (2014) Information on https://itunes.
apple.com/us/app/wild-berries-herbs-nature/id435903166?mt=8
2. Natator Publishing, Herbology On The Go (2014) Information on https://itunes.apple.com/us/
app/herbology-on-the-go/id386509054?mt=8
3. Alphablind Studio, Medicinal Plants Pro—Nature Mobile By (2014) Information on https://
itunes.apple.com/us/app/medicinal-plants-pro-nature/id499824923?mt=8
4. Mohsin HF (2012) Chemical constituents of Pandanaceae species. PhD thesis, Universiti
Teknologi MARA, Malaysia
5. Armayni UA (2014) The isolation of alkaloids from Pandanus species. Master of Science
thesis, Universiti Teknologi MARA, Malaysia (2014) (in print)
[email protected]
6. Md Yusof Y (2013) The hexane extraction of Pandanus Conoideus. Bachelor of Pharmacy

Thesis, Universiti Teknologi MARA, Malaysia (2013)
7. Salim AA, Garson MJ, Craik DJ (2004) New alkaloids from Pandanus amaryllifolius. J Nat
Prod 67(1):54–57
[email protected]
Investigation of Heat Treated
Electrodeposited CoNiFe
on Microstructure and Hardness
Nor Azrina Resali, Koay Mei Hyie, M.N. Berhan and C.M. Mardziah
Abstract In this research, heat treatment is the final finishing process applied on
nanocrystalline CoNiFe to improve microstructure for good hardness property.
Nanocrystalline CoNiFe has been synthesized using the electrodeposition method.
This study investigated the effect of heat treatment at 500, 600, 700 and 800 °C on
electrodeposited nanocrystalline CoNiFe. The heat treatment process was per-
formed in the tube furnace with flowing Argon gas. In this paper, physical prop-
erties such as phase and crystallographic structure, surface morphology, grain size
and hardness of nanocrystalline CoNiFe was studied by changing the heat treatment
temperatures. The nanocrystalline CoNiFe phase revealed the Face Centered Cubic
(FCC) and Body Centered Cubic (BCC) crystal structure. FESEM micrographs
showed that the grain sizes of the coatings were in the range of 78.76–132 nm.
Dendrite shape was found in the microstructure of nanocrystalline CoNiFe. The
nanocrystalline CoNiFe prepared in heat treatment temperature of 700 °C, achieved
the highest hardness of 449 HV. The surface roughness of nanocrystalline CoNiFe
heated at 700 °C was found to be smaller than other temperatures.
Keywords CoNiFe Heat treatment Oxygen content Dendrite Hardness
N.A. Resali (&) K.M. Hyie M.N. Berhan C.M. Mardziah

Faculty of Mechanical Engineering, Universiti Teknologi MARA, 40450 Shah Alam,
Selangor, Malaysia
K.M. Hyie
M.N. Berhan
C.M. Mardziah

[email protected]
228 N.A. Resali et al.
Introduction
Engineering materials are heat treated under controlled sequence of heating and
cooling to alter their physical and mechanical properties to meet desired engi-
neering applications. Heat treatment is a combination of timed heating and cooling
applied to a particular metal or alloy in the solid state to produce certain micro-
structure and desired mechanical properties [1].
Generally, the goal of the sintering process is to produce a coherent body with
controlled microstructure, in some cases with controlled porosity [2]. Sintering
phenomena can be classified into three stages. Initial stage of sintering corresponds
to the situation when necks are forming and growing between grains as shown
schematically in Fig. 1. At the end of this stage, the contact area increases up to
20 % with only a small densification. A marked decrease in the specific surface area
of the compact occurs due to surface smoothing. The second stage of sintering is
characterized by a more or less continuous network of pore channels along the grain
edges. During this stage, the pore channel shrinks and grains grow. Grain growth
can be more extensive in the final stage and difficulties are commonly encountered
in the removal of porosity [3].
Electrodeposition is an electroplating process for producing a dense, uniform
and adherent coating. This process offer many advantages and unique possibilities
in the development of nanocrystalline materials with the low cost processing,
rapidity, higher deposition rate and etc. [4].
A number of electrodeposited alloys are of interest due to various properties.
Ternary alloys of CoNiFe are probably the most promising materials due to their
excellent magnetic and physical properties such as superior corrosion resistance,
low coercivity, and high resistivity. There are several factors including the crys-
tallite size, the crystal orientation, and the alloy composition, that can affect the
mechanical properties of these deposited films [5]. These properties also can be
affected by the finishing the coating with heat treatment process.
During heat treatment, the process parameters such as temperature play an
important role on properties of nanocrystalline CoNiFe. This is due to the grain
boundary diffusion and volume diffusion that depends on temperature, the size,
Fig. 1 Illustration of the sintering stage. a Formation of necks. b Evolution of necks and pores
diffuse through and accumulate at grain boundaries. c Some pores are eliminated and closed pores
appear
[email protected]
Investigation of Heat Treated Electrodeposited CoNiFe … 229
composition and distribution of grains of the material. Sintering process gives a

huge effect on the microstructure, surface roughness and hardness. This is proven
based on the previous researches [6, 7].
Al-Qura’n [7] studied the effect of heat treatment on the hardness of chromium-
nickel steel. The best result of microhardness test was for the samples annealed at
740 °C for a period of 60 min. The authors concluded that heat treatment improves
the quality in terms of microstructures and surface roughness and thus produced the
highest hardness. The applied heat treatment in the final stage of electrodeposition
nickel (EN) coating changes its properties significantly [6]. Mechanical properties
of EN coatings such as hardness and wear resistance are enhanced by heat treat-
ment. When the sintering temperature is increased, the grain size increases thus
make the pore size to decrease [8]. Besides, it is found that the hardness will
increase as the sintering temperature is increased. In contrast, the hardness will
decrease if the grains contain an oxide ion at high temperature.
Remarkable increasing of surface hardness of CoNiFe coating up to 200–
300 HV by elecrodeposition was reported [9, 10]. Among those surface modifi-
cation techniques, electrodeposition is the simplest method with low set-up cost.
However, low adhesion of CoNiFe presents the drawback of this route. The sub-
sequent heat treatment can potentially improve the adhesion of CoNiFe [11].
Furthermore, the intermetallic formed during heating can further enhance the
hardness of the CoNiFe. During heat treatment, the process parameters such as
holding time and temperature play an important role on phase formation of CoNiFe.
In the present work, nanocrystalline CoNiFe were heat-treated under different
temperatures to study the effect of heat treatment on the microstructure, grain size
and hardness.
Experimental
A layer of CoNiFe was applied on a stainless steel substrate by the electrodepos-

ition technique, using graphite as an anode [9, 10]. Figure 2 shows the schematic of
electrodeposition process consists of four elements which are power supply, sample,
graphite and solution. The aqueous bath consists of CoSO4, NiSO4, FeSO4, H3BO3,
and Saccharin. In this experiment, electrodeposition was conducted at 50 °C using a
direct current (DC) condition at 0.133 A/cm2. The holding time for the deposition
was 15 min. 15 min for deposition have ample time to produce fully coating. It is
believed that the coating have sufficient time for the grains to distribute and rear-
range themselves in the microstructure. If the deposition time increases, some of the
grains tend to fused, combining with each other and producing larger grains.
All samples were heat treated in a tube furnace under an argon atmosphere at
500, 600, 700 and 800 °C. The soaking time and heating rate in this experiment
were fixed at 45 min and 8 °C/minute, respectively. After the heat treatment cycle,
the samples were cooled in the furnace.
[email protected]
Following the electrodeposition and heat treatment steps, the samples were
qualitatively characterized by glancing incident angle X-ray diffractometer (XRD)
at the incident angle of 5°. Microstructures of the heat treated samples were
observed by a Field emission scanning electron microscope (FESEM). Hardness of
the samples was measured with a Vicker’s microhardness using 100 g loading.
Physical structure, phase and crystallite sizes, The physical structure, phase and
crystallite sizes of nanocrystalline CoNiFe were characterized by XRD analysis.
XRD patterns of all the nanocrystalline CoNiFe were measured from 2θ angle of
30° to 100° with Cu Kα radiation. XRD patterns of nanocrystalline CoNiFe on
different heat treating temperatures are shown in Fig. 2. All nanocrystalline CoNiFe
showed the similar XRD patterns.
The reflection patterns of these nanocrystalline CoNiFe reveal the characteristics
peaks of CoNi and FeNi. The nanocrystalline have a combination of face-centered
cubic (fcc) structure and body centered cubic (bcc) phase.
After heat treatment at 500 °C, some small peak of FeNi (111) can be observed
with low relative intensities. The peak with low intensities indicate that the volume
fraction of intermetallic phase (FeNi) formed by heat treatment under these con-
ditions are low. By increasing heat treatment temperature to 700 °C, higher relative
intensity peak of FeNi can be observed. This implies that the volume fraction of the
intermetallic formed by heat treatment at 700 °C is significantly higher than that
formed at 500 and 600 °C [12].
There were no significant changes in the XRD patterns at heat treating tem-
peratures below 700 °C, indicating that no phase transition occurred within this
Fig. 2 XRD patterns of nanocrystalline CoNiFe at different heat treating temperatures
[email protected]
temperature range. When the heat treatment temperature was increased to 800 °C,
the structure became crystalline, and new peaks corresponding to crystalline fcc
CoNi (111) appeared. This behavior can be attributed to the crystallization of pure
nickel followed by the precipitation of CoNi from the supersaturated FeNi solid
solution [13].
As the heat treatment was increased, the (111) peak becomes narrow and intense
while other peaks also appeared because of the crystallization. In addition, the
relationship between the FWHM and grain size indicate that the grain size might
have increased with increasing temperature.
FESEM observations, The comparison between the surface morphologies of
nanocrystalline CoNiFe at different heat treating temperatures is presented in Fig. 3.
The average grain sizes of the samples 500, 600, 700 and 800 °C were about 78,
105, 119 and 132 nm, respectively. The increase of grain size may be ascribed to
the grain growth and the precipitation of intermetallic compound resulted from heat
treatment.
At low temperature, the grain is decreased and the grains began to adhere and
grew together. When the temperature has been increased, grain growth began to
occur. Grain boundaries and a metastable pore phase were established. When the
grain growth continued, the cross-sectional area of the pore phase was decreased.
Then, the pore phase becomes discontinuous, and cause the closed pores remained
Fig. 3 Surface morphologies of nanocrystalline CoNiFe at different temperatures
[email protected]
Fig. 4 The average hardness

for nanocrystalline CoNiFe at
different heat treating
temperatures
at the grain boundaries. Complete removal of the remaining pore and leading to a
completely dense material [14].
Nanocrystalline CoNiFe heated at 700 °C was denser and less voids if compared
with other samples. A greater number of grain boundaries with the highest pro-
portion of atoms inside the boundaries created an extremely high volume fraction of
grain boundaries in the CoNiFe microstructure [9, 10]. Therefore, nanocrystalline
CoNiFe were believed to develop a dense and less voids phase.
Hardness, Fig. 4 shows the average hardness of nanocrystalline CoNiFe at
different heat treating temperatures. The hardness was increased with increasing
heat treating temperature. It reached the maximum hardness at 700 °C and drop
drastically after that.
The increase in hardness was associated with a structural change in the coating.
At higher temperatures (>700 °C), the coating began to soften as the grains con-
glomerated and reducing the number of hardening sites. This process also removes
FeNi (111) from the coating, producing a separate phase of soft nickel within the
matrix and further reducing the bulk hardness [15].
Another factor affecting the behavior of hardness of the nanocrystalline CoNiFe
in this study may be attributed to the grain size and porosity effect. The smaller the
grain size leads to higher hardness of the electrodeposits. However, higher heat
treating temperature (>700 °C) was found to significantly reduce the coating
strength. As can be noticed that the hardness increases and reaches the maximum
value of about 449 HV at 700 °C which coincides with less voids.
Conclusion
Nanocrystalline CoNiFe coatings are successfully prepared with surface finish of

heat treatment. The optimum temperature of nanocrystalline CoNiFe was 700 °C.
The grains growth of the nanocrystalline CoNiFe increased with the increment of
heat treatment temperature. Voids phase area was decreased when the grain growth
[email protected]
occurs. The hardness of nanocrystalline CoNiFe increased with the reduction of

void in microstructure. The maximum hardness is 449 HV at 700 °C.
Acknowledgment This work has received financial support by Fundamental Research Grant
Scheme (FRGS) [600-RMI/FRGS 5/3 (19/2013)] and Principal Investigator Support Initiative
(PSI) [600-RMI/DANA 5/3/PSI (155/2013)] from Research Management Institute (RMI) of
Universiti Teknologi MARA (UiTM). The author also would like to thanks to Ministry of
Education Malaysia for financial support.
References
1. Fadare DA, Fadara TG, Akanbi OY (2011) Effect of heat treatment on mechanical properties
and microstructure of NST 37-2 steel. J Miner Mater Charact Eng 10(3):299–308
2. Issariyapat A, Swangsak P, Boonyongmaneerat Y, Visuttipitukul P (2011) Effects of heat
treatment on the interfacial structure of Nickel-Aluminum coating composites. Adv Mater Res
154:1462–1467
3. Kang S-JL (2005) Handbook of sintering: densification, grain growth and microstructure.
Elsevier Butterworth Heineman, Amsterdam
4. Bakonyi I, Toth-Kada E, Tarnoczi T, Varga L K, Cziraki A, Gerocs I and Fogarassy B.,
Structure and properties of fine-grained electrodeposited nickel, Nanostructure Material. 3
(2993) 155
5. Fortas G, Sam S, Fekih Z, Gabouze N (2009) Electrodeposition of CoNiFe alloys on n-type
silicon. Mater Sci Forum 609:207–212
6. Al-Qura’n F (2009) Effect of heat treatment on the microstructure and hardness of Chromium-
Nickel steel. Contemp Eng Sci 8:355–359
7. Liu WL, Hsieh SH, Chen WJ, Hsu YC (2009) Growth behavior of electroless Ni–Co–P
deposits on Fe. Appl Surf Sci 255:3880–3883
8. Al Qura’n F (2009) Effect of heat treatment on the microstructure and hardness of Chromium–
Nickel steel. Contemp Eng Sci 2(8):355–359
9. Hyie KM, Resali NA, Abdullah WNR (2012) Study of alloys addition to the electrodeposited
nanocrystalline cobalt. Adv Mater Res 486:108–113
10. Resali NA, Hyie KM, Abdullah WNR, Ghani MAA, Kalam A (2013) The effect of bath pH on
the phase formation of ternary Co-Ni-Fe nano-coatings. Appl Mech Mater 391:9–13
11. Tamil Arasu P, Dhanasekaran R, Senthil Kumar P, Srinivasan N (2013) Effect of hardness and
microstructure on En 353 steel by heat treatment, research inventy. Int J Eng Sci 2(11):01–05
12. Bordia RK, Camacho-Montes H (2012) Sintering: fundamentals and practice. The American
Ceramic Society, Published 2012 by John Wiley & Sons, Inc.
13. Slokar L, Matković T, Matković P (2011) Alloying and heat treatment effects on the
microstructure and hardness of biomedical titanium alloys. University of Zagreb Faculty of
Metallurgy, Aleja narodnih heroja 3, 44103 Sisak, Croatia
14. Kofstad P (1998) High temperature corrosion. Elsevier Applied Science Publishers LTD, 52
Vanderbilt Avenue, New York, USA
15. Zhou J, Soboyejo WO (2004) Compression–compression fatigue of open cell aluminum
foams: macro-/micro- mechanisms and the effects of heat treatment. Mater Sci Eng A-Struct
Mater 369(1):23–35 (Journal: Materials Science and Engineering A-structural Materials
Properties Microstructure and Processing)
[email protected]
Computational Fluid Dynamics Modeling
of Mercury Emissions in Pulverized Coal
Combustion
Nor Fadilah Mohamad, Noor Hidayu Abdul Rani,

Sherif Abdulbari Ali and Sharifah Aishah Syed A. Kadir
Abstract Mercury is among the main pollutants in the atmosphere. Mercury can be
released to the atmosphere by human activities including coal combustion, cement
production, incineration and others. Coal combustion is primarily contributed to
mercury emissions. This study is aimed at research and implementation of mercury
model in computational fluid dynamics modeling to predict the distribution of
mercury when coal is combusted in the furnace. The mercury model was imple-
mented in CFD Fluent that consists of three models including One Step Model,
Two Step Model and Detail Wilcox Model. The post process modeling is used for
mercury emissions prediction since the low amounts of mercury has negligible
impact on coal combustion. The Detailed Wilcox mercury model was applied and
the result obtained was compared with measured from the plant.
Keywords Mercury Emissions Combustion Modeling
Introduction
Mercury emissions are one of the most problematic environmental concern in the
world because of its toxicity and bioaccumulation within an aquatic food chain [1].
Mercury is very harmful and can damage nerve systems of humans and lead to
neurological syndrome. Mercury has become an issue after the tragedy of methyl
N.F. Mohamad (&) N.H. Abdul Rani S.A. Ali S.A. Syed A. Kadir
Faculty of Chemical Engineering, Universiti Teknologi MARA, 40450 Shah Alam, Malaysia
N.H. Abdul Rani
S.A. Ali
S.A. Syed A. Kadir

[email protected]
236 N.F. Mohamad et al.
mercury poisoning that known as Minimata Disease in the mid-1950s to 1960s. The
first Minamata Disease occur at Minamata City, Kumamoto Prefecture, Japan in
1956, and followed by the disease at Niigata City, Niigata Prefecture, Japan, in
1965 [2]. Mercury was used as catalyst in the production of acetaldehyde. In this
process, the methyl mercury was generated. This methyl mercury is very harmful
and can cause neurological syndrome including ataxia, numbness in the hands and
narrowing the field of vision.
There are two sources of mercury emissions to the atmosphere including
anthropogenic and natural sources. Anthropogenic sources of mercury such as
combustion of fossil fuels, cement production, incineration and metal smelting have
contribute to the emissions of mercury. Among these industries, coal combustion has
major contribution on mercury emissions. About 80 % of total anthropogenic mer-
cury emissions from 1994 to 1995 was from combustion and coal combustion rep-
resents about 33 % of the total emissions [1]. As stated by N. Pirrone et al. (2010),
about 26 % of the global mercury emissions of anthropogenic sources in 2005 are
come from combustion of fossil fuel in power plants and industrials boilers [3].
Mercury found in coal as a trace element and most of them are combined with
sulphide, then release to the air through combustion. Mercury is dominant to
associate with two sulphide minerals that presents in coal-pyrite (FeS2) which are
mercuric sulphide (HgS) or cinnabar. During coal combustion, mercury is volatil-
ized and released to the air along with exhaust gases [4]. Most of mercury (Hg)
released are converted to elemental mercury (Hg0) in high temperature of the boiler.
When flue gas temperature cooled, oxidation can occur that led to the formation of
oxidized mercury (Hg2+) and/or particulate mercury (HgP). Particulate mercury
easily captured by precipitators such as electrostatic precipitator and bag filter.
Oxidized mercury can be removed by wet flue gas desulphurization due to its water
soluble properties. Elemental mercury is hardly to capture because it insoluble in
water. Most of elemental mercury is found in the stack gas. The existence of
chlorine compounds also increase the formation of oxidized mercury thus decrease
the emissions of mercury to the atmosphere. To better understand the distribution of
mercury in coal combustion inside the furnace, a two dimensional computational
fluid dynamics modelling has been model. Besides that, mercury content of coal
also had been determined and the measurement of mercury emissions at stack gas
had been performed to calculate mercury removal efficiency of the plant.
Methodology
Mercury Measurement at Stack Gas
The concentration of mercury was measured at the stack gas of the chimney in coal
fired power plant. The measurement was made online for real data of mercury
emissions using portable mercury analyzer. This portable mercury analyzer used
[email protected]
Computational Fluid Dynamics Modeling of Mercury Emissions … 237
two different types of conversion systems (SnCl2 and thermal) attached in front of
each mercury analyzer. Before measurement was made, a fresh solution was pre-
pared that consist of 10 g of SnCl2 in a mixed solution of 100 ml deionized water
(DI) and 1.8 ml concentrated H2SO4. The probe was placed at stack gas and the gas
stream was heated above 800 °C. At this high temperature, all mercury compounds
were reduced to elemental mercury and the total mercury concentrations was
observed after one hour sampling.
Computational Modeling of Mercury Emissions
The boiler model was developed based on the full scale of 700 MW wall-fired
furnaces. Simplifications was made and a two-dimensional geometry was created
using design Modeler-a Fluent pre-processor. Figure 1 shows the geometry of the
furnace with width of 15.3 m and height of 48.5 m. The qualities of the mesh had
been checked before run the simulation to avoid error and divergence in calculation.
The furnace consists of five inlets of pulverized coals that comes from five mills
which are Mill A, B, C, D and E. Coals from the mill were pulverized to obtains
sizes less than 200 micros in diameter before blow into the furnace. Primary air
used for drying and transporting the pulverized coal into the furnace [5]. Secondary
air used to provide hot air to support combustion of fuel in the furnace. In the
modeling, the mathematical model used is based on the commercial CFD code,
Fluent for momentum, energy, enthalpy and species mass fraction. For turbulence
models, the standard k-ε turbulence model was selected because of its general
applicability, robustness and economy [6]. For radiation model, P1 was used
because P1 radiation model is suitable for combustor larger than 1 m.
The first part of the current work is to establish a base model which will then be
used for prediction of mercury emissions from coal combustion. In ANSYS Fluent,
coal combustion was modeled by using non-premixed combustion model for
reaction chemistry. The values obtained from ultimate and proximate analysis were
Fig. 1 View of the model in

2-D after refinement
[email protected]
Table 1 Operating conditions of the boiler

Input Data Value
Total primary air flow rate (m3/s) 19.95
Temperature of primary air (K) 343
Total secondary air flow rate (m3/s) 89.57
Temperature of secondary air (K) 573
Total coal flow rate (kg/s) 68.7
Table 2 Analysis data on coal

Proximate analysis Ultimate analysis
Elements Weight percentage Elements Weight percentage
%(DAF) %(DAF)
Moisture 3.4 Carbon, C 84.39
Volatile matter 26.4 Hydrogen, H 4.60
Fixed carbon 55.2 Oxygen, O 8.29
Ash 15.0 Nitrogen, N 2.03
Mercury 0.13 ppm Sulfur, S 0.69
Chlorine 0.0004
used as input parameter in non-premixed combustion model. The PDF mixture was
used for species and materials count and PDF table was created that consists of
information on thermo-chemistry and its interaction with turbulence. For coal
injections, a discrete phase model (DPM) was defined and used to solve the
transport equations for continuous phase. The boundary and operating conditions
was set up before the simulation was initiated. Table 1 shows the specifications of
the furnace. The analysis data on coal is given in Table 2.
The finite volume method was used to discretize the governing equations on
computational grids. SIMPLE-based approach was employed for pressure-velocity
coupling scheme. The solution was simulated until convergence is achieved. If the
convergence not achieve, it’s needed to modify the solution parameters and solution
method.
For modelling of mercury emissions from pulverized coal combustion, there are
three mercury models provides in Fluent. Since mercury concentrations in coal have
minimal influence on the predicted flow field, temperature and species mass frac-
tion, therefore, mercury models were used as a post processing after the main
combustion modelling was achieved. During coal devolatilization, mercury is
released and the rate of elemental mercury (Hg) released may be assumed pro-
portional to the rate of devolatilization. Same concept as well goes to chlorine
species. Generally, chlorine was released in the form HCl. Mercury Model in Fluent
consist of One step model, two step model and detailed Wilcox model which are
depending on the mechanism of mercury with chlorine compounds. All reactions in
[email protected]
Table 3 Reaction rate constant for mercury speciation
Reaction Forward A Forward Forward E (kcal/ Reverse A Reverse Reverse E (kcal/
(mol−1 m3 s−1) b mol) (mol−1 m3 s−1) b mol)
HgCl + M = Hg + Cl + M 4.25e13 0 716.13 – – –
HgCl + HCl = HgCl2 + H 4.50e13 0 30.27 – – –
Hg + HCl = HgCl + H 1.93e13 0 93.3 2.55e12 0 13.8
Hg + Cl2 = HgCl + Cl 6.15e13 0 43.3 7.23e12 0 11.8
HgCl2 + M = HgCl + Cl + M 1.35e8 0 76.08 – – –
HgCl + Cl2 = HgCl2 +Cl 1.8115e10 0 0 – – –
HgO + M = Hg + O + M 3.09e9 0 8.8 – – –
Hg + HOCl = HgCl + OH 3.06e13 0 36.6 6.87e11 0 6.2
HgCl + HOCl = HgCl2 + OH 1.827e10 0 0.5 – – –
[email protected]
HgCl2 + M = Hg + Cl2 + M 3.19e11 0 87.0 – – –
Cl2 + M = 2Cl + M 8.5e15 0 55.84 2.23e14 0.0 −1.8
Computational Fluid Dynamics Modeling of Mercury Emissions …
Cl + H2 = HCl + H 1.47e12 1.63 3.16 1.45e12 1.44 2.46

Cl2 + H = HCl + Cl 8.59e13 0 1.17 1.0e17 0.0 47.49
Cl + H + M = HCl + M 7.19e21 −2 0 4.4e13 0.0 81.67
239
the model are assumed to obey Arrhenius type of constant. Equation 1 shows the
Arrhenius equation used in the model. However, in this paper only Detailed Wilcox
Model was discussed.
k ¼ AðT=298Þb expðE=RTÞ ð1Þ
Detailed (Wilcox) Model
In Detailed Wilcox Model, Wilcox proposed the other reactions of mercury chlo-
rination. In this example, Wilcox proposed the other reactions of mercury chlori-
nation because chlorine undergoes a set of reactions on its own and creates a
potential pathway for the reactions to mercury. There are eight pathways for
mercury chlorination/oxidation. There are 12 species involve in the models such as
(Hg, HgCl2, HgO, HgCl, HCl, Cl, Cl2, HOCl, H2, OH and O). The assumption was
made where all the reactions are reversible. Table 3 shows the rate constants for all
the reactions that considered in this model.
Measurement of Mercury Emissions at Stack Gas of Coal

Fired Power Plant
Figure 2 shows the measurement of mercury measurement of mercury emissions at

the stack gas of the coal fired power plant. The stack gas is 110 m height and the
measurement was taken at the sampling point at the height of 80 m. The temper-
ature at the sampling point is 146 °C. Mercury measurement was taken in one hour.
The graph shows that the emission is fluctuating and the maximum mercury con-
centration was obtained with 39.5 µg/Nm3. The average mercury concentration is
Fig. 2 Mercury emissions

from coal combustion at stack
gas
[email protected]
8.5 µg/Nm3. This value is within the range of mercury emissions limit from
Department of Environment (DOE) regulation of coal combustion which is 30 µg/
Nm3. Therefore, mercury emission from this plant under the limit and considered
not harmful to the environment and human surrounding the plant.
Mercury Model: Detail Wilcox Model
In Detail Wilcox Model, there are 14 reactions had been taken into account of
mercury reaction in coal combustion. There are ten reactions of mercury with
chlorine and four reaction of chlorine on its own. Detail Wilcox Model involves 12
species for the reaction of mercury chlorination/oxidation and chlorine specific
reactions. The 12 species in the model are Hg, HgCl2, HgO, HgCl, HCl, Cl, Cl2,
HOCl, H2, H, OH and O. As higher boiler temperature (higher that 1200 °C), the
mercury species vaporized to elemental mercury, which is dominated in the flue
gas. However, with the decreasing of flue gas temperature, the reaction of Hg to
HgCl2 begins. The stable and dominates temperature of HgCl2 is at temperature
below 400 °C [7]. Figure 3 shows the mass fraction of Hg, Cl2, HgCl2 and HgCl.
Fig. 3 a Contour pollutant of Hg. b Contours of pollutant Cl2. c Contours of pollutant HgCl2.
d Contours of HgCl
[email protected]
The mass distribution of these compounds is near the zero at the furnace. This is
because the higher temperature in the furnace not allowed the oxidation of mercury.
The more dominant reaction that occurs in the furnace is the formation for chlorine
compounds. This statement had agreed with the previous study that shows the
chlorination of mercury only occur at temperature below 1200 °C [7]. For the
reactions of mercury that involve with chlorine, the reactions of chlorine itself had
been dominant. Therefore, the formation of HgCl2 and HgCl is low at the furnace
that can be shown in Fig. 3. In this study, it shows that the presence of chlorine in
the flue gas derived from coal combustion has major impact to mercury oxidation
[8]. The presence of chlorine has been inhibiting the mercury oxidation and chlo-
rine specific reactions have been as major reaction in the furnace. From CFD
simulation, after calculated, the prediction of mercury emissions at stack gas is
8.16 µg/m3. After compare with measurement made at the plant, the prediction
gives good agreement with 4 % error.
Conclusion
Two-dimensional simulation of the distribution and flow of mercury emissions

from coal combustion process in the furnace of 700 MW wall fired tangentially
pulverized coal fired power plant was presented in this study. The prediction of
mercury emissions is agreed with the literature surveys. The existence of chlorine
compounds in high temperature of the furnace has decrease the oxidation of mer-
cury from elemental mercury to oxidize mercury. Therefore, most of mercury
leaving the furnace are in elemental form and will be oxidized when flows to air
pollution control devices. Electrostatic precipitator and flue gas desulfurization have
co-benefits to reduce mercury in this plant due to the capture of particulate mercury
by electrostatic precipitator and oxidized mercury by flue gas desulfurization. The
measurement of mercury concentration in stack gas obtained an average of total
mercury released is 8.5 µg/Nm3. The mercury removal efficiency of the plant is
99.32 % that proved the air pollution control devices in the plant can be co-
beneficial for the reduction of mercury to the atmosphere. When compared with
measured concentrations of mercury emissions, the predicted value of the Detail
Wilcox Model has an accurate value with 8.16 µg/Nm3.
Acknowledgment The authors gratefully acknowledgement to Research Management Institute

(RMI) for funding this research through excellence fund (600-RMI/DANA 5/3/RIF (312/2012))
high appreciation to UiTM/MOHE for financial support under Tenaga Pengajar Muda (TPM)
scholarship. The authors also acknowledge to Prof. Masaki Takaoka from Kyoto University, Japan
for his contributions to this study.
[email protected]
References
1. Wu C, Cao Y, Dong Z, Cheng C, Li H, Pan W (2010) Evaluation of mercury speciation and

removal through air pollution control devices of a 190 MW boiler. J Environ Sci 22:277–282
2. Brooks EW (2003) Mercury: U.S geological survey minerals yearbook
3. UNEP (2010) Study on Mercury sources and emissions, and analysis of cost and effectiveness
of control measures. Geneva
4. Jensen RR, Karki S, Salehfar H (2004) Artificial neural network-based estimation of mercury
speciation in combustion flue gases. Fuel Process Technol 85:451–462
5. JIMAH Energy Ventures Sdn. Bhd. (2005) Information memorandum: issuance of up to RM4,
847 million nominal value senior Istisna’ medium term notes facility
6. Kumar R, Pandey KM (2012) CFD analysis of circulating fluidized bed combustion. IRACST-
Eng Sci Technol: An Int J (ESTIJ) 2:163
7. Yokoyama T, Asakura K, Matsuda H, Ito S, Noda N (2000) Mercury emissions from a coal-
fired power plant in Japan. Sci Total Environ 259:97–103
8. Buitrago PA (2011) Gas-phase mercury oxidation: effects of bromine, chlorine and sulfur
dioxide under air firing and oxy-fuel conditions, experimental and modeling study. Doctor of
Philosophy, Department of Chemical Engineering University of Utah
[email protected]
3D Sustainability Analysis of Integrated
Process Design and Control for Production
of Cyclohexanone
Siti Aminah Zakaria, Mohd Jufri Zakaria, Mohamad Rizza Othman

and Mohd Kamaruddin Abd Hamid
Abstract The objective of this paper is to highlight the use of a three-dimensional

(3D) sustainability index in performing a sustainable integrated process design and
control (Sustain-IPDC) for a reactor system. Sustain-IPDC for reactor systems is
formulated as a mathematical programming problem and solved by decomposing it
into six sequential hierarchical sub-problems: (i) pre-analysis, (ii) design analysis,
(iii) controller design analysis, (iv) sustainability analysis, (v) detailed economic
analysis, and (vi) final selection and verification. By using thermodynamics and
process insights, a bounded search space is first identified. Then, the feasible
solution space is further reduced to satisfy the process design, controller design,
sustainability constraint and economical in stage 2, 3, 4 and 5 respectively. Lastly,
all the feasible candidates are ordered according to the defined performance criteria.
In this methodology, the concept of an attainable region (AR) diagram is used in
finding the feasible candidates. The targets for this optimal solution is defined and
selected at the maximum point of the AR diagram. Then, the proposed methodology
is applied to the production of cyclohexanone using a continuous-stirred tank
reactor (CSTR) system. The results show that the proposed methodology is capable
in finding an optimal solution for a CSTR design problem that satisfy design,
control, sustainability and economic criteria in an easy and systematic manner.
S.A. Zakaria M.J. Zakaria M.K.A. Hamid (&)

Process Systems of Engineering Centre, Faculty of Chemical Engineering, Universiti
Teknologi Malaysia, 83130 UTM Johor Bahru, Johor, Malaysia
S.A. Zakaria
M.J. Zakaria
M.R. Othman
Process Systems of Engineering Group (PSERG), Faculty of Chemical and Natural Resources
Engineering, Universiti Malaysia Pahang, 26300 Gambang, Kuantan, Pahang, Malaysia

[email protected]
246 S.A. Zakaria et al.
Keywords Sustainability analysis 3 dimensional Process design Process

control
Introduction
Integrated Process Design and Control (IPDC) methodology was developed which
is able to identify and obtain an optimal solution for the IPDC problem for chemical
processes in an easy, simple and efficient way [1]. However, the developed
methodology for the IPDC did not consider sustainability aspect in the early
chemical processes design stage. Designing controllable and also sustainable pro-
cess is one of the key challenges for sustainable development of chemical pro-
cesses. Chemical process design can be further improved by including sustainability
aspect within the developed IPDC method to ensure that the design is more cost
efficient and controllable, as well as sustainable to meet product quality specifica-
tions. This can be achieved by extending the developed model-based IPDC method
encompasses sustainability aspect.
As we know, sustainability is based on balancing three principal objectives:
environmental protection, economic growth, and societal equity. Metrics and
indicators are used to assess the sustainability performance of a process or a system,
to evaluate the progress toward enhancing sustainability, and to assist decision
makers in evaluating alternatives. Three-dimensional (3D) sustainability index is
one of the easier way in analysing the sustainability aspects since in one single
indicator it will cover all the three aspects of sustainability. There are four indicators
used in 3D sustainability index as mention in Martins et al. [2].
Solving IPDC problem together with sustainability criteria may cause com-
plexity in the optimization problem. To obtain solutions for this problem will
require a huge computational effort which makes this approach impractical for
solving real industrial problems. In order to overcome the complexity of the
Sustain-IPDC problem and obtain an achievable optimal solution, a decomposition
approach is used in this study. The decomposition approach has been applied in
managing and solving the complexity of different optimization problems in
chemical engineering [1]. Here, several feasible targets will be obtained by opti-
mizing the problem with constraints at which the objective function is meet. In the
decomposition-based approach [3] the optimization problem is divided into sub-
problems. Then, the step-by-step analysis is done by reducing the feasible targets in
each stages while considering all the constraints. Then, when all the constraints are
evaluated, the calculation of the objective functions of all the remaining targets is
done in order to locate the optimum solutions. In this work, the decomposition
solution strategy has been adopted to develop a new model-based methodology for
solving Sustain-IPDC problem.
[email protected]
3D Sustainability Analysis of Integrated Process Design … 247
Methodology
Problem Formulation
The Sustain-IPDC problem is formulated as a generic optimization problem where

the optimization of the process design, controllability, sustainability and economic
criteria are subjected to a set of constraints for both dynamic and steady state
process, constitutive (thermodynamics state) and conditional (process, control and
sustainability conditions) constraints. The problem formulation can be referred to
Hamid [1]. The different optimization scenarios can be generated as follows and the
multi-objective function is then formulated as:

1 1
max J ¼ w1 P1 þ w2;1 þ w2;2 P2;2 þ w3 þ w4 P4 : ð1Þ
P2;1 P3
• To achieve the process design objective which is the performance criteria for
reactor design, P1 is maximized.
• To achieve the controllability objectives, P2,1 is minimized and P2,2 is maxi-
mized. P2,1 is the sensitivity of controlled variable, y with respect to disturbance,
d, while, P2,2 is the sensitivity of controlled variable, y with respect to manip-
ulated variable, u.
• To achieve the sustainability objectives, P3 is minimized by minimizing the 1D,
2D and 3D index.
• To achieve the economic objectives, P4 which is a profit function that is need to
be maximized.
Decomposition-Based Solution Strategy
The work flow and steps involved in the decomposition based solution strategy
(refer with: Fig. 1). Accordingly, the Sustain-IPDC problem for reactor system is
decomposed into six step-by-step stages: (1) pre-analysis; (2) design analysis; (3)
controller design analysis; (4) sustainability analysis; (5) detailed economic analysis
and (6) final selection and verification. The sets of constraint in the Sustain-IPDC
problem for reactor system are divided into six sub-problems which correspond into
six stages at which the solution of the decomposed set of sub-problem is equivalent
to the original problem. When step-by-step problem being solved, the infeasible
solutions is eliminated which finally will lead to the final sub-problem which is
smaller and it can be solve easily. Therefore, while the complexity in each stage
increase, the number of feasible solutions is reduced at every stage.
[email protected]
Fig. 1 Sustainable integrated process design and control methodology for reactor systems
[email protected]
Process Description
The multi-step consecutive reaction of cyclohexanone production provides an

interesting challenge for the Sustain-IPDC application. The results will show the
capability of the Sustain-IPDC to handle or solve complex processes as well as the
simple process. Next, the following situation is considered in the CSTR. The pro-
duction of the desired product, cyclohexanone is to be done from the reaction of the
only reactant in the reactor, cyclohexanol. In spite of cyclohexanone, hydrogen is
also produced as a by-product from the first reaction.
Problem Formulation
The Sustain-IPDC problem for the process described above is defined in terms of a
performance objective (with respect to design, controlled, sustainable and detailed
economic), and the three set of constraints (process, constitutive and conditional). It
can be referred in Hamid [1]. For this case, each component used is presented as a,
b, c and d which respect to cyclohexanol, cyclohexanone, high boiler, and hydrogen
respectively.
Stage 1: Pre-analysis. The objective of this stage is to define the operational
window and set the targets for the design-controller solution by using attainable
region diagram [4]. Based on the plotted region, the maximum point at the diagram
(Point A) is chosen as the design target together with the other two points (alter-
native design for verification purposes), (refer with: Fig. 2).
Stage 2: Design analysis, In this stage, all the targets selected in the previous stage
then undergo the validation process for finding its design values by calculating its
residence time and the design variables such as volume and height of the CSTR
(refer with: Table 1).
Fig. 2 Attainable region

diagram for Ca with respect to
Cb at 394.26 K
[email protected]
Table 1 Values of residence time with the corresponding process-controlled and design-
manipulated variables at different CSTR designs
Targets T (hr) Process-controlled Manipulated Design
Ca (kmol/m3) Cb (kmol/m3) F (m3/hr) V (m3) hr (m)
A 0.52 0.175 0.6940 10.0 5.15 2.27
B 0.98 0.100 0.6616 10.0 9.84 4.33
C 0.33 0.250 0.6696 10.0 3.28 1.44
Fig. 3 Attainable region diagram for the desired concentration of b with respect to reactant, Ca,
and its corresponding derivatives of Cb with respect to disturbance Ca at feed (left) and attainable
region diagram for the desired concentration of b with respect to reactant, Ca, and its
corresponding derivatives of Cb and hr with respect to manipulated variable F (right)
Stage 3: Controller design analysis, At this stage, the controllability performance

of each of the feasible candidates is evaluated and validated for the selection of
controller structure. There are two criteria that need to be analyzed which are
sensitivity of controlled variable y with respect to disturbance d, dy/dd and sen-
sitivity of controlled variable y with respect to the manipulated variable u, dy/du.
The results clearly show that design A has the best controller design compared to
designs B and C (Fig. 3).
Stage 4: Sustainability analysis, At this stage, the sustainability index for three-
dimensional (3D) sustainability analysis is based on Martins et al. [2]. Here, the
indexes used are energy intensity, material intensity, potential chemical risk and
potential environmental impact. Energy and material used give negative impact to
the environment (waste generated), positive impact to economy (value creation)
and give both impact to society (higher living standard and future generation
deprive due to depletion). While the other two indexes could have a poor working
environment (environment), financial loss due to shutdown (economic) and illness
(society). In this case, those four indexes are located inside a 3D index that comply
all aspects in sustainability. The result for this analysis shown in table below (refer
with: Table 2).
Stage 5: Detailed Economic analysis, The analysis is continued by identifying the
profit and total cost of each design candidates. Based on the analysis, design at
[email protected]
Table 2 Values of all index in 3D sustainability analysis for each candidates

Candidates Index Potential Potential Material Non renewable
environmental chemical intensity energy intensity
impact impact
A Full 0.5913 57.4522 1.1700 0.0017
index
Scale 1.0000 0.9506 0.8752 0.8947
index
3D 0.9301
index
B Full 0.5888 60.4383 1.3369 0.0019
index
Scale 0.9958 1.0000 1.0000 1.0000
index
3D 0.9989
index
C Full 0.5580 54.2111 1.1040 0.0016
index
Scale 0.9437 0.8970 0.8258 0.8421
index
3D 0.8771
index
Point B is the best in economic analysis because of its high profit. This is due to the
large design volume of CSTR at this point which leads to high production of the
products. Then, it follows by design and Points A and C, respectively. The CSTR
designs will be analyzed by using the profit function in order to analyze the design
that will provide the maximum profit. There are four criteria used to calculate the
profit function, which are profit of product, cost of material, depreciation cost and
operating cost. The economic analysis is calculated by assuming that the CSTR is
operating in 5 years and 340 days per year. From the results, design at Point B has
the highest profit compared to designs at Points A and C (Table 3).
Stage 6: Final selection and verification. The objective of this stage is to select
the best candidate by analyzing the value of the multi-objective function. The
multi-objective function is calculated by summing up the multi-objective function
value. From the tabulated results in Table 4, it is clearly seen that the multi-
objective function, J for CSTR design A is higher compared to other designs.
Table 3 Economic analysis for each candidates

Candidates I1 ($/yr) I2 ($/yr) CR ($/yr) CD ($/yr) CO ($/yr)
Purchase Installed
A 69,402.284 46,254.6 28,901.1 2890.1 5298.5 0.0
B 126,306.297 88,304.1 37,432.2 3743.2 6862.6 0.0
C 42,610.647 29,434.7 24,121.1 2412.1 4422.2 0.0
[email protected]
Table 4 Final result of multi- P1,1 P2,1 P2,2 P3 P4,1

objective function
A 0.694 0.178 2.398 0.930 17,849.2
B 0.662 0.874 2.257 0.999 31,139.6
C 0.670 0.552 2.276 0.877 8753.7
P1,1s P2,1s P2,2s P3,1s P4,1s J
A 1.000 0.203 1.000 0.550 0.573 10.39
B 0.953 1.000 0.941 1.000 1.000 5.89
C 0.965 0.631 0.949 0.337 0.281 7.79
The best candidate is highlighted in bold
Therefore, it is verified that the CSTR design at the maximum point of attainable
region is an optimal solution for Sustain-IPDC of a cyclohexanone production in
CSTR which satisfies the design, control, sustainability and economic criteria.
Conclusion
In this work, a systematic model-based methodology has been developed for sus-
tainable integrated process design and control (Sustain-IPDC) for a continuous-
stirred tank reactor system (CSTR) by using a 3D sustainability index. This
methodology has been applied and verified for a single CSTR system. The main
contribution of this work that has been obtained from this work is that, the meth-
odology applied a step-by-step procedure which allows a systematic analysis being
done at each stage. Every step of the design methodology is clear with respect to
calculations/analysis and generic in terms of application range which makes the
application of the methodology is quite easy.
Acknowledgment The financial support from Zamalah Universiti Teknologi Malaysia is highly
acknowledged.
References
1. Hamid MKA (2011) Model-based integrated process design and controller design of chemical
processes. Ph. D. Thesis, Technical University of Denmark
2. Martins A, Mata T, Costa C, Sikdar S (2007) Framework for sustainability metrics. Ind Eng
Chem Res 46:2962–2973
3. Karunanithi APT, Achenie LEK, Gani R (2006) A computer-aided molecular design framework
for crystallization solvent design. Chem Eng Sci 61:1247–1260
4. Hamid MKA, Sin G, Gani R (2010) Integration of process design and controller design for
chemical processes using model-based methodology. Comput Chem Eng 34:683–699
[email protected]
Three-Dimensional Sustainability Analysis
for Integrated Process Design and Control
for Separation of Benzene and Toluene
Mixture
Mohamad Zulkhairi Nordin, Mohamad Dzulfadzli Jais,

Mohamad Rizza Othman and Mohd Kamaruddin Abd Hamid
Abstract The objective of this paper is to present the development of a sustainable

integrated process design and control methodology for a distillation column system.
The sustainable integrated process design and control methodology for a distillation
column system is developed to ensure the distillation column design is more cost
efficient, controllable and sustainable to meet the product quality. The sustainable
integrated process design and control problem for a distillation column system is
typically formulated as a mathematical programming (optimization with con-
straints) problem, and solved by decomposing it to six sequential hierarchical sub-
problems: (i) pre-analysis, (ii) design analysis, (iii) controller design analysis, (iv)
sustainability analysis, (v) detailed economics analysis and (vi) final selection and
verification. In the pre-analysis sub-problem, the concept of driving force is used to
locate the optimal design-control solution targets, which are defined at the maxi-
mum point of the driving force diagram. The sustainability analysis sub-problem
was calculated by using the three-dimensional sustainability index. The results
through case study of benzene-toluene separation process shows the proposed
methodology is capable to find the optimal solution that satisfies design, control,
sustainability and economic criteria in a simple and efficient way.
M.Z. Nordin M.D. Jais M.K. Abd Hamid (&)

Process Systems Engineering Centre (PROSPECT), Faculty of Chemical Engineering,
Universiti Teknologi Malaysia, 81310, UTM Johor Bahru, Johor, Malaysia
M.Z. Nordin
M.D. Jais
M.R. Othman
Process System Engineering Group (PSERG), Faculty of Chemical and Natural Resources
Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300, Gambang, Kuantan,
Pahang, Malaysia

[email protected]
254 M.Z. Nordin et al.
Keywords Optimal Integrated Design Control Sustainability
Introduction
Integrated Process Design and Control (IPDC) methodology was developed and
able to obtain an optimal solution for the IPDC problem [1]. The IPDC method-
ology has shown that distillation column design at the highest point of driving force
(DF) is the best in terms of design and controllability. However, the developed
methodology for the IPDC did not consider sustainability aspect in the early of
design stage. Distillation column design can be further improved by including
sustainability aspect within the developed IPDC methodology to ensure that the
design is controllable, as well as sustainable.
Sustainability can be defined as the development that meets the needs of the
present without compromising the future generations to meet their own need.
Basically, sustainability is made up from three principles which are environmental,
economic and social. From the three principles of sustainability, three different
sustainability indexes can be obtained which are one-dimensional (1-D) index, two-
dimensional (2-D) index and three-dimensional (3-D) index as shown in Fig. 1
[2–4]. Each index is represented with its own metrics; use to assess the sustain-
ability performance of a process or a system, in order to evaluate the progress
toward enhancing the sustainability [5]. Furthermore, indicator also can be used to
assist decision makers in evaluating the alternatives. This paper will highlight the
implementation of sustainability criteria by using the 3-D index in the IPDC
methodology known as Sustainable Integrated Process Design and Control
(Sustain-IPDC).
Fig. 1 Venn diagram of

sustainability criteria
[email protected]
Three-Dimensional Sustainability Analysis for Integrated … 255
Methodology
The Sustain-IPDC problem is formulated as a generic optimization problem in

which a performance objective in terms of design, control, sustainability and cost is
optimized subject to a set of constraints as Eqs. 1–6.
X
m X
n
max J ¼ Pi;j wi;j ð1Þ
i¼1 j¼1
Subjected to:
Process (dynamic and/or steady state) constraint
dx
¼ f ðu; x; d; h; Y; GÞ ð2Þ
dt
Constitutive (thermodynamic) constraint
0 ¼ g1 ðv; xÞ h ð3Þ
Conditional (process, controllability, sustainability) constraint
0 ¼ h1 ðu; xÞ ð4Þ
0 h2 ðu; x; dÞ ð5Þ
CS ¼ x þ uY ð6Þ
whereas:
x set of process controlled variables
u set of manipulated (design) variables
d set of disturbance variables
θ set of constitutive variables
v set of chemical system variables
G set of independent variables
i category of objective function
j specific term of each category
Y set of binary decision variables
wi weight factor of each objective term
Pi objective term
Sustain-IPDC problem for a distillation column system is formulated as a mixed-

integer dynamic optimization (MIDO). The design with the maximum value of J is
indicated as the best distillation design. The scenarios can be generated as follows:
[email protected]
• To achieve the process design objective, P1 which is the performance criteria for
a distillation design is maximized.
• To achieve the controller design objectives, P2,1 is minimized and P2,2 is
maximized. P2,1 is the sensitivity of controlled variable, y with respect to dis-
turbance, d. While, P2,2 is the sensitivity of controlled variables, y with respect
to manipulated variables, u.
• To achieve the sustainability objectives, P3,j is minimized. There are four
indexes used to achieve the sustainability objective which are P3,1, P3,2, P3,3 and
P3,4. These indexes represent the material intensity, energy intensity, potential
chemical risk and potential environmental impact respectively.
• To achieve the economic objectives, P4 which is net profit function are need to
be maximized.
Then the multi-objective function can be formulated as Eq. 7 (same weight
factor, wi = 1):
1 1 1 1 1
max J ¼ P1 þ þ P2;2 þ þ þ þ þ P4 ð7Þ
P2;1 P3;1 P3;2 P3;3 P3;4
The problem is solved by decomposing it into six stages: (i) pre-analysis,

(ii) design analysis, (iii) controller design analysis, (iv) sustainability analysis,
(v) detailed economics analysis and (vi) final selection and verification, as shown in
Fig. 2. In this way, the solution of the decomposed set of sub-problem is equivalent
Problem Definition and Formulation
Problem Formulation STAGE 1: PRE - ANALYSIS Graphical Methods

1.1 Variable analysis
1.2 Operational window identification
1.3 Design-control target identification
STAGE 2: DESIGN ANALYSIS

2.1 Design variable calculation
NO
Values found?
Within operational window NO
STAGE 3: CONTROLLER DESIGN ANALYSIS

3.1 Sensitivity analysis (Eq. 2)
3.2 Controller structure selection (Eq. 6)
STAGE 4: SUSTAINABILITY ANALYSIS

4.1 Material Intensity
4.2 Energy Intensity
4.3 Potential Chemical Risk
4.4 Potential Environmental Impact
STAGE 5: DETAILED ECONOMIC ANALYSIS

5.1 Profit function
STAGE 6: FINAL SELECTION AND VERIFICATION

6.1 Final selection (Eq. 1)
6.2 Dynamic rigorous simulation: verification of
controller performance
Fig. 2 Sustainable integrated process design and control methodology for distillation column
[email protected]
to the original problem. As each sub-problem is being solved, a large number of

feasible solution within the search space is identified and hence eliminated.
Therefore, while the problem complexity may increase with every subsequence
stage, the number of feasible solution is reduced at every stage.
Application of Methodology for Distillation Column

Systems
The application of the Sustain-IPDC is illustrated by the separation process of

Benzene and Toluene.
Stage 1: Pre-analysis. The objective of this stage is to define the operational
window where the optimal solution is located. The operational window is
identified based on bottom and top product quality. Step-by-step algorithm for a
simple distillation by Bek-Pederson and Gani is implemented in this step [2].
The driving force (DF) diagram is developed by using the Eq. 8, where yi and xi
are phase composition of component i and αij is the relative separability factor
for binary pair.
xi aij
DF ¼ yi xi ¼ xi ð8Þ
1 þ xi aij 1
DF diagram for Benzene-Toluene system at P = 6 atm is drawn as shown in Fig. 3.

The target for optimal sustainable process-control design solution for distillation
column is identified at the maximum point of driving force (Point A). Two other
points are identified as alternative designs (Points B and C) to compare with Point A.
Fig. 3 Driving force diagram for separation of Benzene-Toluene by a distillation column
[email protected]
Table 1 Values of design variable for separation Benzene-Toluene for distillation column design
Point N. Stage N. Feed RBmin RRmin RB RR D (kmol/hr) QC QR
(kW) (kW)
A 20 12 1.70 2.58 2.04 3.09 8.061 217.4 244.9
B 20 14 1.38 2.88 1.66 3.46 8.061 222.6 249.9
C 20 10 2.05 2.33 2.46 2.79 8.061 244.0 272.2
Stage 2: Design Analysis. All targets identified in Stage 1 is validated by

finding the acceptable values of controlled and manipulated variables. Each
points are analyzed by applying the methodology proposed by Bek-Pederson
and Gani [6]. Then, from the results, other design-process variables of distil-
lation column are obtained by using the Aspen HYSYS steady-state process
simulator. The results are summarized in Table 1. It is noted that design A
corresponds to the lowest energy consumption.
Stage 3: Controller Design Analysis. At this stage, the controllability perfor-
mance of each of the feasible candidates is evaluated and validated for the
selection of controller structure. There are two criteria that need to be analyzed
which are sensitivity of controlled variables y with respect to disturbances d and
sensitivity of controlled variables y with respect to the manipulated variables
u. The details procedure for controller design analysis are shown in Hamid’s
work [7]. Figure 4 shows the plot of derivative DF with the respect to xB
(disturbance). A smaller derivative means that the process sensitivity is lower,
hence, from process control of view, distillation column design A is less sen-
sitive to the effect of disturbance thus make it more robust in maintaining its
controlled variable in the presence of disturbances. The controller structure is
selected by calculating the derivatives values of the potential controlled
Fig. 4 Plot of driving force and derivative of driving force with respect to composition as a
function of composition for composition of Benzene
[email protected]
Fig. 5 Driving force diagram for separation of Benzene-Toluene by distillation with

corresponding derivatives of xdB, xdT xbB and xbT respect to L and V
variables (xdB,xdT, xbB, xbT) with respect to manipulated variables (V and L), in
a constant step size. The value of the derivatives are plotted in Fig. 5. Higher
gain in process control means the process will require smaller control action to
maintain the controlled variable at the optimal set point value. So, it can be
concluded that to maintain the purity of the product in the distillation column
can be done by directly manipulating the value of V and L.
Stage 4: Sustainability Analysis. The purpose of this stage is to analyze the
optimal design of distillation column in terms of sustainability. The sustain-
ability of each distillation design is assessed by using 3D metrics. There are four
indexes of 3D metrics used to assess the sustainability analysis which are
material intensity (MI), energy intensity (EI), potential chemical risk (PCR) and
potential environmental impact (PEI) [4]. In order to fulfill the sustainability
criteria, it must have low impacts to the economic losses as well as environ-
mental impacts. Material and energy consumption will be measured to satisfy
sustainability objective functions. The values of each sustainability metric are
tabulated in Table 2. It can be seen that Point A has the lowest index compared
with other points. Thus, it proves that the distillation design at the maximum
point of driving force will have the best objective function in term of
sustainability.
Stage 5: Detailed Economic Analysis. The distillation designs will be analyzed
using profit function in order to analyze the design that will provide the max-
imum profit [8]. There are four criteria used to calculate the profit function,
which are profit of product, cost of material, depreciation cost and operating
Table 2 The sustainability MI EI PCR PEI

analysis values at different
distillation column design Point A 0.9938 2.6393 0.0001 0.0012
Point B 0.9938 2.6955 0.0001 0.0012
Point C 0.9938 2.9477 0.0001 0.0012
[email protected]
Table 3 The detaile economic analysis values at different distillation design

Cost Product Raw material Depreciation Operation Profit, ∅
Point A 108,533,304 81,600,000 1,872,649 1,712,683 23,347,971
Point B 108,533,304 81,600,000 1,896,790 1,749,208 23,287,304
Point C 108,533,304 81,600,000 1,900,070 1,912,927 23,120,306
Table 4 The values of multi-objective function calculation

P1,1 P2,1 P2,2 P3,1 P3,2 P3,2 P3,2 P4,1
A 0.2180 0.0065 0.0720 0.9938 2.6393 0.0001 0.0012 23,347,971
B 0.2098 0.2205 0.0680 0.9938 2.6955 0.0001 0.0012 23,287,304
C 0.2082 0.1558 0.0690 0.9938 2.9477 0.0001 0.0012 23,120,306
P1,1s P2,1s P2,2s P3,1s P3,1s P4,1s J
A 1.00 0.03 1.00 1.00 0.89 1.00 1.00 1.00 40.46
B 0.96 1.00 0.99 1.00 0.91 1.00 1.00 0.99 8.04
C 0.96 0.71 0.98 1.00 1.00 1.00 1.00 0.99 8.34
The best candidate is in bold
cost. The economic analysis is calculated by assuming that the distillation

column is operating in 5 years and 340 days per year. The result of the calcu-
lation is summarized in Table 3. From the results, distillation design at Point A
has the highest profit and this verified that the best design in terms of economic
criteria.
Stage 6: Final Selection and Verification. The objective of this stage is to
select the best candidate by analyzing the value of the multi-objective function.
The multi-objective function is calculated by summing up the objective function
value using Eq. 7. Each objective value is normalized with respect to the highest
value since the range of the objective function can be different. From the result
tabulated in Table 4, it is clearly seen that the multi-objective function, J for
distillation column design A is the highest compared to other designs due to it
has the best value for design, controller design, sustainability and economic.
Therefore, it is verified that the distillation column design at the maximum point
of driving force is an optimal solution for Sustain-IPDC of a Benzene-Toluene
separation process which satisfies the design, controller design, sustainability
and economic criteria.
Conclusion
As a conclusion, a systematic-based methodology has been developed for Sustain-

IPDC for a distillation column system. This methodology has been applied and
verified for a single distillation column. The presented case study of Benzene-
[email protected]
Toluene separation process has shown that the optimal design is at the highest point
of driving force, with respect to design, control, sustainability and economic
criteria. The optimal solutions were obtained in an efficient and systematic way.
Acknowledgment The financial support from Universiti Teknologi Malaysia (RUGS Tier 1
Q.J130000.2509.07H39) and Ministry of Education of Malaysia are highly acknowledged.
References
1. Hamid MKA, Sin G, Gani R (2010) Integration of process design and controller design for
chemical processes using model-based methodology. Comput Chem Eng 34:683–699
2. Shadiya OO, Hogh KA (2013) Sustainability evaluator: tool for evaluating process sustain-
ability. Environ Prog Sustain Energy 32:749–761
3. Uhlman BW, Sailing P (2010) Measuring and communicating sustainability through eco-
efficiency analysis. Chem Eng Process 106:17–26
4. Martins AA, Mata TM, Costa CAV, Sikdar SK (2007) Framework for sustainability metrics.
Ind Eng Chem Res 46:2962–2973
5. Sikdar SK (2003) Sustainable development and sustainability metrics. AIChE J 49(8):1928–
1932
6. Bek-Pedersen E, Gani R (2004) Design and synthesis of distillation systems using a driving
force based approach. Chem Eng Process 43:251–262
7. Hamid MKA (2011) Model-based integrated process design and controller design of chemical
processes, DTU PhD Thesis, ISBN: 978-87-92481-39-9
8. Li CS, Zhang XP, Zhang SJ, Suzuki KZ (2009) Environmentally conscious design of chemical
processes and products: multi-optimization method. Chem Eng Res Des 87:233–243 IChemE
[email protected]
Nucleation Kinetics
of Carbamazepine-Saccharin
(CBZ-SAC) Co-crystal
Khairool Azizul Mohammad, Syarifah Abd Rahim

and Mohd Rushdi Abu Bakar
Abstract In this study the effect of the heating/cooling rate of various CBZ-SAC
mixture in ethanol was investigated in order to deduce the metastable zone width
(MSZW) and the nucleation order. Nývlt’s equation involving the maximum
temperature difference DTmax between dissolution and crystallization temperatures
of CBZ-SAC co-crystals with the heating/cooling rate r was applied for the study of
nucleation kinetics by the polythermal method. A step towards the determination of
dissolution temperature of the crystals has been improved. The results showed that
dissolution temperature for CBZ-SAC co-crystal mixture increased with the
increase in heating rate, CBZ-concentration and the ratio of SAC to CBZ in ethanol
solution. The analysis revealed that a nucleation order for CBZ-SAC co-crystals
increases with the SAC to CBZ ratio and is in the range of 1.58–6.31.
Keywords Crystallization Co-crystal Carbamazepine Metastable zone

width Nucleation kinetics
K.A. Mohammad S. Abd Rahim (&)

Faculty of Chemical and Natural Resources Engineering, University Malaysia Pahang, 26300
Pahang, Malaysia
K.A. Mohammad
M.R. Abu Bakar
Department of Pharmaceutical Technology, Kulliyyah of Pharmacy, International Islamic
University Malaysia, Bandar Indera Mahkota, 25200 Kuantan, Pahang, Malaysia

[email protected]
264 K.A. Mohammad et al.
Introduction
Crystallization from solution has been studied intensively and widely used in many
industries and applications. The main purpose of crystallization is to separate and
purify solids in a suitable solvent which lead to the control of the shape and size of
final crystalline products. This technique provides an end compound with a high
degree of purity that has many desirable properties [1]. A broad knowledge of the
crystallization process is also essential to prevent the undesired forming of crystals
in industrial applications such as a tendency for crystal forming in biodiesel, which
can deteriorate the performance of fuel combustion [2].
There are two steps in forming a crystal/solid from a solution: (i) nucleation,
which the creation of minute nuclei in a supersaturated solution (ii) crystal growth
that transform the stable nuclei into solid crystals [1]. The crystallization occurs in a
metastable region, which exists in a supersaturated solution. Metastable zone width
(MSZW) is located between supersaturation and saturation limit of a system. The
optimization of a crystallization process and the quality of the forming crystal can
be simply controlled when the crystallization runs within the MSZW [2, 3]. The
factors that can affect the MSZW are cooling/heating rates, solution temperature,
impurities or seeds, stirring rate and solvent [2, 4–6]. Therefore, it is particularly
important to identify these parameters precisely for accessing the MSZW. An
analysis of the change in physical quantities such as particle count number, tur-
bidity, intensity (absorbance or transmittance), and electrical conductivity were
used in the research works as an indicator of the nucleation event [7]. Among these,
the turbidity measurement is an inexpensive method which several chemical pro-
cesses are using it [8].
The maximum super-cooling temperature ΔTmax described as the difference
between the dissolution temperature Tdiss and the crystallization temperature Tcrys at
which a detectable crystallisation commences: DTmax ¼ Tdiss Tcrys [5]. Figure 1
illustrates a saturation–supersaturation diagram of the crystals.
A solution of concentration C1 with a temperature Ta, above dissolution tem-
perature Tdiss1 was cooled at a constant cooling rate ‘R’ to a temperature ‘Tb’. First
crystals were detected at the temperature Tcrys1 in the solution. The solution was
Fig. 1 Solubility and supersolubility curves of a solid from solution (figure modified from [1])
[email protected]
Nucleation Kinetics of Carbamazepine-Saccharin … 265
then heated at a constant rate to a temperature Ta. The crystals were fully dissolved
at the temperature Tdiss1. Therefore, the maximum super-cooling ΔTmax at the
concentration C1 was the difference between the dissolution temperature Tdiss1 and
the crystallization temperature Tcrys1, i.e. DTmax ðC1 Þ ¼ Tdiss1 Tcrys1 . There were
saturation concentrations C2 and C3 of solute in the solutions corresponding to
dissolution temperature Tdiss2 and Tdiss3, respectively.
Figure 1 illustrates (i) the stable zone, which is unsaturated, and crystallization
cannot occur in this region; (ii) the metastable zone, which is supersaturated, and
spontaneous crystallization is unlikely to occur, but crystallization can occur by
adding a seed; (iii) the labile zone, which is non-stable zone and spontaneous
crystallization occurs mostly to in this zone [1].
Experimental
In this study, 250 ml vacuum jacketed vessel equipped with a stirrer run with a
speed of 300 rpm, a turbidity probe, and a valve for draining purpose. An analysis
software was connected for the sophisticated automation for control and analysis of
the operations. The turbidity sensor was calibrated before each experiment as 100 %
for a suspension of CBZ- SAC co-crystals and 0 % for a clear solution. Different
mol ratio (3.5, 3.0 and 2.5) of SAC to CBZ with different concentrations of etha-
nolic solutions containing 19.14, 17.96, 17.01 and 15.83 mg/ml CBZ in 200 ml
ethanol were made. Four different heating and cooling rates (0.25, 0.5, 0.75, and
1 °C/min) were applied in this crystallization experiments.
The initial suspension (3.5 SAC/CBZ ratio and 19.14 mg/ml CBZ in 200 ml
ethanol) was heated to 60 °C, which is higher than dissolution temperature. As can
be seen in Fig. 2, the stirred mixture was kept at 60 °C for 40 min to make sure that
the solid was fully dissolved. The mixture solution was afterward cooled at a rate
0.25 °C/min to a temperature T2, which is 2–3 °C lower than the crystallization
temperature Tcrys1. The temperature of the medium at 20 % turbidity serves as
crystallization temperature Tcrys1 of the co-crystal. The solution was again heated to
60 °C and left for 40 min to minimize memory effects and get the complete
dissolution of solid. The steps were repeated with a new cooling rate (0.5, 0.75, and
1 °C/min) to obtain the crystallization temperature (Tcrys2, Tcrys3, Tcrys4) of the co-
crystal and then with the different ratio of SAC to CBZ.
There was some modification in the stage to determine the dissolution temper-
ature Tdiss by comparing with previous researches [6, 9, 10] as illustrated in Fig. 3.
The solution was dissolved at 60 °C for 40 min and cooled with a fixed rate of
0.75 °C/min to a temperature Tb, which is 2–3 °C lower than the crystallization
temperature Tcrys3. The temperature Tb was kept constant for 40 min between point
b and c. The solution was then heated with the different rate of 0.25, 0.5, 0.75, and
1.0 °C/min. Different heating rates resulted in different dissolution temperatures
(Tdiss1, Tdiss2, Tdiss3, and Tdiss4). Solution at point e, f and g follows the same
[email protected]
Fig. 2 Method for determining the crystallization temperature from the solution. Lines AB and CD
are with cooling rate of 0.75 and 0.5 °C/min respectively
Fig. 3 New approached method for determining the dissolution temperature of the crystal. Lines
abc and efg have the same cooling rate and isothermal temperature. Lines cd and gh are with
heating rate of 0.75 and 0.5 °C/min respectively
[email protected]
profiles as point a, b, and c that resulted in the similar properties of co-crystal at a

point c and d. Therefore, the proper comparison can be done for the effect of
different heating rate on MSZW.
The data were processed to obtain the solubility–supersolubility curves and

nucleation. The effect of heating and cooling rate R on Tdiss and Tcrys at different
concentrations and different ratio of SAC/CBZ can be plotted in polynomial form
(T ¼ ar2 þ ba þ c). The c-value is the interception of the curve at zero heating/
cooling rates. This value can be taken as the equilibrium dissolution temperature
Tdiss0 and the equilibrium crystallization temperature Tcrys0 [6]. The results for
Tdiss0 and Tcrys0 are shown in Table 1. The graph of temperature (Tdiss and Tcrys)
versus cooling/heating rate for the CBZ concentration of 19.14 mg/ml and different
ratio of SAC to CBZ is plotted as shown in Fig. 4. The results show that Tdiss for
CBZ-SAC co-crystal increases with the increase of the heating rate and the ratio of
SAC to CBZ in ethanol solution. The graph also represents that Tcrys for CBZ-SAC
co-crystal decrease with increasing heating rate and the ratio of SAC to CBZ in
ethanol solution.
The effect of CBZ concentrations on MSZW is shown in Fig. 5. This figure
shows that as the CBZ-concentration increases, Tdiss increases and Tcrys decreases
with heating/cooling rate. These results are in accordance with previous works
Table 1 Tcrys0, Tdiss0, and dC/dTdiss0 value of CBZ-SAC in ethanol

SAC/CBZ CBZ concentration, C (mg/ml) dC/
mol ratio 19.14 17.96 17.01 15.83 dTdiss0
Tcrys0 Tdiss0 Tcrys0 Tdiss0 Tcrys0 Tdiss0 Tcrys0 Tdiss0
(°C) (°C) (°C) (°C) (°C) (°C) (°C) (°C)
3.5 45.98 55.56 40.15 54.00 39.06 52.05 35.32 43.82 0.252
3.0 44.91 53.49 35.00 50.50 31.50 50.07 32.15 39.70 0.212
2.5 35.62 51.22 30.28 49.75 27.81 47.00 26.61 38.74 0.236
Fig. 4 MSZW for CBZ concentration of 19.14 mg/ml and different ratio of SAC to CBZ
[email protected]
Fig. 5 MSZW for ratio (SAC/CBZ) of 2.5 and different values of CBZ concentration
Table 2 MSZW of CBZ- SAC/CBZ mol ratio MSZW (°C)

SAC co-crystals
CBZ concentration, C (mg/ml)
19.14 17.96 17.01 15.83
3.5 9.57 13.42 12.99 8.50
3.0 8.58 15.07 18.57 7.55
2.5 15.60 19.48 19.19 12.13
[5, 10]. From the Tcrys0 and Tdiss0 results in Table 1, the graph concentration of
CBZ, C against Tdiss0 was plotted and the slope of the graph dC/dTdiss0 was
obtained.
Table 2 shows the MSZW of CBZ-SAC co-crystals. MSZW is measured by the
difference between Tdiss0 and Tcrys0 at equilibrium (zero rates) position. The lowest
concentration of CBZ i.e., 15.83 mg/ml shows the smallest MSZW for all (SAC/
CBZ) mol ratios. This trend is in accordance with previous work [5]. Moreover, for
all CBZ concentration, an increase in (SAC/CBZ) mol ratio resulted in a decrease in
MSZW; thus crystallization is easier to occur.
Using Nyvlt’s approach, the nucleation rate J is defined as J ¼ kðDCmax Þm ,
where k is the mass nucleation rate constant, m is the nucleation order, and DCmax is
the absolute supersaturation (kgm−3). ΔC is equal to ðdC=dTÞDT. As known pre-
viously cooling rate R is equal to DT=Dt. J also can be defined as J ¼ ðdC=dTÞR.
By substitutes all these equation, the following expression is obtained:
lnðRÞ ¼ ðm 1ÞlnðdC=dTÞ þ lnðkÞ þ m lnðDTÞ. By plotting ln(R) versus lnðDTÞ,
the value of m and k can be calculated. Value m is the slope of the graph, and the
interception b is equal to ðm 1Þ lnðdC=dTÞ þ lnðkÞ. Figure 6 shows the graphs of
Fig. 6 Ln R versus ln MSZW for CBZ concentration of 19.14 mg/ml
[email protected]
Table 3 Nucleation order, m and the y-interception value of ln R versus ln MSZW, b

SAC/CBZ CBZ concentration, C (mg/ml)
mol ratio 19.14 17.96 17.01 15.83
m b m b m b m b
3.5 2.75 −9.03 3.51 −11.64 2.61 −9.14 1.59 −5.23
3.0 2.54 −8.43 3.46 −11.88 4.82 −16.77 2.19 −6.83
2.5 3.62 −12.66 4.05 −14.14 6.31 −21.57 3.87 −11.85
Table 4 Nucleation kinetic constant of CBZ-SAC in ethanol

SAC/CBZ mol ratio Nucleation kinetic constant, k
CBZ concentration, C (mg/ml)
19.14 17.96 17.01 15.83
3.5 1.33E−03 2.79E−04 9.88E−04 1.20E−02
3.0 2.39E−03 3.18E−04 1.97E−05 6.92E−03
2.5 1.41E−04 5.95E−05 9.23E−07 4.52E−04
ln(R) versus ln MSZW for CBZ concentration of 19.14 mg/ml and different ratio of
(SAC/CBZ). Values of m and b are shown in Table 3 and the calculated nucleation
kinetics k is given in Table 4. The analysis reveals that the nucleation order m for
CBZ-SAC co-crystals increases with the SAC to CBZ ratio and have the average
value of 3.44. This value is acceptable as reported in the literature that, the values of
m are usually in the range of 1.65–4.9 for organic compounds [11].
Conclusion
In conclusion, this newly improved method is an accurate method for determination

of dissolution temperature of the co-crystals. In this method, the supersaturated
solution is assumed to have the same property and structure of the crystals due to
the same profile of the previous cooling during nucleation event. Therefore, the
proper comparison can be done for the effect of different heating rate on MSZW. It
can also be concluded that the cooling/heating rate can strongly affect the MSZW of
a CBZ-SAC mixture in ethanol solution. The MSZW is wider for a higher heating/
cooling rate. Moreover, the mol ratio of SAC to CBZ and the concentration of CBZ
gave the tremendous effect on dissolution and crystallization temperature of the
CBZ-SAC mixture in ethanol solution. Another advantage of this technique is that
the machine can be pre-set according to desired profile, thus saving the human
labour.
[email protected]
References
1. Mullin JW (2001) Crystallization. Butterworth-Heinemann, London

2. Mota FL, Teychéne S, Biscans B (2014) Measurement of the nucleation and growth kinetics of
some middle distillate fuels and their blends with a model biodiesel fuel. Ind Eng Chem Res
53(7):2811–2819
3. Kannan V, Brahadeeswaran S (2013) Investigations on influence of coloration on growth of
high quality hydrazonium l-tartrate single crystal. J Cryst Growth 374:71–78
4. Kulkarni SA, Kadam SS, Meekes H, Stankiewicz AI, ter Horst JH (2013) Crystal nucleation
kinetics from induction times and metastable zone widths. Cryst Growth Des 13(6):2435–2440
5. Qian Y, Lu G, Sun Y, Song X, Yu J (2014) Metastable zone width of SrCl2·6H2O during
cooling crystallization. Cryst Res Technol 49(1):78–83
6. Corzo DMC, Borissova A, Hammond RB, Kashchiev D, Roberts KJ, Lewtas K, More I (2014)
Nucleation mechanism and kinetics from the analysis of polythermal crystallisation data:
methyl stearate from kerosene solutions. CrystEngComm 16(6):974–991
7. Kubota N (2008) A new interpretation of metastable zone widths measured for unseeded
solutions. J Cryst Growth 310(3):629–634
8. Bernardo A, Giulietti M (2010) Modeling of crystal growth and nucleation rates for
pentaerythritol batch crystallization. Chem Eng Res Des 88(10):1356–1364
9. Bernardes CE, da Piedade MEM (2010) A fully automatic apparatus for thermal analysis of
crystallization from solution and metastable zone width determinations. J Therm Anal Calorim
100(2):493–500
10. Bonnin-Paris J, Bostyn S, Havet J-L, Fauduet H (2011) Determination of the metastable zone
width of glycine aqueous solutions for batch crystallizations. Chem Eng Commun 198
(8):1004–1017
11. de Arunmozhi G, de M Gomes E (2004) Metastability and crystal growth kinetics on
L-arginine phosphate. Cryst Res Technol 39(1):34–39
[email protected]
Part V
Advanced Materials
and Nanotechnologies
[email protected]
Activation of Titanium Dioxide Under
Visible-Light by Metal and Non-metal
Doping
Krishnan Jagannathan, Sikirman Arman

and Nerissa Mohamad Elvana
Abstract Surface modification of the titanium dioxide, TiO2 by doping with

nitrogen or iron in order to produce visible-light active photocatalyst was investi-
gated. Solgel method was used in preparing the modified TiO2. Tetra titanium
isopropoxide, ammonium nitrate and ferric nitrate were used as precursors with the
dopant concentration of 1 % and calcination temperature of 600 °C. The prepared
photocatalyst samples were characterized by XRD, BET and UV-vis DRS in order
to study their physical properties. The results from XRD demonstrated that all the
photocatalysts were anatase phase with high surface area. The reduction of band
gap energy was shown by adding a non-metal dopant into TiO2 lattice. Meanwhile,
FTIR results revealed the presence of N and Fe dopant in the prepared photocat-
alysts. The properties of the photocatalyst affected its effectivity which was tested
by performing a standard batch photocatalytic degradation experiment with meth-
ylene blue as a model pollutant under visible light. The result showed that N-doped
TiO2 and Fe-doped TiO2 photocatalyst degraded the methylene blue within five
hours of irradiation time yielding photodegradation efficiency of 72.3 and 61.98 %,
respectively.

Keywords Doping Titanium dioxide Visible light Solgel Photocatalytic

degradation Nitrogen Iron
K. Jagannathan (&) S. Arman N.M. Elvana

Fakulti Kejuruteraan Kimia, Universiti Teknologi MARA, 40450 Shah Alam, Selangor,
Malaysia
S. Arman
N.M. Elvana

[email protected]
274 K. Jagannathan et al.
Introduction
Photocatalysis is one of the highly effective applications in destructing organic

pollutants to harmless end products at ambient conditions using light and photo-
catalyst [1]. Titanium dioxide (TiO2) is the commonly used photocatalyst in many
environmental applications such as decolorizing of dyes, degradation of organic
pollutants, water purification and hydrogen production from water splitting. It is
due to its desirable characteristics such as economical, environmental friendly,
structural stability and high degradation rate [2, 3]. However, pure TiO2 Titanium
dioxide can be used only under UV light because of its high band gap energy
(3.2 eV). The application of UV is highly energy intensive and hence it makes the
process expensive. Besides, in the solar light spectrum, the UV light exists only 3–
5 % compared to visible-light (45 %) thereby limiting the usage of sun light. In
order to activate TiO2 under visible-light (400 < λ < 500 nm), it is necessary to
modify its surface by doping it with certain elements.nitrogen (N) and iron (Fe)
were selected as dopants in this study because they both could prevent the unde-
sirable recombination of the electron-hole pair, otherwise which would decrease the
performance of the photodegradation [4].
Iron doped TiO2 under visible-light showed better performance in degrading the
toluene gases as compared to pure TiO2 under visible-light [5]. The presence of iron
dopant would incorporate into TiO2 lattice, which made it photo-response to visi-
ble-light and thus increased the photocatalytic activity. Meanwhile, nitrogen dopant
induced reduction of the band gap between conduction and valence band [6].
Besides, the nitrogen dopant absorbed the visible-light to generate electron-hole
pair and thereby finally increasing the photocatalytic activity. In this study, nitrogen
and iron doped TiO2 samples were prepared by using solgel method and their
surface properties were compared. Finally, the effectivity of prepared photocatalysts
was tested by performing a standard batch photocatalytic degradation experiments
under visible-light with methylene blue as a model pollutant.
Materials and Method
Preparation of Photocatalyst, N doped TiO2 was prepared using Solgel method

which had been described in our previous study [2]. Titanium (IV) isopropoxide
(TTIP) (Merck, USA), and Ammonium nitrate (NH4NO3) (Merck, USA) were used
as starting materials in preparing the non-metal doped photocatalyst. These pre-
pared photocatalyst formed were labeled as N-TiO2. For metal doped TiO2, the
photocatalyst prepared were done by changing the precursors to Ferric nitrate (Fe
(NO3)3) (Friendemann Schmidt, Australia) as the source for iron dopant and labeled
as Fe-TiO2.
Characterization of Photocatalyst, The photocatalysts were analyzed using
X-ray diffractometer (XRD, Rigaku Ultime IV) with Cu K∝ radiation sources and
[email protected]
Activation of Titanium Dioxide Under Visible-light … 275
operating at a scanning rate of 8 °/s in a 2θ range of 20–80° to determine the phase

of the crystal formed. The photocatalysts were then analyzed using AUTOSORB-1
Quanta Chrome Instruments (USA) at −196 °C to obtain the BET (Brunauer,
Emmett, and Teller) surface area. Fourier Transform Infrared (FTIR) Spectroscoy
(Perkin Elmer Spectrum One) was used to determine the specific absorption
peak wave number in the range of 4000–500 cm−1 with 4 cm−1 of resolution.
Finally, the band gap of photocatalysts surface was determined using UV-vis
Spectrophotometer (Lambda 35, Perkin Elmer) where BaSO4 was used as a
reflectance standard in UV-vis diffuse reflectance method.
Photocatalytic Degradation, The performance of the photocatalyst was deter-
mined by measuring the removal efficiency of Methylene Blue (MB) under visible
light. The batch experiments were conducted only for a period of five hours of
irradiation time to compare the relative performance of the prepared photocatalysts.
A 25 W Phillips Compact Fluorescence Lamp emitting at an average rate of
2.448 W/m2 (555 nm) was used as the visible-light source. A 0.05 g of N-TiO2
photocatalyst was added into 50 ml of MB solution having a concentration of
10 ppm. The MB solution was stirred for 30 min in the dark condition to ensure
equilibrium adsorption prior to photodegradation. Then, the photocatalytic degra-
dation experiment was run under visible-light meanwhile the samples were col-
lected every 30 min for 5 h for the analysis of its residual concentrations. To ensure
the photocatalyst separated from the MB solution, it was centrifuged for 5 min at
10,000 rpm. The decolorized MB was analyzed using UV-vis Spectrophotometer at
660 nm. The experiment was repeated for Fe-TiO2 photocatalyst.
XRD, Fig. 1 shows the XRD patterns of the photocatalysts. Fe-TiO2 and N-TiO2
photocatalysts approximately fitted with the JCPDS file number (PDF no. 21-1272)
where the samples were of anatase phase of TiO2 with the characteristic features
and diffraction peaks at 25.281° (101), 37.80° (004), 48.05° (200), 53.89° (105),
55.06° (211), 62.688° (204), 68.76° (116), and 75.029° (215). The photocatalysts
were found in anatase phase due to the appropriate combinations of dopant con-
centration and calcination temperature used in the preparation process of photo-
catalysts as discussed in our previous study [2, 3]. The main peak patterns for both
Fe-TiO2 and N-TiO2 were found at 25.361° and 25.379°, respectively. Both metal
and non-metal dopants showed a shift to the right where iron dopants got shifted
more compared to non-metal dopant. This was due to the formation of crystallo-
graphic point defects by the substitution of titania ion, Ti4+ by iron ion, Fe3+ owing
to their similarities in ion radii [7] where most of the iron ions possibly got into
TiO2 Titanium dioxide structure at interstices or the ions might be occupied at the
titanium lattice sites to form a solution with iron-titanium solids [8]. Meanwhile, the
existence of the nitrogen dopant did not change the basic crystalline structure of
TiO2 but had intermingled within TiO2 surface [7].
[email protected]
Fig. 1 XRD patterns of all (101)

PDF#21-1272
prepared photocatalyst
(200)
(204) (215)
Relative Intensity, cps

(004)
(105) (211)
(116)
(220) N-TiO
2
Fe-TiO2
20 30 40 50 60 70 80
2
The presence of dopants was not observed by XRD, which might be due to
either uniform distribution of dopant in TiO2 lattice or the amount of dopant was
below the XRD detection limits [2]. From Fig. 1, as the peak was found slightly
broaden for metal dopants (iron) due to the effect of the crystal size lattice of TiO2
on the dopant material. The particle size based on the main peaks was calculated
using Scherrer equation and shown in the Table 1. Between metal and non-metal
dopant, nitrogen dopant showed bigger particle size compared to iron dopant. This
was due to the functionality of the dopant where Fe3+ might be substituted the Ti4+.
In addition, nitrogen did not substitute any ion and only existed at the surface of the
TiO2 because of its functionality on reducing the band gap energy as discussed in
the next section under the characterization of UV-vis DRS [9].
Surface area, Table 1 presents BET surface area for both the photocatalysts.
The table shows that the BET specific surface area of Fe-TiO2 and N-TiO2 were
42.96 and 74.72 m2/g, respectively. In general, BET surface area of metal or non-
metal can be related to the functionality of the dopant on TiO2 surface. Nitrogen
dopant showed higher surface area compared to iron dopant because it was doped at
the surface of TiO2 providing more surfaces for active sites. Whereas iron dopant
had almost the same BET surface area of pure TiO2 as Fe3+ substituted Ti4+ ion.
This was in good agreement with the XRD data as discussed in the previous
section. The higher surface area resulted in the adsorption of more and more
hydroxyl groups photo-generated by electron-hole pairs thereby enhancing active
surface hydroxyl radicals [10] which are the key component in the photocatalytic
degradation mechanism.
Table 1 Physical properties from XRD data and BET surface area
Photocatalysts Diffraction peak (main) (°) Crystal size (nm) BET surface area (m2/g)
Pure TiO2 25.221 74 42.82
Fe-TiO2 25.379 19 42.96
N-TiO2 25.361 25 74.72
[email protected]
Fig. 2 UV-vis spectra of

prepared photocatalysts
Absorption, a.u
N-TiO2
Fe-TiO2
300 400 500 600 700 800

390 440
Wavelength, nm
UV-vis DRS, Optical absorption spectra for photocatalysts are shown in Fig. 2.
The band gap of the photocatalyst was estimated based on the Kubelka-Munk
equation as follows [11],
1240
Eg ¼ ð2Þ
k
where, Eg is the band gap (eV) and k is the wavelength of absorption edge (nm).
The spectrum of both the photocatalysts showed a red shift indicating that they can
be activated at visible light. The absorption edge and band gap energy of Fe-TiO2
and N-TiO2 were 390 (3.18 eV) and 440 (2.81 eV) nm, respectively. Compared to
Fe-TiO2, N-TiO2 could be a promising photocatalysts which can be applied in
photocatalytic degradation under visible light. Moreover, it was clear that the
function of nitrogen dopant was to narrow down the band gap energy of photo-
catalyst by modifying the electronic structure at the conduction band of TiO2
thereby enhancing the photocatalytic degradation performance under visible-light
[5]. The iron dopant showed a very little reduction of the band gap which might be
due to the small amount of isolated Fe oxides existing in the sample in line with
XRD patterns. [6].
FTIR, The FTIR spectra of the photocatalysts are shown in Fig. 3. N-TiO2
photocatalyst showed the obvious peaks of elements present in the samples. There
were two spectra existed within 1500–1100 cm−1 which can be attributed to the
vibrations of the Ti-N bond [3]. Meanwhile, the Fe-OH band was demonstrated at
about 1200 cm−1 [3]. The existence of Ti-N bond and Fe-OH bond evidenced the N
and Fe species in each sample incorporated into TiO2 lattice, respectively.
However, N-TiO2 photocatalyst showed strong peaks than Fe-TiO2 photocatalyst
which was in accordance with the XRD results as discussed earlier.
Photocatalytic Degradation, The transient photodegradation profiles were
shown in Fig. 4. The percentage removal for Fe-TiO2 and N-TiO2 were 61.98 %,
and 72.30 %, respectively. It demonstrated that the photocatalyst N-TiO2 was more
effective than Fe-TiO2. This result was in good agreement with the XRD, BET,
[email protected]
95.2 95.5
94
94
92
93
3701.42 90
92 3256.15 2160.73
2351.81 88
2376.78
91 2332.15
86
90 84
89 82 1637.73
3714.45
88 80 3701.48
3661.88 2321.65
2375.87 2159.46
87 78 2352.04
2332.20
86 76 3237.60
%T
%T
85 74
84 72
83 70
82 68
81 66
80 64
79 62
60
78
58
77
56
76
75 Fe-TiO2 667.97 54
52
N-TiO2 667.91
74
50
73.0 48.8
4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 515.0 4000.0 3600 3200 2800 2400 2000 1800 1600 1400 1200 1000 800 600 515.0
cm-1 cm-1
Fig. 3 FT-IR spectra of prepared photocatalysts
Fig. 4 Photodegradation 70
profile for sample (A) 1.0 % N-TiO2
Removal Efficiency, %
Fe-TiO2, (B) 1.0 % N-TiO2 60
50
Fe-TiO2
40
30
20
10
0
0 100 200 300
Irradiation time, min
UV-vis DRS and FTIR data where N -TiO2 showed a strong anatase peak, higher
surface area and lower band gap energy respectively which modified the photo-
catalyst to be more activate under visible light. Strong anatase peaks with high BET
surface area represented strong oxygen interaction with the doped TiO2 surfaces,
which resulted in higher photocatalytic activity under visible-light [3]. Generally,
each dopant plays important role in photocatalytic degradation mechanism in
treating MB. Non-metal dopant, nitrogen was responsible for reduction of band gap
energy while the metal dopant, iron was responsible for substituting Ti4+ ion with
Fe3+ in the TiO2 lattice [6].
Conclusion
Non-metal and metal doped photocatalysts were prepared by solgel method and the
photocatalyst with N-TiO2 was found to be very effective in degrading the pollutant
under the illumination of ordinary domestic lamps. The presence of dopant
[email protected]
concentrations was witnessed by FTIR spectra and strong anatase phase transfor-
mation of TiO2 was confirmed by the XRD profile. Besides having a higher BET
surface area, the lower band gap energy of 2.18 eV turned the nitrogen doped
photocatalyst to be active under ordinary visible light. The results from the standard
photodegradation tests with methylene blue showed that photocatalyst, N-TiO2
yielded a higher photodegradation efficiency of 72.30 % in five hours of irradiation
time compared to that of Fe-TiO2.
Acknowledgment The authors thank Research Management Institute (RMI), Universiti

Teknologi MARA (UiTM) and Ministry of Science, Technology and Innovation (MoSTI) of
Malaysia for the financial support through e-Science Grant No: 100-RMI/SF 16/6/2 (19/2012).
References
1. Jagannathan K, Swaminathan T (2011) Removal of gas-phase benzene in an immobilized

photocatalytic reactor. Maced J Chem Chem Eng 30(2):221–228
2. Arman S, Jagannathan K, Junaidah J, Farahziehan S (2014) Preparation, characterization and
effectivity of N, Fe-TiO2 as a visible-light active photocatalyst. Adv Mater Res 894:245–249
3. Arman S, Jagannathan K (2014) Effect of calcination temperature on synthesis of N-Fe-TiO2
for photocatalytic degradation of methylene blue under visible light. Appl Mech Mater
575:55–59
4. Behpour M, Atouf V (2012) Study of the photocatalytic activity of nanocrystalline S, N-
codoped TiO2 thin films and powders under visible and sun light irradiation. Appl Surf Sci
258:6595–6601
5. Sun S, Ding J, Bao J, Gao C, Qi Z, Yang X, He B, Li C (2012) Photocatalytic degradation of
gaseous toluene on Fe-TiO2 under visible-light irradiation: a study on the structure, activity
and deactivation mechanism. Appl Surf Sci 258:5031–5037
6. Sun J, Qiao L, Sun S, Wang G (2008) Photocatalytic degradation of ornage G on Nitrogen-
doped TiO2 catalyst under visible-light and sunlight irradiation. J Hazard Mater 155:312–319
7. Jagannathan K, Elvana MN, Arman S (2014) Effect of calcination temperature of carbon
doped TiO2 (C-TiO2) on photocatalytic activity under visible light. Adv Mater Res 997:
292–296
8. Lin CJ, Liou YH, Zhang Y, Chen CL, Dong CL, Chen SY, Stucky GD (2012) Mesoporous
Fe-doped TiO2 sub-microspheres with enhanced photocatalytic activity under visible-light
illumination. Appl Catal B 127:175–181
9. Shen XZ, Guo J, Liu ZC, Xie SM (2008) Visible-light drien titania photocatalyst co-doped
with nitrogen and ferrum. Appl Surf Sci 254:4726–4731
10. Hu S, Wang A, Li X, Lowe H (2010) Hydrothermal synthesis of well-dispersed ultrafine N-
doped TiO2 nanoparticles with enhanced photocatalytic activity under visible light. J Phys
Chem Solids 71:156–162
11. Li X, Zhang H, Zheng X, Yin Z, Wei L (2011) Visible-light responsive N-F-codoped TiO2
photocatalysts for the degradation of 4-Chlorophenol. J Environ Sci 23(11):1919–1924
[email protected]
Reflectometric Optosensor for Visual
Detection of Ammonia Based on Silica
Pellet Sensing Material
Nur Syarmim Mohamed Noor, Ling Ling Tan,

Siti Nurhafizah Zainuddin, Kwok Feng Chong, Lee Yook Heng
and Saiful Nizam Tajuddin
Abstract Ammonia (NH3) has been widely used in the manufacture of fertilizers
that applied to soil, but the high consumption of fertilizers will end up with water
pollution. Owing to the deleterious effects of NH3 to human and environment, a
new optosensor for NH3 has been fabricated based on silica pellet sensing material.
Microsilica was synthesized by sol-gel method in the presence of cobalt(II) chloride
hexahydrate (CoCl2·6H2O), followed by manual grinding process to obtain micro-
sized silica particles. Due to the non-transparent pellet material used for NH3
sensing, a fiber optic reflectance spectrophotometer was employed for monitoring
of reflectance signal transduction event as the pellet colour changed from pink to
blue hue upon reaction with NH3 at optimum pH 13. Due to the high porosity and
surface area of silica microparticles were used as immobilization matrix, the
immobilized Co2+ ion demonstrated broad dynamic linear range from 18 to
100 ppm NH3 with a fast response time of 3 min. The reflectometric sensing
protocol involves a single-step NH3 assay which merely requires dispensing small
aliquots of NH3 onto the reaction surface of the pellet sensor. This makes on-site
N.S. Mohamed Noor S.N. Zainuddin K.F. Chong S.N. Tajuddin

Faculty of Industrial Sciences and Technology, Universiti Malaysia Pahang,
26300 Gambang, Kuantan, Pahang, Malaysia
S.N. Zainuddin
e-mail: siti.nurhafi[email protected]
K.F. Chong
S.N. Tajuddin
L.L. Tan (&) L.Y. Heng
Southeast Asia Disaster Prevention Research Initiative (SEADPRI-UKM), LESTARI,
Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia
L.Y. Heng

[email protected]
282 N.S. Mohamed Noor et al.
NH3 detection more user-friendly and convenient when compared to traditional

electrochemical-, infrared- and gas chromatography-based methods.
Keywords Optosensor Ammonia detection Reflectance spectrophotometer

Sol-gel
Introduction
NH3 is produced by nitrogen fixation in nitrogen cycle and occurred naturally in

soil from bacterial processes. NH3 is the precursor for fertilizer productions. NH3
can be either converted into solid fertilizers e.g. urea, ammonium nitrate, ammo-
nium phosphate or directly applied to arable soil [1]. The excess of NH3 in sur-
rounding may create potential hazards to human and ecosystems. This indicates that
determination of NH3 is of paramount importance in ecological monitoring.
Ministry of Health Malaysia has regulated the standard limit for NH3 in
Ammonia detection both raw and drinking waters is less than 1.5 ppm [2, 3],
meanwhile, according to WHO [4], NH3 threshold odour concentration at alkaline
pH is approximately 1.5 ppm and a taste threshold of 35 mg/L is advocated for
ammonium (NH4+) ion.
NH3 detection based on wet chemical methods for example titrimetry and gra-
vimetry require long analysis time, high chemical consumption and non-continu-
ous. Several conventional methods for quantitation of NH3 such as fourier
transform infrared spectroscopy (FTIR), non-dispersive infrared gas analyzer,
electrochemical method [5], gas chromatography (GC) [6] and GC coupled with
fluorescence detection [7, 8]. Electrochemical method based on potentiometric
electrodes are sensitivity, yet still they sustain strong interference from alkali metal
ions such as Na+ and K+ ions. Other traditional methods for NH3 assay are gen-
erally cumbersome, time-consuming, consumptive of the analyte, expensive and
rather inaccurate. Therefore, alternative means for rapid detection of NH3 based on
low cost and handy device is highly required.
Over the past two decades, the development and applications of optical chemical
sensors (optodes) have flourished rapidly which applied for determination of var-
ious analytes of interest, including cations, anions, neutral and gaseous species [9].
In terms of portability, low cost and fast assay, optosensor is beneficial for swift
detection of NH3 pollutants in the environment.
Experimental
Reagents
Sol-gel was prepared by mixing tetraethyl orthosilicate (Si(OC2H5)4, 98 %, Acros
Organics), ethanol (C2H5OH, 99.8 %, R&M Chemicals) and deionized water at a
[email protected]
Reflectometric Optosensor for Visual Detection … 283
molar ratio of 1:4:16 followed by addition of 3–4 drops of 12 M hydrochloric acid

(HCl, 37 %, Friendemann Schmidt). The mixture was then stirred until a homog-
enous sol solution obtained at room temperature (25 °C). Phosphate buffer solution
was prepared by mixing monosodium phosphate (NaH2PO4, 99 %, Bio Basic Inc.)
and disodium phosphate (Na2HPO4, 99.8 %, Sigma Aldrich) in deionized water.
The buffer pH was later adjusted with 2 M sodium hydroxide (NaOH, 99 %,
Friendemann Schmidt). NH3 stock solution was prepared by dissolving ammonium
chloride (NH4Cl, 99.8 %, Sigma Aldrich) salt in deionized water.
Instrumentation
Reflectance intensity measurements of pellet sensor were accomplished by using
fiber optic reflectance spectrophotometer (Ocean Optics, USB4000-UV-Vis).
Mettler Toledo pH meter was used to measure buffer pH. Silica pellet was made
using Parr 2811 Pellet Press.
Preparation of Silica-based Pellet Sensor
About 0.049 g of CoCl2·6H2O (99 %, R&M Chemicals) salt was dissolved in silica
sol-gel. The sol solution was stirred for 1 h and kept at room temperature for 24 h.
The solution was then oven-dried at 80 °C for 1 h in a petri dish in order to obtain
the gel monolith. Then, the sol-gel disc immobilized with Co2+ ion was manually
ground by using mortar and pestle and sieved to obtain microsilica at <75 µm. The
CO2+ ion immobilized microsilica was then mixed with methylcellulose (Sigma
Aldrich) at a volume ratio of 1:1 and transferred into the die set of pellet press tool
to be manually compressed to obtain a 12.7 mm diameter × 2 mm height sensor in
circular pellet format.
Optimization on NH3 Pellet Sensor
pH effect on the optosensor was studied by varying the pH of the reaction medium
from pH 1 to pH 13. Dynamic linear range of the NH3 optosensor was determined
at 637 nm using various NH3 concentrations at pH 13. Detection limit of the
optosensor was determined based on the summation of average blank response and
three times the standard deviation of the blank. Response time study of the opto-
sensor was accomplished by dispersing constant NH3 concentration solution on the
pellet sensor and the reflectance response was measured every 30 s for 10 min.
Interference study was conducted by adding elements (e.g. Na+, K+, Ca2+, Mg2+
and SO42− ions) that are commonly co-exist with NH3 in the water system onto the
pellet sensor at various concentrations, and NH3 concentration was maintained at
50 ppm. t-test was used to determine significant difference in the sensor responses.
Characterization of NH3 Pellet Sensor
The immobilized Co2+ ion is pink in colour as it is a transition metal ion with
partially filled d-orbital, and gave maximum reflectance intensity at the wavelength
[email protected]
Fig. 1 Reflectance spectra of

the immobilized Co2+ ion
(a) and its complex with
50 ppm NH3 (b) at pH 13
of 630 nm. When Co2+ ion-immobilized silica pellet reacted with NH3, it changed
to blue colour and gave maximum reflectance at 683 nm. However, the reflectance
intensity was diminished when the bright pink-coloured Co2+ ion silica pellet was
turned to dark blue. Figure 1 depicts the reflectance spectra of immobilized Co2+
ion and its NH3 complex at pH 13.
The chemical reaction taken place between Co2+ ion and NH3 can be described
with Eq. 1:
2þ
CO2þ ðaqÞ þ 6NH3ðaqÞ COðNH3 Þ6 ðaqÞ ð1Þ
In general, NH3 acts as both a base and a ligand. When an aliquot of NH3 is
introduced towards immobilized Co2+ ion, NH3 molecules replace water ligands
attached to the pink hexaaquacobalt(II) [Co(H2O)6]2+ ion and that dark blue colours
of immobilized hexamminecobaltate(II) ([Co(NH3)6]2+) complex are formed. The
formation of dark blue immobilized [Co(NH3)]2+ 6 complex absorbed incident light
transmitted by the bifurcated optical fiber and attenuated the reflectance signal over
the visible wavelength. The immobilized Co2+ ion, on the other hand, demonstrated
higher reflectance intensity as bright colours reflect better than dark. Because
maximum reflectance difference between immobilized reagent and complex was
observed at 637 nm, therefore, this wavelength was used as working wavelength for
subsequent experiments.
pH Effect
Basically, NH3 in water exist in equilibrium with NH4+ ion as shown in Eq. 2:
NH4 þ þ H2 O NH3 þ H3 Oþ ð2Þ
The equilibrium is extremely dependant on the pH where high pH favours the

formation of NH3 and vice versa [10]. Figure 2 shows that the optimum response
for NH3 pellet sensor was obtained at pH 13 and 637 nm, whereby the largest
relative reflectance acquired between immobilized reagent and complex.
[email protected]
Fig. 2 The effect of pH on the immobilized [Co(NH3)6]2+ complex at 637 nm using 50 ppm NH3
at pH 13
From pH 1 to pH 6, the acidic conditions favouring the protonation of NH3,

thereby the formation of blue-coloured [Co(NH3)6]2+ complex was not discernible,
and rendered higher reflectance signal which was relatively similar to that of the
immobilized Co2+ ion with no NH3 was added. As the pH increased from pH 8 and
onwards, the blue colouration of silica pellet was getting obviously developed. The
deprotonation of analyte was high in basic conditions especially from pH 11.6 to pH
12.5, and the blue colour became darkest at pH 13 due to the presence of excess NH3
in basic conditions, which promoted the formation of immobilized [Co(NH3)6]2+
complex [11]. Thus, optimum pH 13 was applied for further optimization analyses.
NH3 Detection
The silica pellet sensor was tested with a range of NH3 concentrations. The
reflectance signal increased linearly with increasing NH3 concentration and grad-
ually levelling off at a steady-state relative reflectance from 500 ppm NH3 onwards
(Fig. 3).
The calibration curve of relative reflectance versus NH3 concentration was found
linear over the range of 18–100 ppm NH3 with a correlation coefficient (R2) of
0.9866. The limit of detection (LOD) of the NH3 pellet sensor was calculated to be
15 ppm NH3. The pellet sensor was found taking 3 min for full colour development
upon addition of NH3 solution. Therefore, 3 min reaction time was fixed for every
optical NH3 detection. The measurement for identical addition of analyte was
reproducible at 0.5 % RSD (n = 3), and the sensor repeatability was obtained at
<5.0 % RSD (n = 5) indicated that the NH3 pellet sensor can be satisfactorily reused
for in situ environmental monitoring of NH3 pollutant. The NH3 optode retained
almost 100 % of its initial reflectance response over 6 months’ storage period at
ambient temperature indicative of a highly stable immobilized reagent. Based on
the linear range obtained, it can be deduced that this silica-based pellet sensor can
be used for continuous monitoring of NH3 in sewage, domestic waste water, landfill
leachate, etc. whereby the typical NH3 levels in those waters are normally between
[email protected]
Fig. 3 The response curve of the pellet sensor generated using NH3 concentrations from 18 to
100 ppm at pH 13. The inset shows the dynamic linear NH3 concentration range for the silica-
based pellet sensing material
10 and 40 ppm. However, it can be excessive up to 100 ppm or more in the highly
polluted waters e.g. river waters.
Optical sensor for NH3 detection in aqueous environment has recently been
developed by Varghese et al. [12] based on evanescent wave fiber optic sensor with
dual wavelength probing for NH3 detection in water at ppb levels. Pre-treatment of
the water sample (i.e. dilution) becomes necessary when the sensor is used for on-
site analysis of sewage for sample concentration >17 ppm, and that direct deter-
mination of NH3 in envronmental samples become impposible. Optical NH3 sensor
for higly polluted water has also been fabricated by Tan et al. [13] with broader
dynamic linear range (*1000–5000 ppm NH3) using reflectance approach.
However, it took some 6 min long for assay of NH3 in water. This suggests the
potential of the proposed pellet-based sensor for in situ analysis of NH3 in aqueous
media, which may be performed in a quantitative or semi-quantitative manner based
on pellet colour change.
Interference
Interference study is crucial to determine the pellet sensor performance for optical
sensing of NH3 in solution. Based on the results shown in Table 1, alkali, alkaline
earth and SO42− ions do not pose significant interference to the determination of
50 ppm NH3 using the developed silica pellet sensor. The presence of those elements
in the concentration higher than NH3 is usually not possible in water samples [14].
[email protected]
Table 1 The pellet sensor response to various interfering agents at different concentration ratios
Interfering ion Pellet sensor response at different interfering ion concentration to NHs concentration ratios
1:0 1:1 1:10 1:100 1:1000
Na+ 38412.61 ± 41.69 38313.90 ± 353.85 31157.45 ± 354.09a
+
K 38412.61 ± 41.69 38301.46 ± 268.26 38321.12 ± 346.82 38402.04 ± 43.19 33288.46 ± 5.53a
Reflectometric Optosensor for Visual Detection …
Mg2+ 38412.61 ± 41.69 38333.95 ± 340.64 38317.78 ± 83.88 33063.76 ± 118.79a

Ca2+ 38412.61 ± 41.69 38426.01 ± 26.12 38417.17 ± 163.53 31245.89 ± 329.83a
[email protected]
SO42− 38412.61 ± 41.69 38423.42 ± 48.62 38404.83 ± 122.15 38333.04 ± 48.15 30133.04 ± 32.89a
a
t−statistic exceeded t−critical value
287
Conclusion
The proposed pellet-based optosensor has the potentiality for visual determination
of NH3 in environmental water samples. The small size of the silica microparticles
presented no significant barrier for diffusion of NH3. The large interfacial contact
area of the immobilized Co2+ ion on the microsilica with NH3 allows mass transfer
of NH3 molecule to the microparticles’ surfaces to effect the immobilized charge-
transfer complex formation, thus faster response time and broader dynamic range
were achieved. The proposed NH3 pellet sensor involved fast detection procedure,
which merely requires dispensing a small aliquot of sample over the sensor surface.
It is also an economical sensing device since no substrate is required to contain the
sensing matrix, whereby the stand-alone reagent phase in its finished form is free
from physical support. The simplicty-based sensing method allows NH3 determi-
nation for a large number of environmental samples to be carried out in a short
period of time.
Acknowledgment We gratefully acknowledge Universiti Kebangsaan Malaysia for the Young

Researchers Encouragement Grant (GGPM-2014-035) and Malaysian Ministry of Higher
Education for Fundamental Research Grant Scheme (FRGS/2/2014/SG01/UKM/02/1) and
Exploratory Research Grant Scheme ( ERGS-RDU120604).
References
1. Appl M (1999) Ammonia: principles and industrial practice. Wiley, New York
2. Hasan HA, Abdullah SRS, Kamarudin SK, Kofli NT (2011) Problems of ammonia and
manganese in Malaysian drinking water treatments. World Appl Sci 12(10):1890–1896
3. Shaposhnik A, Ryabtsev S, Zviagin A, Korchagina S, Meshkova N, Shaposhnik D, Vasiliev A
(2011) Selective detection of ammonia and its derivatives using MOX-sensor and
microreactor. Procedia Eng 25:1097–1100
4. World Health Organization (2011) Guidelines for drinking-water quality, 4th edn. World
Health Organization, Geneva
5. Ni JQ, Heber AJ (2001) Sampling and measurement of ammonia concentration at animal
facilities-A review. American society of agricultural engineers annual international meeting
paper no. 01-4090
6. Aomura Y, Kobayashi Y, Miyazawa Y, Shimizu H (2010) Application of copper sulfate
pentahydrate as an ammonia removal reagent for the determination of trace impurities in
ammonia gas chromatography. J Chromatogr A 1217:1838–1844
7. Nakano N, Sugata K, Nagashima K (1995) Development of a monitoring tape for ammonia
gas in air by fluorescence detection. Anal Chim Acta 302(23):201–205
8. Masserini RT, Fanning KA (2000) A sensor package for the simultaneous determination of
nanomolar concentration of nitrite, nitrate, ammonia in seawater by fluorescence detector. Mar
Chem 68(4):323–333
9. Rastegarzadeh S, Razei V (2007) An optical sensor for zinc determination based on zincon as
sensing reagent. Sens Actuators B 129(1):327–331
10. Staden JF, Taljaard RE (1997) Determination of ammonia in water and industrial effluent
streams with the indophenol blue method using sequential injection analysis. Anal Chim Acta
344:281–289
[email protected]
11. Pranaitytė B, Jermak S, Naujalis E, Padarauskas A (2007) Capillary electrophoretic

determination of ammonia using headspace single-drop microextraction. Microchem J
86:48–52
12. Varghese BP, Pillai AB, Naduvil MK (2013) Fiber optic sensor for the detection of ammonia,
phosphate and iron in water. J Optics 42(2):78–82
13. Tan LL, Ahmad M, Lee YH (2012) A novel optical ammonia sensor based on reflectance
measurements for highly polluted and coloured water. Sens Actuators B: Chem 171–173:994–
1000
14. Azrina A, Khoo HE, Idris MA, Amin I, Razman MR (2011) Major inorganic elements in tap
water samples in Penisular Malaysia. Malays J Nutr 17(2):271–276
[email protected]
Low Noise and Properties of Double Layer
Concrete Paving Blocks
Euniza Jusli, Hasanan Md. Nor, Ramadhansyah Putra Jaya,

Zaiton Haron, Musli Nizam Yahya, M. Azman and Lee Kuan Lim
Abstract Double layer rubberized concrete block (DL-RCPB) was introduced to

effectively make use of waste tyre and to improve the performance of concrete
pavement block. This study was carried out to justify the effectiveness of using
waste tyre rubber as aggregate. In this study, there were eight series of concrete mix
with 10 and 20 % waste tyre rubber replacement level and different thickness of
facing layer; 10, 20, 30 and 40 mm. The compressive strength was reduced when
the percentage of waste tyre rubber was increased. Remarkably, all types of DL-
RCPB with rubber percentage up to 20 % achieved minimum required compressive
strength of 45 MPa. Moreover, DL-RCPB (20 %) with 40 mm top layer was proven
E. Jusli H. Md. Nor (&) R.P. Jaya L.K. Lim

Department of Geothechnics and Transportation, Faculty of Civil Engineering,
Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia
E. Jusli
R.P. Jaya
L.K. Lim
Z. Haron
Department of Structure and Materials, Faculty of Civil Engineering,
Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia
M.N. Yahya
Department of Engineering Mechanics, Faculty of Mechanical and Manufacturing
Engineering, Universiti Tun Hussein Onn, Parit Raja, 86400 Batu Pahat,
Johor, Malaysia
M. Azman
Department of Engineering, Razak School of Engineering and Advanced Technology,
Universiti Teknologi Malaysia, International Campus, Jalan Semarak,

[email protected]
292 E. Jusli et al.
to have the most effective sound absorption concrete mix design with up to 36 % of
sound adsorption recorded. Reduction in block density was found when the per-
centages of rubber were increased. Porosity up to 17 % was recorded for DL-RCPB
(20 %). The rough surface of rubber particles tends to attract air on the rubbers’
surface and the trapped air tends to produce voids once the concrete hardened.
Porosity of DL-RCPB was increased when the rubber was added to concrete mix,
thereby decreasing the density of DL-RCPBs. Low density and high porosity
increased the intensity of sound absorption and relatively gives positive impact on
the development of low noise CBP as long as the compressive strength was
sufficient.
Keywords Double layer Rubberized Concrete Paving block Tyre rubber
Introduction
In general, concrete paving block (CPB) is more dependable as compared to other

pavement materials due to its better attribute in many aspects. Compressive
strength, abrasion, skidding resistance, weathering durability, serving period, aes-
thetic feature and flexibility in terms of installation and removal are example of
aspects that CPB has showed in remarkable performances [1]. Although previous
studies on concrete paving blocks has improved many aspects in the pavement
industry, but there are still some undesirable effects generated. Among the unde-
sirable effects created are noise pollution and poor riding quality. Many researches
have been done on the physical and mechanical properties of different types and
design of concrete paving block [2–4]. These initiatives were taken in order to
widen the usage and improve the performance of CPB. However, it seems that there
were limited researches focusing on the development of low noise CPB. Waste tyre
was well known for its high durability and elasticity properties that make it suitable
materials substituent in concrete to improve sound and energy absorbability.
Apparently, recycling waste tyre rubber as concrete materials seems to be an ideal
solution towards treating this non-biodegradable waste. Therefore to effectively
make use of the waste tyre rubber and increase the performance of CPB, double
layer rubberized concrete paving blocks (DL-RCPBs) was introduced in a hope to
tackle the insufficiencies [5, 6]. DL-RCPBs were designed in order to reduce the
traffic noise due to the interaction between vehicle’s tyre and pavement. Waste tyre
rubber was used as a medium to absorb more sound and vibration in concrete, in
returns to produce low noise CPB. The effect of layer and rubber content on several
physical and mechanical properties of double layer rubberized concrete blocks was
evaluated in this study. The properties that comprised in study include compressive
strength, sound absorption, density, porosity and skid resistance. Moreover, the
results found can also be used as indicators or references for future study as limited
researches are been done on rubberized concrete with double layer characteristic.
[email protected]
Low Noise and Properties of Double Layer Concrete Paving Blocks 293
Materials
Cement, Ordinary Portland Cement (TASEK cement) was used throughout the
experiment. This type of cement was compiled to ASTM C150 Type I [7]. The
cement are composed of SiO2, Al2O3, Fe2O3, CaO, MgO and SO3, respectively [8].
Aggregate, In this study, local natural crushed granite and natural river sand
were used as coarse and fine aggregate. These aggregates are compiled the
requirements stated in ASTM C136-06 [9]. Figure 1 illustrate the size distribution
of coarse aggregates; aggregates fall in the fixed acceptance ranges. The aggregates
used are already in saturated dry condition thus no further drying process is
required.
Sand, For CPB concrete mixture, natural river sand complies with the
requirement in ASTM C136-06 [9] was used. The size distribution of sand particles
was presented in Fig. 2. The sand used was left for air drying for several days
before it been used for the mixing.
Rubber Aggregate, There are two types of rubber aggregate; fine and coarse
aggregates as shown in Fig. 3 has been adopted in this study mainly, each type of
aggregate was classified according to its size. Fine tyre rubber with size ranging
from 1 to 4 mm was used for sand replacement. Meanwhile, 5–8 mm of tyre rubber
was utilized as coarse aggregate replacement. These rubber aggregates were sup-
plied by Yong Fong Industry, Malaysia. Tyre rubber are composed of 48 % styrene-
butadience rubber (SBR), 47 % carbon black, 1.9 % extender oil, 1.1 % zinc oxide,
0.8 %, sulfur, 0.7 % accelerator and 0.5 % strearic acid [10].
Superplasticizer, To maintain the feasibility of the concrete mixtures, a
Glenium C380 superplasticizer was used throughout this study. This superplasti-
cizer is chloride-free and complies with the requirements of the ASTM C494-13
[11].
Mix Proportion, Control mix with mix proportion of cement: sand: aggregate:
superplasticizer; 1:1.5:1.7:0.003 and 0.47 water cement ratio (w/c) was used. The
control mix was designed to carry a characteristic strength of 50 MPa at 28 days.
Fig. 1 Sieve analysis result 120

for coarse aggregate
Percentage of Passing (%)
100
Upper Limit
80
60
Lower Limit
40
20
Coarse Aggregate Grading
0
1 10 100
ASTM SIeve Size
[email protected]
294 E. Jusli et al.
120
Percentage of Passing (%)

100
Upper Limit
80
60
Sand Grading
40
20
Lower Limit
0
0.10 1.00 10.00
ASTM Sieve Size
Fig. 2 Sieve analysis result for sand
Fig. 3 Waste tyre rubber
Eight series of concrete mix with 10 and 20 % of waste rubber aggregate substi-
tutions and different thickness of facing layer was investigated in this study. The
designed concrete samples was shown in Table 1. Coarse and fine tyre rubber was
substituted for top and bottom layer, respectively.
Block Manufacturing, Two series of concrete mixture were produced in the
production of DL-RCPBs. Since the concrete blocks consist of two concrete layers,
thus the mixing process for top and the bottom layer was done in sequence. Series I
(layer 1) was mixed with 5–8 mm rubber granules, whereas series II (Layer 2) was
produced by mixing with 1–4 mm rubber granules. Series II concrete mixture was
poured into steel moulds with 100 × 200 × 80 mm dimension. Then, steel mould
was placed on the vibration table for 5 s. At the meantime, the top layer raw
material mixing begun. Same as mentioned steps were used for bottom concrete
layer preparation. Removal of the excess material and flattening the block surface
[email protected]
Table 1 Concrete mix design

Mix Mix proportion Cement content Water/ Rubber
symbol (kg/m3) cement content (%)
ratio
Series I Series II Series I&II Series I&II Series I&II
(C:A:S) (C:A:S)
CCPB 1:1.7:1.5 1:1.7:1.5 489 0.47 0
DL-RCPB 1:1.5: 1.5 1:1.7:1.35 489 0.47 10
(10 %)
DL-RCPB 1:1.35:1.5 1:1.7:1.2 489 0.47 20
(20 %)
DL-RCPB 1:1.2:1.5 1:1.7:1.05 489 0.47 30
(30 %)
DL-RCPB 1:1.0:1.5 1:1.7:0.9 489 0.47 40
(40 %)
C:A:S Cement: Aggregate: Sand
CCPB Control concrete paving block
DL-RCPB Double layer rubberized concrete paving block
was done after completing vibration process. The features of DL-RCPB were
shown in Fig. 4.
All blocks were covered with moist hessians for a period of 24-h before de-
moulding. Air curing method was adopted as all the blocks were kept in laboratory
under room temperature of 25 °C and 65 % relative humidity. The preparation
method of cylinder samples with 99.5 mm diameter and 80 mm height for sound
absorption test was similar as blocks.
Testing Methods, Sound absorption, compression test, density and porosity
were carried out to investigate the properties of hardened DL-RCPB with different
layer thickness. Sound absorption test was carried out using impedance tube
method; a tube, two microphones and a digital frequency analysis system. In this
test, an array of complex data obtained from the measured transfer function of wave
frequencies would be processed by computer program. The absorption coefficient of
the related material would then be generated later by the program in excel data
form. Cylindrical sample with diameter of 99.5 and 80 mm height were tested
according to ASTM E1050 [12]. The compression test was carried out using
Fig. 4 Double layer

Layer 1
rubberized concrete block
Layer 2
[email protected]
296 E. Jusli et al.
MATEST compression machine with 3000 kN capacity and pace rate of 2.5 kN/s.
Average of three samples each for 7 and 28 days strength with different mix designs
were tested. According to BS 6717 Part 1 [13], 3 mm plywood was used applied on
top and bottom of block surface. Density and prosity properties of DL-RCPBs were
carried out according to the standard procedures stated in ASTM C642 [14].
Sound Absorption, The ability of DL-RCPBs to absorb sound was measured by

the sound absorption coefficient (α). Figures 5 and 6 indicate an improvement of
acoustic performance as the percentage of rubber substitution and thickness of top
layer were increased. Apparently, similar graphs’ pattern was observed for both
graphs in Figs. 5 and 6. It was found that DL-RCPBs for 10 and 20 % rubber
content have better sound absorption properties as compared to CCPB which only
achieve 0.05 maximum α-value. From the results shown in Fig. 5, DL-RCPB
(10 %) with 40 mm layer thickness was noticed to have the highest α-value of 0.253
at 700 Hz. This implied that the thickness of the top layer or amount of coarse
aggregate has an influent on the acoustic performance. Besides, similar result
pattern was shown by 20 % rubber content cylinders. Tyre rubber content of DL-
RCPB (20 %) shown in Fig. 6 recorded the sound absorption coefficient of 0.364
which higher than 0.253 shown by the highest of DL-RCPB (10 %). Indeed, the
damping of sound and reduction in dynamic energy were promoted when the rubber
content increased in advance.
Compressive strength, The results of compression test for DL-RCPB with 10
and 20 % rubber substitution were illustrated in Fig. 7. Each value represents the
average result of three samples. In general, the compressive strength of DL-RCPB
decreases with the increase of rubber tyre substitution and thickness of facing layer.
10 mm 20 mm 30 mm 40 mm
Sound Absorption Coefficient, α
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 200 400 600 800 1000 1200 1400 1600
Frequency (Hz)
Fig. 5 Sound absorption of DL-RCPBs (10 %)
[email protected]
10 mm 20 mm 30 mm 40 mm
Sound Absorption Coefficient, α
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 200 400 600 800 1000 1200 1400 1600
Frequency (Hz)
Fig. 6 Sound absorption of DL-RCPBs (20 %)
Fig. 7 Compressive strength

Compressive strength, σ
65 0% 10 % 20 %
of double layer rubberized σ = - 0.0786t + 61.405
R² = 0.9176
concrete paving blocks 60
55 σ = - 0.1876t + 57.706
(MPa)
R² = 0.984
50
45 σ = - 0.1888t + 52.595
R² = 0.9143
40
10 20 30 40
Thickness of Layer, t (mm)
Compressive strength of DL-RCPB (20 %) with 40 mm thick of facing layer

achieved the minimum requirement of 45 MPa. Besides, the results for DL-RCPB
(20 %) indicate that the highest compressive strength recorded was 51.37 MPa.
From the results, it shows progressive reduction of strength as the thickness of
facing layer increase. This may due to less adhesion between tyre rubber and
cement paste. For DL-RCPBs (10 %), the entire concrete block satisfied with
minimum compressive strength requirement of 45 MPa. However, for DL-RCPB
(20 %), only DL-RCPB with 10 mm top layer that satisfies the minimum strength
requirement.
Density, The density results for DL-RCPB (10 %) and DL-RCPB (20 %) were
presented in Fig. 8. Referring to Fig. 8, the density for both percentages of rubber
substituted blocks decreased progressively as the thickness of the top layer
increased. DL-RCPB (20 %) with 40 mm thickness of facing layer recorded the
lowest density of 2218.30 kg/m3 this value is 3 % lower than the control block. The
density recorded for control block was 2298.51 kg/m3. This phenomenon happened
because of different complexity in block and the presence of the low density
substitution material.
Porosity, A steady increment of porosity pattern was obtained with the increase
of top layer thickness for both percentages of tyre rubber contents as shown in
[email protected]
298 E. Jusli et al.
Fig. 8 Density of DL-RCPBs 10% 20%

2300
2250
Density, ρ (kg/m 3)
2200
2150
2100
2050
2000
10 20 30 40
Thickness, t (mm)
Fig. 9 Porosity of DL- 10% 20%

19
RCPBs
17
Porosity (%)
15
13
11
9
7
5
10 20 30 40
Thickness, t (mm)
Fig. 9. Based on the results, the porosity of different mix designs was ranged from
14.54 to 15.00 % for 7 days, whereas at 28 days the porosity ranged from 15.36 to
17.78 %. From there the lowest porosity value was recorded for the control mix,
whereas the highest porosity was displayed by DL-RCPB (20 %) with 40 mm
facing layer thickness. Non-polar nature of the tyre rubber material was the reason
behind this phenomenon. Regarding to that, air particle would easily attracted
around the rubber aggregate surface during mixing process causing a weaker bond
between non-polar particles and polar particle. According to Sandberg and Ejsmont
[14], the air pumping effect was effectively reduces by porosity, thereby reducing
the tyre-pavement interaction noise.
Conclusion
The conclusions that can be drawn as follows:

(i) Compressive strength of the DL-RCPB tends to decrease when more per-
centage of tyre rubber was substituted.
[email protected]
(ii) Concrete blocks with 10 and 20 % tyre rubber content regardless the thickness
of facing layer was found to comply with minimum requirement of 45 MPa.
(iii) Sound absorption was increased when the percentage of tyre rubber substi-
tution and the amount of coarse rubber aggregates was increased. DL-RCPB
(20 %) with 40 mm facing layer thickness recorded up to 36 % of sound
absorption improvement when compared with the control block.
(iv) Density indicated a reduction when the percentage of the tyre rubber was
increased. Meanwhile, porosity experienced a rise when higher percentage of
the tyre rubber was substituted.
(v) Porosity increased due to the physical characteristics of rubber particle, which
has rough surface that tend to entrap air to their surface.
Acknowledgment The authors would like to express their gratitude toward Ministry of Higher
Education of Malaysia (MOHE) and Universiti Teknologi Malaysia for the financial support by
providing research grant RUG 03H47 for this study.
References
1. Van der Vlist AA (1980) The development of CBP in Netherlands. In: First international
conference on concrete block paving, pp 14–22
2. Ling TC, Nor HM, Lim SK (2010) Using recycled waste tyres in concrete paving blocks. In:
Proceeding of the Institution of Civil Engineers, pp 37–45
3. Poon CS, Chan D (2006) Paving blocks made with recycled concrete aggregate and crushed
clay brick. Constr Build Mater 20:569–577
4. Azman M, Nor HM, Hainin MR, Yaacob H, Ismail CR, Nur Hafizah AK (2013) The effect of
groove-underside shaped concrete block on pavement permanent deformation. Jurnal
Teknologi 61(3):7–14
5. Euniza J, Nor HM, Ramadhansyah PJ, Zaiton H (2014) Effect of using waste tyre rubber on
the properties of double layer rubberized concrete paving blocks. Jurnal Teknologi 71(3):1–5
6. Euniza J, Nor HM, Ramadhansyah PJ, Zaiton H (2014) Double layer concrete paving blocks
using waste tyre rubber as aggregate replacement. Appl Mech Mater 554:128–132
7. ASTM C150-12 Standard specification for Portland cement, American society for testing and
materials. Information on http://www.tasekcement.com/index/cement_facts/cement_quality_.html
8. ASTM C136-06:2011 Standard test method for sieve analysis of fine and coarse aggregate,
Am Soc Testing Mater
9. Euniza J, Nor HM, Ramadhansyah PJ, Zaiton H (2014) Chemical Properties of Waste Tyre
Rubber Granules. Adv Mater Res 911:77–81
10. ASTM C494:2013, Standard specification for chemical admixtures for concrete. Am Soc
Testing Mater
11. ASTM E1050-10.2011 Standard test method for impedance and absorption of acoustical
materials using a tube, two microphones and a digital frequency analysis system. Am Soc
Testing Mater
12. BS 6717-1:1993 Precast concrete paving blocks-Part 1 : specification for paving blocks,
British Standard
13. ASTM C642-06 Standard test method for density, absorption, and voids in hardened concrete.
Am Soc Testing Mater
14. Sandberg U, Ejsmont JA (2002) Tyre/road noise reference book. Informex, Kisa
[email protected]
Synthesis of Modified Covalent Organic
Framework-1 (COF-1) and Its
Characterizations
Muhammad Falaq Muhammad Faisal,

Ahmad Rafizan Mohamad Daud and Kamariah Noor Ismail
Abstract Three samples of covalent organic framework-1 (COF-1) assigned as

COF-1 (S1), COF-1 (S2) and COF-1 (S3) were prepared by varying the initial mass
of 1, 4-benzene diboronic acid (BDBA) used. The samples were characterized using
Nitrogen adsorption for porosity analysis and XRD for crystallinity analysis. All
samples exhibit Type IV isotherm indicating mesoporous materials. The BET
surface area value showed an increasing trend with increasing mass of BDBA used.
The highest BET surface area was recorded by COF-1 (S3) with a value of
107.9 m2/g. This sample has the highest amount of micropore volume of 66 % and
the lowest average pore diameter of 11.05 nm. XRD patterns revealed sharp peaks
indicating a crystalline structure and the peak positions matched well with other
works available in literature. The difference in intensities however was due to the
exposure of the samples to atmosphere resulted in the re-formation of COF-1
structure onto BDBA structure.

Keywords Covalent organic framework-1 Adsorption isotherm Porosity X-ray

diffraction Crystallinity
Introduction
Due to the ever improving industrial developments that necessitate usage of

nanoporous materials with higher loading capacity and enhanced physico-chemical
characteristics, scientists and engineers have focused their effort in developing
M.F. Muhammad Faisal A.R. Mohamad Daud (&) K.N. Ismail

Selangor, Malaysia
M.F. Muhammad Faisal
K.N. Ismail

[email protected]
302 M.F. Muhammad Faisal et al.
alternative nanoporous materials that could be applied for industrial purposes such
as fuel storage [1], catalysis [2], optical sensors [3] and separations [4]. In recent
years, nanoporous Covalent Organic Frameworks (COFs) [5–7] have been syn-
thesized by Yaghi group to act as a possible material for hydrogen gas storage.
COFs was developed to improve the characteristics of the previously developed
Metal Organic Frameworks (MOFs) by the same research group [8]. The metal
content which adds extra weight in MOFs is a drawback [9] and did not signifi-
cantly influence its storage capabilities [10].
The first developed COFs, named COF-1, is synthesized through the molecular
dehydration of 1, 4-benzene diboronic acid (BDBA) and is lightweight, thermally
stable and low in density [5]. Though still in its early stage, COFs have proven to
have potential for gas storage applications such as hydrogen, carbon dioxide and
methane. Several studies revealed that COFs structure is in the microporous range
and high in crystallinity such as COF-LZU1 [11], COF-18 [12] and COF-8 [6].
The aim of this study was to synthesize three modified samples of COF-1 using
different mass of BDBA. All samples were characterized according to porosity and
crystalline structure using nitrogen adsorption and X-Ray Diffraction (XRD)
analyses respectively.
Experimental
Synthesis of COF-1, COF-1 was synthesized in accordance with the general

procedures developed by [5] and methodology adopted from [13]. An ampule glass
tube (Wheaton, I.D. = 22.5 mm) was filled with 25 mg of 1,4-benzene diboronic
acid (Aldrich, ≥95.0 %, USA) and 1 ml of a 1:1 (volumetric ratio) solution of
mesitylene:dioxane (Acros, 99 %, Belgium). The tube was connected to a steel tube
with the other end connected to a vacuum pump. The tube was then flashed frozen
in liquid nitrogen bath, evacuated to an internal pressure of 20 Pa and flame sealed
using a blow torch. The length of the tube was reduced to 120 mm upon sealing and
heated in an oven for 72 h at 100 °C. The resultant fine white powder was then
filtered and washed with 15 ml acetone (Merck, ACS reagent, Germany). Similar
procedure was applied where for the second and third samples, the initial mass of
BDBA used was increased to 50 and 100 mg. The volume of mesitylene:dioxane
solution used was 2 ml for 50 mg of BDBA and 4 ml for 100 mg of BDBA.
Characterizations of COF-1, Porosity analysis was obtained using
Quantachrome Autosorb-1 micropore surface area analyzer. A liquid nitrogen bath
held at 77 K was used to maintain the temperature during analysis and was con-
tinuously monitored using a computerized system. Prior to nitrogen adsorption, 0.2
grams of the sample was degassed at 120 °C for 12 h to remove any impurities and
trapped gas molecules inside the pores.
X-ray diffraction (XRD) patterns were collected on a Rigaku diffractometer
running at 40 kV, 20 mA using a Cu Kα (λ = 0.1543 nm) radiation source, with a
scan speed of 5°/min and a 2θ step size of 0.02° over a range of 3–90°.
[email protected]
Synthesis of Modified Covalent Organic Framework-1 (COF-1) … 303
Isotherm Characteristics, Fig. 1 shows the nitrogen gas adsorption/desorption

isotherm profile for all the synthesized COF-1 at 77 K. All the isotherm patterns
exhibited a well-defined flat line which increases slightly with a sudden surge at the
end. Based on the classification set by the IUPAC system, this isotherm exhibited
Type IV which is in the mesoporous material range. Adsorption occurred at a
relative pressure of 0.1 with a slightly increase uptake as the relative pressure was
increased indicating the absence of many highly microporous materials (pores
smaller than 2 nm).
The initial part of the isotherm followed the isotherm of a Type II material which
is a macroporous material that attributes to monolayer-multilayer adsorption.
(a)
200
Volume, cm 3/g (STP)
175
150
125
100
75 Desorption
50 Adsorption
25
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Partial Pressure, P/Po
(b)
200
175
150
125
100
75 Desorption
50 Adsorption
25
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
(c)
200
175
150
125
100
75 Desorption
50 Adsorption
25
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Fig. 1 Adsorption isotherm profile for a COF-1 (S1), b COF-1 (S2), c COF-1 (S3)
[email protected]
The scarcity of micropores inside the materials could be a drawback in its

application for gas adsorption. The effectiveness of a storage gas material highly
relies on the accessibility of the micropore inside the material to the small gas
particles, which referred as micropore filling [14]. This finding also explains the
lower value of BET surface area obtained for all samples of COF-1 with the highest
achieved BET surface area value was COF-1 (S3) with 107.9 m−2/g. The sudden
enhanced uptake of N2 at a relative pressure after 0.8 suggests the presence of small
quantities of macropores (pores larger than 2 nm).
The isotherm also exhibits a hysteresis loop between adsorption and desorption
lines, indicates the possible presence of many mesoporous structures within the
synthesized COF-1. The hysteresis loop indicates the capillary condensation that is
taking place in the mesopores [14]. So, it can be concluded that all the structures of
COF-1 exhibit the Type IV isotherm which indicates the presence of mesoporous
structures instead of Type I (microporous structures) which was reported by [5].
Porosity Characteristics, Table 1 presents the pore characteristics of all the
samples. The highest BET surface area obtained is sample COF-1 (S3) with
107.9 m2/g followed by COF-1 (S2) with 52 m2/g and COF-1 (S1) with 43.2 m2/g.
Sample COF-1 (S3) also has highest amount of micropores volume which is
0.1980 cm3/g which accounted to 66 % of the total pore volume. Its total pore
volume is the highest at 0.2981 cm3/g and also had the lowest average pore
diameter of 11.05 nm as compared to COF-1 (S1) and COF-1 (S2). The average
pore diameters for all COF-1 samples were in the range of 11.05–19.05 nm. The
lower the average pore diameter resulted in a higher number of micropores avail-
able, thus a higher value of surface area were achieved. It can be further concluded
that porosity for all the samples synthesized in this research can be classified as
mesopores instead of micropores due to the large average pore diameter and low
volume of micropores.
XRD Analysis, Fig. 2 shows the XRD patterns of all samples of COF-1. All
samples showed good crystallinity where sharp peaks occurred while the 2θ
positions were matched well with the works by [5] albeit difference in peak
intensity. This could due to the sample exposure to the atmosphere prior to ana-
lyzing. If the sample was exposed to the atmosphere for a period of time, which was
around ten days in this work there is a change in the crystallinity of the sample and
the XRD pattern does not provide the same peaks intensity as that presented by
Cote et al. [5]. Li and Yang [15] investigated this phenomenon by exposing their
synthesized COF-1 to the atmosphere for seven days where the effect was a
Table 1 Pore characteristics of COF-1

COF-1 Pore volume BET surface Total pore Average pore Micropore
sample (cm3/g) area (m2/g) volume (cm3/g) diameter (Å) (%)
(micropores)
S1 0.1193 43.2 0.2058 190.5 57.97
S2 0.1256 52 0.2331 179.4 53.88
S3 0.1980 107.9 0.2981 110.5 66.42
[email protected]
Synthesis of Modified Covalent Organic Framework-1 (COF-1) … 305
Fig. 2 XRD patterns for a COF-1 (S1), b COF-1 (S2), c COF-1 (S3)
different patterns formed, which correlates well with what was exhibited by the
COF-1 synthesized in this work. Peaks at 12°, 19° and 27° indicated the prominent
structure of COF-1, while peaks at 18°, 23° and 28° revealed the reformation of
BDBA structure. Overall, there are no significant differences in XRD patterns for all
COF-1 samples when compared with respect to initial mass of BDBA.
Conclusions
Modified COF-1 samples were prepared by varying the initial mass of reactant
used. Adsorption isotherm profile and XRD analyses were used to evaluate char-
acteristics of the modified COF-1. Overall, through the isotherm patterns obtained,
the pores developed in all COF-1 samples prepared in this work were in the
mesoporous range of Type IV isotherm. Sample COF-1 (S3) which used 100 mg of
BDBA has the highest BET surface area of 107.9 m2/g. This sample has also
the highest amount of micropore volume and the lowest average pore diameter.
The XRD analysis show good crystallinity structure but different in peak intensities.
This indicates reformation of crystalline COF-1 structure into BDBA due to
exposure to atmosphere. Considering that COF-1 synthesis research is very much at
early stage, findings from this study deemed beneficial in improving the preparation
method of COF-1 for future studies.
[email protected]
Acknowledgment The authors were grateful to the Ministry of Higher Education, Malaysia and
Universiti Teknologi MARA for providing grants 600-RMI/ST/FRGS 5/3/Fst(101/2010) and 600-
RMI/DANA 5/3/RIF(131/2012).
References
1. Benard P, Chahine R, Chandonia PA, Cossement D, Dorval-Douville G, Lafi L et al (2007)

Comparison of hydrogen adsorption on nanaoporous materials. J Alloys Compd
446–447:380–384
2. Pisduangdaw S, Panpranot J, Methastidsook C, Chaisuk C, Faungnawakil K, Praserthdam P
et al (2009) Characteristics and catalytic properties of Pt-Sn/Al2O3 nanoparticles synthesized
by one-step flame spray pyrolysis in the dehydrogenation of propane. Appl Catal A 370:1–6
3. Fernandez-Sanchez JF, Cannas R, Spichiger S, Steiger R, Spichiger-Keller UE (2006) Novel
nanostructured materials to develop oxygen-sensitive films for optical sensors. Anal Chim
Acta 566:271–282
4. Ackley MW, Rege SU, Saxena H (2003) Application of natural zeolites in the purification and
separation of gases. Microporous Mesoporous Mater 61:25–42
5. Cote AP, Benin AI, Ockwig NW, O’Keeffe M, Matzger AJ, Yaghi OM (2005) Porous,
crystalline, covalent organic frameworks. Science 310:1166–1170
6. Cote AP, El-Kaderi HM, Furukawa H, Hunt JR, Yaghi OM (2007) Reticular synthesis of
microporous and mesoporous 2D covalent organic frameworks. J Am Chem Soc 129:12914–
12915
7. El-Kaderi HM, Hunt JR, Mendoza-Cortés JL, Cote AP, Taylor RE, O’Keeffe M, Yaghi OM
(2007) Designed synthesis of 3D covalent organic frameworks. Science 316:268–272
8. Yaghi OM, Li G, Li H (1995) Selective binding and removal of guests in a microporous metal-
organic framework. Nature 378:703–706
9. Klontzas E, Tylianakis E, Froudakis GE (2008) Hydrogen storage in 3D covalent organic
frameworks. A multiscale theoretical investigation. J Phys Chem C 112:9095–9098
10. Assfour B, Seifert G (2010) Hydrogen adsorption sites and energies in 2D and 3D covalent
organic frameworks. Chem Phys Lett 489:86–91
11. Ding SY, Gao J, Wang Q, Zhang Y, Song WG, Su CY, Wang W (2011) Construction of
covalent organic framework for catalysis: Pd/COF-LZU1 in Suzuki-Miyaura coupling
reaction. J Am Chem Soc 133:19816–19822
12. Tilford RW, Gemmill WR, Zur Loye HC, Lavigne JJ (2006) Facile synthesis of a highly
crystalline, covalently linked porous boronate network. Chem Mater 18:5296–5301
13. Musa MAA, Yin CY, Savory RM (2010) Synthesis and textural characterization of covalent
organic framework-1: Comparison of pore size distribution models. Mater Chem Phys 12:5–8
14. Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti RA, Rouquerol A et al (1985)
Reporting physisorption data for gas/solid systems with special reference to the determination
of surface area and porosity. Pure Appl Chem 57–4:603–619
15. Li YW, Yang RT (2010) Hydrogen storage in metal-organic frameworks and covalent-organic
frameworks by spillover. AlChe J 54:269–279
[email protected]
One Pot Synthesis of γ-Valerolactone
from D-Glucose Over H3PW12O40/ZrO2
Catalyst
Dorairaaj Sivasubramaniam and Nor Aishah Saidina Amin
Abstract In this study, two different wt% of H3PW12O40/ZrO2 catalysts were syn-
thesized, characterized and screened for the synthesis of γ-valerolactone (GVL) from
D-glucose. The catalysts involved were 10 % H3PW12O40/ZrO2 and 15 %
H3PW12O40/ZrO2 prepared through wet impregnation method. Catalyst character-
ization using FTIR, BET and NH3-TPD revealed that the catalytic performances were
dominantly influenced by the acid sites of the catalyst in the production of Levulinic
acid (LA) from D-glucose via acid hydrolysis reaction. High acidity greatly influenced
the catalytic performance for the production of GVL. These catalysts were tested based
on five sets of reaction time (8, 10, 12, 14 and 16 h) at a fixed temperature of 180 °C
with 1:1 catalyst to feedstock ratio. 10 % H3PW12O40/ZrO2 exhibited high perfor-
mance at 16 h of reaction time with 22.3 % yield compared to 15 % H3PW12O40/ZrO2
with only 14.8 % yield of GVL.
Keywords γ-valerolactone Solid acid catalyst H3PW12O40 ZrO2 One pot

reaction
Introduction
Biomass can serve as an alternative renewable source for the sustainable

production of energy and chemicals such as LA and γ-valerolactone [1]. The
largest part of biomass is carbohydrates. The carbohydrates are particularly used
for the conversion to platform chemicals [2]. Carbohydrates can be converted to
D. Sivasubramaniam N.A. Saidina Amin (&)

Chemical Reaction Engineering Group (CREG), Faculty of Chemical Engineering,
Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia
D. Sivasubramaniam

[email protected]
308 D. Sivasubramaniam and N.A. Saidina Amin
chemicals including ethanol, n-butanol, sorbitol, hydroxymethylfurfural, and

dimethylfurfural [3–5].
GVL is one of the platform chemicals considered as one of the most dominant
reproducible chemicals for the production of other bio-based chemicals [6]. It is use
as liquid fuel, food additive, solvent and organic intermediate in the synthesis [7].
Generally, there are two steps involved in the production of GVL: (1) dehydration
of C6-sugars to form LA, (2) hydrogenation of LA to form GVL.
GVL has most ideal properties known as sustainable liquid fuel, which is
involved in the production of both energy and carbon-based chemicals such as
pentanoic acid and 5-nonane. It is a renewable, easy and safe to store and for
transportation. It has a low melting point (−31 °C) and a high boiling point (207 °C).
The carbohydrate such as cellulose, hemicelluloses and lignin are polysaccha-
rides utilized in the hydrolysis reaction to form monomer (D-glucose), which is
utilized for the synthesis of many bio-based chemicals. The objective of this paper
is to synthesize GVL using D-glucose in presence of H3PW12O40/ZrO2, prepared
through wet impregnation method, in one pot reaction. The catalysts were char-
acterized to determine the influence of acid sites and surface area on the catalytic
performance [8].
Experimental
Chemicals
All chemicals were supplied from Merck, Germany. The Keggin-type hetero-
polyacid used in the reaction was phosphotungstic acid (H3PW12O40), Zirconium
oxide (ZrO2) was used for the catalyst preparation. Distilled water was used in the
catalyst preparation and as solvent for reactions. Formic acid (FA) was used as
homogeneous acid catalyst and hydrogen supply for the reaction mixture. Standard
solutions of glucose, γ-valerolactone GVL and methanol were used for product
analysis.
Catalyst Preparation
The catalyst was prepared via impregnation method. Intially, zirconium oxide
(ZrO2) was mixed with phosphotungstic acid (H3PW12O40) according to wt% of the
catalyst. 10 g of zirconia (ZrO2) and 20 ml of water followed by 10 or 15 wt% of
H3PW12O40 was added into the mixture. The mixture was then stirred at 60 °C until
water molecules evaporated to form solid catalyst. Finally, the solid form of catalyst
was calcined at 500 °C for 3 h.
[email protected]
One Pot Synthesis of γ-Valerolactone from D-Glucose … 309
Catalyst Characterization
FTIR spectra were used to identify the chemical bonds and functional groups in the
sample. The FTIR spectra (KBr pellets) were recorded using Perkin Elmer in the
range of 400–4000 cm−1. The acidity properties of the two catalysts were analysed
using NH3-temperature programmed desorption (NH3-TPD). Brunauer-Emmet-
Teller (BET) method was applied to determine the surface area, pores volume and
size distribution using Micrometrics ASAP2020 analyzer with isotherm nitrogen
adsorption.
One Pot Reaction
The conversion of D-glucose to GVL was carried out in presence of 48 ml water

and 2 ml FA as hydrogen carrier for the hydrogenation reaction as reported by
Heeres and Chunai [9] and the reaction was performed in a high pressure stirred
autoclave reactor at 180 °C. Finally, the product was analyzed using GC-FID to
identify the yield of products formed.
Catalyst Characterization
Fourier Transform Infrared Spectroscopy (FTIR), Fig. 1 shows the FTIR

spectra for parent ZrO2, 10 % H3PW12O40 and 15 % H3PW12O40. The peak
exhibited at 1077 cm−1 represents the bond between Zr–O–P in its structure after
the impregnation of ZrO2 with H3PW12O40. Moreover, when the loading of
H3PW12O40 increases, weak peak appears at 420 cm−1inferring formation of the
impregnated catalyst for the similar bond.
Brunauer-Emmett-Teller (BET)
The BET values in Table 1 indicate that impregnation of H3PW12O40 on ZrO2

increased the surface area of the solid catalyst but no significant difference between
this two loadings. The surface areas were 14.36 and 14.52 m2/g for 10 and 15 %
H3PW12O40/ZrO2, respectively. Based on the total surface area obtained, it was
confirmed that surface does not influence the catalytic performance for the pro-
duction of GVL.
[email protected]
Fig. 1 FTIR spectra for a ZrO2, b 10 % H3PW12O40/ZrO2, c 15 % H3PW12O40/ZrO2
Table 1 BET results for Catalyst Specific surface area (m2g−1)

H3PW12O40 impregnate on
ZrO2 at different loadings H3PW12O40 4.55
ZrO2 14.95
10 % H3PW12O40/ZrO2 14.36
15 % H3PW12O40/ZrO2 14.52
Ammonia-Thermal Programmed Desorption (NH3-TPD), The acidity of the

catalyst highly influenced the acid hydrolysis reaction for the production of LA
from D-glucose. Table 2 compares the acidity of the parent catalyst with the
impregnated ones. The acidity of the impregnated catalyst was higher compared to
the parent catalysts due to the introduction of H3PW12O40 on ZrO2. Generally,
H3PW12O40, known as super acid [10], contains strong Bronsted acid sites known
as active sites for the catalyst. 10 % H3PW12O40/ZrO2 catalyst displayed the highest
acidity [8].
Dehydration/Hydrogenation of D-glucose to produce γ-valerolactone (GVL),
The reaction temperature was fixed at 180 °C and the reaction time was varied from 8
to 16 h [8]. The reaction was conducted based on five sets of reaction time for
studying the effect of time on GVL yield in presence of catalyst. In this study we
Table 2 NH3-TPD profile Catalyst Acidity (mmol/g)

for the H3PW12O40/ZrO2 with
different loading H3PW12O40 1.68
ZrO2 0.27
10 % H3PW12O40/ZrO2 0.52
15 % H3PW12O40/ZrO2 0.31
[email protected]
found that the GVL yield increased with time [9] due to high amount of LA were
converted to ring closure structure of GVL. When compared this work with the study
demonstrated by Son et al. [11], the yield of GVL obtained was low due to different
feedstock used. Son et al. utilized D-fructose as the feedstock for the production of
GVL and obtained high yield (48 %) but in this study, the feedstock that been utilized
was D-glucose for the one pot reaction. The production of GVL direct from D-
glucose took a long reaction pathway that was already made shorten into one pot.
Hence, the yield of GVL was affected. Figures 2 and 3 exhibit the yield of GVL under
various time taken from 8 to 16 h in presence of 10 and 15 % H3PW12O40/ZrO2.
10 % H3PW12O40/ZrO2 for GVL production, Based on Fig. 2, the highest
yield of GVL was 22.3 % at 16 h of reaction time. This reaction time was suitable to
produce GVL with better yield in one pot reaction. Impregnated 10 % H3PW12O40/
ZrO2 was utilized in the production of GVL from D-glucose. 10 % H3PW12O40/
ZrO2 was an appropriate wt% for the production of GVL. Moreover, it was proven
that this catalyst was capable of forming a ring closure structure of GVL directly
from D-glucose without the need to eliminate LA In this reaction, FA was used as
Fig. 2 Effect of the reaction 25

time to GVL yield in the
presence of 10 %
20
H3PW12O40/ZrO2 at 180 °C
GVL Yield (%)
15
10
5
Yield
0
8h 10h 12h 14h 16h
Reaction Time (h)
Fig. 3 Effect of the reaction Yield

time to GVL yield in the 14.9
presence of 15 %
14.8
H3PW12O40/ZrO2 at 180 °C
GVL yield (%)
14.7
14.6
14.5
14.4
Yield
14.3
8h 10h 12h 14h 16h
Time Take (Hours)
[email protected]
hydrogen source in the conversion of LA (intermediate product) to GVL. Moreover,

FA also was utilized as a homogeneous acid catalyst for the production of LA from
D-glucose via acid hydrolysis.
15 % H3PW12O40/ZrO2 for GVL production, Fig. 3 depicts the GVL (% yield)
from D-glucose in the presence of 15 % H3PW12O40/ZrO2. The highest yield of
GVL was 14.8 % at 16 h. When compared to 10 % H3PW12O40/ZrO2, 15 %
H3PW12O40/ZrO2 produced lower yield of GVL. This maybe attributed to the
acidity of the catalyst, which influenced the production of LA from acid hydrolysis
reaction of D-glucose. Based on the NH3-TPD result, 10 % H3PW12O40/ZrO2 have
higher acidic sites than 15 % H3PW12O40/ZrO2. Recently, 16.6 wt% of GVL via
one pot reaction scheme of giant reed in the presence of Ru/C with niobium
phosphate was reported [12].
Conclusion
GVL can be obtained from one pot reaction of D-glucose. FA was utilized as a
homogeneous catalyst for the dehydration reaction to LA from D-glucose and
hydrogen source for the hydrogenation reaction of LA to produce GVL. It was
found that 10 % H3PW12O40/ZrO2 was suitable for simultaneous dehydration/
hydrogenation of D-glucose which produced 22.3 % of GVL in 16 h reaction when
compared to 15 % H3PW12O40/ZrO2.
Acknowledgment The authors would like to express their gratitude for the financial support from
Universiti Teknologi Malaysia under Research University Grant (RUG) vote no 07H14 and
Ministry of Higher Education (MOHE).
References
1. Huber GW, Iborra S, Corma A (2006) Synthesis of transportation fuels from biomass:
chemistry, catalysts, and engineering. Chem Rev 106(9):4044–4098
2. Deng J, Wang Y, Pan T, Xu Q, Guo QX, Fu Y (2013) Conversion of carbohydrate biomass to
g-valerolactone by using water-soluble and reusable iridium complexes in acidic aqueous
media. Chem Sus chem 6(1163–1167):1163
3. Roman-Leshkov Y, Barrett CJ, Liu ZY, James A. Dumesic JA (2007) Production of
dimethylfuran for liquid fuels from biomass-derived carbohydrates. Nature 447:982–985
4. Huber GW, Cortright RD, Dumesic JA (2004) Renewable alkanes by aqueous-phase
reforming of biomass-derived oxygenates. Angew Chem 116(12):1575–1577
5. Huber GW et al (2005) Production of liquid alkanes by aqueous-phase processing of biomass-
derived carbohydrates. Science 308(5727):1446–1450
6. Horvath IT, Mehdi H, Fábos V, Boda L, Mika LT (2008) c-Valerolactone—a sustainable
liquid for energy and carbon-based chemicals. Green Chem 10:238–242
7. Horvath IT (2008) Solvents from nature. Green Chem 10(10):1024–1028
[email protected]
8. Tang X, Hu L, Sun Y, Zhao G, Hao W, Lin L (2013) Conversion of biomass-derived ethyl

levulinate into γ-valerolactone via hydrogen transfer from supercritical ethanol over a ZrO2
catalyst. RSC. Advances 3:10277–10284
9. Heeres H, Handana R, Chunai D, Rasrendra CB, Girisuta B and Heeres HJ (2009) Combined
dehydration/(transfer)-hydrogenation of C6-sugars (D-glucose and D-fructose) to c-
valerolactone using ruthenium catalysts. Green Chem 11:1247–1255
10. Sharma YC, Singh B, Korstad J (2011) Advancements in solid acid catalysts for ecofriendly
and economically viable synthesis of biodiesel. Biofuels, Bioprod Biorefin 5(1):69–92
11. Son PA, Nishimura S, Ebitani K (2014) Production of [gamma]-valerolactone from biomass-
derived compounds using formic acid as a hydrogen source over supported metal catalysts in
water solvent. RSC Advances 4(21):10525–10530
12. Raspolli Galletti AM et al (2013) From giant reed to levulinic acid and gamma-valerolactone:
a high yield catalytic route to valeric biofuels. Appl Energy 102(0):157–162
[email protected]
Carbon Cryogel from Lignin-Furfural
as Acid Catalyst in Esterification of Oleic
Acid
Muzakkir Mohammad Zainol, Nor Aishah Saidina Amin,

Mohd Asmadi and Norzita Ngadi
Abstract Carbon cryogel (CC) has been synthesized from commercial lignin with
furfural via acid catalyzed sol-gel polycondensation reaction step. Freeze drying
and carbonization were performed on the gels for preparation of the CC. Effect of
water and H2SO4 concentration on carbon synthesis was studied where lignin to
water (L/W) ratio and acid concentration of 1.0 and 8 M, respectively were found
suitable for CC synthesis. The selected CC exhibited higher thermal stability which
can be used in thermal reaction process. CC has potential as super acid solid
catalyst with acidity and surface area of 14.07 mmol/g and 330.35 m2/g, respec-
tively for esterification of free fatty acid to produce 91.3 wt.% FAME yield.

Keywords Carbon cryogel Acid catalyst Lignin-furfural Esterification Oleic

acid Biodiesel
Introduction
Carbon gels are different types of carbons from carbon family. There are three types
of carbon gels known as carbon aerogel, cryogel and xerogel. These types of gels
(or resins) are synthesized through sol-gel polycondensation process by using
phenol or its derivatives with formaldehyde or its derivatives. Direct replacement of
M. Mohammad Zainol N.A. Saidina Amin (&) M. Asmadi N. Ngadi

Chemical Reaction Engineering Group (CREG), Faculty of Chemical Engineering,
Universiti Teknologi Malaysia, 81300 UTM, Johor Darul Takzim, Malaysia
M. Mohammad Zainol
M. Asmadi
N. Ngadi

[email protected]
316 M. Mohammad Zainol et al.
phenol by lignin or modified lignin by phenolation can be used in resin formulation

to form phenolic resin [1]. Lignin has been used in producing resin with aldehyde
derivatives [1–4]. Instead of using formaldehyde, this chemical can be substituted
by using its derivatives such as paraformaldehyde, acetaldehyde, propionaldehyde,
butyraldehyde, glyoxal, trioxane, furfural, and furfurol in producing phenolic resins
[5]. The resin produced can be applied in the production of carbon gel by
employing drying and carbonization processes.
Different types of carbon gels are produced via different drying process. Aerogel,
cryogel and xerogel was prepared via supercritical drying, freeze drying and
evaporative drying (ambient) condition, respectively [6–9]. All these types of
carbon gels have their properties, texture and structure affected by the different
synthesis and preparation process. The carbon aerogel gave best quality and
properties of the product compared to production of carbon cryogel and xerogel but
the process is more expensive. Carbon cryogel (CC) also can give a good quality of
product and have different properties and abilities. The ability to control the pore
volume during the synthesis and drying process makes the CC more flexible with
greater potential for many applications.
Acid catalyst can be used in esterification of free fatty acid (FFA) in methanol to
produce fatty acid methyl esters (FAME). The acid catalyst can replace the usage of
base catalyst to avoid formation of soap during reaction. Heterogeneous acid cat-
alyst becomes alternative to substitute the homogeneous acid catalyst which are
corrosive to the equipment and required good separation process. The CC has
potential application in esterification of FFA as heterogeneous acid catalyst. In
order to fit the material application, the porosity and density of this carbon need to
be adjusted by controlling the synthesis, drying and pyrolysis conditions [6].
In this study, CC was synthesized to produce carbon particle and further utilized
as catalyst in esterification of FFA. The condition of gel synthesis was studied by
using different amounts of water and acid concentrations to synthesize the carbon as
catalyst. The cryogel (dried gel) was carbonized and CC produced was character-
ized using surface area analyzer (BET), temperature programme desorption (TPD),
thermogravimetric analyzer (TGA), and Fourier transform infrared spectroscopy
(FTIR).
Experimental
Chemicals and Reagents
Lignin, alkali and furfural were purchased from Sigma-Aldrich Co., United States
and Merck, Germany, respectively. These materials were used as the main feed-
stocks for resinification process. Sulfuric acid (H2SO4, 95–97 %), ethanol (C2H6O),
oleic acid (C18H34O2) and methanol, (CH4O) were mainly supplied from QRec,
New Zealand. Sulfuric acid was used as catalyst for polycondensation reaction with
[email protected]
Carbon Cryogel from Lignin-Furfural as Acid Catalyst … 317
ethanol as solvent to homogenize the solution. Oleic acid and methanol was used in
esterification reaction to produce biodiesel.
Synthesis of Carbon Cryogel
Commercial lignin (7 g) was used as the main feedstock with furfural at lignin to
furfural weight ratio of 1:1. Different amount of distilled water (by ratio) was used
as diluent and ethanol (7 mL) was used as solvent to homogenize the mixture. The
reaction was conducted in a Schott bottle with mixture of ethanol and water in
different concentrations of H2SO4 as catalyst. The mixture was stirred and heated in
silicon oil bath at 90 °C (oil bath temperature) for 0.5 h. The dark coloured resin
formed was immersed with t-butanol for solvent exchanged. Next, the product was
pre-frozen for 24 h in refrigerator and freeze dried around −60 °C (condenser
temperature) for 8 h in the freeze dryer to form cryogel. Ultimately, the cryogel was
carbonized under nitrogen flow at 500 °C for 5 h to produce carbon cryogel (CC).
The CC was characterized to study its physical and chemical properties.
Characterization
The surface area of cryogel and CC was evaluated using a Micromeritics 3Flex
3.01 instrument according to standard nitrogen adsorption and desorption at 77 K.
The BET was used to measure the surface area of the CC. The acidity of the CC
was analyzed via NH3-TPD using TCD detector by passing 10 % NH3/He gas up to
900 °C. The cryogel and CC was characterized using TGA (Perkin Elmer TGA 7)
from 30 to 900 °C under N2 flow at 10 °C/min to study its composition and product
thermal stability. Detection of chemical bonding for CC was obtained through FTIR
(Perkin-Elmer Spectrum) by using KBr powder technique and the spectra were
recorded in the IR range of 4000–400 cm−1.
Esterification of Oleic Acid
The esterification reaction of oleic acid with methanol was performed in presence of
CC for catalyst selection. The reaction condition was fixed at 65 °C and 5 h with
20:1 molar ratio of methanol to oleic acid and 5 wt.% of catalyst based on oleic acid
loading. The reaction process was conducted in a three-neck flask under reflux
condenser and constant stirring. FAME was characterized using GC-MS and the
yield of FAME was calculated based on the compositions detected through GC-MS.
[email protected]
Effect of Amount of Water on Carbon Synthesis
The preparation of CC depends on the resin (gels) synthesis through the reaction of
lignin with furfural in ethanol by acid catalyst. The amount of water added into the
mixture at constant acid concentration was studied to explore their effect on the gels
structure. Based on the result in Fig. 1, the surface area of CC seems to be affected
by the amount of water. Smaller CC surface area was observed at higher and lower
lignin to water (L/W) ratio with the optimum L/W ratio being 1.0. The CC with
optimum condition can be used as catalyst in esterification of FFA. The L/W ratio
can be adjusted to produce good structure of gels and CC product. As reported by
Babić and co-workers, they have synthesized CC and found that porosity of CC can
be controlled by changing the water ratio [10]. They reported surface area remained
unchanged with water ratio (0.5 and less) using a different feedstock. As for this
study, in contrast, the results show that the surface area was affected by the amount
of water with the feedstocks of lignin and furfural. Besides, the water ratio used was
different and depends on the synthesis process. The L/W ratio was gave a signifi-
cant effect on the lignin-furfural gels therefore consistent with observation of
Tamon et al. [8] who found that the pores in gels synthesis depended on amount of
distilled water used as diluent. They reported that the total pore volume of gels (gel
void) was larger at low resorcinol to water ratio [8]. This explained that high and
enough amount of water are required in the synthesis of gel.
The surface areas of CC are lower at 0.5 and 1.5 of L/W ratio. Smaller CC
surface area is attained at low L/W ratio possibly due to the destruction of gels and
pore structure caused by water. At high L/W ratio, the polymerization process can
take over the reaction that leads to formation of polymer mixture. This situation
affects the formation of gel particles and their surface to form a large particle with
low surface area.
Fig. 1 Effect of lignin to 240 100

water ratio on CC surface area
and FAME yield. (6 M of 200 80
H2SO4)
FAME yield (Wt.%)
S BET (m 2/g)
160
60
120
40
80
40 20
0 0
0.5 0.7 1 1.5
Lignin to water ratio
[email protected]
The production of CC can be affected by gel formation before drying and

carbonization process. Carbon cryogel with surface area of 215.1 m2/g was used as
catalyst and produced 84.3 wt.% of FAME yield. Although the surface area of CC
at 0.5 and 1.5 of L/W ratio are lower but the FAME yield are still decent at 40.0 and
37.5 wt.%, respectively. This is due to the contribution of the total CC acidity, one
of main factors in catalyst performance.
Effect of Acid Concentration on Carbon Synthesis
The effect of acid concentration used in the carbon synthesis was studied to
determine the suitable acid concentration. The amount of water was fixed based on
L/W ratio of 1.0. Figure 2 shows the total acidity of the CC increases as the acid
concentration used on gel synthesis increase. The same trend can be seen for FAME
yield by using CC synthesis from different acid concentration. The high acidity of
CC was achieved at 14.07 mmol/g by using 8 M acid concentration. The CC
produced at 8 M of acid concentration was used as catalyst in the synthesis of
FAME from oleic acid and achieved high yield of FAME at 91.3 wt.%.
At lower acid concentration, the CC produced still has high acidity, but the
FAME yield was lower owing to the smaller surface area obtained (16.2 m2/g). This
may be due to the hydrophilic colloidal particles was hard to be formed and this
situation can lead to polymer formation. Rapid polymer formation is also possible at
higher acid concentration (i.e. more than 10 M) during heating process. The gels
cannot be formed while the polymer structure is difficult to be dried and transferred
into carbon gel. Consequently, low quality CC is formed with less surface area
leading to low FAME yield. Based on the conditions studied for CC synthesis, only
the selected cryogel and CC at L/W ratio of 1.0 and 8 M H2SO4 concentration were
characterized.
Fig. 2 Effect of acid 15 100

concentration on CC acidity
and FAME yield. (1.0 L/W 14
80
Acidity (mmol/g )
ratio)
13
FAME yield (Wt.%)
60
12
11 40
10
20
9
0
2 4 6 8
H2 SO4 Conc. (M)
[email protected]
Characterization of Selected CC
The cryogel synthesized at L/W ratio of 1.0 and 8 M acid concentration has surface
area of 0.61 m2/g only. After carbonization, the surface area increased to 330.4 m2/g.
Based on BET and TPD result, CC with high acidity has demonstrated its potential
as acid catalyst for the esterification reaction to produce high FAME yield.
The thermal stability of the cryogel and CC (Fig. 3) is provided by TG analysis.
The cryogel produced was not stable at high temperature. Cryogel is dried gels and
usually the gels or resin has a lower thermal stability, which can have higher mass
decomposition above temperature of 150 °C. Rapid decomposition of cryogel up to
900 °C explains that cryogels are not thermally stable and cannot be used as catalyst
at high temperature.
The TG curve for CC exhibit higher thermal stability with less than 15 wt.%
mass decomposed for temperature up to 600 °C. As the temperature increases (600–
800 °C), the CC drastically decomposed. The structures of lignin are linked in the
gel structure through the sol-gel polycondensation. Since the carbonization of
cryogel was performed at 500 °C, thus, there is possibility that the lignin structure
was remained in the CC. The decomposition of CC in TG results between 600 and
900 °C is possibly due to the lignin in the CC structure. Based on this, the car-
bonization process is required to change the structure of cryogel into the carbon
structure which has good thermal stability. The high thermal stability of CC makes
it more applicable for high temperature reaction.
The bonding structure of the cryogel and CC was studied through FTIR and the
result is shown in Fig. 4. The broad frequency peak of alcohol group of cryogel
structure around 3300–3600 cm−1. The absorption band at 1500–1700 and 750–
830 cm−1 in the spectrum represent the frequency of aromatic compound and its
Fig. 3 TG curve for cryogel

100
and CC
80
Mass (wt.%)
60
Cryogel
40 Carbon cryogel
20
200 400 600 800

Temperature ( o C )
[email protected]
Fig. 4 FTIR spectra for

cryogel and CC
Cryogel
CC
4000 3500 3000 2500 2000 1500 1000 500

-1
Wavelenght (cm )
substituent types, respectively. There is a small frequency detected for the C–H of
alkyl group (2853–2962 cm−1). The frequency range of C–O (1260–1000 cm−1) is
found to represent the possibility structure of the phenolic group (C–O–H) and/or
ether (C–O–C) in the product. Based on Poljansek and Krajnc, the frequency of C–O
stretch is around 1157 cm−1 where the frequency range of asymmetric stretch of
phenolic C–C–OH is around 1237 cm−1 and the asymmetric stretching vibration of
C–O–C aliphatic ether is around 1100 cm−1 [11]. Thus, the FTIR spectrum detected
the formation bond of C–OH and C–O–C in the structure of cryogel. This explains the
cryogel have formation of resol type resin which is linked through dimethyl ether
bridge between the lignin structure.
The CC structures exhibit some peaks reduction due to the carbonization pro-
cess. The OH functional group has reduced at frequency range of 3300–3600 and
1000–1300 cm−1. There also have some reduction peaks between at 1500–1700 and
1000–500 cm−1 affected by carbonization process which explains the restructuring
of the CC bonding. The sulfuric acid used in preparing the gels can bond together
the gels structure during the synthesis and activate the CC to function as catalyst.
This formation of bond (sulfonates) can be observed through the band spectra
around 1350 cm−1 and around 1150 cm−1 (along with vibration band of C–O–C and
C–OH) with reference to S=O stretching. The weak band appears around 750–
1000 cm−1 corresponds to S–O stretching. The reduction of OH functional group
and the carbon structure of the CC have increased the hydrophobicity. Hydrophobic
substance is usually a non-polar substance which can dissolve or mix with other
non-polar substance. Accordingly, the CC can be mixed with oleic acid in methanol
and the bonding with –SO3H group makes CC more active to catalyze the reaction.
The spectra are consistent with the results in Fig. 2 linked to acidity of CC and
FAME yield.
[email protected]
Conclusion
The sol-gel polycondensation process was conducted to form wet gels and dried
using freeze drying to prepare cryogel. The CC synthesis from commercial lignin
with furfural catalyzed by acid sulfuric has shown the potential in esterification of
FFA. CC can be used as heterogeneous super acid catalyst based on the properties
such as high acidity and surface area. CC synthesis was selected based on L/W ratio
of 1.0 at 8 M acid concentration to give FAME yield = 91.3 wt.%.
Acknowledgment The authors would like to express their sincere gratitude to the Ministry of
Higher Education (MOHE), Malaysia and Universiti Teknologi Malaysia (UTM) for supporting
the project under Research University Grant (vote 04H69) and Fundamental Research Grant
Scheme (vote 4F160).
References
1. Çetin NS, Özmen N (2002) Use of organosolv lignin in phenol–formaldehyde resins for
particleboard production: I. Organosolv lignin modified resins. Int J Adhes Adhes 22:477–480
2. Tejado A, Peña C, Labidi J, Echeverria JM, Mondragon I (2007) Physico-chemical
characterization of lignins from different sources for use in phenol–formaldehyde resin
synthesis. Bioresour Technol 98:1655–1663
3. Vázquez G, Antorrena G, González J, Mayor J (1995) Lignin–phenol–formaldehyde adhesives
for exterior grade plywoods. Bioresour Technol 51:187–192
4. Zhang W, Ma Y, Wang C, Li S, Zhang M, Chu F (2013) Preparation and properties of lignin–
phenol–formaldehyde resins based on different biorefinery residues of agricultural biomass.
Ind Crop Prod 43:326–333
5. Fink JK (2005) 4-phenol/formaldehyde resins, in Reactive polymers fundamentals and
applications. William Andrew Publishing, Norwich, pp 241–281
6. Job N, Théry A, Pirard R, Marien J, Kocon L, Rouzaud J-N, Béguin F, Pirard J-P (2005)
Carbon aerogels, cryogels and xerogels: Influence of the drying method on the textural
properties of porous carbon materials. Carbon 43:2481–2494
7. Pekala RW (1989) Organic aerogels from the polycondensation of resorcinol with
formaldehyde. J Mater Sci 24:3221–3227
8. Tamon H, Ishizaka H, Yamamoto T, Suzuki T (1999) Preparation of mesoporous carbon by
freeze drying. Carbon 37:2049–2055
9. Scherdel C, Reichenauer G (2009) Carbon xerogels synthesized via phenol–formaldehyde
gels. Micropor Mesopor Mat 126:133–142
10. Babić B, Kaluđerović B, Vračar L, Krstajić N (2004) Characterization of carbon cryogel
synthesized by sol–gel polycondensation and freeze-drying. Carbon 42:2617–2624
11. Poljanšek I, Krajnc M (2005) Characterization of phenol-formaldehyde prepolymer resins by
in line ft-ir spectroscopy. Acta Chim Slov 52:238–244
[email protected]
Surface Structure Study of CeZrO2
Nanocatalyst Doped with Different
Transition Metals
S. Hanim Md Nor, M. Nazri Abu Shah, Kamariah Noor Ismail

and Abdul Hadi
Abstract CeZrO2 nanocatalysts doped with three transition metals of Cu, Ni

and Co were synthesized via sequence method of microemulsion and deposition-
precipitation method. The crystallinity phase was studied by using X-ray diffraction
(XRD) while the textural analysis was performed by N2 adsorption desorption
analysis. XRD results exhibited the transition metals peaks were detected on the
CeZrO2 pattern. N2 adsoption–desoption depicted that Ni/CeZrO2 produced higher
BET surface area of 35.19 m2/g as compared to Cu/CeZrO2 and Co/CeZrO2.
Therefore, this result suggested that the CeZrO2 nanocatalyst doped with Ni has a
great potential to enhance the catalytic activity.

Keywords Microemulsion method Deposition-precipitation method Catalyst

properties Transition metals Nanocatalyst Cezro2
Introduction
Cerium oxide (CeO2) is one of the essential material in three way catalyst. It is due
to ability in providing as oxygen storage capacity (OSC), unique redox properties
and low temperature catalytic activity [1–3]. The OSC function of CeO2 able to
control the oxygen concentration during stoichiometric condition [4, 5]. However,
S.H. Md Nor (&) M.N. Abu Shah K.N. Ismail A. Hadi

Faculty of Chemical Engineering, University Teknologi MARA,
M.N. Abu Shah
K.N. Ismail
A. Hadi

[email protected]
324 S.H. Md Nor et al.
the major drawback of CeO2 was easily reduced of thermal stability at high
temperature [1, 6]. In recent years, the addition of zirconium oxide (ZrO2) to CeO2
has gain much attraction among the research in order to improve the structure and
thermal properties of CeO2 [6]. Therefore, the incorporation of CeO2–ZrO2 mixed
oxide produced high surface area, higher thermal stability and performed better at
low temperature catalytic performance [1, 3].
The previous studies reported that the transition metals were able to manipulate
oxygen exchange by incorporating metals ion into ceria lattice [7–9]. Accordingly,
the result of additional transition metals into catalyst support produced better
performance at the same time lowering catalyst cost [10]. Thus, the aim of this
study is to investigate the influence of transition metals such as cobalt oxide, copper
oxide and nickel oxide doped into the CeZrO2 nanocatalyst. The catalysts were
synthesized using combination of microemulsion and deposition-precipitation
methods. The microemulsion method is one of the proven technique to produce
small size nanoparticles and narrowing size distribution of particles meanwhile, the
deposition-precipitation method are well established in producing homogeneously
dispersed the metal particles and narrowing particles size distribution [11].
Experimental
Materials, Cerium nitrate hexahydrate (Ce(NO3)3·6H2O, from Merck, 98.5 %) and

(ZrO(NO3)2·xH2O), were purchased from Merck. Hexadecyltrimethyl ammonium
bromide (HTAB, 99 %) was purchase from SIGMA, and sodium hydroxide
(NaOH), butanol and n-octane were purchased from Fisher. Nickel nitrate hydrate
(Ni(NO3)3·xH2O), Cobalt nitrate hydrate (Co(NO3)3·xH2O) and Copper (II) nitrate
trihydrate (Cu(NO3)2·3H2O) were purchased from Merck, Malaysia.
Preparation of CeZrO2 nanoparticles, CeZrO2 mixed oxide was prepared by
microemulsion method. This method was adapted from Nor et al. [12]. In order to
prepare the CeZrO2 nanocatalyst, two microemulsion solutions was prepared and
assigned as microemulsion A and microemulsion B. The microemulsion A was
consisted of cerium oxide (CeO2) and zirconium oxide (ZrO2) as aqueous solution
while the microemulsion B was consisted of sodium hydroxide (NaOH) as reducing
agent. Both of microemulsion A and B were separately stirred for 30 min until the
transparent solutions were formed. Then the microemulsion B was slowly poured
into the microemulsion A and continuously stirred for 3 h until the precipitated
formed. The precipitated was washed with distilled water and centrifuge at
10,000 rpm for 10 min. The treated precipitated was dried at 110 °C for 24 h and
followed by calcination at 500 °C for 3 h in order to obtain CeZrO2 nanoparticle.
Transition metals doped on the CeZrO2. The preparation of M/CeZrO2
(M = Co, Cu and Ni) with 5 wt% of transition metals was performed through
deposition-precipitation method which was adapted from Hadi et al. [13]. The
prepared CeZrO2 powder was kept in the air tight container at room temperature to
prevent the surface and pores particles from adsorbed moisture and volatiles. 5 wt%
[email protected]
Surface Structure Study of CeZrO2 Nanocatalyst … 325
of M solution was gradually added into the CeZrO2 powder. Then, the mixture was
stirred under magnetic stirring for 3 h until homogeneous mixture was formed.
Then, 0.5 M NaOH solution was slowly dropped into the mixture at 70 °C with
continuous stirring until pH of the dispersion reached 10. The suspension was dried
at 110 °C for 24 h and followed calcined for 3 h at 500 °C. The synthesized samples
of Co/CeZrO2, Cu/CeZrO2 and Ni/CeZrO2 were then characterized using X-ray
diffraction (XRD) and N2 adsorption desorption.
Characterization techniques, The samples crystallites were identified by X-ray
diffraction (XRD). The XRD was analyzed by Rigaku Ultima IV (X-ray
Diffractometer), using Cu Kα radiation (λ = 0.15418) and operated at 40 kV and
20 mA. N2 adsorption-desorption performed by using AUTOSORB-1 Quanta
Chrome Instruments, USA, this instrument was measured average particle diameter,
BET specific surface areas, pore volume and average pore diameter.
Structural properties, Fig. 1 shows the XRD patterns of Cu/CeZrO2, Ni/CeZrO2

and Co/CeZrO2 respectively. The strong peaks at 2θ of 28.8°, 33.4°, 47.9° and
56.8° indicated the CeZrO2 peaks. The peaks of 5 wt% Ni/CeZrO2 at 2θ = 43.3°
correspond to cubic phase type structure for NiO. On the other hand, the diffraction
peaks of 5 wt% of Cu/CeZrO2 occur at 2θ = 35.5° and 38.8° manifest the
monoclinic phase. However, the peaks for Co/CeZrO2 was not detected due to the
concentration of 5 wt% considered as under detection limit for CoO. The presence
of transition metals phases was proved by narrowing scan solution at a small peak
Fig. 1 XRD patterns of Cu/CeZrO2, Ni/CeZrO2, and Co/CeZrO2
[email protected]
of Ni/CeZrO2 and Cu/CeZrO2 detected. The XRD patterns obtained from the
narrowing scan of Ni/CeZrO2 and Cu/CeZrO2 are shown in Figs. 2 and 3.
Figures 2 and 3 show the peaks of NiO was appeared at 2θ of 43.3° while CuO
peaks was detected at 2θ of 35.5° and 38.8°. This finding is in consistent with Hu
et al. [14] which found the corresponding peaks of NiO was located at 2θ of 44.5°
and the peaks was more intense with increasing of Ni content. This result is in
consistent with, Kitla et al. [15] where it indicated narrowing diffraction of
monoclinic tenorite-phase of Cu at 35.7° and 39°.
50
43.3°
45
40
Intensity (CPS)
35
30
25
20
15
10
40 40.5 41 41.5 42 42.5 43 43.5 44 44.5 45
2θ
Fig. 2 The XRD pattern of Ni/CeZrO2 at 2θ between 40° and 45° by using narrowing scan
resolution
35.5°
80 38.8°
70
60
intensity, (a.u.)
50
40
30
20
10
35 35.5 36 36.5 37 37.5 38 38.5 39 39.5 40
2
Fig. 3 The XRD pattern of Cu/CeZrO2 at 2θ range of 35°–40° by using narrowing scan resolution
[email protected]
It can be observed from the XRD pattern in Figs. 2 and 3, the transition metals of
CuO and NiO present in the CeZrO2 structure. These results suggested that the
transition metals were exclusively deposited on the CeZrO2 nanocatalyst.
Textural properties, Fig. 4 shows the N2 adsorption–desorption isotherm
profiles for Cu/CeZrO2, Ni/CeZrO2 and Co/CeZrO2 samples. All samples are
corresponding to the typical type IV of isotherm which indicated the mesoporous
nanocatalyst. This type represents an H1 hysteresis loop, which reported as
cylindrical pore channels with compact of uniform spheres and pores size nano-
catalyst [16].
Table 1 summarizes the average particle diameter, BET specific surface areas,
pore volume and average pore diameter of Cu/CeZrO2, Ni/CeZrO2, and Co/CeZrO2
respectively. Ni/CeZrO2 exhibited higher specific surface area, SBET of 35.19 m2/g
compared to Co/CeZrO2 and Cu/CeZrO2. Theoretically, the specific surface area is
correlated to the pore diameters and size particles [17]. These results show the trend
of average particles size of Ni/CeZrO2, Co/CeZrO2 and Cu/CeZrO2, calculated using
Langmuir equation, increase respectively with the specific surface area decreases.
90
Cu/CeZrO2
80
Ni/CeZrO2
Vol. adsorbed cm 3 /g STP
70
Co/CeZrO2
60
50
40
30
20
10
0
0 0.2 0.4 0.6 0.8 1
Relative Pressure (P/Po)
Fig. 4 N2 adsorption–desorption isotherm of a Cu/CeZrO2, b Ni/CeZrO2 and c Co/CeZrO2
Table 1 Textural properties by N2 adsorption-desorption analysis

Sample Average particle Specific surface Pore volume Average pore
diameter (nm) dLangmuir area SBET (m2/g) (cm3/g) diameter (Å)
Cu/CeZrO2 47.4 12.25 0.06 124.91
Ni/CeZrO2 15.4 35.19 0.12 115.20
Co/CeZrO2 22.5 25.01 0.10 126.98
[email protected]
Conclusion
CeZrO2 nanocatalyst doping with Cu, Co and Ni metals was successfully prepared
by sequence method of microemulsion and deposition-precipitation. The XRD
results found that depositions of CuO and NiO were detected. These findings of
BET analysis results considered the particles size of Cu/CeZrO2 catalysts was
47.4 nm, Ni/CeZrO2 (15.4 nm) and Co/CeZrO2 (22.5 nm). It can be assumed that
the smaller particle diameter with highest particles size will produce greater of
surface area to increase the catalyst activation process. This result confirms that the
formation of Ni metal on the CeZrO2 nanocatalyst shows as a great potential to
enhance the catalytic activity.
Acknowledgment Gratefully thanks to the e-sciencefund-Ministry of Science, Technology and

Innovation, Malaysia (Grant no. 100-RMI/SF 16/16 (20/2012)), Research Management Institute
(RMI) and Universiti Teknologi MARA for supporting this research.
References
1. Li G, Wang Q, Zhao B, Zhou R (2011) Dynamic oxygen mobility and a new insight into the
role of Ni doping in Pd/CeO2-ZrO2 three way catalysts. Catal Today 175:40–47
2. Lamacz A, Krzton A, Djega-Mariadassou G (2011) Catalytic decomposition of nitrogen
oxides from coal combustion flue gases on CeZrO2 supported Cu catalysts. Catal Today
176:126–130
3. Yuechang W, Jian L, Zhen Z, Guiyuan J, Aijun D, Hong H, Xinping W (2010) Preparation
and characterization of Co0.2/Ce1-xZrxO2 catalysts and their catalytic activity for soot
combustion. Chin J Catal 31:283–288
4. Khan A, Sukonket T, Saha B, Idem R (2012) Catalytic activity of various 5 wt% Ni/
Ce0.5Zr0.33M0.17O2-δ catalysts for the CO2 reforming of CH4 in the presence and absence of
steam. Energy Fuels 26:365–379
5. Deraz NM (2009) Characterization and catalytic performance of pure and Li2O-doped CuO/
CeO2 catalyst. Appl Surf Sci 255:3884–3890
6. Neto RCR, Schmal M (2013) Synthesis of CeO2 and CeZrO2 mixed oxide nanostructured
catalysts for the iso-syntheses reaction. Appl Catal A Gen 450:131–142
7. de Leitenburg C, Trovarelli A, Kaspar J (1997) A temperature-programmed and transient
kinetic study of Co2 activation and methanation over CeO2 supported noble metals. J Catal
166:98–107
8. Srinivas D, Satyanarayana CVV, Potdar HS, Ratnasamy P (2003) Structural studies on NiO-
CeO2-ZrO2 catalysts for steam reforming of ethanol. Appl Catal A Gen 246:323–334
9. Liu W, Flytzani-Stephanopoulos M (1995) Total oxidation of carbon monoxide and methane
over transition metal-Flourite oxide composite catalysts. J Catal 153:304–316
10. Matam SK, Otal EH, Aquirre MH, Winkler A, Ulrich A, Rentsch D (2012) Thermal and
chemical aging of model three-way catalyst Pd/Al2O3 and its impact on the conversion of
CNG vehicle exhaust. Catal Today 184:237–244
11. Eriksson S, Nylen U, Rogas S, Boutonnet M (2004) Preparation of catalysts from
microemulsions and their applications in heterogeneous catalysis. Appl Catal A Gen
265:207–219
12. Nor SHM, Shah MNA, Hadi A, Ismail KN (2014) Effect of Co doping on the properties of
Ce0.75Zr0.25O2 nanocatalysts. Appl Mech Mater 575:93–96
[email protected]
13. Hadi A, Yaacob II (2004) Synthesis of PdO/CeO2 mixed oxides catalyst for automotive
exhaust emissions control. Catal Today 96:165–170
14. Kitla A et al (2013) Infrared studies on bimetallic Copper/Nickel Supported on Zirconia and
Ceria/Zirconia. Catal Lett. doi:10.1007/s10562-013-1001-y
15. Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti RA, Rouquerol J, Siemieniewska T
(1985) Reporting physisorption data for gas/solid systems with special reference to
determination of surface area and porosity. Pure Appl Chem 57(4):603–619
16. Jena A, Gupta K (2010) A novel technique for surface area and particle size determination of
components of fuel cells and batteries. Cornell Business and Technology Park, Porous
Materials, Inc., 83 Brown Road, Ithaca, NY 14850, 22 Feb 2010
17. Luo Y, Xiao Y, Cai G, Zheng Y, Wei K (2012) Performance of Ce0.25Zr0.75O2 promoted Pd/
Ag/γ-Al2O3 catalysts for low-temperature methanol oxidation. Fuel 93:533–538
[email protected]
Electrical Properties of Ammonium Iodide
Doped Cellulose Acetate Based Polymer
Electrolyte
N.S. Samsi, R.M. Ali, R. Zakaria, M.Z.A. Yahya and A.M.M. Ali
Abstract Solid Polymer Electrolyte (SPEs) composed of cellulose acetate (CA)

doped with different stoichiometric ratio of ammonium iodide (NH4I) was study.
Alternating current impedance spectroscopy was carried out in the frequency range
between 50 Hz and 1 MHz at room and elevated temperatures. The highest ionic
conductivity obtained from the complex impedance spectra of the order of
10−4 S cm−1 at room temperature with the polymer-salt composition of CA (75
wt%):NH4I (25 wt%). Temperature dependence of the conductivity for all samples
studied was of the Arrhenius type. Modulus formalism analysis verified the samples
are ionic conductor and was confirmed by electronic transference number. X-ray
diffraction spectral analysis showed the introduction of doping salt was decreased
the degree of crystallinity of polymer host.
Keywords Cellulose acetate Solid polymer electrolyte Ionic conductivity
N.S. Samsi (&) R. Zakaria A.M.M. Ali

Faculty of Applied Sciences, Universiti Teknologi MARA,
A.M.M. Ali
R.M. Ali M.Z.A. Yahya
Universiti Pertahanan Nasional Malaysia, Kem Sungai Besi,
A.M.M. Ali
Institute of Science, Universiti Teknologi MARA, 40450 Shah Alam,
Selangor, Malaysia

[email protected]
332 N.S. Samsi et al.
Introduction
Solid Polymer Electrolytes (SPEs) are among the important classes of macromol-
ecules. These electrolytes are charged macromolecules containing a large number of
ionisable or ionic groups [1]. The ionic conductivity of SPEs regularly takes place
through two distinct events: the first is due to the charge migration of ions between
the coordinate sites of the host polymer and the second is associated with the
polymeric chain segmental motion [2]. The charge migration was reported strongly
associated by various factors such as crystallinity of the material, simultaneous
cation and anion motions and the ion pair formation [3]. In general the conductivity
of SPEs coupled to the reduction of the crystalline domain of polymer host with the
high ionic dissociation of salts and also reducing in ion pair. To date, studies on the
electrical properties of SPEs are a must before it will propose for suitable potential
application in electrochemical devices [4]. So far, various type of polymer host has
been explored to study their electrical behaviour including synthetic and natural
polymers. However, research on natural polymers lately become more vital due to
these polymers offer good electrical properties, possess high mechanical strength
and able to solvate salt into ion [5]. Consequently, in this work we used cellulose
acetate (CA) as a polymer host to prepare the SPEs. The selection of this polymer
host is due to the present of C=O functional groups in their structure with the
oxygen atom possess a lone pair of electron [6]. In addition, cellulose is abundant
material, low cost, good mechanical properties and biodegradable.
Experimental
Solid polymer electrolytes (SPEs) were prepared by solution casting method.

Cellulose acetate (CA) with 39.8 wt% acetyl content were used as polymer host,
ammonium iodide (NH4I) salt (Aldrich—purity 99 %) were used as ion sources and
acetone (Aldrich) as a solvent. NH4I was dried for about 2 h at 100 °C prior used.
CA and NH4I with different stichiometric ratio were dissolved in 30 mL acetone for
48 h at room temperature until both the polymer and salt were completely dis-
solved. The homogeneous solutions were then cast onto glass petri dishes and
allowed to dry to form films at room temperature. The films were kept in desiccators
with silica gel desiccants for further drying before being characterized. The ionic
conductivity of the samples was performed by using HIOKI 3532-50 LRC
Hi-Tester in frequency range between 100 Hz and 1 MHz at room and elevated
temperature. The samples were sandwiched between two stainless steel electrodes
(SUS316) with force 9.8 N and placed in a SH-221 Bench-Top type temperature
and humidity chamber at constant humidity 40 % RH. The electronic transference
number were taken using an Auto-Lab PGSTAT30 potentiostat/galvanostat ana-
lyzer with a constant direct current (DC) potential 10 mV applied voltage across a
[email protected]
Electrical Properties of Ammonium Iodide Doped Cellulose … 333
symmetric SUS316 electrodes with cell configuration of SUS316/SPE/SUS316.

The X-ray diffraction (XRD) studies were carried out by means of X’Pert PRO
X-ray diffractometer in 2θ range of 5° ≤ 2θ ≤ 70°.
According to the sample preparation as described above, a series of SPEs film with
NH4I salt of 5–30 wt% was obtained. It was found that the films are opaque and
mechanically stable. The crystalline properties of the films were analyzed by XRD
result. Figure 1 shows the XRD pattern of the CA–NH4I films with different salt
concentration, pure NH4I salt and pure CA. As can be seen in Fig. 1a, the NH4I salt
are in their crystalline form with 2θ peaks at 21.2°, 24.5°, 34.9°, 41.2°, and 56.7° and
as-prepared CA film Fig. 1b also has obviously crystalline with amorphous hump
peaks centre at 2θ = 8.5° and 17.9° respectively. Upon the addition of NH4I onto CA
the XRD spectrum Fig. 1c–h demonstrates all the crystalline peaks possessed by
NH4I and CA has disrupted. Disappearing of all of these peaks indicates that the salt
and polymer is well mixed and form polymer-salt complexes. Since the CA has C=O
functional group in their structure, and the oxygen atom possess a lone pair of
electrons, the formation polymer-salt complexes may takes place via C=O←NH4+
coordination bond. This complexation wills results the dissociation of ammonium
salt into ions and the absent of excess salt can be seen in the CA–NH4I XRD
Fig. 1 XRD patterns of:

(a) pure NH4I, (b) pure CA,
and (c)–(h) different
concentration of NH4I salt
(5–30 wt%) on CA–NH4I
system
[email protected]
spectrum [5]. In contrary the individual characteristic peaks of CA in the CA–NH4I

films are not revealed. The amorphous humps also become less prominent and
broadened upon the addition of NH4I. It can be inferred that all complexes samples
are less crystalline and the amorphous phase become dominant. The reducing of
crystalline nature in CA–NH4I complexes is expected to exhibit the highest ionic
conductivity and the sample with composition of 25 wt% NH4I salt displayed the
most plateaus in the XRD patterns.
The ionic conductivity of SPEs was obtained from impedance measurement.
Figure 2 shows the Arrhenius plots of conductivity depending on the NH4I salt
content. As observed, the ionic conductivity of SPEs increases with the increasing
salt concentration up to maximum value 1.99 × 10−4 S cm−1 at room temperature
when 25 wt% NH4I was added. The increase in conductivity is due to the increase
in the number of mobile ions present in the samples. These mobile ions originate
from the dissociation of NH4I salt also were complexed with the polymer host as a
result was enhanced the amorphousness of sample. Accordingly, the polymer-salt
complexes will favour the ions transport which associated with both the segmental
motion of the polymer chain and the long-range path provided by the ions in the
amorphous phase hence rising in conductivity [7]. In addition the complexation also
will reduce the columbic force between the cation and anion of the salt to form ions
aggregation. However, the ionic conductivity was found to decrease at higher salt
loading (more than 25 wt%). This phenomenon can be attributed to the re-associ-
ation of the ions into natural ions aggregates [8, 9]. It is believed that the
re-association of ions is strongly connected to the complexation site.
When the dissociated salt is high, the coordination between the NH4+ ions and the
polymer host is striking and left less coordination site provided by the polymer.
Therefore, when more salt was added, higher probability of ions association will take
place and form an ions aggregation. The aggregated ion will then consume a large
free volume environment, which is fundamentally needed by the polymer chain to
flexible. As a result the movement of the polymer chain flexibility will reduce and
hence reducing the ionic conductivity. This basis is well agreed with the XRD result,
which higher salt loading the crystallinity of the SPE also was increased.
Fig. 2 Arrhenius plot for -3.80E+00

CA–NH4I SPEs with different -3.85E+00
wt% of salt: (a) 15 wt%,
-3.90E+00
(b) 20 wt%, (c) 25 wt% and (c)
Log σ (S cm-1)
-3.95E+00 R 2 = 0.9942
(d) 30 wt%
-4.00E+00
-4.05E+00 R 2 = 0.9962
(b)
-4.10E+00 (d)
R 2 = 0.9965
-4.15E+00
-4.20E+00 (a)
R = 0.9927
-4.25E+00
2.7 2.8 2.9 3 3.1 3.2 3.3 3.4
1000/T (K-1 )
[email protected]
Usually in the case of SPEs the thermally assisted behaviour is observed for the
temperature dependence conductivity. Therefore, according to the Arrhenius law:

EA
r ¼ ro exp ð1Þ
kT
where σo is the exponential factor, EA is the activation energy, k is the Boltzmann

constant, and T is the absolute temperature) the plots of log σ versus 1000/T will be
manifested the possible mechanism of ionic conduction in the SPEs system. Based
on this plot as shown in Fig. 2, it was observed the conductivity increases mono-
tonically as the temperature rises indicating the temperature play a significant role
to enhance the ionic mobility. This effect can be interpreted as the increments in the
segmental and rotational structural motions of the polymeric chain [8].
Regression values (R2) obtained is close to unity suggesting that all the points lie
on a straight line. Thus, the graph profile can be considered as Arrhenius plot and
subsequently the nature of ionic transport in the SPEs sample is quite similar to that
of in ionic crystals, which the ions able to jump from one vacant site into neigh-
bouring vacant sites and thereby increase conductivity [10, 11]. The EA of the SPEs
was obtained from the slope of the plot and was tabulated in Table 1. The value
shows energy required to transport conducting species in such SPEs dominate
equally with the component present the system which correlates to the ionic
conductivity.
Figure 3 shows the plots of the imaginary electrical modulus (Mi) versus
frequency (f) at ambient temperature for various salt concentrations in CA–NH4I
complexes. In polymeric materials, the complex electric modulus has been used to
investigate the conductivity relaxation phenomena. From physical point of view the
conductivity relaxation process is allied to the applied electric field and be able to
discriminate between electrode polarization and ionic relaxation process. As seen in
Fig. 3, the Mi plot showed a relaxation a peak present in all samples indicates the
SPEs in the present study are ionic conductors [10]. The shifting of these peaks
towards higher frequencies manifested the conductivity increased with salt con-
centration. The frequency range, within which the (Mi)max peaks occur is indicative
of the transition of ions from a long range to a short range mobility at increasing
frequency and can be defined by the condition 2πfmaxτ = 1, where τ is the con-
ductivity relaxation time and fmax is the frequency at which (Mi)max occurred.
Table 1 The ionic conductivity and the electric modulus parameters of CA–NH4I systems
Sample Capacitance, Relaxation time, Conductivity, Activation
name C (F) × 10−15 τ (s) × 10−4 σ (S cm−1) × 10−4 energy, EA (eV)
A 17.6 5.39 0.98 0.3374
B 1.40 0.50 1.61 0.2664
C 0.70 0.24 1.99 0.1673
D 7.03 2.70 1.33 0.2837
[email protected]
Fig. 3 Imaginary part of

electric modulus (Mi) versus
frequency ( f ) for CA–NH4I
complexes with: (a) 15 wt%,
(b) 20 wt%, (c) 25 wt% and
(d) 30 wt% of NH4I content
The equivalent frequency at (Mi)max is known as the relaxation frequency. The

capacitance value of the relaxation peak can be calculated from the equation:
eo
MiðmaxÞ ¼ ð2Þ
2C
These values are tabulated in Table 1. The capacitance value was found to be in
the range 10−14–10−15 F indicates that the dispersion peaks observed are attribut-
able to the bulk effect of the material. This suggests that the relaxation peaks is
dominated by ions and not at the polarization of electrode-electrolytes interface.
To justify the conductivity process mainly contributed by the ions, the trans-
ference number was calculated based on Wagner’s DC polarization technique [12].
This method was used to analyse the mobile species in the electrolyte. The
polarization current was monitored as a function of time on the application of DC
potential across the cell. The current decay immediately and asymptotically
approaches steady state. The total ionic transference number was calculated from
the polarization current versus time plots as shown in Fig. 4 by using the standard
equations are as follows:
Fig. 4 Chronoamperometry
of polarization current as a
function of time for the
sample CA (75 wt%):NH4I
(25 wt%) under constant
voltage of 10 mV at room
temperature
[email protected]
tion ¼ 1 If =Ii ð3Þ
tele ¼ 1 tion ð4Þ
where Ii is the initial current and If is the final residual current. The ionic trans-
ference number of SPE containing 25 wt% salts was found to be 0.916 suggesting
that the charge transport in this electrolyte film confirmed mostly due to the ions.
Thus, it can be conclude that the increase in the conductivity is attributed to the
increase in the number of ions, and only a negligible contribution comes from the
electron.
Conclusion
Solid Polymer electrolytes (SPEs) based on different weight percentage ammonium

iodide (NH4I) doped onto cellulose acetate (CA) were successful prepared and
characterised. The SPE with the composition CA (75 wt%):NH4I (25 wt%) exhibits
the highest ionic conductivity of 1.99 × 10−4 S cm−1 at room temperature. The
electrolyte system has an ionic transference number ion of 0.916 and suggests that
charge transport in this polymer electrolytes system is due to ions. XRD spectral
analysis shows CA was complexed with NH4I and formed polymer salt complexes
with reduce crystallinity of pristine materials.
Acknowledgment The authors would like to thank to the Ministry of Education Malaysia
(KPM), Research Management Institute (RMI), Institute of Science (IOS) and Faculty of Applied
Sciences (FSG), Universiti Teknologi MARA for funding this research.
References
1. Gel YS, Jin SH (2004) A self-doped ionic conjugated polymer: poly (2-ethynylpyridinium-n-
benzoylsulfonate) by the activated polymerization of 2-ethynylpyridine with ring-opening of
2-sulfobenzoic acid cyclic anhydride. Bull Korean Chem Soc 256:777–778
2. Pradhan DK, Choudhary RNP, Samantaray BK (2009) Studies of dielectric and electrical
properties of plastized polymer nanocomposite electrolyte. Mater Chem Phys 115:557–561
3. Pradhan DK, Choudhary RNP, Samantaray BK (2008) Studies of dielectric relaxation and AC
conductivity behaviour of plasticized polymer nanocomposite electrolytes. Int J Electrochem
Sci 3:597–608
4. Armand M (1994) The history of polymer electrolyte. Solid State Ionic 69:309–319
5. Khiar ASA, Puteh R, Arof AK (2006) Conductivity studies of a chitosan-based polymer
electrolyte. Physica B 373:23–27
6. Sreekanth T, Jaipal Reddy M, Subramanyam S, Subba Rao UV (1999) Ion conducting
polymer electrolyte film based on (PEO + KNO3) system and its application as an
electrochemical cell. Mater Sci Eng B 64:107–112
[email protected]
7. Chen-Yang YW, Chen HC, Lin FJ, Chen CC (2002) Polyacrylonitrile electrolytes a novel
high-conductivity composite polymer electrolyte based on PAN, LiClO4 and α-Al2O3. Solid
State Ionic 150:327–335
8. Mellende BE, Albinsson I (1996) Ion association in alkali triflate salt poly(propylene oxide)
polymer electrolytes. In: Chowdari BVR, Dissanayake MAKL, Careem MA (eds) (1996)
Solid state ionics: new developments. World Scientific, Singapore, p 97
9. Rajendran S, Mahendran O, Mahalingam T (2002) Thermal and ionic conductivity studies of
plasticized PMMA/PVdF blend polymer electrolytes. Eur Polym J 38:49–55
10. Khiar ASA, Arof AK (2010) Conductivity studies of starch-based polymer electrolytes. Ionics
16:123–129
11. Ramesh S, Arof SAK (2001) Ionic conductivity studies of plasticized poly (vinyl chloride)
polymer electrolytes. Mater Sci Eng B 85:11–15
12. Wagner H, Richert R (1997) Thermally stimulated modulus relaxation in polymer: method and
interpretation. Polymer 38:5801–5806
[email protected]
Synthesis, Characterization
and Corrosion Inhibition Studies
of 2-Methylbenzoylthiourea Derivatives
Noor Khadijah Mustafa Kamal, Adibatul Husna Fadzil

and Karimah Kassim
Abstract The effect of new corrosion inhibitors, namely N-(2-methoxyphenyl)-N’-

(2-methylbenzoyl) thiourea (A1). N-(3-methoxyphenyl)-N’-(2-methylbenzoyl)
thiourea (A2), N-(4-methoxyphenyl)-N’-(2-methylbenzoyl)thiourea (A3) and N-(2-
methylbenzoyl)-N’-phenylthiourea (A4) on mild steel in 1.0 M H2SO4 with vari-
ation concentrations of inhibitors: 1 × 10−5, 1 × 10−4 and 1 × 10−3 M was tested by
using weight loss method and linear polarization resistance (LPR) technique. These
compounds were characterized using elemental analyser, FTIR, 1H and 13C NMR
spectroscopy. Result obtained for IR showed the important bands of thiourea
compound such as ν(N–H), ν(C=O), ν(C–N) and ν(C=S) which were observed
around 3200, 1670, 1250 and 700 cm−1. The 1H chemical shifts for CONH and
CSNH had been observed around 11 and 12 ppm respectively. For the 13C reso-
nance, the signal of C=O and C=S had been observed at 180 and 206 ppm. The
presence of the inhibitors significantly reduced the corrosion rate of the mild steel.
The most promising inhibitors action is at 1 × 10−3 M and was achieved by
compound A1 where the inhibition efficiency is 82.4–86.2 %. The effectiveness of
the compounds was affected by the presence and absence of methoxy as the sub-
stituents in the compounds.

Keywords Methylbenzoylthiourea Corrosion inhibitor Spectroscopic studies

Corrosion inhibition studies Inhibition efficiency
N.K. Mustafa Kamal (&) K. Kassim

Faculty of Applied Sciences, Universiti Teknologi
MARA, 40450 Shah Alam, Selangor, Malaysia
K. Kassim
A.H. Fadzil
Faculty of Applied Sciences, Universiti Teknologi MARA (Perak),
Kampus Tapah, Tapah Road, 35400 Tapah, Perak, Malaysia
K. Kassim
Institute of Science, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia

[email protected]
340 N.K. Mustafa Kamal et al.
Introduction
Mild steel is highly susceptible to corrosionsince it is widely used as structural

material that involved in industrial processes such as acid pickling, acid cleaning,
acid descaling and oil-well acidizing [1–4]. In order to reduce the corrosion activity,
organic inhibitors containing nitrogen, oxygen, sulphur and/or phosphorus atom in
their structure are the most efficient inhibitor [5, 6]. The inhibitors which have both
nitrogen and sulphur atoms simultaneously like thiourea provide a better inhibition
performance [6]. Apart from that, the presences of aromatic rings in the compounds
make it more effective inhibitors in acid solutions. Heterocyclic ring structure which
contains nitrogen and oxygen atoms can enhance greater adsorption on metal
surface [7]. Therefore, heterocyclic structure of thiourea is the best organic com-
pound that can be used as the corrosion inhibitor. In this study, a series of thiourea
derivatives were synthesized and applied as corrosion inhibitors in order to
investigate the inhibition efficiency of the compounds (Fig. 1).
Experimental
Physical Measurements
This experiment was carried out under an ambient atmosphere and no special
requirements were taken to exclude air or moisture during the experimentation. All
chemicals were purchased from Sigma Aldrich or MERCK and used as received
Fig. 1 The chemical structure of a N-(2-methoxyphenyl)-N’-(2-methylbenzoyl)thiourea, A1; b N-

(3-methoxyphenyl)-N’-(2-methylbenzoyl)thiourea, A2; c N-(4-methoxyphenyl)-N’-(2-meth-
ylbenzoyl)thiourea, A3 and d N-(2-methylbenzoyl) -N’-phenylthiourea, A4
[email protected]
Synthesis, Characterization and Corrosion Inhibition … 341
without further purification. The elemental analyses of compound were performed

by CHNS Analyzer Flash EA 1112 series. Infrared spectra were recorded using
FTIR Perkin Elmer 100 Spectrophotometer in the spectral range of 4000–400 cm−1
by using KBr pellet. 1H and 13C NMR spectra were obtained from Bruker Advance
III 300 Spectrometer at room temperature.
Synthesis of N-(2-Methoxyphenyl)-N’-(2-Methylbenzoyl)
Thiourea, A1
Freshly prepared 2-methylbenzoyl chloride (5 mmol) in 15 ml acetone was added to

equal molar of ammonium thiocyanate. The solution was stirred about 20 min. At
the same time, a solution with o-anisidine and aniline (5 mmol) in acetone were
prepared and added to the previous solution. The reaction solution was heated under
reflux for 2 h. The solution was poured into a beaker containing some ice cubes.
The resulting precipitate was collected by filtration, washed with ethanol and re-
crystallized by using acetonitrile. This compound was obtained as dark brown
solids with percentage yield of 56.1 %. IR (KBr pellet, cm−1): ν(N–H) 3184.23,
ν(C=O) 1678.54, ν(C–N) 1249.76, ν(C=S) 739.52, ν(OCH3) 754.18. 1H NMR δ
2.42 (s, 3H, CH3); 6.90–7.55 (m, Ar–H); 11.70 (d, H, CONH); 12.90 (s, H, CSNH);
3.89 (s, 3H, OCH3). 13C NMR δ 19.89 (CH3); 120.25–136.35 (aromatic ring);
206.97 (C=O); 178.19 (C=S); 56.54 (OCH3). Anal. Calc. For C16H16N2SO: C,
64.00; H, 5.33; N, 9.33; S, 10.67. Found: C, 61.27; H, 3.55; N, 9.01; S, 7.81.
Thiourea, A2
In the similar manner as described above, compound A2 was obtained by the

reaction of 2-methylbenzoyl isothiocyante with m-anisidine, as light yellow solids
(69.5 %). IR (KBr pellet, cm−1): ν(N–H) 3184.23, ν(C=O) 1674.42, ν(C–N)
1279.21, ν(C=S) 739.52, ν(OCH3) 899.51. 1H NMR δ 2.49 (s, 3H, CH3); 7.20–7.60
(m, Ar–H); 11.70 (d, H, CONH); 12.55 (s, H, CSNH); 3.78 (s, 3H, OCH3). 13C
NMR δ 19.93 (CH3); 122.47–136.49 (aromatic ring); 206.99 (C=O); 179.25 (C=S);
54.23 (OCH3). Anal. Calc. For C16H16N2SO: C, 64.00; H, 5.33; N. 9.33; S, 10.67.
Found: C, 63.01; H, 5.31; N, 8.96; S, 7.62.
[email protected]
Thiourea, A3
This compound was prepared as described above using 2-methylbenzoyl isothio-

cyante with p-anisidine, as dark purple solids (67.6 %). IR (KBr pellet, cm−1): ν(N–
H) 3160.02, ν(C=O) 1680.42, ν(C–N) 1238.62, ν(C=S) 737.75, ν(OCH3) 831.09.
1
H NMR δ 2.50 (s, 3H, CH3); 7.25–7.70 (m, Ar–H); 11.65 (d, H, CONH); 12.40 (s,
H, CSNH); 3.78 (s, 3H, OCH3). 13C NMR δ 19.92 (CH3); 114.26–136.44 (aromatic
ring); 206.93 (C=O); 180.01 (C=S); 55.76 (OCH3). Anal. Calc. For C16H16N2SO:
C, 64.00; H, 5.33; N, 9.33; S, 10.67. Found: C, 63.65; H, 4.70; N, 9.12; S, 8.11.
Synthesis of N-(2-Methylbenzoyl) -N’-Phenylthiourea, A4
The compound A4 was obtained by the reaction of 2-methylbenzoyl isothiocyante

with aniline in similar manner as described above, as white solids (69.1 %). IR
(KBr pellet, cm−1): ν(N–H) 3222.11, ν(C=O) 1673.80, ν(C–N) 1253.53, ν(C=S)
729.27. 1H NMR δ 2.45 (s, 3H, CH3); 7.20–7.75 (m, Ar–H); 11.70 (d, H, CONH);
12.60 (s, H, CSNH). 13C NMR δ 19.93 (CH3); 124.70–136.49 (aromatic ring);
206.97 (C=O); 179.43 (C=S). Anal. Calc. For C15H13N2SO: C, 64.91; H, 4.83; N,
10.41; S, 10.67. Found: C, 64.58; H, 3.05; N, 10.09; S, 7.81.
Corrosion Inhibition Studies
Weight loss method, The mild steel coupons were used as specimen with surface
area of 4 cm2. The coupons were abraded with emery paper and washed with
acetone followed by distilled water and finally dried. The coupon were immersed
for a week in 0.1 M sulphuric acid, H2SO4 solution with and without inhibitors. The
concentrations of the inhibitors used for the inhibition studies are 1 × 10−5 M,
1 × 10−4 M to 1 × 10−3 M. After the immersion, the weight loss of each coupon was
recorded. Each samples was triplicate. The IE% was calculated by using the fol-
lowing formula:
ðWL without inhibitor WL with inhibitorÞ

IE% ¼ 100 %
WL without inhibitor
where WL is weight loss of mild steel.

Linear polarization technique, This method was carried out in a three-elec-
trode cells using AUTOLAB instrument equipped with NOVA software. A satu-
rated calomel electrode (SCE) was used as a reference electrode and a graphite
electrode as a counter. The working electrode was prepared by embedding a rod
[email protected]
with mild steel coupon in epoxy resin, with an exposed surface area of 0.065 cm2.
The surfaces were polished with emery paper and washed with distilled water for
electrochemical studies. Prior to analysis, the working electrode was immersed in
the tested solution for 30 minutes to achieve oxidation stability.
The FTIR spectra for all four compounds had shown the expected IR frequencies
for the important functional groups at a range of 3200 cm−1 for N–H, 1670 cm−1 for
C=O, 1250 cm−1 for C–N and 700 cm−1 for C=S. These assignments for thiourea
derivatives were confirmed by [8, 9]. The functional group of methoxy (–OCH3)
were found at different frequencies between compound A1, A2 and A3. A strong
peak was observed at 754.18 cm−1 in A1 indicates the presence of methoxy group
in ortho- positions, a medium intensity peak in A2 at 899.51 cm−1 for meta-
positions and a strong peak was observed in A3 at 831.09 cm−1 for para- positions
[10]. In A4, there is no peak that shows the existence of methoxy since it is a free
phenyl ring.
The 1H NMR data reveal the existence of methoxy proton between 3.78 and
3.89 ppm. One resonance was observed at 2.42–2.50 ppm as singlet shows that
proton of methyl substituents attached to phenyl ring existed. The aromatic proton
was found in the range of 6.90–7.75 ppm as distinctive multiplet due to o-,m- and p-
substituent methoxy groups at the phenyl ring and the overlapping proton in the
aromatic groups. The proton related to CONH and CSNH were observed at 11 and
12 ppm, respectively. The signals are slightly different because of intramolecular
hydrogen bonding in NH and the deshielding of aromating ring [9]. In 13C NMR
spectra, carbon spectra for CONH and CSNH were found around 180 and 210 ppm,
respectively.
For the corrosion inhibition studies, it was found that all four compounds were
able to inhibit corrosion activity in H2SO4 solution. From Table 1 and Fig. 2, it can
be observed that as the concentration of the compounds increased, the inhibition
efficiency (IE%) also increased. The highest IE% is 82.4 % which was achieved at
concentration 1 × 10−3 by compound A1. From the LPR technique, electrochemical
parameters such as corrosion current density (icorr), corrosion potential (Ecorr),
cathodic (bc) and anodic (ba) Tafel slopes and inhibition efficiency (IE%) has been
listed in Table 2. In the presence of the inhibitor, the icorr of mild steel were smaller
than the icorr without the inhibitor. This means that there is an adsorption of the
inhibitor on mild steel surface [11]. It is also found that compound A1 has the
highest inhibition efficiency which is 86.2 % as seen in Table 2. The results from
both weight loss and LPR are in good agreement. Thus, this shows that compound
A1 is a potential corrosion inhibitor. The efficiency of the compounds to act as
inhibitors was affected by the presence of methoxy substituent. The different
positions of the methoxy give different reactivity of the compounds towards the
[email protected]
Table 1 Parameters of mild steel corrosion activities in 1.0 M H2SO4 solution with and without
inhibitors
Compound Concentration Average weight Inhibition Corrosion rate (g/
(M) loss (g) efficiency (IE %) cm2 day)
Blank – 0.0279 – 9.96 × 10−4
−5
A1 1 × 10 0.0227 18.6 8.11 × 10−4
1 × 10−4 0.0191 31.5 6.82 × 10−4
1 × 10−3 0.0049 82.4 2.92 × 10−5
A2 1 × 10−5 0.0201 27.9 7.18 × 10−4
1 × 10−4 0.0167 40.1 5.96 × 10−4
1 × 10−3 0.0086 69.2 3.07 × 10−4
A3 1 × 10−5 0.0238 14.7 8.05 × 10−4
1 × 10−4 0.0222 20.4 7.93 × 10−4
1 × 10−3 0.0069 75.3 2.46 × 10−4
A4 1 × 10−5 0.0247 11.5 8.82 × 10−4
1 × 10−4 0.0202 27.6 7.21 × 10−4
1 × 10−3 0.0083 70.3 2.96 × 10−4
Fig. 2 Graph of inhibitory

efficiency percent (IE%)
versus concentration of
inhibitors by using weight
loss method
Table 2 Corrosion parameters for mild steel in 1.0 M sulphuric acid with and without inhibitor
Compound Ecorr (mV) icorr (mA/cm2) ba (mV/dec) bc (mV/dec) IE%
Blank −425.75 2.61 132.20 38.92 –
A1 −346.62 0.36 30.80 8.93 86.2
A2 −278.92 1.03 222.51 43.77 60.5
A3 −373.60 0.54 58.53 27.81 79.3
A4 −375.94 0.56 63.97 29.34 78.5
[email protected]
Fig. 3 Polarization curves of 1 × 10−3 M A1, A2, A3 and A4 in 1.0 M H2SO4
inhibition activity. As methoxy is electron donating group, it is more favours to o-

and p- positions because it able to donate electrons to the benzene group and
achieve the compounds’ stability to form a protective layer on the metal surface.
The order of the effectiveness of the compounds as corrosion inhibitors are
A1 > A3 > A4 > A2 (Fig. 3).
Conclusion
Four 2-methylbenzoyl thiourea derivatives namely N-(2-methoxyphenyl)-N’-(2-

methylbenzoyl)thiourea (A1), N-(3-methoxyphenyl)-N’-(2-methylbenzoyl)thiourea
(A2), N-(4-methoxyphenyl)-N’-(2-methylbenzoyl)thiourea (A3) and N-(2-meth-
ylbenzoyl) -N’-phenyl thiourea (A4) were successfully synthesized, characterized
and have the potential to inhibit corrosion. Compound A1 was found to be the
potential corrosion inhibitor with the highest inhibition efficiency which is 82.4–
86.2 %.
Acknowledgment The authors are grateful to the Ministry of Higher Education of Malaysia for
the research grant no. 600-RMI/RAGS 5/3 (3/2012)and the Faculty of Applied Sciences, Institute
of Science and Universiti Teknologi MARA for providing research facilities.
[email protected]
References
1. Markhali BP, Naderi R, Mahdarian M, Sayebani M, Arman SY (2013) Electrochemical

impendance spectroscopy and electrochemical noise measurements as tools to evaluate
corrosion inhibition of azole compounds on stainless steel in acidic media. Corros Sci 75:269–
279
2. L. Emilio, Bethni (ed) (2007) Progress in research. Nova Science Publisher Inc., New York,
p 160
3. Kumar CBP, Mohana KN (2013) Corrosion inhibition efficiency and adsorption characteristics
of some Schiff bases on mild steel/hydrochloric acid interface. J Taiwan Inst Chem Eng.
doi:10.1016/j.jtice.2013.08.017
4. Pavitra MK, Venkatesha TV, Kumar MKP, Shylesha BS (2013) Accalyphaterta leaf extract as
green corrosion inhibitor for mild steel in hydrochloric acid. Ind Eng Chem Res 52:728–772
5. Khaled KF (2010) Experimental and molecular dynamic study on the inhibition performance
of some nitrogen containing compounds for iron corrosion. Mater Chem Phys 124:760–767
6. Hosseini SMA, Azimi A (2009) The inhibition of mild steel in acidic medium by 1-methyl-3-
pyridin-2-yl-thiourea. Corros Sci 51:728–732
7. Chauhan JS, Gupta DK (2009) Corrosion inhibition of titanium in acidic media containing
fluoride with bixin. E J Chem 6:975–978
8. Halim NIM, Kassim K, Fadzil AH, Yamin BM (2012). Synthesis, characterisation and
antibacterial studies of Cu(II) complexes thiourea. Malays J Anal Sci 16(1):56–61
9. Yusof MSM, Jusoh RH, Khairul WM, Yamin BM (2010). Synthesis and characterization a
series of N-(3,4-dichlorophenyl)-N’-(2,3 and 4-methylbenzoyl) thiourea derivatives. J Mol
Struct 975:280–284
10. Lampman GM, Pavia DL, Kriz GS, Vyvyan JR (2010) Spectroscopy international edition, 4th
edn. Cengage Learning, Canada
11. Fouda AEE, Hussein A (2012) Role of some phenylthiourea derivatives as corrosion inhibitors
for carbon steel in HCl solution. J Korean Chem Soc 56(2):264–273
[email protected]
Performance Studies of Anode Microbial
Fuel Cells Using Geobacter Sulfurreducens
as a Biocatalyst
Noor Fazliani Shoparwe, Muhammad Hekarl Uzir

and Suhairi Abdul Sata
Abstract Microbial fuel cells (MFC) is a bio-electrochemical device that converts

organic substrate into electricity by electrogenic bacteria. In the MFC, anode
compartment plays an important role in achieving high power density and thereby
improving the cell performance. In this study, Geobacter sulfurreducens was used
as a biocatalyst to catalyze the generation of electricity by using acetate as an
oxidizing agent and sodium fumarate as a reducing agent. The polarization curve
was determined after MFC system was reached at open circuit voltage (OCV)
around 0.80 V and gives maximum power production of about 0.22 mW. The cyclic
voltammetry (CV) method was performed in order to evaluate the oxidation and
reduction processes in a batch of MFC before and after attachment of Geobacter
sulfurreducens cells and it was proved by scanning electron microscope (SEM)
images. Overall, this study provides a convenient way for evaluating the perfor-
mance of MFC.
Keywords Geobacter sulfurreducens Microbial fuel cells Batch Single

chamber Cyclic voltammetry
N.F. Shoparwe M.H. Uzir S. Abdul Sata (&)

School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia,
14300 Nibong Tebal, Seberang Perai Selatan, Pulau Pinang, Malaysia
N.F. Shoparwe
M.H. Uzir

[email protected]
348 N.F. Shoparwe et al.
Introduction
Microbial Fuel Cell (MFC) is a bioreactor that converts chemical energy from
chemical bonds of organic compounds into electricity based on catalytic reactions
of microorganism under anaerobic conditions. Geobacter sulfurreducens are one of
the most effective microorganisms used acetate as a substrate and electrodes as the
sole electron acceptor in MFC to generate electricity. The half and overall reactions
of an MFC using acetate as the substrate can be presented as shown below [1]:
Anodic Oxidation reaction:
CH3 COO þ Hþ þ 2H2 O ! 2CO2 þ 8Hþ þ 8e ð1Þ
Cathodic Reduction reaction:
2O2 þ 8e þ 8Hþ ! 4H2 O ð2Þ
Overall reaction:
CH3 COO þ Hþ þ 2O2 ! 2CO2 þ 2H2 O ð3Þ
Research relating to MFC is rapidly increasing and involves researchers from

several different fields, due to its bioelectrochemical nature. Many researches have
been conducted are looking at ways in which to evaluate and increased MFC
performance. An improved understanding of the catalytic activity of the bioanode
as a function of solution conditions would contribute to enhanced MFC operation
and performance. However, the electrochemistry of the bioanode is complex due to
changing microbial communities and interactions within these communities in the
anode biofilm. One of the techniques that have been focused by researchers now-
adays is using electrochemical techniques. Cyclic voltammetry (CV) is gains in
popularity as a rapid method of evaluating the performance of electron transfer
mechanisms that are used in the fuel cell. CV is an electroanalytical technique in
which a varying voltage potential is applied to an electrode in an electrochemical
system while the resulting current is measured [2]. CV has been applied to MFC in
an increasing number of studies and has been widely used for characterization of
electrochemically active compound in MFC [1, 3–8]. The strategy to applied CV
for the study of anodic bioelectroactalytic electron transfer in MFC has been
described well by several authors [7, 9–11]. In this study, overall performance
studies of anode batch microbial fuel cells were evaluated using Geobacter sul-
furreducens as a biocatalyst and graphite felt as an anode electrode in a single-
chamber MFC. The performance of MFC was determined based on the interaction
between the cell biofilm of Geobacter sulfurreducens and anode electrode surface
via polarization curve and CV method.
[email protected]
Performance Studies of Anode Microbial Fuel Cells … 349
Experimental
Geobacter sulfurreducens strain PCA (ATCC 51573), used in this study was
obtained from the American Type Culture Collection, ATCC (Rockville, MD). The
strain was subcultured in the laboratory at and incubated under anaerobic condition
at 32 °C. The growth medium was prepared according to the protocol supplied by
the ATCC consisted the following (per liter): 1.5 g of NH4Cl, 0.6 g of NaH2PO4,
0.1 g of KCl, 2.5 g of NaHCO3, 10.0 ml Wolfe’s vitamin mix solution and 10.0 ml
Modified Wolfe’s mix mineral solution, respectively. The medium pH was initially
adjusted to 6.8 and was autoclaved at 121 °C for 15 min. The anode compartment
of MFCs was flushed with N2/CO2 (80–20 %) stream before inoculation at least 1 h
and a mild flow was then maintained during the experiment. The electrochemical
reactor was inoculated (10 % v/v) with cell suspensions that had an optical density
at 620 nm around 0.2–0.3 during the late of exponential phase. The single-chamber
MFC was designed and fabricated using glass (Pyrex). Schematic diagram of the
design used in this research are depicted in Fig. 1. Single chamber design consist of
graphite felt as anodic electrode, where graphite as a cathode electrode was
deposited to one side to ensure the directly exposed to air. The working volume of
each chamber was about 500 ml. A proton exchange membrane (PEM) (Nafion
117, DuPont Co., USA) was used to separate the anode and cathode compartment
and allowing transfer of H+ ions through the membrane was held in place by rubber
gaskets.
In all experiments, the MFC was connected to the multi-potentiostat (VMP3
Biologic SA). Three-electrode systems (working electrodes (WE) and counter
electrodes (CE) and reference electrode (RE)) were implemented in this study. For
the polarization experiment, the value of current and voltage were measured at
variable external resistance, Rext between (50–150 Ω). These values were measured
Computer
Filte red External
sterilized resistance
(for N2) Rext
MULTI-POTENTIOSTAT
Reference electrode (SCE)
Cathode
Anode
Membrane
Fig. 1 Schematic diagram of the glass single-chamber MFC connected to multi-potentiostat and
computer
[email protected]
after open circuit voltage (OCV) reached a steady state optimum voltage. CV
experiment was performed under OCV conditions and the voltage was changed
from −1.0 to 1.0 V against standard calomel electrode (SCE) in forward and reverse
scans 10 mVs−1. Samples were withdrawn at regular time intervals and analyzed for
cell and acetate concentrations. Cell growth was monitored using spectrophotom-
eter (Thermo Spectonic Genesys 20) at an optical density of 660 nm (OD660). A
correlation was established relating OD660 to dry weight per liter of culture broth
(gl−1). One unit of OD660 corresponds to 0.532 ± 0.0035 gl−1 of dry cell weight
(DCW). Acetate concentration was performed using High Performance Liquid
Chromatography (HPLC). HPLC analyses were performed with a Chromatography
(HPLC) instrumentation system equipped with Supelcolis CL-SAXI column
(250 × 4.8 mm). The mobile phase was acetonitrile-water (3:97 % (v/v)) at pH 3.0
adjusted with H3PO4. The surface morphology of anode electrode before and after
attachment of Geobacter sulfurreducens cells were observed using scanning elec-
tron microscope (SEM) (Oxford INCA/ENERGY.350). After the end experiment,
the anode electrode was collected and fixed overnight with paraformaldehyde and
glutaraldehyde in a buffer solution at pH 5.4 and 4 °C. The sample was coated with
gold and observed with a SEM.
In this study, the electricity was produced by the Geobacter sulfurreducens growth
in the anaerobic anode compartment by using acetate as a substrate. Initially, in
order to determine the optimum voltage inside the single-chamber of MFC system,
the open circuit voltage (OCV) (without any load) was determined (Fig. 2). It was
found that the maximum value of OCV was achieved around 0.80 V after 4 days of
fermentation time. This voltage was slightly lower than the theoretical value
obtained by Logan et al. [1]. However, this value was significant might due to
energy utilization by bacteria, electrode overpotential, and internal resistance [12].
Fig. 2 Open circuit voltage 1.0

(OCV)
0.8
Voltage, E (V)
0.6
0.4
0.2 Open circuit voltage (OCV)
0.0
0 20 40 60 80 100 120 140
Time, (hr)
[email protected]
Fig. 3 Time-course profile of growth Geobacter sulfurreducens and acetate utilization in a batch
anode MFC
A time-course profiles of anaerobic cultivation of Geobacter sulfurreducens using

20 mM of initial acetate concentration in anode compartment of MFCs is shown in
Fig. 3. A classical sigmoidal growth trend for the Geobacter sulfurreducens cells, in
which, an initial lag phase (*24 h), which varied in duration were observed. This
was followed by the exponential growth phase, where cell growth increased
exponentially and a stationary phase accordingly. Polarization curve as a function
of current, potential and power production was obtained and shown in Fig. 4.
The polarization curve is the most common method of representing the perfor-
mance of fuel cells. The polarization curve can be divided into three main region;
activation, ohmic and mass transport losses [13]. The power performance curve was
calculated from the polarization curve, which represents the relationships between
power generation and given current. The polarization curve as a function of current,
potential and power measured at variable external resistance Rext between
Fig. 4 Polarization curve of batch anode MFC
[email protected]
Fig. 5 Cyclic votammogram

at scanning rate 10 mVs−1
(before inoculation and after
6 days of anaerobic Oxidation
fermentation)
Reduction
(50–150 KΩ). The dominance of activation loss was observed from the initial step
decrease in the voltage from OCV value at 0.80 to 0.39 V. The subsequent slope of
the voltage decreased from 0.18 to 0.9 mA and was almost linear, which might
indicate the dominance of ohmic losses. The relatively lower drop in voltage from
0.18 to 0.05 V and maximum current obtained at lower resistance reveals a lower
potential drop and lower mass transfer losses at the electrode. The effective electron
discharge observed at lower resistances is due to the further potential drop and slow
stabilization of the voltage at lower resistance [14]. The curve depicts a maximum
power of 0.22 mW at the resistance 350 Ω. After reaching the optimum power
production, power generation showed a decreasing trend with an increases in Rext.
This is consistent with the literature [15–19], an indicated typical MFC behavior.
CV is extensively used to examine the performance of MFC throughout the
oxidation and reduction properties. It also have been used to evaluate the bio-
electrocatalytic activity of cell attachment on the anode surface of electrode toward
substrate oxidation. In this study, CV experiment was carried out after the current
reached its maximal value in a batch MFC process (Fig. 5). As can be shown CV in
Fig. 5, none of the redox peaks were obtained in the sterile culture medium before
the inoculation. However, after six days of anaerobic fermentation, the cyclic
voltammogram obtained showed very sharp oxidation and reduction peaks at
potential of 0.61 and 0.32 V (vs SCE) which indicated the electrochemical activity
of the attached cell on the anode electrode.
The results before and after attachment of cell Geobacter sulfurreducens were
proved by the SEM (Fig. 6a, b). The SEM images indicated surface and mor-
phology information about the used electrode in the anode compartment before and
after attachment of cell on the anode surface. The obtained images demonstrated
microorganisms grown on the graphite felt surface. Based on Logan and co-worker
studies [1, 20, 21], when microorganisms interact with conductive surface (elec-
trode), they formed electroactive microbial biofilms and transfer the electron to
generate electricity.
[email protected]
Fig. 6 SEM results a before attachment and b after attachment at 15,000 x magnification
Conclusion
In this study performance of anode MFC was evaluated in a single-chamber design

using Geobacter sulfurreducens as a biocatalyst. The polarization curve analysis
was performed in order to characterize the voltage losses involved in the system.
The optimum power production was reached about 0.22 mW at an external resis-
tance of 350 Ω. The oxidation and reduction processes occurring at the anodes were
confirmed using the CV method and SEM analysis. This research provided a
potential approach to evaluate the overall performance of MFC.
Acknowledgment This research was supported by funding through the Fundamental Research
Grant Scheme (FRGS) (Acc Number: 6071232) by Ministry of Higher Education, Malaysia and
Postgraduate Research Grant Scheme (PRGS) (Acc Number: 8046005) by Universiti Sains
Malaysia (USM).
References
1. Logan BE, Hamelers B, Rozendal R, Schröder U, Keller J, Freguia S, Aelterman P, Verstraete

W, Rabaey K (2006) Microbial fuel cells: methodology and technology. Environ Sci Technol
40:5181–5192
2. Nicholson RS (1965) Theory and application of cyclic voltammetry for measurement of
electrode reaction kinetics. Anal Chem 37:1351–1355
3. Zhou M, Chi M, Luo J, He H, Jin T (2011) An overview of electrode materials in microbial
fuel cells. J Power Sour 196:4427–4435
4. Liu X-W, Sun X-F, Huang Y-X, Sheng G-P, Zhou K, Zeng RJ, Dong F, Wang S-G, Xu A-W,
Tong Z-H (2010) Nano-structured manganese oxide as a cathodic catalyst for enhanced oxygen
reduction in a microbial fuel cell fed with a synthetic wastewater. Water Res 44:5298–5305
5. Kim JR, Jung SH, Regan JM, Logan BE (2007) Electricity generation and microbial
community analysis of alcohol powered microbial fuel cells. Bioresour Technol 98:2568–2577
6. Rabaey K, Ossieur W, Verhaege M, Verstraete W (2005) Continuous microbial fuel cells
convert carbohydrates to electricity. Water Sci Technol 52:515–523
7. Marsili E, Rollefson JB, Baron DB, Hozalski RM, Bond DR (2008) Microbial biofilm
voltammetry: direct electrochemical characterization of catalytic electrode-attached biofilms.
Appl Environ Microbiol 74:7329–7337
[email protected]
8. Pham CA, Jung SJ, Phung NT, Lee J, Chang IS, Kim BH, Yi H, Chun J (2003) A novel
electrochemically active and Fe (III)-reducing bacterium phylogenetically related to
Aeromonas hydrophila, isolated from a microbial fuel cell. FEMS Microbiol Lett 223:129–134
9. Fricke K, Harnisch F, Schröder U (2008) On the use of cyclic voltammetry for the study of
anodic electron transfer in microbial fuel cells. Energy Environ Sci 1:144–147
10. Marsili E, Sun J, Bond DR (2010) Voltammetry and growth physiology of Geobacter
sulfurreducens biofilms as a function of growth stage and imposed electrode potential.
Electroanalysis 22:865–874
11. Harnisch F, Freguia S (2012) A basic tutorial on cyclic voltammetry for the investigation of
electroactive microbial biofilms. Chem Asian J 7:466–475
12. Oh S, Kim J, Joo J, Logan B (2009) Effects of applied voltages and dissolved oxygen on
sustained power generation by microbial fuel cells. Water Sci Technol 60:1311
13. Logan BE (2008) Microbial fuel cells. Wiley, New York
14. Nimje VR, Chen C-C, Chen H-R, Chen C-Y, Tseng M-J, Cheng K-C, Shih R-C, Chang
Y-F (2012) A single-chamber microbial fuel cell without an air cathode. Int J Mol Sci
13:3933–3948
15. Bond DR, Lovley DR (2003) Electricity production by Geobacter sulfurreducens attached to
electrodes. Appl Environ Microbiol 69:1548–1555
16. Cho HK, Trinh NT, An JY, Kim BW (2008) Optimizing the power density generated from a
microbial fuel cell using Geobacter sulfurreducens with acetate. J Biotechnol 136
(Supplement):S485
17. Nevin KP, Richter H, Covalla S, Johnson J, Woodard T, Orloff A, Jia H, Zhang M, Lovley D
(2008) Power output and columbic efficiencies from biofilms of Geobacter sulfurreducens
comparable to mixed community microbial fuel cells. Environ Microbiol 10:2505–2514
18. Kim M-S, Lee Y-J (2010) Optimization of culture conditions and electricity generation using
Geobacter sulfurreducens in a dual-chambered microbial fuel-cell. Int J Hydrogen Energy
35:13028–13034
19. Trinh NT, Park JH, Kim B-W (2009) Increased generation of electricity in a microbial fuel cell
using Geobacter sulfurreducens. Korean J Chem Eng 26:748–753
20. Logan BE (2009) Exoelectrogenic bacteria that power microbial fuel cells. Nat Rev Microbiol
7:375–381
21. Logan BE, Regan JM (2006) Electricity-producing bacterial communities in microbial fuel
cells. Trends Microbiol 14:512–518
[email protected]
Effects of Chain Length
on the Thermotropic and Lyotropic
Phase Behaviours of Maltosides
by Small-angle X-ray Diffraction Study
Hairul A.A. Hamid, Rauzah Hashim, John M. Seddon

and Nicholas J. Brooks
Abstract Industrially useful surfactants such as alkyl polyglucosides (APGs) and

synthetic branched-chain glycosides are interesting biosurfactants and offer an
interesting insight into life-science applications. Hence, research presented here a
study of the thermotropic and lyotropic phase of 2-hexyldecyl β-D-maltoside
(β-Mal-C10C6) and the corresponding 2-decyltetradecyl β-D-maltoside (β-Mal-
C14C10) over a range of temperatures (0–80 °C). Small-angle X-ray diffraction
(SAXS) and optical polarizing microscopy (OPM) have been used to characterize
the mesophase behaviour. This study has shown that in dry conditions β-Mal-C10C6
and β-Mal-C14C10 adopt a lamellar crystalline phase (Lc), on the other hand, in
excess water these compounds form a fluid lamellar phase (Lα) upon heating and
cooling. The study also reported the effects of temperature (0–80 °C) on the d-
spacing value using SAXS, and we estimated the surface area per molecule and
water content of the fluid lamellar phase (Lα) in the fully hydrated state for both
β-Mal-C10C6 and β-Mal-C14C10. The former gave the surface area per molecule and
water content of 57 Å2 and 25 wt % respectively, however, the latter forms 69 Å2
and 20 wt %.
Keywords Thermotropic phase Lyotropic phase Lamellar phase
H.A.A. Hamid (&)

School of Chemistry and Environment, Faculty of Applied Sciences,
Universiti Teknologi MARA, 40450 Shah Alam, Malaysia
H.A.A. Hamid R. Hashim
Department of Chemistry, Faculty of Science, University of Malaya,
J.M. Seddon N.J. Brooks
Department of Chemistry, Imperial College London, London SW7 2AZ, UK
N.J. Brooks

[email protected]
356 H.A.A. Hamid et al.
Introduction
New generation sugar-based surfactants (glycolipids) such as alkyl glycosides are

made from renewable resources. Biodegradability, highly value-added and material
versatilities are, hence, obtained [1, 2]. These surfactants, therefore, exhibit a varied
potential application and are gaining importance progressively especially in cellular
membrane’s fields [3, 4]. Membranes mainly composed of a variety of lipids and
most lipid membranes are built from phospholipids [5]. An interesting membrane
component which is glycolipids has recently received much attentions. Its pervasive
presence in nature has made tantalizing hints that the glycolipids play a pivotal role
the same as phospholipids. Both component, are amphitropic liquid crystals which
contain hydrophilic and hydrophobic parts. As such, these characteristics have
enabled them to self-organize in a wide variety of phases showing lipid polymor-
phisms in both thermotropic (dry state) and lyotropic (presence of solvent) condi-
tions [6–9].
In addition to provide more understanding of membrane functions study, the
present work reports the amphiphilic liquid crystals Guerbet branched-chain
maltosides with two different types of hydrophobic chains known to produce
phases in the absence (thermotropic) and presence of excess water (lyotropic) as
shown in Fig. 1. Combination of two techniques which were small-angle X-ray
diffraction and optical polarizing microscopy had been used during investigations.
Moreover, the study also presented the d-spacing value at different temperatures
(0–80 °C) for both compounds in dry and excess water states. The area per
molecule and water content were then estimated for both compounds.
Synthesis of β-D-Maltosides
Chemically synthesized β-maltosides were prepared [10] and a higher purity (98 %)
of these compounds was estimated by 1H-NMR and TLC techniques. The HPLC-
grade water was used in all excess water sample preparations.
Fig. 1 Chemical structures of the two surfactants: a β-Mal-C10C6, b β-Mal-C14C10
[email protected]
Effects of Chain Length on the Thermotropic … 357
X-Ray Diffraction
Philips generator was used for X-ray measurements. It was operated at 40 kV and
30 mA that capable to generate X-rays. Line X-ray beam was focused due to the
isolation of Cu radiation by quartz crystals monochromator. Copper block, which
consisted of X-ray capillary for holding the prepared-sample [11], was heated and
cooled at temperatures in the range of 0–80 °C. The heating or cooling rate was at
10 °C per hour. The diffraction pattern was recorded on a Kodak Biomass MS
autoradiography film. The film was inserted into a film holder which was able to
move at the rate of 0.06 mm per minute. It was developed using standard procedure
and analysed using Axcess software [12].
Calculation of Fully Hydrated Lα Phase
The area per molecule [13] and volume fraction of water of Lα phase from the
measured d-spacings in dry and excess water conditions, and density of the
maltosides [14] can be estimated by using the following formula: Calculation of
estimated volume fraction of water is:
tW dW Dd
¼ ¼ ð1Þ
ttotal d0 dexcess
where tW and ttotal are the partial specific volumes of water and total (lipid + water),
dw and Δd are the water layer thickness, and d0 and dexcess are d-spacings in dry and
excess water conditions. Volume percent of water is approximately the same as
weight percent of water if the density of the maltosides is equal to the density of water.
Optical Polarizing Microscopy
Thermotropic and lyotropic liquid crystal phases were determined by optical

polarizing microscopy (OPM). Nikon Eclipse E600 polarizing microscope com-
bined with a Nikon DXM1200F digital camera was used to take sample images. As
Nikon Act-1 (version 2.70) recorded the actual images, Linkam heating stage
controlled by Linkam Controller software was utilised to heat and cool the sample.
Dry Sample Preparation
A dry sample was carefully transferred to a glass microscope slide and covered
with a glass cover slip. The sample was placed on the hot stage and further
investigated by controlling the temperature.
[email protected]
Hydrated Sample Preparation
Sample was placed on the glass microscope slide and a drop of water was added.
The mixture was then covered with a glass cover slip and sealed with a silicone
sealant. Heating and cooling were employed on the sample under investigation.
Small-Angle X-Ray Diffraction
The X-ray of β-maltosides consist of chain length C16 (β-Mal-C10C6) and C24
(β-Mal-C14C10) were identified. As can be seen from Fig. 2a, c, only the first-order
peak of the lamellar phase is visible for both compounds in the dry state upon heating
and cooling at temperatures ranging from 0 to 80 °C. This phase is a lamellar liquid
crystalline phase (Lc) with d-spacings of 30 and 36 Å, respectively. The occurrence
of the same phase over the temperature range is confirmed by an optical polarizing
microscopy results (Fig. 3a, c). It is well known that chain branching stabilizes non-
lamellar curved phases [15]; in contrast, we observe that the longer chain lipids
(β-Mal-C10C6 and β-Mal-C14C10) form only lamellar Lc phase.
Figure 2b, d shows the X-ray diffraction results of β-Mal-C10C6 and β-Mal-
C14C10 compounds in excess water with the d-spacings of 40 and 45 Å,
(a) Dry sample (b) Excess water sample

Intensity (Arbitrary Units)
1
1 1 1 2
2
1/d-spacing (Å) 1/d-spacing (Å)

(c) Dry sample (d) Excess water sample
1 1
1 1 2
2
1/d-spacing (Å) 1/d-spacing (Å)
Fig. 2 Small-angle X-ray diffraction patterns (right-hand side) and the corresponding intensity
versus 1/d spacing (left-hand side) of β-Mal-C10C6 (a and b) and β-Mal-C14C10 (c and d). X-ray
image integration is performed horizontally via AXcess software program [12].
[email protected]
(a) Dry sample (b) Excess water sample
Lc Lα
L c phase (heating, T = 50 °C, Mag = 10x) L α phase (T = 50 °C, Mag = 10x)
(c) Dry sample (d) Excess water sample
Lc Lα
L c phase (heating, T = 60 °C, Mag = 20x) L α phase (T = 30 °C, Mag = 10x)
Fig. 3 Optical textures of β-Mal-C10C6 (a and b) and β-Mal-C14C10 (c and d)
respectively. The equally spaced peaks (ratio 1 and 2) fluid lamellar phases (Lα) are
detected in both compounds. The same type of phase remains in existence
throughout the variation of temperatures which is confirmed by the optical polar-
izing microscopy as indicated in Fig. 3b, d.
Optical Polarizing Microscopy
Birefringent optical textures were investigated for both compounds with the tem-
peratures ranging from 0 to 80 °C. Fig. 3 illustrates sample of the dry and excess
water conditions. On heating and cooling, both β-Mal-C10C6 and β-Mal-C14C10
exhibit a lamellar crystalline phase (Lc) in the dry conditions and a fluid lamellar
phase (Lα) in the samples containing excess water.
Hydration of the Fluid Lamellar Lα Phase
In the calculation of the estimated area per molecule and weight percent of water of
the fully hydrated maltoside bilayers, the X-ray d-spacings of the dry and excess
water conditions are plotted as functions of temperatures upon heating and cooling
between 0 and 80 °C (data not shown) in order to determine the water layer
thickness (Δd) for each maltoside.
[email protected]
Table 1 The structural parameters of the fluid lamellar Lα phase of β-maltosides at full hydrations
varying at temperatures 0–80 °C (heating and cooling)
Lipid Phase Chain length Water layer Area per Weight %
thickness (Δd) [Å] molecule [Å2] of water
β-Mal-C10C6 Lα C16 9 57 25
β-Mal-C14C10 Lα C24 10 69 20
In excess water, for both maltosides, upon heating and cooling between 0 and
80 °C the only stable state formed is a fluid lamellar Lα phase as demonstrated in
Table 1. This can be explained by the increased chain length region surface area
which lowers the sugar domain packing stress [14]. Eventhough, the type of phase
occurred is the same for both β-Mal-C10C6 and β-Mal-C14C10, the concominent
increases in area per molecule are observed from 57 (former) to 69 Å2 (latter).
These are considerably lower than the obtainable value 72.2 Å of fully hydrated 1,2-
dioleoyl-sn-glycero-3-phosphatidylcholine (DOPC). Indeed, the larger molecular
weight of DOPC (786) indicates the greater molecular volume (1303 Å3) [ 16 ] and
thus, requires greater an area per molecule. This result is consistent with those
β-Mal-C10C6 (M: 567 g/mol) and β-Mal-C14C10 (M: 679 g/mol) which exhibit an
increment in the area of the molecule when the chain length of the hydrocarbons
and the molecular volumes are increased [14].
One of the important factors that affects the water spacing is the fluctuation
pressures of the interbilayer interactions exist in the fully hydrated Lα phase [16].
Moreover, McIntosh and Simon [17] found that the flexible Lα lipid with bigger
fluctuation pressure resulted in a larger water spacing in comparison with rigid gel
or subgel lipid. As shown in Table 1, the water layer thickness Lα phase of β-Mal-
C10C6 and β-Mal-C14C10 are 9 and 10 Å, respectively. This small increment in
Δd arises probably due to the higher interbilayer fluctuation pressures as mention
earlier and a change in free energy leads to the layer to bend [18].
Twenty five weight percent of water of β-Mal-C10C6 and 20 wt % of water of
β-Mal-C14C10 indicate the solvation of both maltosides. It is interesting to note
that, the latter shows a reduction in weight percent of water as the chain length
increases. This suggests that the interaction factors needed to be considered are the
freely water entering the fluid space between bilayers, the intercalated water into the
head groups bilayers [19]; and the interbilayer interactions namely van der Walls
interaction [20], fluctuation pressure [16], hydration force [21] and the steric
thickness [22].
Conclusions
Knowledge of the fundamental parameters, for instance, the surface area per molecule
and amount of water per lipid molecule [ 11 ] are essential to biomembrane of lipids.
Hence, in this work, the thermotropic and lyotropic phase behaviours of β-Mal-C10C6
[email protected]
and β-Mal-C 14C10 upon heating and cooling (0–80 °C) have been determined . In dry
conditions, at all temperatures, both maltosides exhibit Lc phase with d-spacing values
30 and 36 Å, respectively. On the contrary, in excess water the former (d-spacing
40 Å) and the latter (d-spacing 45 Å) produce a fluid lamellar phase. Values of area per
molecule and water content which are 57 Å, 25 wt % (β-Mal-C10C6) and 69 Å, 20 wt %
(β-Mal-C14C10) have been estimated in the fully hydrated Lα phase.
Acknowledgement This work was supported by UiTM-scholarship, UM.C/HIR/MOHE/SC11 and

EPSRC (UK)-Platform grant EP/G00465X. We thank them for all financial and moral support given.
References
1. Andreozzi P, Gente G, Mesa CL (2009) Solution behavior of alkyl glycosides and related
compounds. In: Ruiz CC (ed) Sugar-based surfactants fundamentals and applications. CRC
Press Taylor & Francis, New York, pp 21–59
2. Vill V, Hashim R (2002) Carbohydrate liquid crystals: structure–property relationship of
thermotropic and lyotropic glycolipids. Curr Opin Colloid Interface Sci 7:395–409
3. Hill K, LeHen-Ferrenbach C (2009) Sugar-based surfactants for consumer products and
technical applications. In: Ruiz CC (ed) Sugar-based surfactants fundamentals and
applications. CRC Press Taylor & Francis, New York, pp 1–20
4. Lopez O, Cocera M, Coderch L, Barbosa-Barros L, Parra JL, Maza ADL (2009) Applications
of alkyl glucosides in the solubilization of liposomes and cell membranes. In: Ruiz CC (ed)
Sugar-based surfactants fundamentals and applications. CRC Press Taylor & Francis, New
York, pp 585–626
5. Karp G (2010) Cell biology, six edn. Wiley, New Jersey
6. Mannock DA, McElhaney RN (2004) Thermotropic and lyotropic phase properties of
glycolipid diastereomers: role of headgroup and interfacial interactions in determining phase
behaviour. Curr Opin Colloid Interface Sci 8:426–447
7. Dumoulin F, Lafont D, Boullanger P, Mackenzie G, Mehl GH, Goodby JW (2002) Self-
organizing properties of natural and related synthetic glycolipids. J Am Chem Soc 124:13737–
13748
8. Minden HMV, Branburg K, Seydel U, Koch MHJ, Garamus V, Willumeit R, Vill V (2000)
Thermotropic and lyotropic properties of long chain alkyl glycopyranosides. Part II.
Disaccharide head group. Chem Phys Lipids 106:157–179
9. Brooks NJ, Hamid HAA, Hashim R, Heidelberg T, Seddon J, Conn CE, Husseini SMH, Zahid
NI, Hussen RD (2011) Thermotropic and lyotropic liquid crystalline phases of Guerbet
branched-chain β–D-glucosides. Liq Cryst 38:1725–1734
10. Hashim R, Abdalla Hashim HH, Rodzi NM, Hussen RSD, Heidelberg T (2006) Branched
chain glycosides: Enhanced diversity for phase behavior of easily accessible synthetic
glycolipids. Thin Solid Films 509:27–35
11. Hamid HAA, Hashim R, Seddon JM, Brooks NJ (2014) Lyotropic phase behaviour and
structural parameters of monosaccharide and disaccharide Guerbet branched-chain β-D-
glycosides. Adv Mater Res 895:111–115
12. Seddon JM, Squires AM, Conn CE, Ces O, Heron A, Mulet X, Shearman GC, Templer RH
(2006) Pressure-jump X-ray studies of liquid crystal transitions in lipids. Philos Trans Roy Soc
A 364:2635–2655
[email protected]
13. Seddon JM, Templer RH, Warrender NA, Huang Z, Cevc G, Marsh D (1997)
Phosphatidylcholine-fatty acid membranes: effects of headgroup hydration on the phase
behaviour and structural parameters of the gel and inverse hexagonal (HII) phases. Biochim
Biophys Acta 1327:131–147
14. Nguan HS, Heidelberg T, Tiddy G, Hashim R (2010) Quantitative analysis of the packing of
alkyl glycosides: a comparison of linear and branched alky chains. Liq Cryst 37:1205–1213
15. Nilsson F, Soderman O, Johansson I (1998) Four different C8G1 alkylglucosides: anomeric
effects and the influence of straight vs branched hydrocarbon chains. J Colloid Interface Sci
203:131–139
16. Tristam-Nagle S, Petrache HI, Nagle JF (1998) Structure and interactions of fully hydrated
dioleoylphosphatidylcholine bilayers. Biophys J 75:917–925
17. McIntosh TJ, Simon S (1993) Contributions of hydration and steric (entropic) pressures to the
interactions between phosphatidylcholine: experiments with the subgel phase. Biochemistry
32:8374–8384
18. Seddon JM, Templer RH (1993) Cubic phases of self-assembled amphiphilic aggregates.
Philos Trans Roy Soc Lond A 344:377–401
19. McIntosh TJ, Simon S (1986) Area per molecule and distribution of water in fully hydrated
dilauroylphosphatidylethanolamine bilayers. Biochemistry 25:4948–4952
20. Costigan SC, Booth PJ, Templer RH (2000) Estimation of lipid bilayer geometry in fluid
lamellar phases. Biochim Biophys Acta 1468:41–54
21. McIntosh TJ (1996) Hydration properties of lamellar and non-lamellar phases of
phosphatidylcholine and phosphatidylethanolamine. Chem Phys Lipids 81:117–131
22. Kucerka N, Tristam-Nagle S, Nagle JF (2005) Structure of fully hydrated fluid phase lipid
bilayers with monounsaturated chains. J Membr Biol 208:193–202
[email protected]
Thermal Properties Comparison Between
Alumina Filled and Organic Nano-crystal
Filled UPR/EPS Composite
Syed Anas Bin Syed Mustafa, Rahmah Mohamed

and Husni Bin Rustam
Abstract The preparation and characterization of Unsaturated Polyester Resin

(UPR) filled recycled Expanded Polystyrene (EPS) composite systems were sys-
tematically investigated. Additives such as Alumina and Organic Nano-crystal
(ONC) were added to the composite for infrared reflection enhancement. The effect
of different weight percentage of additive on the thermo-physical properties, e.g.
thermal conductivity and physical density were determined. The fabricated com-
posite undergoes thermal analysis by using a hand-held thermal probe which yield
thermal conductivity, thermal resistivity and heat capacity data for analysis. The
results obtained from comparing the use of ONC additive against alumina addition
in the composite system revealed and concluded that ONC could impart better heat
insulation to the studied composite using a lower concentration. It can be inferred
that this is might be due to alumina being a metal oxide, which are more heat
conductive compared to ONC. Further studies on the material can be done to
determine the actual mechanism of how ONC and Alumina enhances thermal
insulation by reducing thermal conductivity and increasing heat capacity of the
composite.
Keywords EPS UPR Alumina Organic nano-crystal Thermal properties
S.A.B. Syed Mustafa (&) H.B. Rustam

Faculty of Applied Science, Universiti Teknologi MARA (UiTM), 40450 Shah Alam,
Selangor, Malaysia
H.B. Rustam
R. Mohamed
GREEN Polymer Research Group, Faculty of Applied Science (FSG), Universiti Teknologi
MARA (UiTM), 40450 Shah Alam, Selangor, Malaysia

[email protected]
364 S.A.B. Syed Mustafa et al.
Introduction
Unsaturated Polyester Resin (UPR) matrix composites remains a common material

in use under a broad technology fields such as land and naval construction appli-
cations. Wan Nur Bahirah [1] postulated that this is due to UPR had proven itself an
economical thermo-set material with excellent process ability, good cross-linking
tendency as well as mechanical properties once fully cured.
Expanded Polystyrene (EPS) is a common polymeric material used in impact
cushioning purposes. According to Vaidya et al. [2], EPS sheets has been used as
core sandwich material for composite core door shutters, an alternative aimed to
replace the use of wooden door shutters in building.
Studies had been done on the use of EPS incorporated into UPR matrices with
relevant diluents to enhance its impact performance. The need to toughen UPR is
very important for its efficient use in construction applications as indicated by
Gryshchuk et al. [3].
Fahlman [4] described a nano-crystal as any single-crystalline nano-sized
materials with at least one of its dimension be less than 100 nm. Song et al. [5] had
indicated that metal oxides such as alumina are currently the commonly utilized
nano-crystals in imparting certain spectral and thermal properties in polymeric
materials, and are even constituted as an important ingredient of certain protective
thin films. However, organic nano-crystals (ONC) had the potential to replace their
metal oxide counterparts as noted by Aloui et al. [6], and hence prompting their use
was employed in this study.
This paper was written to compare the effects of ONC and metallic alumina
integration had on the thermo-physical properties of the polymeric composite
systems. By comparing the resultant changes appearing in each addition, we may
determine which additives results in the best thermo-physical enhancement.
Experimental
Materials and Methodology
The selected thermo-set matrix material used for fabricating these composite sys-
tems is Reversol UPR system, consisting of vinyl ester oligomers with 41–44 %
styrene content, bulk density of 1.12 g/cc, and the viscosity of 450–600 cps. Methyl
Ethyl Ketone Peroxide (MEKP) and Cobalt solution used as initiator and promoter
respectively, were supplied together with the UPR resin by Revertex Sdn Bhd.
These ingredients are similar to materials used by Rashidan et al. [7]. The EPS or
styrofoam filler was obtained from waste material; alumina and the silane coupling
agent was supplied by Merck KGaA.
This paper was produced from a study conducted via qualitative analysis of the
sample manufactured. The fabricated sample were tested via thermal probe to
[email protected]
Thermal Properties Comparison Between Alumina Filled … 365
determine their thermal conductivity, thermal diffusivity and volumetric heat

capacity; whereas the density of the sample were determined by a densitometer.
Composite Fabrication and Testing
Recycled EPS, fixed at 10 % weight percent (10 wt%) of UPR resin, was blended
by utilizing a high speed agitation mixer until complete dissolution was attained.
This is to enable the gaseous and solid contaminates present are easily eliminated
from the mixture via gravity settling in 24 h.
Additives were added before the samples were prepared through utilizing the
hand lay-up processes onto an aluminium mould. The silica aerogel content was
fixed at 0.2 wt% while the ONC content was varied up to 0.2 wt%. The alumina
impregnated samples followed similar procedures where silica aerogel content was
fixed at 3 wt% while alumina content was varied up to 5 wt%.
The process was followed by adding 3 wt% of MEKP initiator and suitable
amount of cobalt as accelerant to cure the EPS/UPE composite. Gelation time was
set to be about 5 min. Once fully gelled, complete curing was done under com-
pression for 1 h at 70 °C. These steps are similar to the process used by Rashidan
et al. [7].
The thermal property of samples was then determined by using a KD2 Pro Hand-
held Thermal Probe which adheres to ASTM D5334-08 and IEEE Standard 422-03
[8, 9]. The density of the samples was obtained using a digital densitometer.
The thermal conductivity showed a minimum in both systems at 0.122 and

0.147 W m−1 K−1 respectively, which are results from UPR-EPS samples. Aerogel
introduction in the ONC system generates a significant thermal conductivity spike
compared to the alumina system. ONC addition increases the thermal conductivity,
k above the baseline value in the ranges between 0.131 and 0.227 W m−1 K−1;
alumina addition also produces similar trends which ranges between 0.227 and
0.313 W m−1 K−1 (Tables 1 and 2).
The heat capacity (Cp) values obtained from thermal probe technique showed
some increase for the silica aerogel sample; the ONC system experienced some
significant increase before stabilizing, while the alumina system reports gradual
increment before fluctuating. The thermal diffusivity reported from the probe also
showed similar trend, where ONC system was relatively stable compared to the
alumina system. The bulk density of samples from both systems was uniform
despite differences in additives used (Fig. 1).
Cutnell and Johnson [10] explains how the thermal properties in this study are
related. Thermal conductivity relates to the amount of thermal energy could pass
[email protected]
Table 1 KD2 pro handheld probe data for ONC filled

Samples Conductivity Diffusivity (mm2/s) Volumetric heat capacity Density
(W/m K) (MJ/m3 K) (g/cm3)
UP 0.122 0.092 1.112 1.230
UPA 0.2 0.254 0.122 1.810 1.210
UPAO 0.05 0.183 0.105 1.853 1.220
UPAO 0.1 0.131 0.098 1.546 1.228
UPAO 0.2 0.227 0.102 1.657 1.240
Table 2 KD2 pro handheld probe data for alumina filled

Samples Conductivity Diffusivity Volumetric heat capacity Density
(W/m K) (mm2/s) (MJ/m3 K) (kg/m3)
UE 0. 147 0.093 1.231 1.221
UEA 3.0 0. 165 0.103 1.434 1.206
UEAL 0.5 0.227 0.146 1.547 1.237
UEAL 1.0 0.300 0.286 1.050 1.234
UEAL 3.0 0.234 0.160 1.467 1.252
UEAL 5.0 0.313 0.283 1.105 1.266
0.35
K(W/m.K) D (mm^2/s)
0.30
0.25
0.20
0.15
0.10
0.05
0.00
UP UPA UPAO UPAO UPAO UE UEA UEAL UEAL UEAL UEAL
0.2 0.05 0.1 0.2 3.0 0.5 1.0 3.0 5.0
Fig. 1 Comparative thermal conductivity and resistivity for systems tested
through a material for a given length and temperature difference; a higher value
indicates how much easier and/or larger amount of heat may be transferred.
Volumetric heat capacity relates to how much a material could store heat for a given
volume and temperature change; the higher value indicates a larger heat retention
capability.
Diffusivity however is a product of thermal conductivity against heat capacity,
and it indicates the thermal inertia of a material or how fast a material flows heat
[email protected]
Thermal Properties Comparison Between Alumina Filled … 367
compared to storing it. High diffusivity causes a material to move heat across better
than storing it. The benefit of this attribute depends heavily on the material’s
application. Applications requiring quick removal of heat favors materials with high
diffusivity while heat shielding applications favors the opposite. Metals are usually
associated with high level of diffusivity.
As illustrated by the two charts, few patterns had emerged. Addition of Aerogel to
the base UPR-EPS mix (UP and UE respectively) increases all three thermal prop-
erties, quite significantly at the lower concentration of 0.2 wt% compared to 3.0 wt%.
This might indicate that either the higher concentration have better dispersal and
spread, or that the lower concentration is more volatile to the composite system.
Addition of ONC seems to lowers the conductivity to a certain level while
maintaining a stable diffusivity level. alumina addition however gradually increases
both conductivity and diffusivity in lock-step. Both system as a result had their
respective heat capacity level reacted to produce the correlation evident in Fig. 2.
This effect might be caused by the nature of the additives used; alumina is a metal
oxide which are more heat conductive compared to ONC which are organic in
nature, which then was imparted to the systems.
It could be inferred that the ONC system are suitable for use in heat shielding
applications since their stable diffusivity lessen temperature increase of layers
below the material. Hence, the capability of ONC to reduce heat transfer might also
be inferred to be better than alumina while maintaining stability of the system.
Conclusion that could be derived that ONC had shown to impart better thermal
insulation in the form of lower thermal conductivity and higher heat capacity, all at
a lower concentration compared to alumina. ONC also had an advantage of
retaining the diffusivity of the composite at low, stable level.
2.0
Cp (MJ/m^3.K) Density (g/cm^3)
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
UP UPA UPAO UPAO UPAO UE UEA UEAL UEAL UEAL UEAL
0.2 0.05 0.1 0.2 3.0 0.5 1.0 3.0 5.0
Fig. 2 Comparative heat capacity and bulk density for systems tested
[email protected]
Conclusion
It could be concluded that the addition of alumina and ONC have mixed effects on
the thermal characteristics of the composites tested. ONC had shown to be able to
impart better thermal insulation capability to the composite at a lower concentration
compared to alumina integration. More research are required to fully understand its
true mechanism, but the aim of the paper is to compare the effect of ONC and
alumina addition on the thermal performance of the compounded composite.
Acknowledgment I would like to acknowledge the technicians, and staffs for their assistance
during the composite fabrication and testing phases; and to my colleagues and lecturers for
discussions related to this work. Special acknowledgement for Assoc. Prof. Rahmah Mohamed for
her guidance and direction in producing this publication. I also wish to thank Universiti Teknologi
MARA (UiTM) for their research funding under the Research Intensive Faculty (RIF) fund
[File:600-RMI/DANA 5/3/RIF (64/2012)].
References
1. Wan Nur Bahirah W (2011). Effect of curing and flame retardant system on properties of
polyester-filled EPS composite (Master's thesis). March
2. Vaidya UK, Hosur MV, Earl D, Jeelani S (2000) Impact response of integrated hollow core
sandwich composite panels. Compos Part A: App Sci Manuf 31(8):761–772
3. Gryshchuk O, Jost N, Karger-Kocsis J (2002) Toughening of vinylester–urethane hybrid
resins by functional liquid nitrile rubbers and hyperbranched polymers. J Polymer 43:4763–
4768
4. Fahlman BD (2007) Materials chemistry, vol 1. Springer Mount Pleasant, Mt. Pleasant,
pp 282–283
5. Song JE, Kim YH, Kang YS (2006) Preparation of indium tin oxide nanoparticles and their
application to near IR-reflective film. Curr Appl Phys 6(4):791–795
6. Aloui F, Ahajji A, Irmouli Y, George B, Charrier B, Merlin A (2007) Inorganic UV absorbers
for the photo-stabilisation of wood clear-coating systems: comparison with organic UV
absorbers. Appl Surf Sci 253(8):3737–3745
7. Rashidan K, Rahmah M, Nor ZIZ, Mohd HM (2009) Effect of antioxidant and flame retardant
additives on thermal degradation and flammability of SBR filled EPS/UPR composite. In:
Malaysia polymer international conference (MPIC 2009)
8. ASTM D5334-08, Standard test method for determination of thermal conductivity of soils and
rock by thermal needle probe procedure
9. IEEE Standard 422-03, IEEE thermal conductivity/resistivity measurement standards
10. Cutnell JD, Johnson KW (2004) Cutnell & Johnson physics, 6th edn. Wiley, NEW York
[email protected]
Part VI
Safety, Policies and Regulations
[email protected]
Hierarchy of Controls Analysis
for Equipment Failures Prevention
Nor Afina Eidura Hussin, Anwar Johari, Kamarizan Kidam

and Haslenda Hashim
Abstract Equipment failures play significant roles in most accidents that occur and
recur in the chemical process industry. Many risk reduction strategies were applied
to prevent these accidents but the accidents keep on happening. In this study, 50
equipment-related accidents are analyzed based on their comprehensive accident
investigation reports, retrieved from several accident databases to determine the
preferred accident prevention strategies of the industry. Based on the analysis, 91 %
out of 590 recommended corrective actions are procedural strategies. Other layers
of hierarchy of controls are only 9 % i.e. active-engineered (3.4 %), inherently safer
(3.1 %), and passive-engineered (2.5 %). An analysis using Analysis of Variance
(ANOVA) shows a significant relationship exists between the groups of hierarchy
of controls. By emphasizing on the most reliable group of hierarchy of controls for
equipment failures prevention such as inherently safer, passive-engineered, and
active-engineered accident prevention strategies, the preference for procedural
strategies in the industry may be reduced.
Keywords Accident prevention Equipment failures Hierarchy of control

Risk reduction
N.A.E. Hussin A. Johari K. Kidam

Faculty of Chemical Engineering, Institute of Hydrogen Economy, Universiti Teknologi
Malaysia, 81310 Johor Bahru, Johor, Malaysia
e-mail: afi[email protected]
A. Johari (&) K. Kidam H. Hashim
Department of Chemical Engineering, Faculty of Chemical Engineering, Universiti
Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia
K. Kidam
H. Hashim

[email protected]
372 N.A.E. Hussin et al.
Introduction
The history of accident prevention in the chemical process industry (CPI) shows
that different approaches of risk reduction strategies have been implemented.
Previously, accident prevention emphasizes either on design, technical, or proce-
dural strategies. In the 1800s, a CPI plant with little instrumentation and means of
protection only emphasized on procedural aspects for accident prevention. One of
the procedural risk reduction examples is the one-legged stool strategy in nitro-
glycerin production. In this case, the operators had to sit on one-legged stools while
watching over the production of highly exothermic nitroglycerin in large-stirred
pots. If the heat was not removed by cooling and stirring, the reaction became
uncontrollable and might lead to an explosive decomposition of the nitroglycerin.
Hence, the operators had to watch the temperature closely. If they fell asleep, they
fell off and injured themselves or at worst, could lead to fatality [1].
In 1960s, a great change in CPI occurred with process operating conditions (e.g.
temperature, pressure etc.) became more severe; the energy stored in the process
increased; problems in areas such as material construction and process control
became more taxing; and the plants grew in size with a factor of about 10, were
often single stream. Relatively sophisticated instrumentation provision was devel-
oped to run a process under extreme conditions and close to the limits of safety thus
causing higher accident rates [2]. The focus of accident prevention shifted to
technical and design-oriented. Later, from 1980s onwards, the trend in accident
prevention was mostly utilizing the outer layers of protection by adding add-on
engineered either passive or active; and procedural control strategies. However, the
risk reduction approaches were only effective to a certain extent. Nowadays, the
focus of loss prevention is human and organizational-related which emphasis on the
safety management system and safety culture to overcome fluctuating accident rate
issues. The accident rates however remain persistently high. Thus, this paper
analyzes the application of risk reduction strategies in solving equipment failure
problems that commonly lead to the occurrence and recurrence of accidents in the
industry.
Hierarchy of Controls
Risk reduction strategies can be classified into four different layers i.e. layers of
protection (LOP). The layers are inherently safer, passive-engineered, active-
engineered, and procedural. In managing risk, the most reliable LOP is the inner
most layer which is the inherently safer, followed by passive-engineered, active-
engineered, and procedural strategies, respectively. Thus, the priority in risk
management strategy is inherently safer > passive > active > procedural. By
changing the design and operation at the earlier stages, the capital and operating
costs required are much cheaper than the latter stages. Only, procedural control
[email protected]
Hierarchy of Controls Analysis for Equipment Failures Prevention 373
strategies require low relative costs compared to other stages but the reliability of the
strategies is the lowest and the modification is difficult to mark since the complexity
increases throughout the process lifecycle [3].
Inherently Safer
Inherently safer is the premier strategy for hazard avoidance and control at its
source through design changes. By inherent safety principles, elimination is used to
avoid hazard by design; intensification, substitution or attenuation is used to reduce
the severity of the hazard; and simplification of process or plant is used to reduce
the likelihood of the hazard occurring. Using moderation principle, the existing
processing condition is changed to an inherently safer by manipulating tempera-
tures, pressures, concentrations and physical states of the chemicals. Substitution is
done by selecting safer and compatible chemicals. Use of safer chemical reduces
the severity of accidents.
A process plant can be simplified using credible equipment and establishment of
a reliable or self-regulating system. The system reduces the need for redundant
systems and complicated controls. Thus, a simpler and more robust design is the
key for reliable and safer chemical plant operations. Meanwhile, error tolerance is
effectively being used as inherently safer strategy to solve problem related to wrong
material selection, corrosion, chemical reactivity, incompatibility, and sub-standard
equipment application. Majority of the corresponding corrective actions amend the
existing technical and design deficiencies that led to process failure. The idea is to
redesign the default part of the plant (or equipment) to a robust and accident-
resistible design.
Passive-Engineered
Add-on layers are mainly installed as passive or active engineered safety protection
systems. Passive strategy employs systems that remain static and do not perform
any fundamental operations. This passive-engineered risk control further reduces
the likelihood and consequences of accident by using passive safety protection such
as dikes, containment and fire wall. The passive-engineered modifications are
mostly related to layout, mechanical/physical aspects, design specification changes,
additional equipment, equipment modification, and friendlier design.
The common errors in plant layout are the incorrect arrangement and safety
distance between main processing equipment which eventually increase the severity
of damage due to domino effect. The detailed layout changes for safety distance
normally involve redesigning and repositioning of piping system, and reshaping of
specific equipment or parts. Other safety issues related to plant layout are
[email protected]
equipment accessibility and visual obstacles are improved using proper organiza-
tion of plant layout. The mechanical and passive engineering control changes
material of construction for a better robust, and corrosion-and-high-pressure-
resistant process equipment. The design specification is applied to equipment with
changes in process condition, fire and explosion ratings. Additionally, plant and
equipment modifications are carried out in order to improve human-machine
interface leading to user-friendlier process which reduces the occurrence and
recurrence of human-related errors.
Active-Engineered
Active add-on engineered strategies use active systems that depend on timely
hazard detection and initiation (i.e. utilizes safety devices that respond to process
changes) to further reduces the accidents using relief valves, controllers, detectors
and alarms. For controlling risk, active-engineered control requires additional
devices to sense and indicates process variables, valves, etc. either by adding or
removing the instrumentation and automation of the equipment. Among the com-
mon active-engineered control strategies are modifying control setting specification,
supplementing additional instrumentation, and introducing new protection systems.
To specify control setting, majority of actions are conducted to accurately set the
safety limits for flow rate, temperature, pressure, density and speed of the auto-
mation system. The correct number of control instrumentations is important for
early detection of process deviation. A number of temperature and pressure-related
accidents have been reported due to lack of sensors or detectors e.g. chemical
reactors require adequate number of detectors with correct positioning. In addition,
the process equipment needs relief and mitigation systems to avoid accidents such
as Seveso and Bhopal. The accidents had severe consequences as a result of under-
designed protection and mitigation systems. Thus, suitable protection and mitiga-
tion systems based on the worst-case scenario with adequate design capacity are
essential to manage process hazards and reduce risks.
Procedural
Procedural or human and organizational-oriented risk control usually focuses on

safe operation including training, supervision, procedure, work instructions,
inspection and maintenance. This operator and maintenance procedures should be
the last resort, especially for control and mitigation where the chance of errors or
failure is high. Among the highest procedural corrective actions were proper doc-
umentation, improved management system, continuous monitoring/supervision,
and training. The gathered data indicates that poor organizational support of work
system and ineffective management systems have potential to cause accidents.
[email protected]
Confusion might arise from poor or unclear documentation and increases chances
of operators to perform incorrect work sequences and taking shortcuts. To deal with
these attitudes, effective safety management system and safety culture education are
essential in promoting safety awareness among the CPI players.
Research Approach
In this study, 50 comprehensive accident investigation reports from 1998 to 2012

were extracted from U.S. Chemical Safety and Hazard Investigation Board (CSB)
and U.S. National Transportation Safety Board (NTSB) databases [4, 5]. The data
mean was 2005. In the paper, the emphasis is on the recommended corrective
actions of the reports. Statistical analysis on the hierarchy of controls is conducted.
Then, the preferred risk reduction criteria are discussed and lessons learned are
established for effective accident prevention of the CPI.
Analysis of Results
The corrective action section of the accident reports was analyzed to determine the
applied risk reduction strategies of the CPI. The analysis has showed that the
industry normally takes several corrective actions for multiple causation accidents.
However, procedural strategies were mostly recommended as the corrective actions.
In this study, 590 corrective actions had been suggested by the boards. Out of these
590 corrective actions, 91 % were procedural-based. From all the recommended
corrective actions, active-engineered was 3.4 %, followed by inherently safer
(3.1 %). Passive-engineered strategies were the least options used with only 2.5 %
of the total recommended corrective actions. Further details on the applied hier-
archy of controls are summarized in Table 1.
Process hazards and their risks can be managed effectively through layer of
protection. The statistics indicated that procedural strategies were most preferred by
the industry. The layer is less reliable than inherent safety, passive-engineered, and
active-engineered controls [5]. Based on the analysis, 3.1 % of the corrective
actions were inherently safer strategies. Inherently safer approach uses material and
process conditions that are less hazardous to eliminate and mitigate hazard. Four
types of inherently safer used were minimization (44.4 %), moderation (27.8 %),
simplification (16.7 %), and substitution (11.1 %). Minimization was used to limit
energy generation capabilities by using smaller amount of hazardous substances.
Among others recommended inherently safer strategies were replacing a hazardous
substance with a less hazardous one such as the use of air or pigging with air instead
of natural gas blow for cleaning fuel gas piping (substitution) and sodium
[email protected]
Table 1 Classifications for the applied hierarchy of controls

Percentage
Inherently safer Passive-engineered Active-engineered Procedural
3.1 % out of 590 2.5 % out of 590 3.4 % out of 590 91 % out of 590
Minimization 44.4 Protective 60.0 Instrumentation 55.0 Communication 22.9
system
Moderation 27.8 Design 13.3 Mitigation 30.0 Regulations 15.3
changes system and guidance
Simplification 16.7 Layout 6.7 Protective 15.0 Training and 12.7
system education
Substitution 11.1 Sizing 6.7 Inspection 11.5
Equipment 6.7 Management 10.1
modification system
Additional 6.7 Emergency 5.4
equipment preparedness
Work 4.8
mechanism
Documentation 3.2
Enforcement 2.6
Cooperation 2.4
Expertise and 2.0
consultation
Management of 1.9
change
Maintenance 1.5
Monitoring and 1.1
supervision
Research 1.1
Contractor 0.9
performance
Cleaning and 0.6
housekeeping
hypochlorite as a biocide in cooling water treatment instead of chlorine (substitu-

tion); and the use of appropriate materials for wastewater treatment (moderation),
and revision of operating conditions to reduce risk of over-chlorination (modera-
tion) for safer process condition.
The add-on engineering controls are used to further reduce the likelihood and
consequences of accident. Passive-engineered controls do not require any device to
sense or actively respond to process changes but active-engineered controls require
devices to monitor and mitigate a hazard. In the study, 2.5 % passive-engineered
controls were used and 3.4 % were active-engineered controls. For passive-engi-
neered, protective system (60 %), design changes (13.3 %), layout (6.7 %), sizing
(6.7 %), equipment modification (6.7 %), and additional equipment (6.7 %) were
recommended. Passive-engineered controls further reduced hazard and risk by
[email protected]
using firewalls and blast-resistant construction; adding protective fireproofing for

fire rack support steel near process unit containing highly pressurized flammables;
and ensuring that penetrations of partitions, floors, walls, and ceiling were sealed
dust-tight. Meanwhile, instrumentations were mostly used as active-engineered
controls (i.e. 55 %). Other active-engineered controls were mitigation system
(30 %), and enhanced protective system (15 %). Active-engineered controls
established adequate layers of protection system by using additional interlock and
shutdown, air monitoring devices, automated audible alarms, level indicators,
automatic controls, post-ignition deflagration detection, and damage control
devices.
Procedural safety systems are administrative controls which include standard
operating procedures, safety rules and procedures, operator training, management
systems, and emergency response procedures. Human and organizational-oriented
strategies for safe operation were mostly recommended by the study. The proce-
dural strategies were 91 % of all the corrective actions suggested. Communication
was the highest strategies recommended by the boards i.e. 22.9 %, followed by
development or amendment of safety regulations and guidance (15.3 %). Training
and education (12.7 %), inspection (11.5 %), and management system (10.1 %)
were also parts of these implemented corrective actions. Other less prioritized
procedural strategies were emergency preparedness (5.4 %), work mechanism
(4.8 %), documentation (3.2 %), enforcement/implementation (2.6 %), cooperation
(2.4 %), expertise and consultation (2 %), management of change (1.9 %), main-
tenance (1.5 %), monitoring/supervision (1.1 %), research (1.1.%), contractor safety
performance (0.9 %), and cleaning and housekeeping (0.6 %).
Discussion
In the study, communication issues were commonly addressed to ensure timely

transmission of critical safety information to responding personnel. Therefore,
safety alerts and health bulletin were published to warn owners and operators on
potential hazards and risks of the industry. The safety alerts advised them on their
responsibilities for accident prevention such as recommending the use of inherently
safer design features, describing sufficient security measures, and recommending
the use of hazards signs to identify the fire and explosion hazards.
The importance of implementing hierarchy of controls includes to comprehen-
sively control all potential ignition sources and continuously monitor hazards at
appropriate locations and elevations; to train and certify emergency response per-
sonnel; to publish the technical guidance addressing the safe operating procedures;
to avoid occurrence and recurrence of accidents by analyzing the key findings,
cause, recommendations of the reports to shareholders, membership, and work-
force; and to establish a timely notification community procedures in the event of
chemical release that save life and public properties.
[email protected]
Establishment of safety regulations and guidance are important in preventing

accidents. The involvement of several governmental agencies such as the
Occupational Safety and Health Administration (OSHA), the Environmental
Protection Agency (EPA), National Fire Protection Association (NFPA), and other
regulatory bodies in the reports has contributed to these issues. Cooperation
between the agencies is needed to develop comprehensive regulations and stan-
dards, and avoid any misunderstandings. Training and education are recommended
to the industry players to increase safety awareness and skills of their personnel. By
training and education, accidents caused by human errors could be reduced. Regular
inspections are applied to ensure that the facilities and work mechanism are com-
plied with safety regulations and guidance. Management systems are updated to
cope with several issues such as changing of work mechanism, modified equipment
or facilities, new staffing, and other organizational changes.
In general, the emphasis on procedural accident prevention strategies does not
provide adequate hazard elimination and risk reduction. The human reliability is not
high enough and often leads to improper problem-solving, inappropriate actions,
and ill-timed responses. The CSB and the NTSB suggested various procedural
corrective actions with the involvement of governmental agencies, industry players,
research institutes, and other non-governmental agencies. The industry should
reconsider the implementation of the inner most layers of hierarchy of controls to
prevent accidents thus resulting in safety and cost benefits of the CPI. Moreover,
one-way ANOVA using Excel on the data shows that Fstatistic ≫ Fcritical that indi-
cates a significant relationship exists between the groups of hierarchy of controls
(137.52 ≫ 2.653). The probability is small that the difference or relationship
happened by chance. In the analysis, p-value is 0, less than the critical alpha level
(p < 0.05).
Conclusion
In the CPI, the preference for outer layer of hierarchy on controls results in
unbalanced risk management approach. Only 9 % of the 590 recommended cor-
rective actions were applying the inner layer controls compared to 91 % of the outer
layer countermeasures. As the reliabilities of the hierarchy of controls are
decreasing from inherently safer to procedural strategies i.e. from inner to outer
layer, the accident rate of the CPI is still high due to the less reliable corrective
actions been implemented. To reduce the accident rate, the CPI should focus on the
inherently safer and passive-engineered i.e. the inner layer accident prevention, thus
avoiding human unreliability and creating safer work environment.
Acknowledgment Special thanks to Zamalah Scholarship of Universiti Teknologi Malaysia

(UTM), Malaysia for funding this research.
[email protected]
References
1. Kletz TA, Amyotte P (2010) Process plant—a handbook for inherent safer design. CRC Press,
Boca Raton
2. Lees FP (1996) Loss prevention in the process industries, 2nd ed. Butterworth Heinemann,
London
3. CCPS (1998) Guidelines for design solutions for process equipment failures. AICheE, New
York
4. Information on http://www.csb.gov/investigations/completed-investigation
5. Information on http://www.ntsb.gov/investigations/reports
[email protected]
Analysis the Effect of Explosion Efficiency
in the TNT Equivalent Blast Explosion
Model
Zulkifli Abdul Rashid, Azil Bahari Alias, Ku Halim Ku Hamid,

M. Shahnor Bani and Mohanad El Harbawi
Abstract In this study, the analysis will focus firstly on the analysis of the effect of
an increment of explosion yield, η value between 2 and 9 % and its correlation
toward the parameter peak side-on overpressure po, the duration of the positive
phase td compared to diagnostic features of explosion damage results translated
according to criteria developed by V.J. Clancey. Secondly, the aim is to validate the
recommended methods of yield factor by Less [1] with estimated results (η between
2 and 9 %) by using TNT model in comparison with the physical observation
damage of VCE incident impact at Flixborough in 1974.
Keywords Investigation effect Explosion efficiency TNT equivalency
Introduction
Explosions are the most serious hazards which may occur in petrochemical
industries. Historical evidence has shown that these explosion incidents in petro-
chemical and refining industries caused a high level of damage scale. For example,
Flixborough explosion in1974, uncontrolled leakage of about 30 tons of cyclo-
hexane causing death to 28 men and serious damage to onsite infrastructure.
Another serious incident also happened at PEMEX LPG Ixhuatepec, Mexico City,
which over 500 people died and over 4000 people injured and also resulted in
Z.A. Rashid (&) A.B. Alias K.H. Ku Hamid

Faculty of Chemical Engineering, Universiti Teknologi MARA, 40450, Shah Alam, Malaysia
M.S. Bani
Petronas Carigali Sdn Bhd, Miri, Sarawak, Malaysia
M.E. Harbawi
Faculty of Chemical Engineering, King Saud Riyadh, Riyadh, Saudi Arabia

[email protected]
382 Z.A. Rashid et al.
severe damage to the immediate surroundings with window break-age up to 4 km

from the source [1–4]. Other examples are incident at BP Texas, 2005, Bukom
refinery, Singapore 2011, Chevron refinery, Richmond United States, 6 August
2012, Amuay refinery and Venezuala. Based on previous incidents analysis, even
though precaution has been taken to reduce accident, when a new petrochemical
plant was built or an existing plant was undergoing process maintenance or mod-
ification process route, these types of devastating accidents were still happened. In
this paper, the consequences impact from the magnitude of explosion will be
analyzed especially on the effect of explosion efficiency in the TNT equivalent
model.
Method of Analysis
An explosion is a process involving the production of blast wave resulting from a

rapid release of energy. The result from this blast wave, a pressure disturbance is
generated into the surrounding medium and air becomes heated due to its com-
pressibility. This phenomena leads to an increase in the velocity of sound and
causing the front disturbance becomes steepen as it travels through the air. As a
results the pressure and the density of air is increased until a maximum pressure
wave is developed at reflected incident distance. For cases of unconfined vapor
cloud explosion, the energy release in the blast wave generally is only small fraction
of the energy theoretically available from the combustion of all the material in the
cloud. The ratio of the actual energy in the explosion to that theoretically available
is termed the ‘explosion efficiency’. Typically the value of explosion efficiencies is
in the range of 1–10 %. For an explosion such as the bursting of a pressure vessel
due to high gas pressure at ground level the explosion energy absorbed by the
ground is generally smaller than for a condensed phase explosion. Thus the value of
the yield ratio lies nearer to 2. However, it is not always clear what assumption
should be made when concerning the yield ratio, and therefore it may be necessary
to exercise experience judgment. To investigate the effect of explosion efficiency,
TNT equivalent model is applied. TNT equivalency explosion model is the most
popular model for estimating far field damage because it is simple and the corre-
lation strength among the parameters in the model was improvised with a detailed
experimental process from trials done using explosives. In TNT equivalency model,
value from the fraction of total energy of the explosion used in the shock wave will
be calculated before it is converted into the equivalent mass of TNT, WTNT (kg):-
Wgas gDHcðgasÞ
WTNT ¼ ð1Þ
EðTNTÞ
Where WTNT is the equivalent mass of TNT (kg) that would produce the same
effects as the explosion, η represents the explosion yield (dimensionless); it is
[email protected]
Analysis the Effect of Explosion Efficiency in the TNT … 383
generally accepted that, taking as a basis for calculation the total quantity of vapor
in the cloud. Wgas is the total mass of flammable gas in the cloud, ΔHc(gas) is the
lower heat of combustion of the flammable material (kJ/kg), and E(TNT)is the energy
of explosion of TNT (approximately 4680 kJ/kg). However, the TNT equivalent
models have tendency to differ in few conditions, as discussed by several
researchers such as Brasie and Simpson 1968, Harris and Wickens 1989 [5, 6] at
conditions are (1) the mass of vapour participating in the explosion, (2) the yield
factor of the explosion and (3) the value used for the energy of explosion of TNT.
For instance, the value of energy of explosion of TNT quoted by Brasie and
Simpson [5], varied between 1800 BTU/lb (4190 kJ/kg) and 2000 BTU/lb (4650 kJ/
kg). However, Harris and Wickens [6] quoted the energy level of TNT explosion
limited to 4.6 MJ/kg. This is also observed in method used in TNT model equiv-
alent by Eichler and Napadensky (1977), Prugh (1987) [7, 8]. Finally, CCPS [9] in
1994, reviewed and stated that the values used for the explosion energy of TNT
should lie in the range of 4140–4680 kJ/kg. The value of equivalent TNT mass, was
subsequently used to calculate the equivalent effects of explosions occurring at the
same normalized distance, as expressed in Eq. 2:-
R
Ze ¼ ð2Þ
ðWTNT Þ1=3
where Ze is the scaled distance (m/kg1/3) and R is the real distance (m). To get a
better estimation of potential structural and building damage and human casualties
from air shock waves, there are parameters which are principally use to explain the
explosion scenario. The principal parameters of the blast wave from a TNT
explosion are the peak side-on overpressure po, the impulse of the positive phase is,
the duration of the positive phase td and the arrival time ta. Overpressure Po, (kPa)
was calculated using the equations as in Eqs. (3) and (4), where WTNT = TNT
equivalency (kg), Ps = scaled side on overpressure, Pa = atmospheric pressure and
Ze (scaled distance) = actual distance (m)/W1/3, The appropriate are included in
these two equations.
Po
Ps ¼ ð3Þ
Pa
Where Po is calculated by using an Eq. 4

2
1616 1 þ 4:5 Ze
P0 ðkPaÞ ¼ Pa : rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 ð4Þ
1þ Z e
1 þ Ze 1 þ Ze
0:048 0:32 1:35
[email protected]
Table 1 Estimation of common structures damage based on overpressure result (these values are
only approximations)* [10]
Pressure (kPa) Damage
0.21(E1) Occasional breaking of large glass windows already under strain
0.69(E2) Breakage of small windows under strain
1.03(E3) Typical pressure for glass breakage
2.07(E4) “Safe distance” (probability 0.95 of no serious damage below this value);
projectile limit; some damage to house ceilings; 10 % window glass broken
2.76(E5) Limited minor structural damage
3.4–6.9(E6) Large and small windows usually shatter; occasional damage to window
frames
4.8(E7) Minor damage to house structures
6.9(E8) Partial demolition of houses, made inhabitable, corrugated asbestos shatters;
corrugated steel or aluminium panels, fastenings fail, followed by buckling;
wood panels (standard housing), fastenings fail, panels blow in
6.9–13.8(E9) Partial collapse of walls and roofs of houses, concrete or cinder block walls,
not reinforced, shatter
13.8(E10) 50 % destruction of brickwork of houses
13.8–20.7 Frameless, self-framing steel panel buildings demolished; rupture of oil
(E11) storage tanks
17.2(E12) Cladding of light industrial buildings ruptures
20.7–27.6 Wooden utility poles snap; tall hydraulic presses (40,000 Ib) in buildings
(E13) slightly damaged
34.5–48.2 Nearly complete destruction of houses
(E14)
68.9(E15) Probable total destruction of buildings; heavy machine tools (7000 Ib), moved
and badly damaged, very heavy machine tools (12,000 Ib) survive
Positive phase duration t, (ms) is calculated using Eq. 5

h Ze 10 i
980 1 þ 0:54
td ðmsÞ ¼ W 1=3 : h Z 3 i h Z 6 i h Z 2 i1=2 ð5Þ
1 þ 0:2
e
1 þ 0:74
e
1 þ 6:9e
Values peak overpressure will be interpreted according to the criteria developed

by V.J. Clancey [10] as shown in Table 1.
For the above analysis, flammable propane was used as the study material with
the data heat of combustion for propane at 46013 kJ/kg.
[email protected]
In this analysis, the quantity of propane involved in the explosion incident was
13,000 kg and the value of explosion efficiency used was within 1–10 % [1, 9, 11].
Even though majority of explosion in chemical process plant occur at ground level
with 2 % of explosion efficiency (η) value, as recommended by Brasie and Simpson
[5], the probability to use η, of 2 % directly in TNT equivalent model for VCE
incident consequences can be argued. According to Brasie and Simpson [5] the
value of 2 % (η) or less in the range 1.52 %, is applied when the energy in the blast
wave is expended and the detonation impact may result crater formation at the
ground level of explosion source. For that reason, the recommended guidelines to
determine yield factor values used in the TNT equivalent model as proposed by
Less [1] still need to be improved based on the suitability of method selection with
the incident scenario. Meanwhile according to Phillips H. [12] the significant
investigation of an explosion impact incident for the equivalent mass of TNT is
obtained by site examination on the damage effects. Therefore the past field
investigation report for vapor cloud explosion incident cases such as Flixborough in
1974 [1, 9, 13], Port Hudson Missouri in 1970 [13], Ludwigshafen, at BASF plant
Germany in 1948 [13] and several others VCE incident were referred [1, 9, 13]. It is
observed that the severity of damage on the structural building around the explosion
is obviously varies [1, 10, 13–15]. The correlation trend impact on the structural
damage differs depending on the type of explosion as observed in the method
developed by Merrifield and Mackenzie [16], in which the researchers have
established a relationship between damage and distance, also known as destruction
profile. Mendonca et al. [17] also used the same method introduced by Merrifield
[16] to estimate the TNT equivalence of a 15-ton single base powder explosion and
predict the possible damage in buildings and structures over distance. He is also
observed that the effect of an explosion was not only caused by the blast damage,
but is also by the thermal effect, flying fragment, ground shock and from crater
formation. However, not all of the above effects occur in every explosion. Many of
the data on the effects of explosions gathered from studies of military and industrial
explosives [1]. From these studies, the pressure waves created by explosive
materials were measured at a different distance and parallel to the recorded building
location. To test field analysis, parameter correlation between the size of explosive,
distance, flame speed and the effect of obstacle is optimized into simplified value
impact represent as peak overpressure [18]. Therefore in this analysis, the effect of
explosion yield on the peak overpressure (kPa) results is obtained based on the
visual simulation of structural building damage as developed by V. J. Clancey [10].
The results analysis is better, since previously the impact for all VCE incidences
and the first indicator for all VCE incidences were derived from the physical
structure damage observation. Even though the recommended guideline by Less [1]
is available to choose an appropriate explosion yield applied to different scenario,
this guideline still have deficiency. The usage of explosion yield still becomes a
[email protected]
major discussion in TNT equivalent model calculation. For instance in CCPS [9],
Crowl and Louvar had discussed in their book [19], for reactive source such as
flammable cloud of propane, the recommended explosion yield was 0.05. However,
the recommended η value, still does not yield an accurate indicator. It is well known
that, not every flammable mass in the cloud involved in the explosion. This indicate
that the higher the η value, the potential fraction mass involved in the explosion will
be also higher. However this is also depending on how effective and reactive the
flammable material mixing with air to produced a complete combustion. The
confinement and turbulent will caused the flame velocity to speed up the unburnt
fuel and air homogenous mixing process and these will increase the cloud deto-
nation. Therefore in this study, the analysis will focus firstly on the analysis of the
effect of increment of η value between 2 and 9 % and its correlation toward the
parameter peak side-on overpressure po, the duration of the positive phase td
compared to visual damage developed by Clancey [10] as shown in Table 2.
Secondly, the aim is to validate the recommended guidelines by Less [1] with
estimated results (η between 2 and 9 %) by using TNT model in comparison with
the physical observation damage of VCE incident impact di Flixborough, [1, 13,
14]. Table 2 shows the results from the above analysis.
Based on Table 2, it is shown that the values of peak overpressure receptor at
points 50, 100, 150, 200 to 500 m (from the source of explosion) increases
exponentially when η (%) increases. Meanwhile, time duration phase for the value
of Po, (kPa) at the same points do no show obvious changes, for example at the
value of η (between 2 and 9 %) at impact distance of 50 m is between 0.048 and
0.032 s, compared to the positive duration phase measured in other model such as
Lees 2005. When the denominator value is [1 + (z/0.02)3][1 + (z + 0.74)6][1 + (z/
6.9)2]1/2, this will give the td value between 27 and 29 s with 2 s difference. Even
though, this value is still short for any emergency plan to be undertaken by the
expected victim to find a safe shelter, the possibility of escape is better compared to
the model proposed by Naval Weapon Center of United States. This probability
might be happened since more time is needed for vapor formation when the
quantity of material involved is large [η (%) increase], and also is due to the
duration to achieve the explosion. Meanwhile, at the formation of explosion and its
destructive effect, the magnitude of effect is shown to be gradually increase from E4
till E15 due to various factors such as fraction to flash off to form vapour cloud
detonation is increase with the increment of η value between 2 and 9 %, effect of
temperature due to exothermic reaction is probably occur with the increment of η
value and has a tendency for flammable cloud to initiate the burn faster due to the
decreasing of lower limit of flammability (LEL) value of the cloud. Since the case
incident occur in open air so the chances of mass to be unsaturated condition is less
therefore when [η (%) increase] the flammable range to mix with air will be also
increase, thus the possibility of energy transfer to initiate reaction in fresh air by
compression in high velocity of shock wave is increase over a very short time. The
energy released in the combustion process will be longer and maintains under the
magnitude of shock front [1, 13] if the duration of overpressure is increased.
[email protected]
Table 2 Peak overpressure η (%) R (m) Po, (kPa) Td (s)Less/USEPA Diagnose

versus scaled distance for
blast wave pressure from an 2 50 53.51 0.054/26.95 E14
explosion of 13,000 kg of 100 15.44 0.143/41.83 E9
propane 150 8.82 0.200/48.66 E8
200 6.22 0.236/51.97 E7
300 3.96 0.274/54.80 E6
400 2.92 0.293/55.90 E5
500 2.32 0.305/56.44 E4
3 50 71.27 0.048/27.7 E15
100 19.08 0.141/44.7 E10
150 10.52 0.208/53.4 E8
200 7.30 0.251/57.9 E8
300 4.59 0.300/61.9 E6
400 3.37 0.325/63.5 E5
500 2.66 0.341/64.3 E4
4 50 87.94 0.044/28.1 E15
100 22.36 0.139/46.7 E11
150 11.99 0.212/56.8 E8
200 8.21 0.262/62.2 E8
300 5.10 0.319/67.3 E7
400 3.73 0.350/69.4 E6
500 2.94 0.369/70.4 E5
5 50 103.85 0.041/28.4 E15
100 25.41 0.136/48.1 E11
150 13.32 0.214/59.4 E8
200 9.02 0.269/85.7 E8
300 5.54 0.334/71.7 E7
400 4.03 0.369/74.2 E6
500 3.18 0.391/75.5 E5
7 50 134.04 0.035/28.8 E15
100 31.11 0.130/50.0 E12
150 15.73 0.214/63.1 E9
200 10.44 0.278/70.9 E8
300 6.31 0.356/78.7 E7
400 4.56 0.399/82.1 E6
500 3.58 0.426/83.8 E6
9 50 162.60 0.032/29.0 E15
100 36.43 0.124/51.3 E13
150 17.91 0.213/65.9 E10
200 11.70 0.282/74.9 E8
300 6.96 0.371/84.2 E8
400 5.00 0.421/88.4 E7
500 3.92 0.453/90.5 E6
[email protected]
To get the most appropriate effect of explosion yield in TNT model, since the
recommended methods (which is a compilation of analysis result by few
researchers) as proposed by Less [1] still have some uncertainty, therefore the
analysis VCE which occurred in Flixborough in 1974 as published by Hoiset et al.
[20] and Sadee et al. [14] including site incident observation [1, 14] were referred.
Figure 1 shows the comparison of equivalent analysis results of TNT on the
effect of vapor cloud explosion which occurred in Flixborough by both researchers
[14, 19]. Based on the analysis done, both researchers used value η = 2 to 3 % in the
TNT model by considering the explosion occurred near the field scenario using
Brasie and Simpson methods [5] and ACMHH as recommended in the selection of
explosion yield table by Less [1]. Meanwhile the value of physical and chemical
properties of cyclohexane used for both analysis are the same, except for the
different quantity of cyclohexane involved in VCE, whereby Hoiset et al. [20]
assumed that 70 % of 40 tons of cyclohexane released was contributed to the
formation of the flammable cloud (based on local observations). For Sadee et al.
[14], the quantity of cyclohexane released was based on the flash fraction of fuel
(0.17 at 423 K) on the basis of actual thermodynamics data as recorded from daily
activity of plant operation before incident which is estimated 34,000 kg og
cyclohexane. The outcome of results comparison for overpressure by Hoiset et al.
[20], Sadee et al. [14] and site incident observation [1, 14] is shown in Fig. 1.
Figure 1 shows that value of overpressure from TNT model method estimated by
Sadee et al. [14] which was observed to have a tendency to provide the actual
condition as occurred in Flixborough in 1974. However the value is still far from
the actual picture of damage on the structure building in the plant vicinity and its
Fig. 1 The comparison of equivalent analysis results of TNT model on the effect of vapor cloud
explosion which occurred in Flixborough by Hoiset et al. [19] and Sadee et al. [14] and site
observation [1, 14]
[email protected]
Fig. 2 The comparison of equivalent analysis results of TNT model on the effect of an increment
of η value between 2 and 10 % and site observation [1, 14] which occurred in VCE Flixborough
with correlation to structure damage by V.J Clancey [10]
surrounding area; within 120 m (offices), 130 m (plant canteen), 220 m, 230 m
(machine shop), 290 m (storage tank farm), 535 m (houses) and 945 m (houses).
For Fig. 2, the value of explosion efficiency used in TNT model is measured at yield
factor (η (%) between 6 % and limit to 10 % [1, 13, 19],) at 535 m and 945 m within
the actual range of structural building damage as occurred in Flixborough in 1974.
As a results, at η (10 %) factor the overpressure value is closer to the actual scenario
of structural damage as referred by site observation between distance 100–300 m.
From the results in Fig. 2, also shows that the possibility of mass formation mix
to air can happened very fast and lead to a large flammable cloud, therefore the high
temperature will accelerate the effect of evaporation rate and finally the flame
speed velocity will reach the unburnt material very fast and the pressure effect
received by the structural building will be very strong. The site observation
structure damage can be seen in agreement with the damage indicator as developed
by VJ Clancey [10].
Conclusion
For high yield explosion, the area of destruction scales becomes lesser in direct
proportion to the energy release. As a result, the average blast wave arrival time is
increasing and allowing more time for escape between the flash, the blast and the
sound arrival. The most suitable zone for any land use activities, which is allowed
[email protected]
to operate for major hazard installation is proposed at 600 m and above. It is also
observed that beyond this zone, the damage impact may contribute to a very minor
house damage.
References
1. Lees FP (2005) Loss prevention in the process industries. Butterworth Edited by Sam Mannan
2. Zulkifli RA, Mohanad HE, Rashid Sharif A (2010) Assessment on the consequences of liquefied
petroleum gas release accident in the road transportation. J Appl Sci 10(12):1157–1165
3. Zulkifli RA, Mohanad HE, Shariff AR (2009) Development and design of smart advisory
system in the accident of transportation of hazardous material via quantitative risk approach.
J Occup Safety Health 6(12):48–67
4. El-Harbawi M, Mustapha S, Choong TSY, Abdul Rashid Z, Abdul Rashid S, Sherif AA
(2010) SCIA: GIS-based software for assessing the impacts from chemical industrial
accidents. ASCE J Pract Periodical Hazard Toxic Radioactive Waste Mgmt 14:104
5. Brasie WC, Simpson DW (1968) Guidelines for estimating damage explosion. Loss Prev 2:91
6. Harris RJ, Wickens MJ (1989) Understanding vapour cloud explosions an experimental study.
Comm 1408. Instn Gas Engrs, London
7. Eichler TV, Napadensky HS (1977) Accidental vapor phase explosions on transportation
routes near nuclear plants. Rep. J6405 ANL-K77-3776-1. Argonne Nat Lab, Argonne, IL
8. Prugh RW (1991) Quantify BLEVE hazards. Chem Engng Prog 87(2):66
9. Center for Chemical Process Safety (CCPS) (2000) Guidelines for chemical process
quantitative risk analysis. 2nd Edition, American Institute of Chemical Engineers, New York
10. Clancey VJ (1972) Diagnostic features of explosion damage. In: Paper presented at the sixth
international meeting of forensic sciences, Edinburgh
11. Health and Safety Executive Report (1979) Advisory committee on major hazards ‘second
report, HMSO publication
12. Phillips H (1981) Unconfined vapour cloud explosions: a new look at Gugan’s book. Chem
Engr Lond 369:286
13. Center for Chemical Process Safety (CCPS) (2000) Guidelines for evaluating the
characteristics of vapor cloud explosions, flash fires and BLEVEs. 1st Edition, American
Institute of Chemical Engineers, New York
14. Sadee C, Samuels DE, O’Brien TP (1974) The Characteristics of the explosion of cyclohexane
at the NYPRO (UK) Flixborough plant on 1st June 1974. J Occup Accid 203
15. Parker RJ (1975) The Flixborough disaster. Report of the Court Inquiry London
16. Merrifield R, MacKenzie JF (1997) Methodology for estimating the explosion yield of
incidents involving conventional or improvised explosives. In: Proceedings of the eighth
international symposium on interaction of the effects of munitions with structures
17. Mendonça-Filho LG, Bastos-Netto D, Guirardello R (2008) Estimating the TNT equivalence
of a 15-ton single base powder explosion through damaged building profiles analyses.
J Hazard Mater 158:599–604
18. Gould KE, Kempo (1971) High explosive field test: explosion phenomena and environmental
impacts. Defence Nuclear Agency, USA
19. Crowl DA, Louvar JF (2002) Chemical process safety: fundamental with application, 2nd edn.
Prentice Hall, New Jersey
20. Hoiset S, Hjertager BH, Solberg T, Malo KA (2000) Flixborough revisited—an explosion
simulated approach. J Hazard Mater 77:1–9
[email protected]
Disparities in Generated Noise Between
Predictions and Measurements
from Construction Sites
Zaiton Haron, Nadirah Darus and Khairulzan Yahya
Abstract In order to avoid excessive noise exposure and to improve construction

environmental performance, noise should be predicted at the planning stage and
appropriate mitigation measures should be implemented. There are various noise
prediction methods that have been practiced by the respected parties. The simplest
and most recommended method of noise prediction is the procedure of BS 5228-
1:2009. However, previous studies have claimed that this method was inaccurate
because of several factors. Therefore, this study attempts to assess the difference
between the prediction using this method and the noise obtained from the mea-
surement. The study was conducted by measuring noise emissions from earthwork,
piling work and structural work at selected measurement stations, investigating the
individual noise emission of construction equipment and predicting construction
noise using the measured data. Several related variables were also measured to
identify their effects on outdoor sound propagation. The results of t-test showed that
all the stations have the significant disparities between prediction and measurement.
The highest over prediction of 5 dBA was due to the use of several moving
machines during operation processes and high noise emission level of machines.
Consequently, this affects the distance between the sound source and the
measurement station (geometry factor).
Keywords Construction noise Noise prediction Noise measurement

Construction sites
Z. Haron (&) N. Darus K. Yahya

Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 Skudai,
Johor Bahru, Malaysia
N. Darus
K. Yahya

[email protected]
392 Z. Haron et al.
Introduction
In most developing countries including Malaysia, construction industry is one of

the major contributors to economic growth. Construction is a major contributor to
the environmental impact and pollution [1]. There are several initiatives have been
taken to cater the problem but the effect remains the same. The environmental
impacts are including effects on human health [2] due to the noise of the con-
struction process, among others. This is because many plants with loud sound are
extensively used without considering the effect on social problems [3].
Early prediction of construction noise at planning stage and implementation of
any appropriate noise mitigation measures may help to avoid excessive noise
exposure to the public and to improve the environmental performance and con-
struction activity process. Therefore, sound of construction is one of the key ele-
ments that are subjected to the environmental impact assessment [4]. The prediction
of construction noise will be included in the EIA report by an EIA consultant.
However, the methods of construction noise predictions vary between different
consultants since there is no established method to be adopted.
The most recommended prediction method is the BS 5228-1:2009 procedures.
However, previous studies have claimed that this method is not accurate and a
stochastic model was proposed to simulate noise arising from construction activities
[5]. It was due to the nature of the sound fluctuations coming from the construction
process. Previous studies are very limited and not conclusive. Thus, this paper studies
the noise emissions from construction activities, investigates the individual noise
emission level of construction equipment, predicts noise from construction activities,
assesses the disparity between the noise level prediction and the measured noise
level, and subsequently determines the cause of the disparity between the results.
Noise is a set of undesired sound [6–9], and from the construction viewpoint,
noise is characterised by the noise emission level from machines, acoustic noise
emission characteristics, number of machines concurrently in use and distance
between the receiver and machines, and existence of obstruction or reflection
between the receiver and the machines. These include the changes in acoustic
power during full or idle working condition and movement of the machine when
working in a workspace. Furthermore, each stage of construction has different
spectrum levels [10], so the noise characteristics must be considered in estimating
the impact to humans [4]. The impacts cannot be detected spontaneously but will
escalate in the long-term period. Effects of noise on animals should not be taken
lightly because it may disturb the ecosystem where they live.
Currently, there are four methods of construction noise prediction in the BS
5228-1:2009 procedures [11]. The first two methods for the stationary machine are
activity LAeq and sound power, and the other two methods for mobile machines on-
site are on-site (limited area) and hauling in the streets. Basically, noise levels
(LAeq) at the receiver is predicted by combining three basic elements such as
acoustic power of the machine, emission model, and propagation model. The
construction noise predictions as described in Appendix F of BS 5228-1:2009,
[email protected]
Disparities in Generated Noise Between Predictions … 393
which take into account the sound power machine, the operating facility, the
distance between the source and the receiver, the presence of the inspection, the
reflected sound, and the attenuation due to the earth’s surface. The result of this
method is accurate, provided that all the factors mentioned earlier, which are
regarded as an input to the model, are accurate. Temperature and wind speed are not
included in the BS 5228-1:2009 procedures, unless the distance is greater than
300 m. The wind speed is less than 5 km/h and also had a great impact on the
accuracy of the measurement noise.
Without any obstruction between receiver and operation activities, and also no
reflection factor, the noise level of the prediction arising from the two plants
(excavator and dump truck) of a small site (50 × 50 m) had an insignificant dif-
ference or an over prediction of approximately 1 dBA from the measurement [12].
However, based on similar comparison on a larger site and found that, using various
plants (five to six plants), there were significant differences in the noise level
predictions compared with measurements (an average of 4 dBA, higher than the
measured value) because of the effects of plant movement [13]. Over prediction will
be felt by contractors who enter a project tender. Although predictions are used just
to get approval from the local authorities, the responsibility for compliance will be
transferred to the contractor. The over prediction of LAeq will result in an exces-
sively high bid price and lessen the chances of gaining the contract [14].
Methodology
Figure 1 shows the three selected construction stages of earthworks, substructure

(piling), and superstructure located in Kempas and Skudai, Johor. The study was
carried out in three phases as in Fig. 2. Phase 1 involves the measurement noise
levels from construction activities (LAeq), and noise level of individual machines
involved in construction activities. Because of safety reasons, the average noise
emission level was obtained during the machine working at its full load. The
instrumentations used are the sound level meter (type 1 and type 2), anemometer,
and distometer.
Measurements were carried out accordingly at the selected stations for 1 h with
15 min interval. Stations are free from reflection that can influence the sound
transmission. The sound level meter was calibrated before and after measurements,
and it was set at 1.2 m above the ground. Noise level (dBA), height of the machine
Fig. 1 Selected sites a Earthworks b Substructure (piling works) and c Superstructure works
[email protected]
394
(a) (b) (c)

Selected Construction Sites: Individual Construction Machinery or Plant a) Null hypothesis, H0 and
Site 1: Earthworks - Kempas nth item of individual construction machinery alternative hypothesis, H1
Site 2: Substructure (Piling Works) - Kempas H0, µ measured = test value and
Site 3: Superstructure Works - Skudai H1, µ measured test value
Step 1: Select the sound power level (Lw) of nth item where, µ measured = mean real measurements and
test value = prediction
Instrumentations: Step 2: Calculate the distance adjustment (Kd)
1) Sound level meter (SLM) Type 1 and Type 2 Kd = (20 log10 R) – 8 where, R = distance (m)
2) Distometer
3) Anemometer b) Two-tailed test
Step 3: Make a screening allowance (Ks)
5 dBA: partial, 10 dBA: full
Variables:
1) Background noise levels c) Critical significant or alpha level value
Step 4: Make a reflection allowance (Kr) = 0.05
2) Noise emission levels (individual machinery)
+ 3 dBA (receiver - 1 m from reflecting surfaces)
3) Construction noise emisssion levels
4) Distance between SLM and any reflecting surface
5) Distance between receiver and sources
Step 5: Obtain the average percentage on-time (KT) d) Calculate t-value
6) Height of the SLM microphone
[email protected]
(Refer Figure F.5 BS 5228 - Part 1: 2009) ttest = (test value – µ measured)/ (S/ )
7) Height of sources
where, S = variance and
8) Temperature (°C)
= sample size
9) Wind speed (m/s)
Step 6: Calculate equivalent continuous sound
pressure level (Lp @ LAeq) Add or subtract from the
value in Step 1
Results: e) Decide to reject H0 or fail to reject H0
1) Distribution of sound pressure levels Desicion rule is by comparing the calculated values
2) Histogram of sound pressure levels Step 7: Calculate the combined LAeq with critical region values
Fig. 2 Methodology a Phase 1 b Phase 2 and c Phase 3

Z. Haron et al.
(m), noise sources (m), temperature (°C), and wind speed (m/s) were measured.
Measurements were performed at a minimum wind speed and temperature variation
to reduce the differences between predictions and measurements.
Meanwhile, Phase 2 involves the noise predictions by using the BS 5228-1:2009
procedures that were carried out for all stations. For this study, the data from Phase
1, that is, the noise emission by individual plant and the average distance between
the receiver and the machine were used. For Phase 3, the disparities of the results of
noise prediction and the mean values of measurement for all stations were assessed.
A t-test was conducted to test the significant difference in the mean values of real
on-site measurement results and noise prediction. H0 was rejected if the p value was
less than 0.05 or if the t value from calculation was higher than the t value from the
t table (critical region).
Table 1 shows the values of all variables measured for all measurement stations.
These data were obtained to identify the variables that affect sound propagation
outdoors. It can be seen that the change in wind speed and temperature is relatively
small during the measurement.
Meanwhile, Table 2 summarises the distribution and histogram of the sound
pressure level. The highest and lowest noise emissions were generated from the
substructure (piling) and superstructure works, respectively. It was caused by
several individual construction plants involved for each construction activity.
Meanwhile, the level of noise generated from earthworks was within the range of
noise level generated from substructures (piling) and structural works. However, the
noise level at station E2 (earthwork) was less than the noise from the installation of
the main structure because of the fewer number of individual machines operated
during noise measurement.
Table 1 Computation of variables measured on-site

Stages of Stations Background Noise Temperature Wind
construction noise levels sources (°C) speed
(dBA) (m) (m/s)
Earthworks E1 56.20 1.2–2.4 29.8 1.5
E2 48.90 1.2–2.4 29.6 2.5
E3 61.50 1.2 30.8 1.8
Substructure P1 66.30 1.2–2.4 27.7 1.5
(piling works) P2 63.80 1.2–2.4 32.8 2.0
P3 61.20 1.2–2.4 31.8 1.8
Superstructure S1 53.80 1.2–2.4 32.4 0.5
works S2 55.50 1.2–2.4 33.6 0.7
S3 60.40 1.0–1.2 29.0 0.5
[email protected]
396
Table 2 Computation of sound pressure levels for 1 h (5-min time interval)

Stages of construction Stations Distributions of sound pressure Histogram of sound pressure levels (dBA)
levels (dBA)
Highest Lowest Range Highest frequency of occurrence Highest Lowest
Earthworks E1 74.904th 5min 64.709th 5min 65.00–75.00 70.00–71.00 3.143 70.82
E2 69.503rd 5min 50.2011th 5min 50.00–70.00 55.00–60.00 6.225 59.61
E3 76.306th 5min 62.0010th 5min <60.00–>80.00 65.00–67.50 4.809 67.94
Substructure (piling works) P1 77.605th 5min 71.802nd 5min 71.00–78.00 75.00–76.00 1.612 74.94
P2 77.606th 5min 69.6012th 5min 68.00–80.00 74.00–80.00 3.401 75.65
P3 79.7011th 5min 69.008th 5min 68.00–80.00 78.00–80.00 3.323 75.96
[email protected]
Superstructure works S1 67.607th 5min 59.3011th 5min 58.00–68.00 64.00–68.00 3.098 63.72
S2 70.203rd 5min 56.7010th 5min <55.00–>75.00 60.00–61.00 3.899 62.44
S3 68.802nd 5min 57.5010th 5min <58.00–>70.00 67.50–70.00 3.619 65.56
Z. Haron et al.
The results of the noise emissions from individual construction equipment were
used to calculate their sound power level (Lw) as input data to the noise prediction
model. Eight machineries involved in earthwork: two crawler excavators, a bull-
dozer, a dump truck, and four compactors. Meanwhile, ten machines: three piling
machine, three crawler cranes, and four crawler excavators were used in the sub-
structure (piling). There were only five machines used for main structural works:
two mobile cranes, a backhoe, a concrete mixer truck, and a bar bending and a
cutting machine. The highest and the lowest Lw values were generated from the
piling machine of the substructure (piling) (2–115 dBA) and the concrete mixer
truck (1–92 dBA) from the structural works, respectively. Meanwhile, Table 4
shows the noise predictions for all measurement stations of three construction
stages. On the basis of Table 3, station E3 produced the highest noise emissions for
work related to earthwork because all machines were operated simultaneously.
Meanwhile, stations P3 and S3 generated the highest level of noise emissions to the
substructure (piling) and major structural works, respectively.
Table 4 shows the results for all measurement stations construction. On the basis
of Table 4, the critical t value (tcritical) for all measurement stations was 1.812.
However, the t value observed for all measurement stations exceeded the tcritical.
Therefore, all H0 were rejected. Thus, the t-tests showed that there were significant
disparities between the noise level of the actual on-site measurements and the noise
prediction results for stations E1, E2, and E3.
All stations at the three stages of construction measurement have higher noise
prediction results compared with the results of the actual on-site measurements. The
disparities between the prediction and the measurement t values of stations E1, E2,
and E3 (earthworks) were 3.00, 6.70, and 5.92 dBA, respectively (an over pre-
diction). This result is higher than the previous study [13] because the number of
moving machines in this study was more than that of the study. Meanwhile, for
stations P1, P2, and P3 in the substructure (piling) stage, the over prediction values
were 1.14, 2.40, and 2.45 dBA, respectively. For stations S1, S2, and S3 in the
superstructure works stage, the over prediction values were 5.73, 3.15, and
4.24 dBA, respectively. For t-test, the observed values of tnoise for all stations were
greater than the tcritical value of 1.812. Therefore, the hypothesis that measured and
predicted noises were equal was rejected. There were significant disparities between
real on-site measurement and noise prediction.
It can be seen that the static machines with high noise emission level produced a
smaller disparity noise level at the receiver, whereas the construction operation
involved several moving plants that produced the largest disparities. The large
disparities may be due to a gross simplification in the data input related to the
distance between the station and the moving plants. Because the plants have to
move around the construction site in real life and because of the level of noise
generated from on-site construction equipment, individuals have relied on the
variation of the acoustic power of machines during heavy load. In this study, the
average acoustic power of machines was used in the prediction.
[email protected]
Table 3 Computation of LAeq obtained using the BS 5228-1:2009 procedures
398
Stages of Station Sub Construction Lw Distance Corrections Individual LAeq Combined

construction site plants (dBA) (m) Kd Ks Kr KT LAeq sub LAeq
sites
Earthworks E1 A CE 1 108 36.5 −39.25 0 0 0 69.75 70.74 73.82
CE 2 106 20.3 −34.15 0 0 −8.0 63.85
B DT 1 107 56.7 −43.07 0 0 −2.0 63.93 68.05
DT 2 107 45.1 −41.08 0 0 0 65.92
C BD 1 101 60.5 −43.64 0 0 −2.0 55.36 67.66
DT 3 109 65.8 −44.36 0 0 0 64.64
DT 4 108 62.2 −43.88 0 0 0 64.12
E2 A CE 2 106 55.0 −42.81 0 0 −2.5 60.69 60.69 66.31
B DT 1 107 50.0 −41.98 0 0 −2.5 62.52 63.78
CP 1 98 40.7 −40.19 0 0 0 57.81
C CE 1 108 88.5 −46.94 0 0 −2.5 58.56 58.56
E3 A CE 1 108 45.2 −41.10 0 0 −0.2 66.70 73.04 73.86
[email protected]
CE 2 106 20.9 −34.11 0 0 0 71.89
B BD 1 101 20.6 −34.28 0 0 −0.5 66.22 66.22
Substructure P1 A PM 1 111 42.7 −40.61 0 0 −0.2 70.19 74.23 76.08
(piling works) CE 1 104 16.5 −32.35 0 0 0 71.65
CC 1 101 37.4 −39.46 0 0 0 61.54
B PM 2 115 55.5 −42.89 0 0 −2.5 69.61 71.49
CE 2 105 52.5 −42.40 0 0 0 62.60
CE 3 105 40.0 −40.04 0 0 −1.0 63.96
CC 2 100 50.0 −41.98 0 0 0 58.02
P2 A PM 1 111 50.0 −41.98 0 0 0 69.02 69.30 78.05
(continued)
Z. Haron et al.
Table 3 (continued)
Stages of Station Sub Construction Lw Distance Corrections Individual LAeq Combined
construction site plants (dBA) (m) Kd Ks Kr KT LAeq sub LAeq
sites
CC 1 101 55.0 −42.81 0 0 57.19
B PM 2 115 30.4 −37.66 0 0 −0.5 76.84 76.84
C PM 3 114 65.0 −44.26 0 0 −1.5 68.24 68.46
CC 3 99 60.0 −43.56 0 0 0 55.44
P3 B PM 2 115 90.0 −47.08 0 0 0 67.92 67.92 78.41
C PM 3 114 30.6 −37.71 0 0 −0.5 75.79 78.00
CE 3 105 23.2 −35.31 0 0 0 69.69
CE 4 106 20.3 −34.15 0 0 −1.0 70.85
CC 3 99 18.6 −33.39 0 0 0 65.61
Superstructure S1 A MC 1 99 30.4 −37.66 0 0 0 61.34 69.45 69.45
works MC 2 109 37.5 −39.48 0 0 −2.5 67.02
Disparities in Generated Noise Between Predictions …
CM 1 92 40.6 −40.71 0 0 −0.2 51.63
[email protected]
BH 1 98 20.5 −34.24 0 0 −0.2 63.56
S2 A MC 2 109 60.5 −43.64 0 0 0 65.36 65.44 65.59
CM 1 92 49.7 −41.93 0 0 −2.0 48.07
B BH 1 98 90.7 −47.15 0 0 0 50.85 50.85
S3 A MC 2 109 38.1 −39.62 0 0 0 69.38 69.62 69.80
BH 1 98 44.8 −41.03 0 0 0 56.97
B BB 1 93 28.3 −37.04 0 0 0 55.96 55.96
399
400
Table 4 Test of significant difference in mean equivalent noise levels

Stages of Stations Test value Means, M No. of Standard p t Test hypotheses
construction (dBA) (µmeasured) samples(η) deviation H0: µmeasured = test
value
H1: µmeasured ≠ test
value
Earthworks E1 73.82 70.82 11 3.143 0.010 3.168 H0 was rejected*
E2 66.31 59.61 11 6.225 0.005 3.570 H0 was rejected*
E3 73.86 67.94 11 4.809 0.002 4.086 H0 was rejected*
Substructure (piling P1 76.08 74.94 11 1.612 0.040 2.353 H0 was rejected*
works) P2 78.05 75.65 11 3.401 0.042 2.336 H0 was rejected*
[email protected]
P3 78.41 75.96 11 3.323 0.035 2.441 H0 was rejected*
Superstructure works S1 69.45 63.72 11 3.098 0.000 6.137 H0 was rejected*
S2 65.59 62.44 11 3.890 0.023 2.683 H0 was rejected*
S3 69.80 65.56 11 3.619 0.003 3.882 H0 was rejected*
* At a 95% significance level.
Z. Haron et al.
Conclusion
This study has evaluated the difference in noises predicted using BS 5228-1:2009
procedures and measured at several stations of three selected stages: earthworks,
substructure (piling), and structural works. The t-test showed significant differences
for all stations. In addition to the high noise emission level of machines, the highest
difference (5 dBA) has been shown as the result of the activities that involve several
moving machines. Consequently, this affects the distance between the sound source
and measurement station. Therefore, in this study, the difference is caused by the
movement of the plant which affects the distance between the sound source and
measurement station (geometry factor), number of moving machines, and high
levels of machine noise emissions.
Acknowledgment The authors would like to express their appreciation for the support of the
sponsors with Project No. Q.J130000.2522.03H43 and UTM Zamalah Scholarship.
References
1. Fuertes A, Casals M, Gangolells M, Forcada N, Macarulla M, Roca X (2013) An

environmental impact causal model for improving environmental performance of construction
processes. J Clean Prod 52:425–437
2. Li X, Zhu Y, Zhang Z (2010) An LCA-based environmental impact assessment model for
construction process. Built Environ 45:766–775
3. Manatakis E, Skarlatos D (2002) A statistical model for evaluation and prediction of the noise
exposure in a construction equipment area. Appl Acoust 63:759–773
4. Department of Environment (2007) The planning guidelines for environmental noise limits
and control. Department of Environment Malaysia, Malaysia
5. Carpenter F (1997) Construction noise prediction at the planning stage of new developments.
Build Acoust 3(4):239–249
6. Edworthy J (1997) Noise and its effect on people: an overview. Int J Environ Stud 51:335–344
7. Muzet A (2007) Environmental noise, sleep and health. Sleep Med Rev 11:135–142
8. Hamoda MF (2008) Modelling of construction noise for environmental impact assessment.
J Constr Developing Countries 13(1):79–89
9. Fernandez MD, Quintana S, Chavarria N, Ballesteros JA (2009) Noise exposure of workers of
the construction sector. Appl Acoust 70:753–760
10. Ballesteros MJ, Fernandez MD, Chavarria N, Ballesteros JA, Gonzalez I (2010) Noise
emission evolution on construction sites-measurement for controlling and assessing its impact
on the people and on the environment. Build Environ 45:711–717
11. British Standards Institution (2008) BS 5228—Part 1: 2009 code of practice for noise and
vibration control on construction and open sites. BSI Group Headquarters, The United
Kingdom
12. Idris N (2012) Automated prediction of noise from construction site using stochastic approach.
Master thesis, Universiti Teknologi Malaysia
13. Jahya Z (2014) Automated construction noise prediction by considering the variability of
noise sources and outdoor sound propagation. Master thesis, Universiti Teknologi Malaysia
14. Haron Z (2007) Probability approach in environmental acoustics. Ph.D. thesis, University of
Liverpool
[email protected]
Flame Retardancy of Polymeric Building
Material with Recycled Expanded
Polystyrene Filler
Syed Anas Bin Syed Mustafa, Rahmah Mohamed, Wan Bahira

and K. Rasidan
Abstract Flame retardancy is an important characteristic in the flame and fire

safety field to both prevent and limit the effects of material ignition and flame
consumption. The preparation and characterization of Unsaturated Polyester Resin
(UPR) filled recycled Expanded Polystyrene (EPS) composite systems were sys-
tematically investigated. Additives such as organic nanocrystal (ONC) and silica
aerogel powder were added to the composite for imparting suitable characteristics
to the composite. The result obtained from comparing the current study against its
predecessor reveals Liquid Natural Rubber (LNR) is significantly more flammable
than Styrene Butadiene Rubber (SBR) by more than doubled but aerogel utilization
can negate this effect. It can be concluded that the use of additive could affect flame
retardancy and thermal properties. Further studies of the material can be done in
determining and/or confirming the actual mechanisms which ONC and aerogel
produces this effect on flammability.
Keywords EPS UPR Recycled composite Flame retardancy Thermal

characteristics
S.A.B. Syed Mustafa (&) W. Bahira K. Rasidan

Faculty of Applied Science (FSG), Universiti Teknologi MARA (UiTM), 40450 Shah Alam,
Selangor, Malaysia
R. Mohamed
GREEN Polymer Research Group, Faculty of Applied Science (FSG),
Universiti Teknologi MARA (UiTM), 40450 Shah Alam, Selangor, Malaysia

[email protected]
Introduction
Unsaturated Polyester (UPE) matrix composites have been used for many years in a
broad technology fields such as naval construction, offshore applications, water-
lines, and building construction. UPE is an economical thermo-set material that is
widely used due to its excellent process ability, good cross-linking tendency as well
as mechanical properties when cured. Expanded Polystyrene (EPS) sheet has been
used as core material for sandwich core composite door shutter to replace wooden
door shutters in building [1]. Studies had been done on use of EPS incorporated into
Unsaturated Polyester Resin (UPR) with relevant diluents. Toughening of the
unsaturated polyester resin is very important to increase impact performance
especially for building structures. [2].
Flammability could be defined as the capability of a material sustaining a flame
when it is ignited. Flammability poses an ever-present concern in fire safety and
prevention; where reducing the flammability, or increasing the flame retardancy, of
materials are a priority to help prevent fires by hardening them against ignition, and
reduces a flames spread once it occurs.
The preparation and characterization of UPR filled recycled EPS composite
systems were systematically investigated. Additives such as ONC and silica aerogel
powder were added to the UPR-EPS blend for imparting suitable characteristics to
the composite. The effect of differing weight percentage of additives on the thermal
properties and flammability of the material were determined. The fabricated com-
posites undergoes linear flame propagation test to determine its level of
flammability.
The use of rubber as a toughening agent with UPE had been investigated [2, 3]
and use of aerogel in paint for increasing thermal insulation and flammability had
also been performed [4–6]. Integration of nanocrystal and aerogel to UPE/EPS
blend was found to lower the thermal conductivity of the material, which is a
characteristic required for insulating panels. From our previous paper [7], we found
that the thermal conductivity, k value had reduced by more than half upon adding
the additives as prescribed.
This paper emphasises on comparing the flammability of various UPR/EPS
composite systems impregnated with some selected additives by evaluating the
changes in flame propagation speed as effect of the additives utilized.
Experimental
Materials and Method, The selected thermo-set matrix material used for fabri-
cating these composite systems is the Reversol P9780 UPR resin, consisting of
vinyl ester oligomers having density of 1.12 g/cc, viscosity of 450–600 cps with
41–44 % styrene content. Methyl Ethyl Ketone Peroxide (MEKP) and Cobalt
solution used as initiator and promoter each, were supplied together with the UPE
[email protected]
Flame Retardancy of Polymeric Building Material … 405
resin by Revertex Sdn Bhd. The EPS or Styrofoam filler was obtained from waste
material. These ingredients are similar to materials used by Rashidan et al. [1] and
Wan Bahira [8].
This paper was produced by applying an analytical quantitative methodology
where the sample produced are subjected repeated testing on its flammability via
ASTM D635-10 method, and measurement of its thermo-physical properties via a
KD-2 Pro thermal probe and a digital densitometer.
Composite Fabrication and Testing, Recycled EPS, fixed at 10 % parts by
weight (%wt) of UPR resin, was blended utilizing a high speed agitation mixer until
complete dissolution was attained. This is to enable the gaseous and solid contam-
inates present are easily eliminated from the mixture via gravity settling for 24 h.
Additives were added before the samples were prepared utilizing hand lay-up
processes onto an aluminium mould. The additives such as ONC content was added
at specific quantities as a percentage of blend weight. The process was followed by
adding 3 %wt of MEKP initiator and suitable amount of cobalt as accelerant to cure
the EPS/UPE composite. Gelation time was set to be about 5 min. Once fully
gelled, complete curing was done under compression for 1 h at 70 °C. These steps
are also similar to the process used by Rashidan et al. [1] and Wan Bahira [8].
The produced sample were tested by using ASTM D635-10 method, where
samples were lit while suspended in a horizontal position to which flame charac-
teristic and propagation speed were noted. The samples were also subjected to
thermo-physical measurements via a KD-2 Pro thermal probe and density deter-
mination via a digital densitometer.
The additives utilized across the samples are elastomers (SBR or LNR), antioxidant
(AO), flame retardant (FR), silica aerogel (Ag) and organic nanocrystal (ONC).
Table 1 shows the time required for burning the samples at the fixed distance,
indicated as elapsed time (s), and the rate of combustion is shown by the linear
burning formulation, V(mm/min),
Linear Burning; V ¼ 60 L=t
where
L the burn length, in millimetres between reference marks (100 − 25 = 75 mm)
t the time(s) (elapsed time from the 25 to 100 mm reference mark)
Both Fig. 1 and Table 1 shows a few key trends; an integration of elastomer into
the system promotes flame propagation, a composite utilizing SBR is shown to be
[email protected]
Table 1 Comparison of combustion elapsed time in different additive systems

Additive in UPE/EPS composite (per 100 % Designation Elapsed Linear
weight mixed UPE/EPS) time, t burning, V
(s) (mm/min)
20 % SBR − 0.5 % AO S1 85 52.94
20 %SBR − 0.5 % AO − 0.5 % FR S2 105 42.86
4 % SBR − 0.5 % AO SA1 511 8.81
4 % SBR − 0.5 % AO − 2.5 % FR SA2 814 5.53
4 % SBR − 0.5 % AO − 5 % FR SA3 1433 3.14
4 % LNR − 0.5 % AO − 5 % FR L1 542 8.30
4 % LNR − 0.5 % AO − 5 % FR − 0.2 % L2 617 7.29
(ONC)
4 % LNR − 0.5 % AO − 5 % FR − 0.2 % Ag L3 1196 3.76
4 % LNR − 0.5 % AO − 5 % FR − 0.2 % L4 698 6.45
ONC − 0.2 % Ag
Fig. 1 Flame Propagation Speed for all Systems
more flame retardant compared to the group utilizing LNR, and the use of aerogel
and ONC additives have mixed effect on flame propagation speed.
By going along the set definition of flammability, faster flame propagation
speeds correlates to better flammability or lower flame retardancy of a material.
Hence, a 75 % reduction of SBR content had reduced the flame speed by 80 % as
shown by comparing sample S1 and SA1 of similar content. This could also be
inferred that excessive amount of elastomer might be a preferred fuel for the flame
as shown by how slightly the flame retardant tempers the flame propagation speed
at higher elastomer content; at 20 wt% SBR, 5 wt% FR only reduces the flame
speed by 19 %, whereas at 4 wt% SBR, similar additions reduces the flame speed
by 64 %.
[email protected]
On another note, comparison between sample SA3 and L1 which differs only on
the type of elastomer used, it is inferred that SBR is the more flame-retardant of the
two. This is shown by monumental increase in flame speed reported from the L1
sample despite using similar flame retardant amount. This phenomena could be
attributed to the different chemical composition between the two elastomer as
resulted from the manufacturing processes.
The addition of aerogel and ONC into the composite systems had some effects
on the flammability of the materials. The aerogel integrated L3 sample shows
significant reduction in flame speed (−55 %), while the ONC integrated L2 sample
report slight reduction (−12 %). The L4 sample containing both shows value
slightly above the expected average of the L2 and L3 (actual speed of 6.45 mm/min
against the expected 5.53 mm/min).
The significant reduction imparted by aerogel might be inferred to be attributed
to its inherent Silicon, Si based mineral which also serves as thermal insulating
material. Silica as a mineral is fairly resistant to ignition and burning which might
inhibit flame propagation. The excellent thermal insulating property of aerogel was
found to inhibit flame propagation by shielding the more flammable contents from
the high temperature and heat necessary for ignition.
In order to further determine the suitability of utilizing these additives systems,
these additive systems should be investigated based on another avenue, namely
their effects upon the thermo-physical properties of the resultant material. The
chosen measured thermo-physical properties are their thermal conductivity, thermal
diffusivity, volumetric specific heat, and overall density.
As Table 2 suggested, adding only the flame retardant as in L1 reduces the
thermal conductivity and specific heat, thus maintaining the diffusivity which is the
ratio of conductivity to specific heat. Diffusivity is an important ratio in due part to a
higher value denotes a material which transfer heat better than retaining it, and
maybe suitable for certain applications.
Introducing ONC into the system as in L2 reduces all three thermal properties.
Adding aerogel instead such as L3 increases them instead. Finally, adding both
additives into L4 balances the two systems preceding it.
The results shows that aerogel enhances specific heat capacity as is expected of a
material commonly used as a heat insulator, despite going against the same notion
by also increasing its thermal conductivity as well. It could be theorized that this is
Table 2 Comparison of thermal properties of different additive systems

Samples Conductivity Diffusivity volumetric Specific heat Density
(W/m K) (mm2/s) (MJ/m3 K) (g/cm3)
UPR/EPS (L0) 0.165 0.093 1.764 1.221
L1 0.099 0.092 1.112 1.230
L2 0.090 0.089 0.983 1.217
L3 0.254 0.122 1.810 1.205
L4 0.227 0.102 1.657 1.240
[email protected]
caused by the silica within the cured material, which as a material, has a fairly good
thermal retention. With its minute particle size and proper dispersion across the
material, this in turn helps enhance heat transfer through the prompt heat adsorption
of adjacent aerogel particles across the material.
The slight thermal properties reduction effect of ONC might be explained by its
proposed use as an Infra-Red (IR) radiation reflector. IR radiation generated by a
heating source were reflected from the sample surface and thus not absorbed by the
material, which in turn reduces the amount of heat that could be measured. This
reduction was then manifested as a reduction in the specific heat and thermal
conductivity.
Density was also affected by the additives system as shown above. Inclusion of
FR which was used in liquid form might had increased the density by due of the
denser liquid integrated. Both ONC and aerogel however might had reduced the
density by having some of material volume displaced by the suspension solvent
used to integrate either additive into the material.
Aerogel is also quite significantly the least dense additive used, hence explaining
further density reduction upon its integration. However, the integration of both
additives had slight increased the density instead, which may be caused by an
undetermined mechanism or reaction to enable the integration, thus requiring fur-
ther study.
Despite all explanations given, these descriptions were merely deductions that
are speculated and interpolated from the inherent properties of the material and
additives used. In short, further studies are required to fully determine whether the
proposed explanations given are true or another different mechanism are at work.
Conclusion
It could be concluded from the results that incorporation of SBR instead of NR

reduced the flammability of the composite; significant reduction in flammability
could also be brought by reduced amount of SBR. The increase in FR used proved
to inhibit flammability while the use of additives such as ONC and Aerogel have
some mixed effects on flame retardancy; the incorporation of aerogel reduces flame
retardancy significantly compared to ONC. The integration of aerogel also
improves volumetric specific heat while reducing overall density. ONC integrations
have slight effect on overall thermal properties. Combining both additives however
produces slightly optimized properties. Further research is required to confirm the
proposed explanation of the result.
Acknowledgment I would like to acknowledge Universiti Teknologi MARA (UiTM) for their
research funding under the Research Intensive Faculty (RIF) fund [File:600-RMI/DANA 5/3/RIF
(64/2012)]
[email protected]
References
1. Rashidan K, Rahmah M, Nor ZIZ, Mohd HM (2009). Effect of antioxidant and flame retardant
additives on thermal degradation and flammability of SBR filled EPS/UPR composite. In:
Proceedings of Malaysia polymer international conference (MPIC 2009)
2. Bonnia NN, Surip SN, Ratim S, Mahat MM (2012) Mechanical performance of hybrid
polyester composites reinforced Cloisite 30B and kenaf fibre. In: Proceedings of 2nd ASEAN-
APCTP workshop on advanced materials science and nanotechnology
3. Redzuan AA, Bonnia NN, Siti Shakirah S, Mohamed MRH, Surip SN (2013) Mechanical
properties of rubber toughened polyester filled carbon black. Adv Mater Res 701:37–418
4. Cuce E, Cuce PM, Wood CJ, Riffat SB (2014) Optimizing insulation thickness and analysing
environmental impacts of aerogel-based thermal superinsulation in buildings. Energy Build
77:28–39
5. Papadopoulus AM (2005) State of the art in thermal insulation materials and aims for future
developments. Energy Build 37(1):77–86
6. Abdou OA, Murali K, Morsi A (1996) Thermal performance evaluation of a prefabricated fiber-
reinforced plastic building envelope system. Energy Build 24(1):77–83
7. Mustafa SABS, Mohamed R, Zakaria NZI, Rustam HB (2013) Thermal characterisation of
thermoset polyester resin filled recycle expanded polystyrene composite with aerogel and
alumina additives. Adv Mater Res 664:600–604
8. Bahirah WN (2011) Effect of curing and flame retardant system on properties of polyester-filled
EPS composite (Master's thesis). March
[email protected]
Grey Model for Accident Prediction
in Data-Scarce Environment
Ali Al-shanini, Arshad Ahmad and Faisal Khan
Abstract Effective accident prediction is needed in the chemical process industries

to facilitate risk management during plant operations. The unavailability of data
impeded the process for accident modelling. This article discusses the use of grey
model in predicting accidents in data-scarce environment. The model used is a first
order single variable, i.e., G(1, 1) and is applied to a vessel and a CSTR case
studies. The results obtained are compared to that of Poisson–Gamma model.
Results show that G(1, 1) is more accurate compared to the Poisson model that
exhibited underestimation in prediction. It is concluded that the use of G(1, 1) can
represent the CPI deviations with better accuracy, and it can replace the use of
Poisson model especially when some of its weaknesses are overcome.
Keywords CPI accident prediction Poisson–Gamma model First order grey

model Precursor data
A. Al-shanini A. Ahmad (&)

Institute of Hydrogen Economy, Universiti Teknologi Malaysia,
A. Al-shanini
A. Al-shanini A. Ahmad
Faculty of Chemical Engineering, Universiti Teknologi Malaysia,
F. Khan
Faculty of Engineering and Applied Science, Memorial University of Newfoundland,
St. John’s, NL A1B 3X5, Canada

[email protected]
412 A. Al-shanini et al.
Introduction
Efforts towards minimizing risks of accidents in the chemical process industry

(CPI) can be effectively facilitated by the ability to predict the expected future
incidents. These predictions provide valuable information to support risk-based
decision to prioritize plant maintenance and management of change needed during
operations. A common approach is to use probabilistic methodology by mapping
data points onto a distribution function that is assumed to represent the process
behavior to generate predictions. These stochastic models e.g. Poisson model is
widely applied to predict accidents in the CPI [1–4] since only one observed data is
needed to predict the future. However, it has low sensitivity to the observed data
which leads to poor prediction with highly significant error when observed data
have increasing or decreasing trends [3, 5, 6].
Alternatively, it is also common to classify process behavior according to the
order of the process. In this case, the CPI is often represented by a first order,
second order or higher order, the latter typically a combination of first order pro-
cesses. Mathematically, a first order process is modeled by a first order differential
equation, and its salient features is embedded in the model parameters. In this case,
the challenge is to obtain accurate model parameters with limited amount of data.
This article discusses accident prediction issues in data-scarce environment using
grey model. The model is based on a first order differential equation with specific
parameter estimation method using only four data points. The scheme is imple-
mented on two case studies and the results obtained are compared to that of
Poisson-Gamma model.
Prediction Models
Poisson–Gamma Model
Poisson–Gamma model is a class of stochastic forecasting models that is widely

used to predict the number of unwanted consequences in the CPI. The model
predicts well when the mean and the variance of the data are in close proximity,
whereas in cases where variance exceeds the mean, negative binomial distribution
models can provide better prediction [7]. Details about Poisson model can be
obtained in the literature e.g., Refs. [3–5, 8].
Time Series Grey Forecasting Model
First introduced by Deng [9], grey model is an M-order partial differential equation
of N variables that belongs to time series forecasting family, denoted by G(M, N).
[email protected]
Grey Model for Accident Prediction in Data-Scarce Environment 413
The name of “Grey” represents that the model consists of both components that are
founded on fundamental of science and those that have no theoretically explanation,
often fitted using experimental data. The model has been applied in various
applications including agriculture, environmental, and electronic industry due to its
capability to reasonably forecast the unknown events using as few as four data
points [10–14].
Grey model has the ability to enlarge few data points into a monotonically
increasing data series using an accumulated generating operation (AGO), which
approximately would exhibit the exponential behavior. This leads to reduce the
noise from the original data series [15]. The solution of the differential equation of
Grey model has an exponential form similar to the AGO, and it is used to model the
data series from AGO and then to forecast the future events.
First-Order Grey Model G (1, 1)
Although various forms of models are possible, G(1, 1) is the most commonly
adopted [10]. It is a single variable first-order grey model is used to predict future
events for non-negative data points series X(0) = {x(0)(1), x(0)(2), … x(0)(n)|n ≥ 4}
[16] that is taken in consecutive order and at equal time interval, where X(0) is the
time series data at time i, and n is the number of observation that must be equal or
greater than 4 observations. Applying G(1, 1) to predict the future value
x(0)(n + k) for k ≥ 1 is summarized as in the following steps [17–19].
Step 1: the new sequence of data X(1) is generated by AGO and is represented as:
n o
X ð1Þ ¼ xð1Þ ð1Þ; xð1Þ ð2Þ; . . .; xð1Þ ðnÞ ð1Þ
P
Here, X ð1Þ ð1Þ ¼ X ð0Þ ð1Þ; and X ð1Þ ðkÞ ¼ ki¼1 xð0Þ ðiÞ; k ¼ 2; 3; . . .; n. There is a
constraint associated with the original historical data that the ratio test σ = [σ(2),
σ(3),…, σ(n)], hence rðkÞ ¼ xð0Þ ðk 1Þ=xð0Þ ðkÞ; for k ¼ 2; 3; . . .; n has to fall
within 0.1345–7.389, and this indicates that X(0) satisfy the grey model [20].
Step 2: establish the first-order grey differential equation as follows:
xð0Þ ðkÞ ¼ aZ ð1Þ ðkÞ þ u; k ¼ 2; 3; . . .; n ð2Þ
where, a is the developing coefficient, u is the grey input, and z(1)(k) =

0.5x(1)(k) + 0.5x(1)(k − 1), for k= 2, 3, …, n. Equation 2 can be rewritten to a matrix
form as follows:
[email protected]
2 3
Z ð1Þ ð2Þ 1
6 7
6 Z ð1Þ ð3Þ 17
6 7
6 7
Xn ¼ B â; where B ¼ 6 7;
6 .. .. 7
6 . .7 ð3Þ
4 5
Z ð1Þ ðnÞ 1
h iT
Xn ¼ xð0Þ ð2Þ; xð0Þ ð3Þ; . . .; xð0Þ ðnÞ ; and â ¼ ½a; uT
Step 3: estimating a and u coefficients in Eq. 3 using the ordinary least-squares

(OLS) method which is represented as:
1
½â; ûT ¼ BT B BT Xn ð4Þ
Step 4: form the first-order grey reflection differential equation as follows:
dxð1Þ ðkÞ
¼ axð1Þ ðkÞ þ u ð5Þ
dt
Equation 5 is called the shadow equation of the source model

xð0Þ ðkÞ ¼ aZ ð1Þ ðkÞ þ u. By solving this equation, with initial condition of
X(1)(1) = X(0)(1), the following is obtained:
Kð1Þ h ui u
x ðkÞ ¼ xð0Þ ð1Þ eaðk1Þ þ ; k ¼ 2; 3; . . .; n ð6Þ
a a
Here, the ^ donates the grey prediction value.

Step 5: restore the demand prediction value by inverse AGO which is defined as:
xð0Þ ðk þ 1Þ ¼ xð1Þ ðk þ 1Þ xð1Þ ðkÞ ð7Þ
Substituting Eq. 6 into Eq. 7, the grey prediction of future events will be obtained as
follows:
Kð0Þ h ui
x ðkÞ ¼ xð0Þ ð1Þ ð1 ea Þ eaðk1Þ ; k ¼ 2; 3; . . .; n ð8Þ
a
Grey Model Configurations
The models are applied to two case studies, i.e., a CSTR and a Process Vessel [1,
21]. The data sets employed are shown in Table 1. Two different configurations of
Grey prediction models applied with these data sets which are as follows:
[email protected]
Table 1 Historical real time Time interval System 1 System 2

data of the case studies
CSTR precursor data Vessel precursor data
1 6 7
2 9 7
3 12 8
4 13 8
5 15 5
6 11 12
7 16 14
8 10 16
Data of this table has been taken from [1] and [21] for CSTR and
vessel processing respectively
• One step prediction using a moving window technique that considers only the
last four observed data.
• One step prediction using observed data population increasing technique in
which after the predicted event becomes known, it is added to the data points to
predict the next event.
However, the second configuration is applied in this article due to space limitation.
According to prediction criteria introduced by Lewis [22], typically a prediction is

considered excellent if (MAPE < 10), good if (10 < MAPE < 20), reasonable if
(20 < MAPE < 50), and incorrect if (MAPE > 50). The results obtained in revealed
that both models are considered reasonable with MAPE of 20–30 %. These are
shown in Tables 2 and 3. The results also illustrated the better prediction provided
by the G(1, 1) model for both the CSTR and the process vessel data.
Table 2 Predictive models’ outputs and errors CSTR process with observed data population
configuration
Time Actual no. of Poisson model G(1, 1)
interval consequences Model Absolute Model Absolute
value error (%) value error (%)
1 6 – – – –
2 9 6 33.33 9 0
3 12 7 41.67 11 8.33
4 13 9 30.77 13 0
5 15 10 33.33 16 6.67
6 11 11 0 18 63.64
7 16 11 31.25 14 12.5
8 10 12 20 16 60
MAPE (%) – 27.2 – 21.6
[email protected]
Table 3 Predictive models’ outputs and errors of tank process with observed data population
configuration
Time Actual no. of Poisson model G(1, 1)
interval consequences Model Absolute Model Absolute
value error (%) value error (%)
1 7 – – – –
2 7 7 0 7 0
3 8 7 12.5 8 0
4 8 7 12.5 8 0
5 5 7 40 9 80
6 12 6 50 6 50
7 14 8 42.86 11 21.43
8 16 9 43.75 15 6.25
MAPE (%) – 28.8 – 22.5
Figures 1 and 2 show the validation of the prediction models to the actual
observed data. The lack of sensitivity of the Poisson model to the observed data has
led to underestimation especially when the observed data are more excited. The
Grey model has some difficulties in dealing with fast changes. This has caused large
prediction error on 6th and 8th intervals for the G(1, 1). The 6th interval prediction
output has extended the data trend from the 1st to the 5th interval. Likewise, the 8th
interval prediction also followed the decreasing trend, which indicate consistent
18
Actual
Poisson
16 G(1,1)
14
No. of deviations
12
10
6
1 2 3 4 5 6 7 8
Time interval
Fig. 1 Prediction models validation of CSTR case study
[email protected]
16
Actual
Poisson
14 G(1,1)
12
No. of deviations
10
4
1 2 3 4 5 6 7 8
Time interval
Fig. 2 Prediction models validation of vessel tank case study
weakness on the G(1, 1) model. In the case of the Poisson model, in all cases it is
underestimating the prediction. It consistently showed weak response in the manner
of excessive filtering of all observed data.
Conclusion
Grey model has been studied to predict CPI consequences in scarce data envi-
ronment. In both case studies considered, the Grey model based on G(1, 1) con-
figuration produced higher prediction accuracy than the commonly used Poisson
model. However, the implementation of Grey model requires a minimum of four
data points. The capability of the original G(1, 1) model in providing reasonably
good prediction has inspired further extension by using hybrid approach using
Bayesian network, which is currently on-going. It is believed that by producing
better predictions, effective risk based decision supports methodology can be
established to facilitate the maintenance planning and management of change.
References
1. Meel A, Seider WD (2006). Plant-Specific Dynamic Failure Assessment Using Bayesian

Theory. Chem Eng Sci 61:7036–7056
2. Rathnayaka S, Khan F, Amyotte P (2011). SHIPP methodology: Predictive accident modeling
approach. Part I: Methodology and model description. Process Saf Environ Prot 89:151–164
[email protected]
3. Rathnayaka S, Khan F, Amyotte P (2012). Accident modeling approach for safety assessment
in an LNG processing facility. J Loss Prev Process Ind 25:414–423
4. Meel A, O’Neill L, Levin J, Seider WD, Oktem U, Keren N (2007). Operational risk
assessment of chemical industries by exploiting accident databases. J Loss Prev Process Ind
20:113–127
5. Rathnayaka S, Khan F, Amyotte P (2011). SHIPP methodology: Predictive accident modeling
approach. Part II. Validation with case study. Process Saf Environ Prot 89:75–88
6. Al-shanini A, Ahmad A, Khan F (2014). Accident Modelling and Analysis in Process
Industries. J Loss Prev Process Ind 32:319–334. doi:10.1016/j.jlp.2014.09.016
7. Bradlow ET, Hardie BG, Fader PS (2002). Bayesian inference for the negative binomial
distribution via polynomial expansions. J Comput Graph Stat 11:1
8. Al-shanini A, Ahmad A, Khan F (2014). Accident Modelling and Safety Measure Design of a
Hydrogen Station. Int J Hydrogen Energy 39(35):20362–20370. doi:10.1016/j.ijhydene.2014.
05.044
9. Ju-Long D (1982). Control problems of grey systems. Syst Control Lett 1:288–294
10. Deng J-L (1989). Introduction to grey system theory. J Grey Syst 1:1–24
11. Tseng F-M, Yu H-C, Tzeng G-H (2001). Applied hybrid grey model to forecast seasonal time
series. Technol Forecast Soc Chang 67:291–302
12. Hsu L-C (2003). Applying the grey prediction model to the global integrated circuit industry.
Technol Forecast Soc Chang 70:563–574
13. Lin C-T, Yang S-Y (2003). Forecast of the output value of Taiwan's opto-electronics industry
using the Grey forecasting model. Technol Forecast Soc Chang 70:177–186
14. Chang N-B, Tseng C (1999). Optimal evaluation of expansion alternatives for existing air
quality monitoring network by grey compromise programing. J Environ Manage 56:61–77
15. Huang Y-P, Wang S-F (1997). The identification of fuzzy grey prediction system by genetic
algorithms. Int J Syst Sci 28:15–24
16. Zhou P, Ang B, Poh K (2006). A trigonometric grey prediction approach to forecasting
electricity demand. Energy 31:2839–2847
17. Hsu C-C, Chen C-Y (2003). Applications of improved grey prediction model for power
demand forecasting. Energy Convers Manag 44:2241–2249
18. Yao AW, Chi S, Chen J (2003). An improved grey-based approach for electricity demand
forecasting. Electr Power Syst Res 67:217–224
19. Hsu L-C, Wang C-H (2007). Forecasting the output of integrated circuit industry using a grey
model improved by the Bayesian analysis. Technol Forecast Soc Chang 74:843–853
20. Wang Y-F (2002). Predicting stock price using fuzzy grey prediction system. Expert Syst Appl
22:33–38
21. Kalantarnia M, Khan F, Hawboldt K (2009). Dynamic risk assessment using failure
assessment and Bayesian theory. J Loss Prev Process Ind 22:600–606
22. Lewis CD (1982) Industrial and business forecasting methods: a practical guide to exponential
smoothing and curve fitting. Butterworth Scientific, London
[email protected]
Index
A Construction noise, 392

Accident prevention, 372, 377, 378 Construction site, 397
Acid catalyst, 308, 312, 318, 322 Corrosion inhibition studies, 343
Acid hydrolysis, 124, 127–129, 310 Corrosion inhibitor, 340, 345
Activated carbon, 70, 116–120 COSMO-RS prediction, 31
Adsorption, 5, 8, 26, 34, 35, 37–39, 65, 71, Covalent organic framework-1, 302
106, 109, 111, 116–119, 168, 275, 292, CPI accident prediction, 412
302, 317, 340, 343 Critical radius, 141, 144
Adsorption isotherm, 305 Crystallinity, 302, 304, 305, 323, 332, 337
Adsorption process, 35, 71, 111, 168 Cyclic voltammetry, 186, 348
Agarwood, 169, 171, 173, 174
Alumina, 364, 367 D
Amine, 34–37, 52, 150 3-Dimensional, 254
Ammonia detection, 282 Dendrite, 228
Diethanolamine (DEA), 88, 89, 91, 92
B Digital image processing, 158
Batch, 97, 133, 136, 137, 207, 273, 275, 348 Doping, 274, 328
Bentonite, 169, 170, 172–174 Downdraft gasifier, 208, 209
Benzyltrimethylammonium hydroxide, 132, Drilling fluid, 178, 196–200
137 Dyes, 62, 70, 106, 107, 112, 274
Biodegradable, 5, 12, 150, 332
Biodiesel, 131, 134, 264, 317 E
Blending ratio, 209 Electron lone pair, 84, 85, 332
Emissions, 70, 116, 168, 214, 216, 217, 236,
C 238, 242, 392, 395, 397
Castor, 95, 96, 98 Encapsulation, 62–64, 66, 67
Carbon cryogel, 316, 317 Energy storage, 186, 192
Carbon steel, 82 Equipment failure, 372
Chemisorptions, 84 Esterification, 316, 317, 320
CO2, 34–36, 38, 39, 150, 168, 170, 214, 217 Ethylene-co-vinyl acetate (EVA), 88, 89, 91
Coagulant, 71, 72, 74, 76, 151 Expanded polystyrene (EPS), 364, 404, 405
Co-gasification, 204, 207, 208 Explosion efficiency, 382, 385
Cold finger, 89, 91, 92
Combustion, 26, 34, 116, 205, 207, 208, 212, F
236, 238, 241, 242, 264, 405 Fatty acid methyl esters, 137, 316
CoNiFe, 228, 229, 231, 232 Fermentation, 124, 178–182, 350

M.A. Hashim (ed.), ICGSCE 2014, DOI 10.1007/978-981-287-505-1
[email protected]
420 Index
First order grey model, 412 O

Fish scale, 71, 72 Oil palm empty fruit bunch, 120
Fluidization, 159 Oleic acid, 316, 317, 319
Fruit skin, 126, 129 One pot reaction, 308, 311, 312
Optimization, 102, 150, 246, 247, 255, 285
G Optosensor, 282, 283, 288
Gas permeation, 150, 153, 155, 169, 170 Organic corrosion inhibitor, 81
Geobacter sulfurreducens, 348–350, 352 Organic nano-crystals, 364
Ground rotten cabbage, 179 Oxygen content, 124
H P
H3PW12O4/ZrO2, 308 Palm kernel oil, 196–198, 200
Hardness, 229, 232, 233 Palm methyl ester, 196, 198
Heat treatment, 137, 228–230, 232 Palm oil, 12, 42, 43, 52, 55, 131, 196, 204
Hierarchy of control, 375, 377 Pandanaceae, 222, 225
Hydrolysates, 127, 128 Phase transfer catalysis, 132
Photocatalytic degradation, 108, 274, 275, 277
I Photodegradation, 107, 109, 112, 274, 277,
Impregnation, 35, 37, 38, 307, 308 279
Inhibition efficiency, 340, 343 Physisorption, 84, 85, 119, 120
Intensity, 65, 160, 164, 250, 264, 283, 304, 343 Poisson–gamma model, 412
Interfacial tension, 4, 141, 143 Polymer composite membrane, 169
Investigation effect, 385 Polymeric material, 335
Ionic liquid, 4, 26, 28, 31 Polymorphism, 145, 356
iPhone, 222, 225 Polysaccharides, 124, 127, 129, 178, 308
Iron, 70, 116, 274–277 Polysulfone, 169
Isododecane, 26, 27, 31 Porosity, 152, 153, 158, 228, 281, 292, 297,
302, 304, 316, 318
J Precursor data, 415
Jatropha curcas l, 132, 133, 137 Pressure swing adsorption (PSA), 35
Process control, 258, 372
K Process design, 246–248, 254, 256
Kenaf, 34, 36, 38, 39 PVA alginate, 62, 64, 65
Pyrrole, 26–28, 31
L
Lacey mixing index, 160, 164 R
Lignin-furfural, 318 Reduced graphene oxide, 186
Reflectance spectrophotometer, 283
M Response surface methodology, 102, 117
Mercury, 116–120, 235, 236, 238, 241, 242 Rheological test, 199
Methylbenzoylthiourea, 342, 343 Ricinoleic acid, 95
Microbial fuel cells, 348 Risk reduction, 372, 375, 378
Micro-crystalline cellulose, 125
Microwave irradiation, 132, 133, 137 S
Modelling, 236, 238 Segregation, 158, 162, 163
Simulated fuel gas, 116, 119
N Single chamber, 349
Nitrogen, 13, 26–28, 30, 37, 39, 117, 205, 214, Sol-gel, 274, 278, 282, 315, 320, 322
274, 276, 279, 301, 340 Solid acid catalyst, 42
Noise measurement, 395 Solvent extraction, 96, 98, 99, 102
Noise prediction, 392, 395, 397 Spectroscopic studies, 340
Non-covalent functionalization, 186, 192 Spherical polystyrene, 158, 159
Nucleation rate, 143, 144, 146 Static saline environment, 81
[email protected]
Index 421
Superabsorbent polymer, 12, 13, 17 V

Superabsorbent polymer composite, 12, 17 Visible light, 275, 277, 278
Sustainability analysis, 247, 250, 256, 259
W
T Wax deposition, 88, 90
Temperature profile, 204, 206, 208
Textile wastewater, 70–72 X
Thermal properties, 324, 364, 365, 367, 404, Xanthan gum, 178, 180–182
407, 408 Xanthomonas Campestris, 178, 179
Titanium dioxide, 106, 107, 109, 112, 274, 275 X-Ray diffraction, 63, 66, 302, 323, 325, 333,
TNT equivalency, 382, 383 358
Xylocarpus moluccensis, 62–64, 66
U
Ultrasonic cleaner, 62 Y
Unsaturated polyester resin (UPR), 364, 365, YDC, 179, 180, 182
404 γ-valerolactone, 307, 308, 310
[email protected]
View publication stats

ICGSCE2014

Uploaded by

Copyright:

Available Formats

ICGSCE2014

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ICGSCE2014

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Solvent Extraction of Castor Beans Oil: Experimental Optimization via

Chapter · June 2015

Jibrin Mohammed Danlami Agus Bin Arsad

SEE PROFILE SEE PROFILE

Muhammad Abbas Ahmad Zaini

Original Article View project

Investigation on Silver Nanoparticles in Water Environment View project

The user has requested enhancement of the downloaded file.

ISBN 978-981-287-504-4 ISBN 978-981-287-505-1 (eBook)

Library of Congress Control Number: 2015939164

Springer Singapore Heidelberg New York Dordrecht London

Printed on acid-free paper

On behalf of the editorial members, I would like to convey my heartiest congrat-

Abdul Aziz Ishak, Universiti Teknologi MARA

Dr. Nurul Fadhilah Kamalul Aripin, Universiti Teknologi MARA

Part I Green and Sustainable Technologies

Overview on Chemical-Based, Bio-based and Natural-Based

OPEFB Filler from Biomass in Superabsorbent Polymer Composite

Exploring Non-wood Plants as Alternative Pulps: From the Physical

Separation of Pyrrole from Isododecane Using Imidazolium

The Development of Amine-Based Potential Kenaf Sorbent

Potential Effect of Palm Oil Fuel Ash as Micro-Filler

Liquid Membrane Formulation for Succinic Acid Extraction

Preservation of Natural Textile Colorant from Xylocarpus

Coagulant from Chemically Modified Fish Scale for Textile

Part II Chemical and Biotechnology Engineering

An Investigation on N-Containing Organic Compounds

Evaluation on the Effect of Inhibitor Ratio to the Wax Deposition

Solvent Extraction of Castor Beans Oil: Experimental Optimization

Immobilization of Nano-sized TiO2 on Glass Plate for the Removal

Mercury Removal in Simulated Flue Gas by Oil Palm EFB Activated

Production of Glucose, Galactose and Mannose from the Skins

Reaction Kinetics Study for Microwave Energy Pretreated

Induction Time of L-Isoleucine Crystallization with the Presence

Gas Permeation Performance of Poly(lactic acid) Asymmetric

Particles Mixing in a Fluidized Bed by Using Digital Image

Polymeric Composite Membrane for CO2/CH4 Separation. . . . . . . . . . 167

Comparative Study Between Isolated Xanthomonas Campestris

Part III Energy and Renewable Energy

1-Pyrenebutyric Acid Functionalized Reduced Graphene Oxide

Clarifying the Palm-Based Drilling Fluids Potentials . . . . . . . . . . . . . . 195

The Study of Temperature Profile and Syngas Flare in Co-gasification

Evaluation of Energy Cost Saving and Pollutants Emission Reduction

Part IV Process Control, Modelling, Simulations, Data Mining

Introducing Pandanaceae Via IOS-Based Software Application . . . . . . 221

Investigation of Heat Treated Electrodeposited CoNiFe

Computational Fluid Dynamics Modeling of Mercury Emissions

3D Sustainability Analysis of Integrated Process Design

Three-Dimensional Sustainability Analysis for Integrated Process

Nucleation Kinetics of Carbamazepine-Saccharin

Part V Advanced Materials and Nanotechnologies

Activation of Titanium Dioxide Under Visible-Light by Metal

Reflectometric Optosensor for Visual Detection of Ammonia

Low Noise and Properties of Double Layer Concrete

Synthesis of Modified Covalent Organic Framework-1 (COF-1)

One Pot Synthesis of γ-Valerolactone from D-Glucose

Carbon Cryogel from Lignin-Furfural as Acid Catalyst

Surface Structure Study of CeZrO2 Nanocatalyst Doped

Electrical Properties of Ammonium Iodide Doped Cellulose