Bioresource Technology 199 (2016) 113120

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Supercritical fluids as a green technology for the pretreatment

of lignocellulosic biomass
L.V. Daza Serna a, C.E. Orrego Alzate a, C.A. Cardona Alzate a,
a
Instituto de Biotecnologa y Agroindustria, Universidad Nacional de Colombia at Manizales, Km 07 va al Magdalena, Colombia

h i g h l i g h t s

The effect of moisture content on supercritical pretreatment of rice husk was evaluated.

Delignification yields of supercritical pretreatment for rice husk were found at the selected conditions.

It was assessed and compared the crystallinity index of rice husk after supercritical and diluted acid pretreatments.

Supercritical and conventional pretreatment effects for enzymatic hydrolysis were compared.

Techno-economic and environmental performance for supercritical and conventional dilute acid pretreatment was developed.

a r t i c l e i n f o a b s t r a c t

Article history: One of the main drawbacks for using lignocellulosic biomass is related to its recalcitrance. The pretreat-
Received 30 June 2015 ment of lignocellulosic biomass plays an important role for delignification and crystallinity reduction
Received in revised form 21 September 2015 purposes. In this work rice husk (RH) was submitted to supercritical pretreatment at 80 C and 270 bar
Accepted 22 September 2015
with the aim to determine the effect on lignin content, crystallinity as well as enzymatic digestibility.
Available online 9 October 2015
The yields obtained were compared with dilute sulfuric acid pretreatment as base case. Additionally a
techno-economic and environmental comparison of the both pretreatment technologies was performed.
Keywords:
The results show a lignin content reduction up to 90.6% for the sample with 75% moisture content using a
Supercritical pretreatment
Lignocellulosic biomass
waterethanol mixture. The results for crystallinity and enzymatic digestibility demonstrated that no
Recalcitrance reductions were reached. Supercritical pretreatment presents the best economical and environmental
Crystallinity performance considering the solvents and carbon dioxide recycling.
Digestibility 2015 Elsevier Ltd. All rights reserved.

1. Introduction purposes. Second generation raw materials are mainly composed

of lignocellulosic materials (cellulose, hemicellulose and lignin rich
Many studies have demonstrated the potential use of biomass materials) produced during different extraction or transformation
as raw material for bioenergy, biofuels, food, feed, fiber needs as stages such as seeding, cropping, harvesting and processing
well as a wide family of biomolecules under biorefinery approach (Rincn et al., 2014). Other sources of second generation raw mate-
(Moncada et al., 2014). According to the information reported by rials are the non-edible crops. Finally, Microalgae are recognized as
the World Energy Council it is presumed that biomass can provide third generation raw materials (Ahmad et al., 2011).
about 16% of the expected energy consumption of 17208 Mtoe in Second generation raw materials do not directly compete with
2020 (World Energy Council, 2013). food production which is a positive attribute. These materials are
The term biomass is defined by ASTM as any material, exclud- organized in a structure in which lignin and hemicellulose form a
ing fossil fuel which was a living organism that can be used as a complex through covalent cross-linked bounds encapsulating and
fuel either directly or after conversion process (ASTM, 2002). This giving support to the cellulose. Despite the high availability of lig-
term includes a huge range of materials which can be classified in nocellulosic biomass the main drawback to be overcome for its
first, second and third generation raw materials. First generation practical use is related to biomass recalcitrance. Recalcitrance
raw materials refer to edible crops used for food or agribusiness refers to an arrange of features (epidermal plant issues, structural
heterogeneity, constituent complexity and degree of lignifications)
Corresponding author. Tel.: +57 (6) 8879300x55354. which plays a structural protection role against microbial attack
E-mail address: [email protected] (C.A. Cardona Alzate).
(Himmel et al., 2007). Another important aspect contributing to

http://dx.doi.org/10.1016/j.biortech.2015.09.078
0960-8524/ 2015 Elsevier Ltd. All rights reserved.
114 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120

the biomass recalcitrance is represented by the cellulose crys- (above 100 C) and the biomass recovering costs. Other types of
tallinity. Cellulose crystallinity provides to biomass fibers resis- pretreatments have been explored as alternatives to reduce opera-
tance against chemical and biochemical attack. tional conditions in terms of reactions severity, residence time,
equipment deterioration and its associated costs. Supercritical pre-
1.1. Biomass pretreatments treatment (L et al., 2013) among others are innovative proposals
recently discussed. This paper studies the effect of supercritical
A number of different processes exist to reduce the recalcitrance pretreatment as a green technology (based on an inexpensive
and cellulose crystallinity of lignocellulosic biomass. Numerous solvent as CO2) on the cellulose crystallinity and lignin removal
pretreatment technologies are used to make value-added products in rice husk. In this sense the description of the effect is presented
from biomass including physical, chemical, hydrothermal and below as specific introduction to this work.
biological methods. This topic is so widely covered in open source
literature, that it is difficult to establish all the primary authors
for each method. 1.2. Supercritical pretreatment
Physical pretreatments are not widely used given the limited
effectiveness. However if some chemicals are added as for example A supercritical fluid (SF) is any substance above critical condi-
sodium azide together with a buffer solution it is possible to tions (temperature and pressure), exhibiting properties of liquids
decrease the crystallinity of raw materials through size particle such as density and gases such as compressibility. Carbon dioxide
reduction. Kim et al. (2013) reported crystallinity reduction in is the most used supercritical fluid due to its non-toxic, recyclable,
orders of 32% until 78% using this approach. Chemical pretreat- low-cost, environmentally friendly characteristics (considering it
ments consider the use of acid, alkaline and organic solvents in a closed loop as a solvent), and low critical temperature and
(organosolv) solutions to increase the raw materials digestibility. pressure (31.1 C and 7.36 MPa respectively).
A summary of typical effects and conventional reagents used in Supercritical and subcritical fluids have been used in
chemical pretreatments is presented in the Table 1. pretreatments since seventies as was reported by Bludworth and
Hydrothermal pretreatments (Zhang et al., 2012) involve higher Knopf (1993) and Reyes et al. (1989) among others. It is known
temperatures with the aim to fractionate the hemicellulose as well the delignification effect produced by this kind of pretreatments
as to increase the material porosity. Steam explosion, steam treat- in different types of raw materials such as corn stover
ment and liquid hot water treatment are included in this pretreat- (Narayanaswamy et al., 2011), switchgrass (Narayanaswamy
ment category. These pretreatments use temperatures in the range et al., 2011), rice straw (Gao et al., 2010), sugarcane bagasse
of 160240 C. The main effect observed is related to the hemicel- (Pasquini et al., 2005a) among others (Pasquini et al., 2005b). The
lulose hydrolysis and lignin breakdown (Ruiz et al., 2008). The most important parameters taken into account are the moisture,
shorter reaction times can be recognized as main advantage of temperature, pressure and pretreatment time. Water in the bio-
these types of pretreatments. Nevertheless, pH value represents mass combined with supercritical carbon dioxide generates a car-
an important constraint: for liquid hot water (LWH) the pH should bonic acid mixture which generates a kind of weak acidic
be restricted to 4.07.0 to avoid monomeric sugars release during environment preliminary promoting the hemicellulose hydrolysis
this pretreatment stage (Ruiz et al., 2008; Kim et al., 2009). and intensifying the mass transfer conditions (Narayanaswamy
Finally, biological pretreatments use different fungi species et al., 2011). Additionally, when water is mixed with other type of
such as white-rot, brown-rot and soft-rot to produce enzymes solvents the delignification capability is enhanced; for instance,
degrading lignin and cellulose. Among these enzymes it is ethanol improves the solubility of lignin fragments (L et al., 2013).
identified lignin peroxidase, lacasse, manganese peroxidase According to authors Pasquini et al. (2005a,b) high pressure
(Wang et al., 2013). However, long times (from days to weeks) enhances the interaction between fluids or solvent mixtures and
are required for these methods to get the desired impacts. lignocellulosic biomass increasing and promoting the rupture of
Despite the higher yields obtained through conventional pre- the bonds connecting lignin with the lignocellulosic matrix. In
treatments, a huge range of disadvantages are presented. Among addition, authors as Pasquini et al. (2005a,b) and Puech et al.
them, it is possible to highlight the release of inhibitory com- (1990) concluded that short reaction times of 3015 min including
pounds such as furfural, hydroxymethyl furfural, acetic acid among heating or pressurizing were enough to obtain good delignification
others, the higher energetic costs associated to high temperatures results in presence of water:ethanol mixture.
The aims of the present study were to evaluate the effects of using
Table 1 supercritical carbon dioxide pretreatment of RH and to perform the
Chemical pretreatments summary.
techno-economic and environmental assessment of the process. To
Pretreatment Reagents Treatment effects achieve this, sulfuric dilute acid was considered as the base case to
Alkalinea Sodium hydroxide, potassium Lignin removal compare the composition and digestibility of RH after pretreatment.
hydroxide, calcium hydroxide, All the operational conditions were selected taking into account the
ammonium hydroxide, lime most sensitive parameters. In this case short pretreatment times as
among others well as relatively moderate temperatures and pressures were used.
Dilute acidb Sulfuric acid, hydrochloric Hemicellulose fractionation
acid, nitric acid, phosphoric
Both configurations (supercritical pretreatment and the base case)
acid among others were simulated using commercial simulators and programming
Organosolvc Ethanol, acetic acid, formic Lignin removal and packages (mainly Aspen Plus and MatLab) to calculate the energy
acid, peracetic acid with hemicellulose fractionation and mass balances used to assess the production costs (based on
organic and inorganic
the Colombian context) and the environmental behavior.
catalysts
Ionic liquidsd Anions from chloride, Cellulose crystallinity
formate, acetate or reduction and partial
alkylphosphonate hemicellulose and lignin
2. Methods
removal
a
2.1. Raw material
Park and Kim (2012).
b
Chandel et al. (2012).
c
Amiri et al. (2014). RH is a non edible agricultural residue generated during the
d
Li et al. (2010). whole grain de-husking process. Approximately 0.23 ton of RH
 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120 115

can be generated per ton of processed rice (Soltani et al., 2014). RH CO2 is put in contact with the sample in the extraction or reaction
samples from Oryza sativa variety were obtained from a rice mill chamber (equipment 3, made of stainless steel of 10,000 psi). In this
company located at Tolima region (central zone of Colombia, equipment approximately 20 g of RH get in contact with the super-
2 520 5900 and 5 190 5900 N, and 74 240 1800 and 76 060 2300 W) critical carbon dioxide. This reactor is immersed in a water bath at
between the Magdalena and Cauca rivers. The RH was cut and the desired temperature (atmospheric water bath that can reach
sieved using a US Standard Sieve No. 4 (4.75 mm). The samples 93 C). The CO2 then endures a pressure drop while it is conducted
were not dried and stored at atmospheric conditions due to the to the collector (equipment 4). In this stage due to the pressure drop
low equilibrium moisture content (11%). the CO2 returns to gaseous state (all the solid and liquid compounds
extracted or dragged by the CO2 are kept in the collector) and then it
2.2. Experimental characterization can be recycled or purged returning or leaving the system practi-
cally clean, with no additional recovering steps (Smith et al., 1999).
RH was chemically characterized to determine its extractives, The samples were submitted to supercritical pretreatment. The
cellulose, hemicellulose, lignin and ash content. The samples were conditions considered were 80 C, 270 bar with a pretreatment
dried using a convective furnace at 50 C and then milled to reduce time of 10 min. The moisture content was adjusted to desired
the particle diameter below 0.85 mm. The initial moisture content levels by spraying a calculated amount of water or cosolvent
was determined using a Shimadzu Moisture Balance MOC-120H. mixture according to the following conditions and notations:
The extractives determination were developed following the ASTM
Standard Test Method E 1690 Determination of Ethanol Extrac- (i) M1: moisture content 25% and water as selected cosolvent.
tives in Biomass (Han and Rowell, 1995) by using ethanol at (ii) M2: moisture content 75% and water as selected cosolvent.
45 C for 24 h. Holocellulose (cellulose and hemicellulose fractions) (iii) M3: moisture content 25% and water:ethanol mixture as
was determined through the chlorination method described by selected cosolvent.
Rabemanolontsoa and Saka (2012). The alpha-cellulose fraction (iv) M4: moisture content 75% and water:ethanol mixture as
was determined from the holocellulose residue through sodium selected cosolvent.
hydroxide and acetic acid treatment (Han and Rowell, 1995). The
total lignin (soluble and non-soluble fractions) was determined The ratio of the mixture water:ethanol was 2:1. It was selected
by acid treatment with H2SO4. Finally the ash content was deter- based on the literature (Puech et al., 1990; Pasquini et al., 2005a,b).
mined by sample ignition at 575 C according to the TAPPI stan- The ethanol used in experiments was reactive grade (96%)
dard T211 (Han and Rowell, 1995). All the experimental purchased from Panreac.
procedures were carried out by triplicate.
2.4. Dilute acid pretreatment
2.3. Supercritical pretreatment
The base case considered for this study was the pretreatment
All the experiments were performed in a laboratory unit, con- with dilute sulfuric acid (2% v/v) as has been reported by Martin
sisting of a line for conditioning CO2 composed by a cooler and a et al. (2007). The pretreatment was developed at 120 C with a
pump, a reaction chamber and a collector vessel The Fig. 1 shows solid to dilute acid ratio 1:10 (w/w) during 1 h. After pretreatment,
a schematic representation of the unit used. the solid and filtrate fraction were separated by vacuum filtration.
The CO2 is stored in a cylinder, according to the scheme pre- The solid fraction was washed twice using warm distillated water.
sented in the Fig. 1. The gas is passed through a cooler system
(equipment 1) using ethylene glycol as cooling fluid. In this stage 2.5. Structural characterization
the CO2 is carried to liquid state at 20 C. Then, the liquid CO2 is
pressurized by a pneumatic pump until the desired pressure To evaluate the effect as well as the extent of pretreatment a
(equipment 2, a Dual Seal pneumatic plunger pump WILLIAMS chemical characterization and crystallinity determination were
Chemical Metering Pumps.) The maximum endured pressure is performed for the pretreated samples (including supercritical
350 bar and the reactor capacity is 254 mL. Later on, the pressurized pretreated samples and the dilute acid pretreated sample) to be

Fig. 1. Supercritical pretreatment device scheme.

116 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120

compared with the initial sample. The delignification yield was the systems first unit operation (i.e., heat exchanger, compressor,
calculated according to the expression of Eq. (1): reactor etc.) and continues through the process, illustrating all
intermediate unit operations and the interconnecting streams.
Lignin1  Lignin2
Delignification yield  100% 1 When the calculations are complete, the software lists the results,
Lignin1 stream by stream and unit by unit, so it is possible to know not
where the sub index 1 and 2 represents the initial and final lignin, only the change in species and also on properties of the flows
before and after each pretreatment respectively. including energy.
Some physicochemical properties of non-included compounds
were obtained from the National Institute of Standards of Technol-
2.5.1. Chemical composition
ogy. Aspen Plus database for biofuels components developed by
The pretreated raw materials were submitted to re-
the National Renewable Energy Laboratory was used for properties
characterization processes to determine the new distribution of
specifications. The thermodynamic models used for simulation
cellulose, hemicellulose and lignin. For this purpose the used
were the Non-Random Two-Liquid (NRTL) for activity calculations.
methods were the same for the initial characterization above men-
The total production costs were estimated using the commercial
tioned. Furfural production was determined by spectrophotometry
software Aspen Process Economic Analyzer V8.2 (Aspen Technol-
methods at 277 nm of wavelength (Chi et al., 2009).
ogy Inc., USA) assuming the depreciation of capital for 12-year per-
iod (8000 h/period) beneath the straight-line method. The
2.5.2. Crystallinity
economic parameters such as income taxes, interest rate, labor sal-
X-ray diffraction (XRD) is a technique widely used for crys-
aries, electricity and water costs corresponded to the Colombian
tallinity characterization of lignocellulosic biomass and composites
context. From this stage, the capital costs as well as the operating,
(Tolosa et al., 2014). This characterization was performed using a
raw materials, utilities, equipment and other general and adminis-
RIGAKU MINIFLEX II diffractometer, using Cu-Ka radiation at
trative costs were estimated.
30 kV and 15 mA with a scan rate of 5 (2h) min1, a sampling
width of 0.02 (2h), and over a range between 5 and 50 (2h).
2.8. Environmental assessment
The crystallinity index defined as the crystalline to amorphous
ratio was calculated based on the method proposed by Segal
The environmental evaluation was made through the method-
et al. (1959) according to Eq. (2)
ology proposed by the Environmental Protection Agency (EPA)
I002  Iam denominated Waste Algorithm Reduction (WAR). This tool evalu-
Crystallinity index %  100 2
I002 ates from the mass and energetic balance the potential environ-
mental impact (PEI) of process generated by kilogram of product.
I002 represents the maximum intensity (measured in arbitrary units)
The bounds of the systems were defined from the inlet to the outlet
of the diffraction from the plane at 2h  22.6 for cellulose I and Iam
plant door, in this sense the PEI contributions of the manufacture
represents the intensity of the background scatter measured at
process for each input were ignored in this evaluation. The selected
2h  19.
product stream was the dilute glucose stream while the other
streams were considered as non-product streams.
2.6. Enzymatic hydrolysis

3. Results and discussion

The pretreated samples were submitted to enzymatic hydroly-
sis with the aim to compare the saccharification performance for
3.1. Raw material characterization
each case (pretreated and non-pretreated samples). The protocol
followed was based on the Laboratory Procedure 009: Enzymatic
The experimental results for RH characterization are consigned
Saccharification of Lignocellulosic Biomass developed by the
in the Table 2. It is confirmed that RH is mainly composed by cel-
National Renewable Energy Laboratory (Selig et al., 2008). The
lulose, hemicellulose and lignin. Despite the fact that the ash con-
samples were hydrolyzed by the Celluclast 1.5 L purchased from
tent is lower than the those reported by other authors in the
Novozymes which is a cellulose cocktail. Additionally a supple-
Table 2, this amount is higher in comparison to lignocellulosic resi-
mentary beta-glucosidase used to enhance the hydrolysis. The
dues such as sugarcane bagasse, empty fruit bunches and coffee
samples were dissolved in sodium citrate 0.1 M (pH 4.8). The enzy-
cut stems whose ash contents are 4.17%, 1.60% and 2.27% respec-
matic hydrolysis was carried out at 50 C and 100 rpm during 72 h.
tively (Quintero et al., 2013). Nevertheless, the same compositional
After enzymatic hydrolysis the sugar content was determined in
trend for cellulose, hemicellulose and lignin content was observed.
each sample by the colorimetric DNS method at 540 nm of wave-
Differences in compositions are characteristic problems for
length. A standard curve was prepared using anhydrous glucose
these feedstocks, given the diversity of soils, harvesting methods,
(Wood et al., 2012).
characterization methods and even conservation or standardiza-
tion of the storage. In this case the characterization was made by
2.7. Techno-economic assessment
triplicate, ensuring the quality of the data fed to the simulations.

With the aim to contrast the techno-economic performance of

the pretreatments analyzed in this work the processing of Table 2
10 ton/h of RH were assessed. The simulations included the com- RH physicochemical composition.
parison of the dilute acid and supercritical pretreatment until the Feature This Quintero et al. Blandino et al.
saccharification stage. Energy, economic and environmental work (2013) (2014)
assessment were performed. For simulation purposes the biomass Extractives(db) (%) 71 3.35 N.A
load and the experimental yields for pretreatment stage (supercrit- Cellulose(db) (%) 40 2 26.45 36
ical and base case) and the enzymatic hydrolysis were used. The Hemicellulose(db) (%) 16 3 27.29 12
Lignin(db) (%) 26 7 28.03 16
mass and energy balance were calculated in Aspen Plus V8.2
Ash(db) (%) 11 1 14.89 20
(Aspen Technology Inc., USA) based on models for every step of
the flowsheet showing one or more inlet streams entering into db: Dry basis.
 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120 117

3.2. Supercritical pretreatment pretreatment were kept between 34.4% and 90.6% which are con-
sidered very similar to the average values reported for this type
The results for holocellulose content (moisture and extractives of pretreatment (Pasquini et al., 2005a,b). This means that the
free) gave extents of 78.7 3.88%, 81.23 1.72%, 83.58 4.54%, characteristics of the pretreated materials can be modulated by
96.55 3.57% and 73.87 1.99% for M1, M2, M3, M4, and M6 different variables in the supercritical pretreatment (type of sol-
respectively. Holocellulose content of pretreated solids increases vents and cosolvents, ratios, temperature, pressure among others)
compared to initial raw material (68.00 5.00%) because of the to get the preferred effect.
delignification effect. For supercritical pretreatment the moisture
content has minor impact using only water as solvent (M1 and 3.3. Crystallinity index
M2 samples) compared to the use of waterethanol mixture as
pretreatment solvent (M3 and M4). However, among the samples The unpretreated sample (M5) as well as the acid dilute pre-
M1, M2 and M3 there are not statistical differences due to the error treated sample (M6) and the supercritical pretreated sample
values. (M4) which exhibited the best delignification performance were
It was observed that lignin content was altered by varying the submitted to X-ray diffraction analysis to determine the crys-
solvents amounts for each treatment. The supercritical pretreated tallinity index. The obtained crystallinity index values were 31%,
samples present lower lignin content in each case considering only 33% and 35% for M4, M5 and M6 respectively.
one solvent or a solvent mixture. The higher crystallinity index was presented by diluted acid,
For supercritical pretreatment the results indicated that deligni- followed by the unpretreated sample and last the supercritical
fication can be controlled by varying the solvent characteristics, sample. This fact allows to conclude that supercritical pretreat-
particularly the water:ethanol content. The sample M4 presented ment provides the best performance of this pretreatment accord-
the highest holocellulose content meaning that ethanol presence ing to crystallinity modification. A similar behavior to dilute acid
enhances the delignification process. This effect can be interpreted was reported by Srinivasan & Ju, in which pretreated samples
as a kind of organosolv pretreatment using ethanolwater as showed higher crystallinity index compared to unpretreated ones.
solvent mixture catalyzed by some carbonic acid which could be These authors concluded that even if the pretreatments were cap-
formed due to a reaction between the carbon dioxide and water able to break the lignin bounds, they were not able to modify the
present. Additionally, the ethanol as polar co-solvent (with polar crystallinity of the cellulose (Srinivasan and Ju, 2012). Others
characteristics) can promote the pretreatment based on the authors as Fazelin et al. affirmed that the increase of crystallinity
cleavage of the lignocellulosic matrix. The last statements agree is due to the removal of lignin and hemicellulose existing in the
with those discussed by other authors (L et al., 2013; Puech amorphous region. The residual cellulose corresponds to the crys-
et al., 1990). Moreover, an intensification effect on the hydrolysis talline fraction which could imply a cellulose purification (Fazelin
reactions should be expected. et al., 2014).
Residual lignin contents of dilute acid pretreatment can become
in a digestibility problem due to the capability that those residues
have to depositing on the crystalline cellulose surface as has been 3.4. Enzymatic hydrolysis
defined by Li et al. (2010).
For this case the temperature and pressure were kept constant The Table 3 shows the results for enzymatic hydrolysis after
in 80 C and 270 bar considering the specifications of the used pretreatments. Considering that endoglucanases, exoglucanases
laboratory equipment. The literature reports huge ranges of and b-glucosidases develop specific roles in enzymatic hydrolysis,
temperatures and pressures of 100198 C and 7300 bar (Gao the lack of one of them would affect the hydrolysis yields. In this
et al., 2010; L et al., 2013; Narayanaswamy et al., 2011; case, the hydrolysis was provided in the first case just with
Pasquini et al., 2005a,b). Indeed, in those works the increase of endoglucanases (to attack the amorphous cellulose) and exoglu-
the pressure gave higher delignification yields and an increase in canases (to attack the tight crystals) enzymes. This fact together
surface area was also observed (Narayanaswamy et al., 2011). with the overall high crystallinity indexes of the hydrolyzed sam-
Nevertheless those authors reported a degradation of cellulosic ples are the main reasons of the low yields (g reducing sugars/
matrix at high temperatures values (above 190 C). g cellulose) observed in the Table 3. According to Singh et al.
On the other hand, furfural concentration presented different (2014) low enzymatic yields could be due to the cellobiose accu-
orders of magnitude for each pretreatment. The concentrations mulation. In absence of b-glucosidases enzymes it is possible that
were 0.002 g/l and 1.015 g/l for supercritical (M4) and dilute acid cellobiose accumulation cause severe feedback inhibition to the
(M6) respectively. Those amounts are related to the severity factor cellulase reactions. Similar yields were reported by Wei et al.
obtained due to the temperatures and pretreatment times. (2009) for rice hulls pretreatment in absence of b-glucosidase
The delignification yields were calculated according to the enzymes. The concentration range of reducing sugars reported in
Eq. (1). The obtained results using water as cosolvent were their work was 0.38 0.073.37 0.1 g/L. However, for compar-
34.4% 5.4 and 40.6% 3.67 for M1 and M2 respectively. When ison purposes these yields are enough to understand the positive
using ethanolwater mixture as cosolvent the delignification or negative effect of the pretreatments and the capability of each
yields reached 50% 4.54 and 90.6% 2.39 for M3 and M4 respec-
tively. The yield for dilute acid pretreatment was 18.8% 4.12 Table 3
which is significantly lower than all the supercritical values. These Enzymatic hydrolysis results.
results agree with the results obtained by other authors for raw
Sample Reducing sugars (g/L) Yield (g sugar/g cellulose)
materials as sugarcane bagasse (Pasquini et al., 2005a), Pinus taeda
M1 1.56 3.89%
wood chips (Pasquini et al., 2005b) where the delignification
M2 1.82 4.54%
yields were kept between 50.5% and 94.6% respectively. These M3 2.54 6.35%
values were obtained using different co-solvent composition as M4 2.87 7.16%
well as pressure and temperature variations. In the cited works, M5 1.65 4.11%
for all the cases the supercritical carbon dioxide pretreatment rep- M6 2.60 6.50%

resents an improvement in delignification process. The obtained in M1: 25% water, M2: 75% water, M3: 25% waterethanol, M4: 75% waterethanol,
this work delignification results for supercritical carbon dioxide M5: unpretreated, M6: dilute acid.
118 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120

pretreatment and pretreatment conditions to enhance the deligni- for dilute acid and supercritical pretreatment cases respectively by
fication or cellulose digestibility. b-glucosidase addition. It is reported in literature a conventional
Comparing the samples pretreated with supercritical CO2, it pilot or near to industry cases (Davis et al., 2013) costs close to
was evidenced a slightly difference between the moisture content 0.27 USD$/kg of sugar. Last show a possible feasibility, but addi-
levels. However the samples pretreated using only water to adjust tional research is needed. Once again the tendency of supercritical
moisture content presented lower yields compared to the samples pretreatment is to be lower in costs than dilute acid.
pretreated with ethanolwater mixture. The sample M1 produced The scale (fed flow of RH to be transformed) has a huge
lower sugar yield compared to the unpretreated sample while the influence on the sugars production costs. For the case of dilute acid
M2 was slightly higher meaning that for this case the only use of pretreatment the effect caused by the higher amounts of inputs
water did not enhanced the enzymatic conversion in comparison represents the main disadvantage. For the supercritical pretreat-
to unpretreated cases. This fact allows concluding that a kind of ment, the highest impact was on the utilities costs, due to the
enhancement for enzymatic yields is provided by the mixture car- energy needed for CO2 conditioning. As a positive aspect of this
bon dioxide as solvent and ethanol as cosolvent. The sample M4 technology is the possibility of recycling the solvent and cosolvent
presented the highest conversion yields (7.16%) notwithstanding for delignification purposes. The evaluated features present differ-
this value is a bit higher than conventional pretreated sample. ent results showed easily in the Fig. 2.
Two additional enzymatic hydrolysis were performed to M4 In the Fig. 2 in the raw material costs, for the dilute acid pre-
and M6 samples using Celluclast 1.5L and Viscozyme (cellulolytic treatment there are involved reagents that cannot be recycled or
enzyme mixture which includes b-glucosidase enzymes) donated reprocessed (acid and reagents for detoxification). In the supercrit-
by other lab. These samples were selected considering the holocel- ical pretreatment the reuse of solvents is not so difficult (carbon
lulose content and the changes performed to crystallinity index. dioxide and waterethanol recycling). Total utilities costs are lar-
The use of this enzyme mixture increased the conversion yields ger for supercritical pretreatment due to the operational conditions
up to 55% 1.99 and 53% 0.56 for M4 and M6 respectively con- (high pressures and energy required for carbon dioxide liquefac-
firming the importance of b-glucosidase in saccharification pro- tion) nevertheless the raw materials costs and the general and
cesses. It is important to note that the addition of b-glucosidase operating costs are higher for dilute acid pretreatment due to the
to the samples with best performances improves clearly the inputs and the devices used in this pretreatment associated to
digestibility, based on the cellobiose conversion to glucose, but additional stages (detoxification and biomass recovering). The sim-
the obtained results preserve the same trends reported before for ulation doesnt consider (due to the complex task to include it into
the celluclast case. simulation) the recovering of the residual energy contained in car-
bon dioxide (still at high pressures that can be used for example in
3.5. Techno-economic results a turbine to generate electricity) leaving the chamber after pre-
treatment. This would represent perhaps, the main challenge to
The Table 4 presents the techno-economic results per period. It improve the economic performance of this type of nonconven-
considers the inputs, raw materials, the total utilities costs and tional pretreatment.
other administrative costs associated to the processes considering
saccharification yields for M4, M4, M6 and M6. Initially, the pre- 3.6. Environmental results
treatment, detoxification and enzymatic hydrolysis stages were
evaluated for the base case (dilute sulfuric acid). For the second The environmental assessment was made to determine in a
case were evaluated the pretreatment (including the supercritical comparative scenario the potential environmental impact (PEI) of
carbon dioxide and 2:1 waterethanol solvent mixture as well as each pretreatment. The PEI is related to the effect that a deter-
the recovering stages) and lately, the enzymatic hydrolysis stage. mined material or process would have on the environment, being
However, even if the costs are high based on the selected flow an interesting tool and approximation to understand the environ-
(10 ton/h) and technologies used, it is clear that the best results mental efficiency of the process. The PEI is measured by eight
can be found for supercritical pretreatment. It is important to note dimensionless impact categories: HTPI (Human Toxicity Potential
that generally, the high costs presented in Table 4 are related to the by Ingestion), HTPE (Human Toxicity Potential by Exposure), ATP
sugar yields obtained in the experiments. Based on the 10 tons/h (Aquatic Toxicity Potential), TTP (Terrestrial Toxicity Potential),
case the costs were reduced until 1.88 and 0.20 USD$/kg of sugars GWP (Global Warming Potential), ODP (Ozone Depletion Potential)
and AP (Acidification Potential). The results presented in the
Table 5 were classified in PEI leaving the system and PEI generated
by the system per kilogram of product generated.
Table 4 According to the Table 5 the dilute acid pretreatment as process
Techno-economic results. generates the highest PEI. The major contributions are based on the
Feature Acid Supercritical toxicity potential by ingestion and the terrestrial toxicity potential.
pretreatment pretreatment The HTPI is related to the lethal dose that produce death in 50% of
(USD/period) (USD/period) rats by oral ingestion meaning that are needed lower amounts of
Total raw materials cost 34,400,000 3,165,400 products or wastes produced in dilute acid pretreatment to affect
Total utilities cost 468,260 202,665 a determinate population in comparison with the supercritical pre-
Operating labor cost 85,680 85,680 treatment. Thereby the values for HTPI and TTP (both toxicological
Maintenance cost 130,000 53,000
Operating charges 21,420 21,420
potentials) should be explained by the nature of inhibitory com-
Plant overhead 107,840 69,340 pounds released by this conventional pretreatment (HMF, furfural
G and A cost 2,820,970 287,800 and acids) and the residues formed during the detoxification stage.
DEP (depreciation expense) 723,699 440,398 The Table 5 also presents the results of PEI generated per each
Total (USD/period)a $ 38,638,045 $ 4,325,703 kilogram of product generated. The PEI leaving the system is
Production cost (USD/kg)a $ 15.32 $ 1.56 similar in magnitude than the generated. It means that the major
Production cost (USD/kg)b $ 1.88 $0.20
contribution to the PEI is represented by the pretreatment
a
Based on experimental yields from the present work without b-glucosidase. technology mainly by the different waste streams generated in
b
Based on experimental yields from the present work using b-glucosidase. every pretreatment technology.
 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120 119

100%
90%
80%
70%
60% General and Operating Costs
50% Total Utilities Cost
40% Total Raw Materials Cost
30%
20%
10%
0%
Acid Pretreatment Supercritical
Pretreatment

Fig. 2. Cost distribution.

Bludworth, J., Knopf, F.C., 1993. Reactive extraction of lignin from wood using
Table 5 supercritical ammoniawater mixtures. J. Supercrit. Fluids 6, 249254.
Potential environmental impact-PEI. Chandel, A.K. et al., 2012. Dilute acid hydrolysis of agro-residues for the
depolymerization of hemicellulose: state-of-the-art. In: da Silva, S.S., Chandel,
PEI generated PEI leaving A.K. (Eds.), D-Xiylitol: Fermentative Production, Application and
Acid Supercritical Acid Supercritical Commercialization. Springer, Berlin Heidelberg, Berlin, Heidelberg, <http://
link.springer.com/10.1007/978-3-642-31887-0> (accessed 20.11.13).
HTPI 2.37E01 2.54E05 2.55E01 9.65E04 Chi, C. et al., 2009. Determination of furfural and hydroxymethylfurfural formed
HTPE 1.62E02 7.97E07 5.90E03 1.88E04 from biomass under acidic conditions. J. Wood Chem. Technol. 29 (November
TTP 2.37E01 2.54E05 2.55E01 9.65E04 2013), 265276.
ATP 8.67E04 1.07E05 5.49E04 1.16E05 Davis, R., et al., 2013. Process Design and Economics for the Conversion of
GWP 1.74E04 6.20E05 1.74E04 6.20E05 Lignocellulosic Biomass to Hydrocarbons: Dilute-Acid and Enzymatic
ODP 6.59E10 2.35E10 6.59E10 2.35E10 Deconstruction of Biomass to Sugars and Biological Conversion of Sugars to
PCOP 1.27E07 2.34E04 1.27E07 1.24E05 Hydrocarbons Process Design and Economics for the Conversion, Denver, USA.
AP 1.80E03 6.43E04 1.80E03 6.43E04 Fazelin, N. et al., 2014. Preparation and characterization of cellulose and
nanocellulose from Pomelo (Citrus grandis) Albedo. J. Nutr. Food. Sci. 5 (1),
Total 4.59E01 4.30E04 5.18E01 2.85E03 1013.
Gao, M. et al., 2010. Effect of SC-CO2 pretreatment in increasing rice straw biomass
conversion. Biosyst. Eng. 106 (4), 470475.
4. Conclusions Han, J.S., Rowell, J.S., 1995. Chemical composition of fibers. In: Rowell, R., Young, R.,
Rowell, J. (Eds.), Paper and Composites from Agro-based Resources. CRC Press,
Taylor & Francis Group.
One important aspect evidenced in this work corresponds to the Himmel, M.E. et al., 2007. Biomass recalcitrance: engineering plants and enzymes
different effects of pretreatments on lignin content, crystallinity for biofuels production. Science (New York, NY) 315, 804807.
and digestibility of the rice husk. Supercritical pretreatment Kim, H.J. et al., 2013. Environmentally friendly pretreatment of plant biomass by
planetary and attrition milling. Bioresour. Technol. 144, 5056, <http://www.
presents the best economic and environmental performance. ncbi.nlm.nih.gov/pubmed/23867527> (accessed 31.07.14).
However, different drawbacks should be overcame such as residual Kim, Y. et al., 2009. Liquid hot water pretreatment of cellulosic biomass. In: Mielenz, J.
energy use to reduce costs. The perspectives of the supercritical (Ed.), Biofuels: Methods and Protocols. Humana Press, Oak Ridge, USA, pp. 93102.
Li, C. et al., 2010. Comparison of dilute acid and ionic liquid pretreatment of
pretreatment using the carbon dioxide as a green solvent for switchgrass: biomass recalcitrance, delignification and enzymatic
continuous processing of biomass depends on the results for other saccharification. Bioresour. Technol. 101 (13), 49004906, <http://www.ncbi.
lignocellulosic residues with different composition characteristics. nlm.nih.gov/pubmed/19945861> (accessed 13.07.14).
L, H. et al., 2013. Pretreatment of corn stover using supercritical CO2 with water
Then it is possible to understand the real possibilities of this
ethanol as co-solvent. Chin. J. Chem. Eng. 21 (5), 551557. http://dx.doi.org/
pretreatment in the future. 10.1016/S1004-9541(13)60508-X.
Martin, C. et al., 2007. Dilute sulfuric acid pretreatment of agricultural and agro-
industrial residues for ethanol production. Appl. Biochem. Biotechnol. 136,
Acknowledgements 339352.
Moncada, J., Tamayo, J.A., Cardona, C.A., 2014. Integrating first, second, and third
The authors express their gratitude to the Universidad Nacional generation biorefineries: incorporating microalgae into the sugarcane
biorefinery. Chem. Eng. Sci. 118, 126140.
de Colombia at Manizales, to the Biotechnology and Angroindustry Narayanaswamy, N. et al., 2011. Supercritical carbon dioxide pretreatment of corn
Institute, to the Nano-structured and Functional Materials Labora- stover and switchgrass for lignocellulosic ethanol production. Bioresour.
tory, as well as to COLCIENCIAS for the financial support of this Technol. 102 (13), 69957000.
Park, Y.C., Kim, J.S., 2012. Comparison of various alkaline pretreatment methods of
research through the program Jvenes Investigadores, Convocato- lignocellulosic biomass. Energy 47 (1), 3135, <http://linkinghub.elsevier.com/
ria 617 de 2013. retrieve/pii/S0360544212006226> (accessed 5.08.14).
Pasquini, D., Pimenta, M.T., et al., 2005a. Extraction of lignin from sugar cane
bagasse and Pinus taeda wood chips using ethanolwater mixtures and carbon
References dioxide at high pressures. J. Supercrit. Fluids 36, 3139.
Pasquini, D., Pimenta, M.T.B., et al., 2005b. Sugar cane bagasse pulping using
Ahmad, A.L. et al., 2011. Microalgae as a sustainable energy source for biodiesel supercritical CO2 associated with co-solvent 1-butanol/water. J. Supercrit.
production: a review. Renewable Sustainable Energy Rev. 15 (1), 584593. Fluids 34, 125131.
http://dx.doi.org/10.1016/j.rser.2010.09.018. Puech, J.-L. et al., 1990. Delignification of oak wood with an ethanolwater solution
Amiri, H., Karimi, K., Zilouei, H., 2014. Organosolv pretreatment of rice straw for in a flow-through reactor. Holzforschung 44 (5), 367371.
efficient acetone, butanol, and ethanol production. Bioresour. Technol. 152, Quintero, J., Moncada, J., Cardona, C., 2013. Techno-economic analysis of bioethanol
450456, <http://www.ncbi.nlm.nih.gov/pubmed/24321608> (accessed production from lignocellulosic residues in Colombia: a process simulation
3.06.14). approach. Bioresour. Technol. 139, 300307.
ASTM, 2002. Standard terminology. Relating to biotechnology. Practice 95, 16. Rabemanolontsoa, H., Saka, S., 2012. Holocellulose determination in biomass. In:
Blandino, A., Belleli, C., Caro, I., 2014. An effective process for pretreating rice husk Zero-Carbon Energy Kyoto 2011: Special Edition of Jointed Symposium of
to enhance enzyme hydrolysis. Ind. Eng. Chem. Res. 53, 1087010875. Kyoto, pp. 135140.
120 L.V. Daza Serna et al. / Bioresource Technology 199 (2016) 113120

Reyes, T., Bandyopadhyay, S.S., McCoy, B.J., 1989. Extraction of lignin from wood Srinivasan, N., Ju, L.K., 2012. Statistical optimization of operating conditions for
with supercritical alcohols. J. Supercrit. Fluids 2, 8084. supercritical carbon dioxide-based pretreatment of guayule bagasse. Biomass
Rincn, L.E. et al., 2014. Techno-economic analysis of the use of fired cogeneration Bioenergy 47, 451458.
systems based on sugar cane bagasse in south eastern and mid-western regions Tolosa, R.A. et al., 2014. Cementicious materials reinforcement using Angustifolia
of Mexico. Waste Biomass Valorization 5 (2), 189198. kunth bamboo fiber covered with nanostructured manganese oxide. Ind. Eng.
Ruiz, E. et al., 2008. Evaluation of steam explosion pre-treatment for enzymatic Chem. Res. 53 (20), 84528463.
hydrolysis of sunflower stalks. Enzyme Microb. Technol. 42, 160166. Wang, F.-Q. et al., 2013. Biological pretreatment of corn stover with ligninolytic
Segal, L. et al., 1959. An empirical method for estimating the degree of crystallinity enzyme for high efficient enzymatic hydrolysis. Bioresour. Technol. 144, 572
of native cellulose using the X-ray diffractometer. Text. Res. J. 29 (10), 786794. 578, <http://www.ncbi.nlm.nih.gov/pubmed/23896439> (accessed 15.08.14).
Selig, M., Weiss, N., Ji, Y., 2008. Enzymatic Saccharification of Lignocellulosic Wei, G.Y. et al., 2009. Pretreatment and saccharification of rice hulls for the
Biomass Laboratory Analytical Procedure (LAP) Issue Date: 3/21/2008 production of fermentable sugars. Biotechnol. Bioprocess Eng. 14, 828834.
Enzymatic Saccharification of Lignocellulosic Biomass Laboratory Analytical Wood, I.P. et al., 2012. Rapid quantification of reducing sugars in biomass
Procedure (LAP), (March). hydrolysates: improving the speed and precision of the dinitrosalicylic acid
Singh, A., Bajar, S., Bishnoi, N.R., 2014. Enzymatic hydrolysis of microwave alkali assay. Biomass Bioenergy 44, 117121, <http://linkinghub.elsevier.com/
pretreated rice husk for ethanol production by Saccharomyces cerevisiae, retrieve/pii/S0961953412002097> (accessed 16.11.13).
Scheffersomyces stipitis and their co-culture. Fuel 116, 699702. http://dx.doi. World Energy Council, 2013. World Energy Resources: 2013 Survey, Available at:
org/10.1016/j.fuel.2013.08.072. <https://www.worldenergy.org/publications/2013/world-energy-resources-
Smith, R.L. et al., 1999. Energy analysis of supercritical carbon dioxide extraction 2013-survey/>.
processes. J. Supercrit. Fluids 15, 145156. Zhang, Y., Fu, X., Chen, H., 2012. Pretreatment based on two-step steam explosion
Soltani, N. et al., 2014. Review on the physicochemical treatments of rice husk for combined with an intermediate separation of fiber cells optimization of
production of advanced materials. Chem. Eng. J. 264, 899935, <http:// fermentation of corn straw hydrolysates. Bioresour. Technol. 121, 100104,
linkinghub.elsevier.com/retrieve/pii/S1385894714015150> (accessed 4.12.14). <http://www.ncbi.nlm.nih.gov/pubmed/22858472> (accessed 14.08.14).