Accepted Manuscript

Biodiesel production from canola oil using novel Li/TiO2 as a heterogeneous catalyst
prepared via impregnation method

Mariam Alsharifi, Hussein Znad, Sufia Hena, Ming Ang

PII: S0960-1481(17)30745-0
DOI: 10.1016/j.renene.2017.07.117
Reference: RENE 9092

To appear in: Renewable Energy

Received Date: 24 April 2017

Revised Date: 17 July 2017
Accepted Date: 29 July 2017

Please cite this article as: Alsharifi M, Znad H, Hena S, Ang M, Biodiesel production from canola oil
using novel Li/TiO2 as a heterogeneous catalyst prepared via impregnation method, Renewable Energy
(2017), doi: 10.1016/j.renene.2017.07.117.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Biodiesel production from canola oil using novel Li/TiO2 as a heterogeneous catalyst

prepared via impregnation method

PT
RI
SC
Mariam Alsharifi, Hussein Znad*, Sufia Hena, Ming Ang

U
Department of Chemical Engineering, Curtin University, GPO Box U 1987 Perth WA 6845,
AN
Australia
M
D
TE
C EP
AC

* Corresponding author

Email: [email protected]

1
 ACCEPTED MANUSCRIPT
Abstract

This study focused on the development of heterogeneous catalyst for transesterification

process by implanting lithium onto TiO2 by wet impregnation process. A series of Li/TiO2

was prepared with different amounts of Li (20, 30, 40 wt. %) and at different calcination

temperatures (450, 600, 750). The Li/TiO2 catalysts were characterized by several

PT
spectroscopic and analytical techniques like XRD, FT-IR, BET, TG-DSC and FESEM. The

RI
XRD study revealed that the insertion of Li improved the catalyst efficiency without any

alteration in structure of TiO2. Li/TiO2 catalysts with 30%w/w Li and calcined at 600 ºC was

SC
found to be the most efficient with 98% transesterification yield. The best performance of

catalyst was achieved with methanol to oil ratio of 24:1, 5 wt. % of catalyst loading, at 55ºC

U
reaction temperatures for 3 hours of reaction time. The kinetic studies revealed the
AN
transesterification process was compatible with the pseudo-first order model. However the
M

reusability decreases after every successive use.

D
TE
C EP
AC

Keywords: Li/TiO2, transesterification, biodiesel, heterogeneous catalyst

2
 ACCEPTED MANUSCRIPT
1. Introduction

In recent years, increasing global consumption of fossil based fuel results in the

environmental issues associated with the emission of greenhouse gases such as NOx, SOx,

CO, CO2 and shortage in the conventional fossil fuel sources [1]. This problem has promoted

researchers to take steps to substitute the fossil fuels with a renewable and sustainable fuel.

PT
Biodiesel, which is chemically a term of the mono-alkyl esters of long chain fatty acids and is

RI
mainly synthesized by transesterification of vegetable oils or animal fats with alcohol

(methanol or ethanol), has appeared as an alternative, biodegradable, renewable, and non-

SC
toxic fuel [2]. Conventionally, homogeneous catalysts such as sodium hydroxide (NaOH) are

widely used in industrial processes for biodiesel production. However, the product of these

U
catalysts should be washed and purified to remove the soluble catalyst and to agree with the
AN
European standards for the biodiesel specifications. Thus, the separation and purification of
M

the product which increase the process cost and affect the environment are the drawbacks of

homogenous catalysts. To solve these problems, researchers have extensively focused on

D

replacing homogeneous catalysts with different types of heterogeneous (solid) catalysts

TE

which are insoluble and recyclable. Eliminating the additional purification steps and the

ability to reuse the catalyst several times result in reducing the synthesis cost of biodiesel and
EP

environment concerns [3]. Currently, various heterogeneous catalysts such as mixed oxides,
C

alkaline-earth metals oxides, alkali metals supported on zeolite, alumina and solid acid
AC

catalysts have been reported at variable operating conditions. Among a great variety of solid

catalyst, TiO2 has been utilized widely as photo catalyst. Separate studies have documented

the use of TiO2 and Lithium oxides as catalyst for biodiesel production, as compile in Table

1. However, to the best of our knowledge, no study has been reported biodiesel synthesis

utilizing Li corporate with titanium oxide catalysts. Xie et al. [12] have reported the

transesterification of soybean oil in the presence of Lithium-Doped ZnO Catalysts. Kumar

3
 ACCEPTED MANUSCRIPT
and Ali [1] have demonstrated the catalytic activity of Li, Na, and K impregnated on CaO for

the transesterification of used cotton seed oil. Recently, Castro et al. [13] have investigated

the significant effect of the supporting material of the catalyst and they found lithium-based

mixed oxides have been strongly affected by the supporting materials. However, these

catalysts exhibited weak stability due to the losing active sites from the catalyst into the

PT
solution. In fact, enhancing the catalyst stability leads to improve the catalyst performance for

RI
increasing the biodiesel production. Hence, the development of a heterogeneous catalyst with

a high stability, and ability to produce FAME under mild temperature and atmospheric

SC
pressure, are of great interest for industrial scales. Generally, the reaction temperature could

be controlled by incorporating alkali earth metals in catalysts, which results in enhancing the

operating conditions [14].

U
AN
Table 1
M

Plant oils have attracted a considerable attention as an alternative fuel source which is

exposed to some chemical modification methods like transesterification (alcoholysis),

D

pyrolysis (thermal cracking) and emulsification. However, Transesterification has a

TE

prominent position in the synthesis of higher quality biofuel from vegetable oils in the

presence of a catalyst. The process involves the formation of methoxide ions (CH3O-M+)
EP

from the proton adsorption (H+) in the methanol’s hydroxyl group by the basic sites (i.e. M:
C

metal cations, Li and Ti) on the catalyst’s surface then these ions attack (nucleophilic attack)
AC

the carbonyl groups in the triglycerides (TAG).

Nucleophilic attack leads to form three molecules of fatty acid methyl esters (biodiesel) and

one molecule of glycerol through a three steps of alkoxy species interchange as shown in

Scheme. 1.

Scheme 1

4
 ACCEPTED MANUSCRIPT
In this study, Li-based TiO2 catalyst as a solid catalyst has been adopted for biodiesel

production from canola oil. The influence of calcination temperature and lithium loading

amounts were investigated in details and the reaction kinetics was studied. Moreover, the

reaction parameters such as methanol/ oil ratio, catalyst amount, and reaction temperature and

time have been optimized. Furthermore, the reusability and the alkali metal leaching were

PT
studied.

RI
SC
2. Method

2.1 Materials

U
Anhydrous methanol (99.8%) and n-hexane (95%) were purchased from Thermo Fisher
AN
Scientific Australia Pty Ltd. Canola oil was obtained from local shopping centre (Coles),
M

while the cooking canola oil was collected from local restaurant and their physio-chemical

properties are tabulated in Table 2. Analytical standards, Titanium (II) oxide and Lithium
D

nitrate were obtained from Sigma-Aldrich Pty Ltd. All the reagents were used without further
TE

purification.

Table 2
EP

2.2 Catalyst preparation

C

Lithium impregnated TiO2 catalysts for transesterification reactions were prepared by a wet
AC

impregnation method. In a typical catalyst preparation procedure, the required amount of

lithium nitrate (LiNO3) was suspended in deionized water using ultrasonic bath, and titanium

dioxide was slowly added into the aqueous solution. The resulted slurry was mixed for 4 hr at

room temperature using magnetic mixer, and then evaporated to dryness by hot plate and

heated overnight at 120ºC and finally calcined at 450, 600 and 750ºC for 4hr under air flow

in muffle furnace. Similarly, a series of catalysts with different amounts of LiNO3 (20, 30, 40

5
 ACCEPTED MANUSCRIPT
wt. %) were prepared in order to get the final lithium concentration on TiO2 in the range of

20- 40 wt. %. The obtained samples were calcined at different temperatures (450, 600, and

750°C) and denoted as x-LT-y where, x symbolizes lithium amounts and y symbolizes

calcination temperature, for example 30LT600 means 30 wt. % of lithium doped on TiO2 and

calcined at 600 ºC.

PT
2.3 Catalyst characterization

RI
The surface area of samples were measured by N2 adsorption- desorption isotherm using the

BRUNAUER- EMMER- TELLER (BET) surface area method with Micromeritics, Tristar II

SC
Surface area and Porosity analyser. Before measurements, all samples were degassed at

200°C for overnight. The crystallography of the synthesized catalyst samples was identified

U
by XRD. The X-ray diffraction analysis was performed on a Bruker D8 diffractometer using
AN
Cu-Kα radiation electrons (λ= 1.5418 0A) which were accelerated at 40mA and 40Kv over 2θ

angle range from 10 to 800 and scanning rate of 28min-1.

M

Uncalcined sample was analysed by the thermo gravimetric (TGA) analyser to investigate the
D

thermal transition of the prepared catalyst. The analysis was performed using METTLER
TE

TOLEDO TGA/DSC 1 STARe System under the Argon flow rate of 60 ml/min over the

temperature range of 35 to 900 ºC at a ramping rate of 10 ºC/min. The functional groups of

EP

the prepared catalyst were determined using Fourier-transformed infrared (FTIR)

C

spectroscopy. The spectra were obtained via (Perkin Elmer spectrum 100) FT-IR

spectrometer in the wavelength range from 400 to 4000 cm-1. Field emission scanning
AC

electron microscope (FESEM) measurements were recorded with Zeiss Neon 40EsB FESEM

with an Oxford Instruments Inca x-act SDD x-ray detector and Inca software.

2.4 Transesterification reaction and kinetic study

6
 ACCEPTED MANUSCRIPT
The experiments were carried out in a three- necked round bottom flask of 200 ml equipped

with a condenser and a magnetic stirrer. The flask was immersed in water bath to control the

temperature through the reaction. After the temperature of oil was reached a suitable degree

(35-75°C), a measured amount of catalyst (1-7% of oil weight) and different volumes of

methanol (different molar ratios: 9:1-28:1) were added. During transesterification reaction

PT
the, samples were withdrawn for analysis at every hour for 5 hours. After the completion of

RI
reaction, samples were centrifuged to separate the catalyst and the excess methanol was

evaporated by a rotary evaporator under vacuum condition to get the FAMEs.

SC
The produced FAMEs were analysed by gas chromatograph (Agilent Technologies 7890B)

equipped with a mass spectroscopy (5977A MSD) and connected to a FAME-Wax capillary

U
column (30m, 0.25mmID, 0.25um). The carrier gas was Helium and methyl octanoate with
AN
concentration of 10mg/ml (diluted in hexane) was the internal standard. The volume of
M

injected sample was 1 µl and the FAME conversion was estimated according to the following

equation (Eq. 1):

D

FAMEs conversion (%) = ∑A− A S

×
CS − VS
× 100% (1)
TE

AS m

Where ƩA is a sum of the areas of all peaks (C4:0 to C24:1); AS is the area of the peak of
EP

methyl octanoate used as the internal standard; CS is concentration of the methyl octanoate

solution (mg/ mL); VS is a volume of the internal standard solution (mL); and m is a sample
C

weight (mg).
AC

In order to understand the relationship between the reaction time, temperature and, their

influences on the transesterification reaction, the kinetics of biodiesel production process

were investigated at different temperatures (45, 55, 65 ºC) using zero order and pseudo-first

order model and the best fitted model was selected based on the higher correlation coefficient

R2 value.

7
 ACCEPTED MANUSCRIPT

3. Results and Discussion

3.1. X-ray diffraction and surface area analysis

The XRD patterns of the titanium- based catalysts at different LiNO3 loading is shown in

PT
Figure 1a. The X-ray diffraction peaks of TiO2 were sharp and intense at 2θ of 25.41°, 37.93°,

RI
48.14°, 53.86°, 55.05° and 62.57° which mainly correspond to the anatase phase with

appearance of small peaks of rutile at 2θ of 68.36°, 69.97° and 74.95°. New diffraction peaks

SC
were observed when the LiNO3 was added (samples were calcined at 600 ºC) which are

assigned to the Li2TiO3 phase at 2θ of 18.60°, 36.29°, 43.47°, 57.77° and 63.14°. Li2O

U
diffraction peaks were not detected for Li loaded samples. The finding is consistent with
AN
findings of past studies by many researchers [10, 12, 13], where lithium was used with
M

different supports such as MgO, Mg(Al)O and ZnO.

D

Fig. 1
TE

Furthermore, since the calcination temperature impacts the structure of catalyst and the

catalytic activity, the catalyst with different calcination temperatures was characterized by
EP

XRD, as illustrated in Figure (1b). Since the catalyst with 30% LiNO3 content gave the
C

highest conversion (78% refer Figure. 6), this amount was fixed for further XRD
AC

characterization at different calcination temperatures. The calcination at low temperature

(450 ºC) showed a low crystallization intensity of Li2TiO3 phase while the intensity was

increased at calcination temperature of 600 ºC (without affecting the TiO2 structure).

However, at 750 ºC of calcination the intensity of TiO2 crystalline phases reduced. In

addition, the diffraction lines at 2θ of 18.60°, 36.29°, 43.47°, 57.77° and 63.14 were shifted

to produce a new phase which was assigned to the LiTiO4. These results could be attributed

8
 ACCEPTED MANUSCRIPT
to the incidence of interaction between metals which led to alteration in the crystalline

structure [15].

Table 3 presents the results of surface area, pore volume and pore size of catalyst with

different loading amounts of lithium and calcination temperatures. It is revealed from Table 3

that the surface area of pure TiO2 decreased with increase of Lithium from 20 to 40 loading

PT
wt. %.

RI
Table 3

SC
However, unexpected increase in the surface area of catalyst was noticed with loading 30wt.

U
% of lithium, comparing with other two catalysts (20 and 40 wt %), as shown in Table 3. This
AN
finding reveals that lithium was combined with the supporting materials and penetrated
M

through the surface of supporting material (TiO2), which led to an increment of the surface

area [16]. However, the surface area reduced from 9.25m2/g to 3.06 m2/g with increase of the
D

calcination temperatures from 450 to 750 °C respectively (Table 3). Based on the results,
TE

increasing the calcination temperature induces phase crystallization changing and sintering

process, which dramatically reduces the surface area and pore volume [15]. Kaur and Ali [10]
EP

and Goncalvesa et. al [4] have also noticed these changes of Li addition on the morphological
C

properties of magnesium and calcium oxides, respectively. Moreover, the average pore size
AC

(~11 nm) of prepared samples was found in the mesoporous range (2-50nm) which is larger

than triglyceride molecule (~0.6nm) to diffuse through the catalyst [17].

3.2. FT-IR analysis

The chemical groups of the prepared catalyst were investigated by employing FTIR. Figures

2a and b present the FTIR spectra of the catalysts prepared with different lithium amounts

9
 ACCEPTED MANUSCRIPT
loading and at different calcination temperature. For pure TiO2, the spectrum at 729.1cm-1

was characterized to the Ti-O vibration. As depicted in Figure 2a, the spectrum at the reign

400 – 860 cm-1 are attributed to the presence of M+-O bond structure [14]. The band at 866

cm-1 corresponds to the formation of mixed metal oxides (Li2TiO3) which was undetectable

for pure TiO2 and noticeable for catalysts possessing Li. These results are analogous with

PT
XRD data where Li2TiO3 formation was observed. The sharp spectrum at 1440 cm-1 and

1505 cm-1 are recognized to the C=O bond from carbonates which is attributed to the

RI
adsorption of CO2 from the ambient atmosphere and the split peak is an evidence of the

SC
presence of CO2 as it is a common feature in carbonate compounds [6].

U
Fig. 2
AN
The C=O peaks were remarkable for Li samples since Li+ ion is sensitive to CO2 at high

temperature [4]. Moreover, a broad spectrum which is shown at around 3200 cm-1, is
M

associated to the O-H stretching vibration mode of water species [18]. The FTIR spectrum for
D

the catalysts at different calcination temperature (Figure 2b) shows appearance of a new peak
TE

at 1131 cm-1 for 750 ºC calcinate catalyst due to the formation of LiTiO4. In 30LT450 sample

the peaks appeared at 1645cm-1 and 1347 cm-1 due to ᵟOH bending vibration and N-O bond
EP

vibration of LiNO3 respectively were vanished as the temperature increased in case of

C

30LT600 and 30LT750 because of the loss of water from the catalyst structure and
AC

decomposing of LiNO3. The FTIR findings indicate that the impregnation of Li metal on

TiO2 was successfully linked.

3.3. TGA/DSC analysis

10
 ACCEPTED MANUSCRIPT
Figure (3) illustrates the thermogravimetric curves of Li/TiO2. From the Figure 3, it is

apparent that two stages of mass loss were observed for 30LT sample. In first stage, a little

weight loss, around 100 ºC, which was correspond with endothermic peak in DSC curve at

same temperature, which were attributed to the escape of adsorbed water species.

PT
Fig.3

RI
While, the second stage occurred at 600 ºC due to the thermal decomposition of LiNO3 that

associates with endothermic process observed in the DSC curve. Xie et al. [12] have also

SC
reported an endothermic signal at 600ºC for LiNO3 decomposition. DSC curve shown an

endothermic peak between 200-300ºC correspond to no weight loss in TGA curve which

U
probably could be assigned to the interference between LiNO3 and TiO2. Moreover, no
AN
weight loss was noticed above 600ºC on the TGA profile which means the decomposition
M

process was completed. Consequently, from this analysis, 600ºC was considered to be the

consistent temperature to calcine 30LT catalyst.

D
TE

3.4. FE-SEM analysis

EP

The FE-SEM was used to study the changes in morphology of prepared catalysts at different
C

amounts of lithium loading and calcination temperatures. The FS-EM image of pure TiO2
AC

particles is originally spherical in form with a relatively smooth surface as shown in Figure

4a. Figure (4b-d) shows the effect of lithium loading on the morphology of TiO2 at 20, 30,

and 40 wt. %. The FE-SEM images of the surface of the synthesized catalysts exposed an

obvious transformation of the spherical particles of the TiO2 into assorted particle shapes

between round and angled corner particles with inhomogeneous size distribution. With

11
 ACCEPTED MANUSCRIPT
increasing addition of lithium (40 wt.% of LiNO3), the catalyst particles became like molten

with a dense appearance.

Fig. 4

Figure (5a-c) shows the FE-SEM images of the synthesized samples at different calcination

PT
temperatures (450, 600, and 750 ºC). Increasing the temperature from 450 to 600 ºC resulted

RI
in changing the particle size with emergence of agglomerations and at high temperature (750

ºC), it turned to smooth large particles which was attributed to small surface area as a

SC
consequence of sintering. These results are in agreement with the results discussed before

using XRD and BET analysis.

U
AN
Fig. 5
M

3.2. Effect of the Li to TiO2 percentage weight ratio

In order to verify the influence of impregnated amount of lithium ion on the activity of the
D

catalyst to obtain the highest FAMEs conversion, three different concentrations of lithium
TE

were tested in respect with the TiO2 weight (wt. %). Figure 6 has shown that impregnation

with 30 wt. % LiNO3 exhibited the highest conversion under the reaction conditions of 12:1
EP

MeOH:oil ratio, for reaction time of 3 hr, at reaction temperature 65 ºC and a 5% of catalyst.
C
AC

Fig. 6

The high surface area of 30LTthat calcined at 600 C could be the reason of the high catalytic

activity [16].

3.3. Effect of the calcination temperature

12
 ACCEPTED MANUSCRIPT
Figure 7 depicts the results for the effect of calcination temperature on the performance of

30LT catalyst.

Fig. 7

PT
This study has found that the FAME conversion increases with the increase of temperature up

RI
to 600ºC then the conversion goes down upon increasing the calcination temperature. The

finding suggests that Li2TiO3 phase which was observed in the 450 and 600ºC calcined

SC
catalyst samples as mentioned in XRD analysis together with TiO2 phases might be the

reason of increasing the conversion. The calcination at low temperature (450 ºC) showed a

U
low crystallization intensity of Li2TiO3 phase while the intensity was increased at calcination
AN
temperature of 600 ºC (without affecting the TiO2 structure) as shown in Fig 1. While
M

increasing heating up to 750ºC caused decreasing TiO2 phase intensity and emergence of

LiTiO4 phase which could be the reason after losing the activity of catalyst. These results
D

were supported by XRD and BET analysis too. It seems that TiO2 and Li2TiO3 phases have
TE

mainly affected the catalyst’s efficiency. Consequently, 600 ºC was a suitable calcination

temperature for the prepared catalyst as supported by TGA/DSC analysis.

C EP

3.4. Effect of transesterification conditions

AC

To evaluate the efficiency of the prepared catalyst, it was applied to the canola oil

transesterification under various reaction parameters. Since the loading amount of 30wt. %

and calcination temperature at 600 ºC gave the highest production value, all investigations

were conducted with 30LT catalyst calcined at 600 ºC.

3.4.1 Effect of Methanol to oil ratio

13
 ACCEPTED MANUSCRIPT
The influence of Methanol to oil ratio was studied at the reaction conditions of 65ºC, for 3

hours reaction time and 5 wt. % catalyst. Transesterification is reversible reaction so excess

amount of methanol is required to drive the reaction towards biodiesel formation. Thus, the

reaction was carried out by varying the MeOH:oil ratio from 9:1 to 28:1 to assess the

optimum value of methanol. Figure 8a shows increasing the methanol to oil ratio increases

PT
the FAMEs conversion. However, the excess amount of methanol could lead to reduce the

RI
yield as a result of dissolving of glycerol (by-product) and shifting the reaction towards the

reactants [19]. Therefore, the optimum ratio of methanol to oil was 24:1.

SC
3.4.2 Effect of Catalyst amount

U
The effect of catalyst dosage was investigated on the activity of 30LT600 by changing the
AN
loading amounts of catalyst to the reactor based on the oil amount. The reaction was carried
M

out under 24:1 methanol:oil ratio at 65 ºC for 3 hours. Figure 8b shows with increasing the

amount of catalyst, the FAMEs conversion was increased and beyond 5wt. %, the conversion
D

was remained same. Hence, 5wt% was seemed to be reasonable. It is apparent from this
TE

figure that the reaction reached to equilibrium state [20] at 5wt% and beyond of catalyst.
EP

3.4.3 Effect of Reaction temperature

C

The reaction temperature has a significant effect on carrying out the transesterification
AC

reaction. Figure 8c, shows the influence of reaction temperature on the performance of

catalyst from 35 – 75 ºC at 24:1 methanol:oil ratio with catalyst amount of 5% and a reaction

time for 3 hours.

As it can be seen, increasing the temperature is favourable condition for FAMEs conversion.

Referring to the kinetic performance, increasing reaction temperature enhances the reaction

rate leading to raise the molecular collision and decline the activation energy limitation [14].

14
 ACCEPTED MANUSCRIPT
However, after 65ºC the conversion was reduced due to the possibility of evaporating the

methanol resulting in reducing the availability of methanol for the reaction [21].

Consequently, the reaction temperature at 65 ºC was chosen to carry out the rest experiments.

Fig. 8

PT
3.4.4 Effect of Reaction time

RI
One of the important variables in the biodiesel production is a reaction time. As

transesterification is equilibrium reaction, thus adequate time is required to reach equilibrium.

SC
In addition, a well contact time between reactants is necessary so that reagents can reach the

active sites of catalyst and the conversion is occurred [22]. The effect of reaction time is

U
shown in the Figure 8d where the other parameters were remained constant as follow: 24:1
AN
methanol:oil molar ratio, 5%wt of catalyst and a reaction temperature of 65ºC.
M

As it is cleared from Fig.8d, the FAMEs conversion was increased with increasing the

reaction time up to 3 hours and achieved the highest conversion, then the conversion was
D

reduced with time. This behaviour might be explained by the possibility of occurring
TE

reversible reaction [23, 24].

EP

3.4.5 Kinetic studies

C

In order to understand the relationship between the reaction time and temperature and, their
AC

influences on the transesterification reaction, the kinetics of biodiesel production process

were investigated. The order of the reaction was varied to find the model that fit best to the

reaction using regression coefficient (R2).Since the transesterification is a reversible reaction,

excess amount of methanol is needed to shift the reaction towards esters formation.

Therefore, the reverse reaction could be neglected and the reaction rate constants could be

independent on the methanol concentration. Furthermore, ignoring the intermediate reactions

15
 ACCEPTED MANUSCRIPT
could be assumed and considering the transesterification reaction occurs in one step. To

verify this hypothesis, firstly the zero order was assumed and the equation as follows:

d [Me] dx
= K Me ( 2)
dt dt

Where, [Me] is concentration of FAMEs, KMe is the rate constant, x is the FAMEs

PT
conversion. The rate constant was obtained by drawing the FAMEs conversion (x) against the

reaction time (t).

RI
The second model which could be considered for transesterification is a pseudo-first order

SC
reaction rate and the equation is expressed as below:

d [TG ] d [Me]
r=− = = k ′ [TG ][ROH ]
3
(3)

U
dt dt AN
Where r and k’ are the reaction rate and reaction rate constant, respectively, [TG], [Me] and

[ROH] are the triglyceride, methyl esters and methanol concentration , respectively and t is
M

the reaction time. As the assumptions of the reaction are occurring in the triglyceride

molecule and the methanol is in excess, the above equation (3) could be re-written as follows
D

[25]:
TE

d [TG ]
r=− = k [TG ] ( 4)
dt
EP

Where k = k ′ [ROH ]
3
C

By Re-arrearage the above equation and integrate both sides (taking on the account when the
AC

initial time (t) = 0, [TG] =[TG]o and when t = t, [TG] = [TG]), the equation could be

expressed as follows:

d [TG ]
− = k dt (5)
[TG ]

− ln
[TG ] = kt ( 6)
[TG ]o
Mass balance equation is as below,

16
 ACCEPTED MANUSCRIPT
[TG ] = (1 − X Me ) (7 )
[TG ]o
Where XMe is the FAMEs conversion.

So from substituting the equation (6) in (7), the rate reaction constants could be calculated by

fitting the FAMEs conversion and reaction time to the below equation:

PT
− ln (1 − X Me ) = k t (8)

In addition, because the transesterification reaction was conducted at different reaction

RI
temperatures, the activation energy was determined through applying Arrhenius equation to

SC
investigate the effect of temperature on the particular reaction rate [17, 18, 26] as shown

below:

k = A e RT
− Ea

U
AN
(9)
M

After simplifying the above equation, we obtain the following equation:

Ea
D

ln k = ln A − (10)
RT
TE

Where Ea is the activation energy (kJ/mol), R is the universal gas constant (8.314 kJmol-1K-
1
), k is the reaction rate constant (hr-1), A is a frequency factor (hr-1) and T is the reaction
EP

temperature (K). From plotting the (ln k) vs. (1/T), the intercept corresponds to (ln A) and the
C

slope represents ( ).
AC

The experimental data were conducted to the equations (2 and 8) and shown a good

agreement with a zero order model. Table 4 demonstrates that the values of correlation

coefficients R2 indicated a better fit of zero order model for canola oil transesterification as

well the increment in the temperature and duration of reaction, increase FAMEs conversion

in both models. Table (4) also presents the reaction rate constants (k) and R2 values with

17
 ACCEPTED MANUSCRIPT
reaction temperatures represents that the reaction rate constants increase with increasing the

reaction temperatures.

Table 4

The findings suggest that increasing the temperatures promote the molecules collision and

thus decreasing the mass transfer limitation. The activation energy, which is defined as the

PT
minimum quantity of energy needed for activating atoms or molecules to undergo a chemical

RI
reaction, (Ea) was calculated and it was found to be 39.366 and 16.461 kJ/mol for Pseudo

first order and zero order models respectively. The findings provide evidence that the normal

SC
conditions of temperature and pressure could be employed to carry out the reaction.

3.5 Reusability of catalyst

U
AN
The reusability is one of the important characters of the heterogeneous catalysts. To study the
M

possibility of regenerating the catalyst, after the completion of reaction, the catalyst was

separated by filtration and washed three times with methanol and once with hexane then dried
D

overnight at 120°C after that, it was directly subjected to the transesterification reaction under
TE

the optimum conditions (24:1 MeOH:oil, 5% of catalyst, 65ºC reaction temperature and 3 hr

of reaction time). Figure 9 shows the FAMEs conversion was gradually declined with
EP

increasing the number of reuse recycle.

C
AC

Fig. 9

The decreasing catalyst efficiency might be due to leaching of Li+ ion from catalyst surface or

diffusion of canola oil and FAME molecules into the catalyst by blockage the catalyst pores

and remain persist even after washing the catalyst with organic solvents. Therefore, to

evaluate the catalyst deactivation due to the blocking of pores by organic material, TG/DSC

analysis was done. Figure 10 presents TG/DSC result for the fourth reuse of catalyst, as it can

18
 ACCEPTED MANUSCRIPT
be observed that the stages of decomposition are similar to the fresh one except appearing of

an exothermic peak at 300ºC which corresponds to the decomposition of residual organic

material. Thus, the accessibility of the reactants to the active sites was hindered due to the

blockage of catalyst pores which led to decline the FAMEs conversion. However the

deactivation of catalyst due to pores filling can be resolved and it is suggested to regenerate

PT
the catalyst for every consecutive cycle of usage, via recalcitrant it at 300 for 2 hrs, which

RI
would remove all the materials blocked in the catalyst pores. Moreover, to assess the losing

of catalyst activity as the leaching of Li+ ion and the catalyst stability, the catalyst was reacted

SC
with methanol under the optimum conditions (without oil) then the methanol was decanted

and has been reacted with oil in the absence of catalyst.

U
AN
Fig. 10
M

The latter experiment revealed there is just 26% of FAMEs conversion, which revealed that,

there was leaching but the leached Li+ ions have insignificantly contributed to prompt the
D

homogeneous reaction, hence, the catalyst deactivation was mainly by pore-filling with
TE

organic moieties rather than by leached ions.

EP

3.6 Investigation the performance of 30LT600 for waste and fresh oil
C

The biodiesel production from fresh edible oil is always being criticized; hence finding the
AC

alternate feedstocks is introduced and tested by many researchers. Consequently usage of

waste cooking oil was proposed as feedstocks for biodiesel production [27-29], which

possess many advantages such as, it reduces the harmful environmental impacts caused by

the disposal of used cooking oils directly into drainage which causes sever blockage too and

on the other hand biodiesel production from used cooking oil significantly reduced the total

production cost of biodiesel [30]. However, the optimum condition for biodiesel production

19
 ACCEPTED MANUSCRIPT
from fresh oil and used oil are unpredictable, thus here in this research the same optimum

conditions were used to compare the capability of FAME conversion for used or waste

cooking oil. It was found that the catalyst can be used for biodiesel production from waste

cooking oil with 91.73% FAME conversion.

PT
4. Conclusion

RI
The purpose of the current study was to enhance the activity of TiO2 for biodiesel production

by inclusion of Lithium ions onto it using impregnation method and evaluate the efficiency of

SC
prepared catalyst under mild reaction conditions. A series of LiNO3/TiO2 catalysts were

synthesized at different calcination temperatures which were investigated and conducted for

U
the transesterification process using canola oil for biodiesel production. The study showed
AN
that addition of lithium has improved the TiO2 reactivity towards biodiesel production which
M

could be attributed to enhance the surface properties of TiO2. However, 30 wt%, of LiNO3

and 600ºC of calcination temperature has shown the highest activity for FAMEs formation
D

due to the formation of Li2TiO4. The obtained 30LT600 catalyst exhibited high catalytic
TE

performance for biodiesel production under the optimum transesterification conditions of

24:1 methanol:oil ratio, with 5% catalyst dosage, for the reaction time of 3 hours at 65ºC of
EP

reaction temperature and up to 98% FAMEs conversion was achieved for fresh canola oil.
C

The kinetics of reaction provided evidence that the normal conditions of temperature and
AC

pressure could be employed to carry out the transesterification process by 30LT600. The

synthesized catalyst was used for transesterification process of waste cooking oil also and

obtained 91.73% of FAMEs conversion. The catalyst reusability studies showed that the

activity of catalyst decreased upon successive runs of used mainly due to pore-filling by

organic materials, i.e. triglyceride and glycerol as approved by TGA/ DSC analysis and also

because of Li+ leaching.

20
 ACCEPTED MANUSCRIPT
References

[1] D. Kumar, A. Ali, Nanocrystalline lithium ion impregnated calcium oxide as

heterogeneous catalyst for transesterification of high moisture containing cotton seed

oil, Energy Fuels, 24 (2010) 2091–2097.

[2] T. Saba, J. Estephane, B. E. Khoury, M. E. Khoury, M. Khazma, H. E. Zakhem, S.

PT
Aouad, Biodiesel production from refined sunflower vegetable oil over KOH/ZSM5

RI
catalysts, Renewable Energy, 90 (2016) 301-306.

[3] E. Bet-Moushoul, K. Farhadi, Y. Mansourpanah, A. M. Nikbakht, R. Molaei, M.

SC
Forough, Application of CaO-based/Au nanoparticles as heterogeneous nanocatalysts

in biodiesel production, Fuel, 164 (2016)119–127.

[4]
U
A. M. Goncalves, R. A. B. Lima-Correa, J. M. Assaf, A. R. A. Nogueria, Lithium and
AN
calcium based perovskite type oxides for ethylic transesterification, Catalysis Today,
M

279(2016)177-186.

[5] D. Salinas, S. Guerrero, A. Cross, P. Araya, E. E. Wolf, Potassium titanate for the
D

production of biodiesel, Fuel, 166 (2016)237-244.

TE

[6] E. A. Z. Gonzalez, M. Garcia-Guaderrama, M. R. Villalobos, F. L. Dellamary, S.

Kandhual, N. P. Rout, H. Tiznado, G.G.C. Arizaga, Potassium titanate as

EP

heterogeneous catalyst for methyl transesterification, Powder Technology, 280 (2015)

C

201-206.
AC

[7] Y. Dai, J. Wu, C. Chen, K. Chen, Evaluating the optimum operating parameters on

transesterification reaction for biodiesel production over a LiAlO2 catalyst, Chem Eng

J, 280 (2015) 370-376.

[8] B. Lertpanyapornchai, C. Ngamcharussrivichai, Mesostructured Sr and Ti mixed

oxides as heterogeneous base catalysts for transesterification of palm kernel oil with

methanol, Chem Eng J, 264 (2015) 789-796.

21
 ACCEPTED MANUSCRIPT
[9] N. Hindryawati, G. P. Maniam, M. R. Karim, K. F. Chong, Transesterification of used

cooking oil over alkali metal (Li, Na, K) supported rice husk silica as potential solid

base catalyst, Engineering Science and Technology, an International Journal, 17

(2014)95-103.

[10] N. Kaur, A. Ali, Lithium ions-supported magnesium oxide as nano-sized solid catalyst

PT
for biodiesel preparation from muton fat, Energy Sources, Part A: Recovery,

RI
Utilization, and Environmental Effects, 35:2(2013)184-192.

[11] Z. Wen, X. Yu, S. Tu, J. Yan, E. Dahlquist, Biodiesel from waste cooking oil

SC
catalyzed by TiO-MgO mixed oxides, Bioresource Technology, 101(2010) 9570-9576.

[12] W. Xie, Z. Yang, H. Chun, Catalytic properties of Lithium-Doped ZnO catalysts used

U
for biodiesel preparations, Ind. Eng. Chem. Res., 46(2007)7942-7949.
AN
[13] C. S. Castro, C. Ferreti, J. I. D Cosimo, J. M. Assaf, Support influence on the basiscity
M

promotion of lithium-based mixed oxides for transesterification reaction, Fuel,

103(2013) 632-638.
D

[14] H. M. Amani, Asif, B. H. Hameed, Transesterification of waste cooking palm oil and
TE

palm oil to fatty acid methyl ester using cesium-modified silica catalyst, Journal of the

Taiwan Institute of Chemical Engineers, 58(2016) 226–234.

EP

[15] D. Salinas, P. Araya, S. Guerrero, Study of potassium-supported TiO2 catalysts for the
C

production of biodiesel, Applied Catalysis B: Environmental, 117– 118(2012) 260–

AC

267.

[16] X. Yu, Z. Wena, H. Li, S. Tu, J. Yan, Transesterification of Pistacia chinensis oil for

biodiesel catalyzed by CaO–CeO2 mixed oxides, Fuel 90 (2011)1868–1874.

[17] T. Maneerung, S. Kawi, Y. Dai, C. Wang, Sustainable biodiesel production via

transesterification of waste cooking oil by using CaO catalysts prepared from chicken

manure, Energy Conversion and Management, 123(2016)487–497.

22
 ACCEPTED MANUSCRIPT
[18] N. Hindryawati, G. P. Maniam, Novel utilization of waste marine sponge

(Demospongiae) as a catalyst in ultrasound-assisted transesterification of waste

cooking oil, Ultrasonics Sonochemistry, 22 (2015) 454–462

[19] M. Farooq, A. Ramli, A. Naeem, Biodiesel production from low FFA waste cooking

oil using heterogeneous catalyst derived from chicken bones. Renewable Energy, 76

PT
(2015)362-368.

RI
[20] Y. Lu, Z. Zhang, Y. Xu, O. Liu, G. Qian, CaFeAl mixed oxide derived heterogeneous

catalysts for transesterification of soybean oil to biodiesel, Bioresource Technology,

SC
190(2015)438- 441.

[21] F. Li, H. Li, L. Wang, Y. Cao, Waste carbide slag as a solid base catalyst for effective

U
synthesis of biodiesel via transesterification of soybean oil with methanol, Fuel
AN
Processing Technology 131(2015)421–429.
M

[22] M. A. Olutoye, C. P. Wong, L. H. Chin, B. H. Hameed, Synthesis of FAME from the

methanolysis of palm fatty acid distillate using highly active solid oxide acid catalyst,
D

Fuel Processing Technology, 124(2014)54-60.

TE

[23] M. A. Olutoye, S. W. Wong, L. H. Chin, H. Amani, M. Asif, B. H. Hameed, Synthesis

of fatty acid methyl esters via the transesterification of waste cooking oil by methanol
EP

with a barium-modified montmorillonite K10 catalyst, Renewable Energy, 86(2016)

C

392-398.
AC

[24] Y. Syamsuddin, M. N. Murat, B. H. Hameed, Synthesis of fatty acid methyl ester from

the transesterification of high- and low-acid-content crude palm oil (Elaeis guineensis)

and karanj oil (Pongamia pinnata) over a calcium–lanthanum–aluminium mixed-

oxides catalyst, Bioresource Technology, 214(2016) 248–252.

[25] G. R. Moradi, M. Mohadesi, M. Ghanbari, M. J. Moradi, S.H. Hosseini, Y.

Davoodbeygi, Kinetic comparison of two basic heterogeneous catalysts obtained from

23
 ACCEPTED MANUSCRIPT
sustainable resources for transesterification of waste cooking oil, Biofuel Research J, 6

(2015).236-241.

[26] I. C. Emeji, A. S. Afolabi, A.S. Abdulkareem, J. Kalala, Characterization and kinetics

of biofuel produced from waste cooking oil. Proceedings of the World Congress on

Engineering and Computer Science 2015 Vol II WCECS 2015, October 21-23, 2015,

PT
San Francisco, USA.

RI
[27] K. G. Georgogianni, M.G. Kontominas, E. Tegou, D. Avlonitis, V. Vergis, Biodiesel

production: reaction and process parameters of alkali-catalysed transesterification of

SC
waste frying-oils, Energy Fuels, 21(2007)3023–3027

[28] Z. Utlu, Evaluation of biodiesel obtained from waste cooking oil, Energy Sources,

Part A, 29 (2007)1295–1304
U
AN
[29] A. A. Refaat, N. K. Attia, H. A. Sibak, S. T. El Sheltawy, G. I. El Diwani, Production
M

optimisation and quality assessment of biodiesel from waste vegetable oil, Int J

Environ Sci Technol, 5 (2008)75–82

D

[30] A. N. Phan, T. M. Phan, Biodiesel production from waste cooking oils, Fuel, 87
TE

(2008) 3490–3496
C EP
AC

24
 ACCEPTED MANUSCRIPT

Tables captions
Table 1: TiO2 and Lithium oxides as catalyst for biodiesel production
Table 2: Physico-chemical properties of fresh and waste cooking canola oil

Table 3: Results of BET surface area, pore size and pore volume for various Li loading

PT
amounts and calcination temperatures

Table 4: Reaction temperatures, rate constants and R2 values of canola oil transesterification

RI
by 30LT600 catalyst

U SC
AN
M
D
TE
C EP
AC

25
 ACCEPTED MANUSCRIPT
Table 1: TiO2 and Lithium oxides as catalyst for biodiesel production
Catalyst type Preparation Feedstock Operating Yield/ Ref.
method conditions conversion
(%)
Ca0.9Li0.6Zr0.9O3 nitrate-b- soybean Ethanol/oil= 12:1, 60 [4]
alanine oil Temp.=50ºC,
solution t=30min, cat%=10
combustion
method

PT
Potassium titanate TiO2 was canola oil MR=36:1, cat%=6, ~100 [5]
20%K/TiTH impregnated t=2hr, temp.=65ºC
with KNO3

RI
after being
hydrothermally
treated

SC
Potassium titanate hydrothermal safflower MR=1:1, Cat%=3, 100 [6]
synthesis oil temp.=50ºC
method

U
LiALO2 The solid-state soybean MR=24:1, 97.5 [7]
reaction oil cat%=8,temp.=65ºC,
AN
t=2h
Mesostructured Sr 1.sol-gel palm Temp 99.9 [8]
and Ti mixed combustion kernel oil (SGC)=170ºC,
oixdes (SGC) method. temp(EISA)=
M

MST-SGC 2.evaporation- 150ºC, MR=20:1,

MST-EISA induced self - cat%=10, t=3h
assembly
D

(EISA)
method.
TE

Alkali metal (Li, Wet used MR= 9:1, cat%=3, 96.5-98.2 [9]
Na, K) supported impregnation cooking t=1h, temp.=65ºC
EP

rice husk method oil

5-Li/MgO-250 Impregnation animal MR=12:1, cat%=5, 99 [10]
method (mutton) t=40min,
fat temp.=65ºC
C

TiO2–MgO mixed sol-gel method waste MR=50:1, cat%=10, 92.3 [11]

AC

oxides (MT-1-923) cooking t=6hr, temp.=433K

oil

26
 ACCEPTED MANUSCRIPT
Table 2: Physico-chemical properties of fresh and waste cooking canola oil

Physico-chemical property Fresh canola oil Waste cooking oil

Saponification value (mgKOH/g) 187 234
Acid Value (mgKOH/g) 0.072 3.67
Specific gravity 0.9457 0.9388

PT
Water content (wt.%) 0.95 1.6
Oleic acid (wt%) 56 21.9

RI
Linoleic acid (wt%) 26 54
Linolenic acid (wt%) 10 6

SC
Palmitic acid(wt%) 4 8
Stearic acid (wt%) 2 6.1
Erucic acid (wt%) 2 4

U
AN
M
D
TE

Table 3: Results of BET surface area, pore size and pore volume for various Li loading
amounts and calcination temperatures.
EP

S. No. Catalyst type BET surface area (m2/g) Pore size(nm) Pore volume (cm3/g)
1. TiO2 10 7.32 0.018
C

2. 20LT600 5.57 12.87 0.011

AC

3. 30LT600 7.63 11.94 0.02

4. 40LT600 5.53 13.04 0.011
5. 30LT450 9.25 9.72 0.024
6. 30LT750 3.06 6.45 0.006

27
 ACCEPTED MANUSCRIPT
Table 4: Reaction temperatures, rate constants and R2 values of canola oil transesterification
by 30LT600 catalyst

Temperature Rate constant (k) R2

(ºC ) Zero order Pseudo first order Zero order Pseudo first order
45 0.2368 0.3714 0.8862 0.7727
55 0.3289 0.7077 0.9341 0.8025

PT
65 0.3456 0.9577 0.9422 0.8061

RI
U SC
AN
M
D
TE
C EP
AC

28
 ACCEPTED MANUSCRIPT
Scheme Caption

Scheme.1. The mechanism of transesterification process

Figures Captions

Fig.1. XRD patterns of the catalysts: (a) different LiNO3 loading calcined at 600 °C; (b)

30LT at different calcinations temperatures

PT
Fig. 2. FT-IR spectra of (a) different Li impregnated amount; (b) different calcination

temperatures

RI
Fig. 3. Thermogravimetric TGA/DSC profile of 30LT

SC
Fig. 4. FE-SEM micrographs of (a) TiO2; (b) 20LT600; (c) 30LT600; (b) 40LT600

Fig.5. FE-SEM micrographs of (a) 30LT450; (b) 30LT600; (c) 30LT750

U
Fig. 6. Effect of impregnated lithium ion on the FAME production
AN
Fig. 7. Effect of calcination temperature on the FAME production

Fig.8. Effect of (a) methanol: oil ratio (b) catalyst amount (wt.%) (c) reaction temperature (d)
M

reaction time on the FAMEs conversion

D

Fig.9. Reusability of 30LT600 catalyst under the optimum reaction conditions

TE

Fig.10. TGA/DSC analysis of 4th used of 30LT600

C EP
AC

1
 ACCEPTED MANUSCRIPT

(a)

PT
RI
U SC
AN
(b)
M
D
TE
C EP
AC

2 theta (º)

Fig.1. XRD patterns of the catalysts: (a) different LiNO3 loading calcined at 600 °C; (b)
30LT at different calcinations temperatures

2
 ACCEPTED MANUSCRIPT

PT
RI
U SC
(a)
AN
M
D
TE
C EP
AC

(b)
Fig. 2. FT-IR spectra of (a) different Li impregnated amount; (b) different calcination
temperatures.

3
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M

Fig. 3. Thermogravimetric TGA/DSC profile of 30LT

D
TE
C EP
AC

4
 ACCEPTED MANUSCRIPT

(a) (b)

PT
RI
SC
(c) (d)

U
AN
M
D
TE

Fig. 4. FE-SEM micrographs of (a) TiO2; (b) 20LT600; (c) 30LT600; (b) 40LT600
C EP
AC

5
 ACCEPTED MANUSCRIPT

(a)

PT
RI
SC
(b) (c)

U
AN
M
D
TE
EP

Fig.5. FE-SEM micrographs of (a) 30LT450; (b) 30LT600; (c) 30LT750

C
AC

6
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
Fig. 6. Effect of impregnated lithium ion on the FAME production
M
D
TE
C EP
AC

Fig. 7. Effect of calcination temperature on the FAME production

7
 ACCEPTED MANUSCRIPT
110 110

105 a 105 b
100 100

95 95
FAMEs conversion (%)

FAMEs conversion (%)

90 90

85 85

80 80

75 75

PT
70 70

65 65

60 60
5 10 15 20 25 30 0 1 2 3 4 5 6 7 8

RI
Methanol:oil ratio Catalyst amount (wt%)

110 110

c d

SC
100
100
90
90
FAMEs conversion (%)

FAMEs conversion (%)

80

80 70

U
70 60

50
AN
60
40
50
30

40 20
M

30 40 50 60 70 80 0 1 2 3 4 5 6

Reaction temperature (ºC) Reaction time (hrs)

D
TE

Fig.8. Effect of (a) methanol: oil ratio (b) catalyst amount (wt.%) (c) reaction temperature (d)
EP

reaction time on the FAMEs conversion

C
AC

8
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP

Fig.9. Reusability of 30LT600 catalyst under the optimum reaction conditions

C
AC

9
 ACCEPTED MANUSCRIPT

PT
RI
U SC
Fig.10. TGA/DSC analysis of 4th used of 30LT600
AN
M
D
TE
C EP
AC

10
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
C EP
AC

Scheme.1 The mechanism of transesterification process

12
 ACCEPTED MANUSCRIPT

Highlights

 The activity of TiO2 was enhanced for biodiesel production by inclusion Li.

 Li/TiO2 catalysts with 30%w/w Li and calcined at 600 ºC has shown 98% yield.

 The formation of Li2TiO4 represented the major active sites of the prepared catalyst.

PT
 Mild conditions of temperature and pressure were sufficient for transesterification.

RI
SC
 

 
U
AN
 

 
M
D
TE
C EP
AC