Received: 24 May 2021 Revised: 14 October 2021 Accepted: 16 October 2021

DOI: 10.1002/er.7432

RESEARCH ARTICLE

Transesterification using isopropanol as a co-solvent for the

production of green biodiesel fuel

Thao Pham Thi Phuong1,2 | Uyen Nguyen Huynh Phuong3 |

Kiyoshi Imamura1 | Masakazu Furuta1 | Yasuaki Maeda4

1
Department of Quantum and Radiation
Engineering, Graduate School of Summary
Engineering, Osaka Prefecture University, Previously, we have conducted the transesterification using isopropanol (IPA) as
Sakai City, Japan
a co-solvent for biodiesel fuel (BDF) production in the homogeneous condition
2
Center for Research and Technology
(a single fatty acid methyl ester [FAME] phase). In this report, we investigated
Transfer, Vietnam Academy of Science
and Technology, Hanoi, Vietnam the transesterification kinetics by varying IPA amount and the optimum condi-
3
Sanicon Co. Ltd., Sakai City, Japan tions for BDF production with 10 wt% IPA. In the IPA amount of 5 to 25 wt%,
4
Graduate School of Humanities and the transesterification proceeded in a heterogeneous condition (FAME and glyc-
Sustainable System Sciences, Osaka
erol [GL] phases) as the GL phase was formed and separated from the FAME
Prefecture University, Sakai City, Japan
phase. The transesterification rate increased with the increase of IPA amount,
Correspondence and the correlation coefficient (r) between the rate constant and IPA amount
Thao Pham Thi Phuong, Department of
was 0.97 (P < 0.05). The optimal conditions of transesterification for biodiesel
Quantum and Radiation Engineering,
Graduate School of Engineering, Osaka production from canola oil with 10 wt% IPA were as follows: 1.0 wt% KOH,
Prefecture University, 1-2 Gakuen-cho, MeOH/oil molar ratio of 6:1, and a reaction temperature of 30
C. The quality of
Naka-ku, Sakai City, Osaka 599-8570,
Japan.
biodiesel satisfied the JIS K2390 and EN 14214 standards. Transesterification
Email: [email protected] using 10 wt% IPA co-solvent could facilitate the phase separation after the reac-
tion, and the amount of waste was reduced. The recovery of IPA from FAME
Funding information
This research did not receive any specific
phase was >96%, hence IPA could be reused. These results indicated that IPA
grant from funding agencies in the public, was a superior co-solvent for the BDF production from plant oil feedstocks.
commercial, or not-for-profit sectors.
Novelty statement
• Isopropanol as a co-solvent for biodiesel production from canola oil was
examined.
• Kinetic parameter of reaction rate was estimated.
• The transesterification proceeded rapidly with increase of isopropanol.
• The transesterification conditions were optimized for biodiesel production
with 10 wt% of isopropanol.

KEYWORDS
biodiesel production, co-solvent method, isopropanol, transesterification

1 | INTRODUCTION research for alternatives to fossil fuel. Biodiesel fuel

(BDF) is a renewable, biodegradable, noncorrosive, non-
Currently, the awareness of environmental issues and the toxic, and aromatics and sulfur-free fuel. Compared with
energy crisis related to fossil fuels have prompted petrol diesel, it has a higher cetane number, higher

Int J Energy Res. 2021;1–10. wileyonlinelibrary.com/journal/er © 2021 John Wiley & Sons Ltd. 1
2 PHAM THI PHUONG ET AL.

combustion efficiency, and lower emission of pollutants isopropyl ether, tert-butyl methyl ether, and acetone.
such as carbon dioxide and particulates.1–5 Using DEE and THF co-solvent, the FAME yields were
According to American Society for Testing and Mate- 97% to 98%. Recently, the alkaline (KOH) catalyzed trans-
rials (ASTM) standards, BDF is defined as the mono alkyl esterification using an acetone co-solvent was applied to
esters of long-chain fatty acids derived from vegetable oils Jartropha curcas oil,13,29 candlenut oil,15 Vernicia mon-
or animal fats, for use in compression-ignition (diesel) tana oil,39 and waste cooking oil,40 and the trans-
engines.6 It is produced by the transesterification of tri- esterification using FAME co-solvent to rubber seed oil41
glycerides (TG) with a primary alcohol in the presence of and canola oil.12 The FAME yields were >98% at low
a catalyst.2 Among the alcohols, methanol (MeOH) is the reaction temperature.
most suitable for transesterification due to the high reac- Maeda et al38 reported that IPA is one of the potential
tivity (the shortest alkyl alcohol), cost-effective, and low co-solvents due to the solubility of the related compounds
boiling point. Furthermore, the excess amount of MeOH and examined the optimum condition for trans-
can be recovered from the glycerol (GL) after reaction.7 esterification in homogeneous (a single phase) condition.
There are various kinds of catalysts used for trans- However, the detailed transesterification kinetics has not
esterification of raw materials such as oil and waste been well known. In this study, we examined the behav-
cooking oil. The catalysts are classified as homogeneous, ior of IPA as a co-solvent for producing BDF from canola
heterogeneous, and enzyme type.7–9 The homogeneous oil. The effect of IPA amount on reaction rate and the
(soluble in solution) alkaline type such as NaOH, KOH, optimal conditions for transesterification such as
and CH3ONa is the most efficient for amounts of catalyst, MeOH-to-oil molar ratio, and reac-
transesterification,10,11 and the system can be expanded tion temperature were investigated.
in the industrial scale.12 On the other hand, the homoge-
neous acid one such as H2SO4, HCl, and sulfonic acid is
used for esterification of free fatty acid (FFA): the high 2 | MATERIAL A ND METHODS
content (>5%) of FFA in raw materials such as Jatropha
oil,13,14 candles nuts oil,15 and rubber oil16 can be 2.1 | Materials
converted to FAME due to the soap formation in the pro-
cess of the alkali catalysis transesterification.17,18 The het- Canola oil was produced by Tsuji Oil Mills Co., Ltd.
erogeneous catalysts are solid bases of CaO, Ca2Al2O5, (Matsusaka City, Mie, Japan). The chemical and physi-
Ca2Fe2O5, KOH/Al2O3, ETS-10 zeolite, Li4SiO4, BaO- cal properties of canola oil are listed in Table 1. IPA
MoO2, BaSnO3,19–23 and acid ones of ZnO/I2, ZrO2/SO24, (99%), MeOH (99%), KOH (85%), H3PO4 (85%), GL, eth-
niobic acid, sulfated zirconia.24,25 As for enzymatic cata- anol, 1-propanol, and 1-butanol were analytical grade
lysts, many studies have currently been carried out using and supplied by Fujifilm Wako Pure Chemical Co. Ltd.
lipase immobilized on magnetic nanoparticles for the pro- (Osaka, Japan). Tri-olein, di-olein, mono-olein, and
duction of biodiesel.26–28 methyl oleate (MO) were purchased from Sigma-
The transesterification rate of alkali catalysis is low as Aldrich (Japan). Solvents such as acetonitrile (99.8%)
the reaction occurs at the interface between oil and and acetone (99.7%) were high-performance liquid
MeOH phases.29 In order to accelerate the reaction rate, chromatography (HPLC) grade and obtained from
the selected methods of raising the reaction temperature,
mixing using mechanical stirring,30 ultrasonic
irradiation,31–33 and supercritical method34,35 were con- TABLE 1 Chemical and physical properties of canola oil used
ducted. However, these procedures consume much more Property Unit Average ± SD
energy.36 On the contrary, the co-solvents decrease the 3
Density kg m 915 ± 20
mass transfer resistance between the oil and the alcohol 1
Water mg kg 295 ± 4.5
during the chemical reaction.37 Maeda et al38 proposed
various kinds of co-solvents such as tetrahydrofuran Acid value mg KOH g1 0.80 ± 0.12
(THF), diethyl ether (DEE), acetonitrile, acetone, and iso- Oleic acid (C18:1)a wt% 63.0 ± 0.3
propanol (IPA) that dissolve the related materials of Linoleic acid (C18:2) a
wt% 19.4 ± 0.2
transesterification to produce a homogeneous reaction Linolenic acid (C18:3) a
wt% 8.6 ± 0.2
system (a single phase). Using acetone and IPA co-sol- Palmitic acid (C16:0)a wt% 4.4 ± 0.05
vent, the reaction was completed in a short time and the a
Stearic acid (C18:0) wt% 1.7 ± 0.1
FAME yield was >98% at a low temperature. Encinar
Other fatty acids wt% 2.8 ± 0.1
et al1 also compared the FAME yield using various kinds
of co-solvents such as THF, DEE, di-butyl ether, di- a
Carbon atom number: double bond number.
PHAM THI PHUONG ET AL. 3

Fujifilm Wako Pure Chemical Co. Ltd. (Japan). The 2.3 | Analytical method
purified water of Milli-Q grade was supplied from
Osaka Prefecture University. During the reaction processes, the concentrations of TG
and products, such as di-glycerides (DG), mono-
glycerides (MG), and FAME, were monitored at certain
2.2 | Transesterification procedure intervals. An aliquot of 1 mL reaction mixture was trans-
ferred to 2 mL centrifuge vial containing 100 μL of 5% H3
A flow diagram of the production of BDF from canola PO4 solution to quench the reaction. After centrifuging
oil using a co-solvent of IPA is shown in Figure 1. A the mixtures, 40 μL of the FAME phase was dissolved
100 g of canola oil and the desired amount of IPA were with 4 mL of acetone/acetonitrile (70:30 vol/vol) for anal-
put into a 300 mL three-necked round bottom flask ysis. The analytical method for determination of TG, DG,
placed in a water bath controlled at the desired temper- MG, and FAME was described in previous work.12 These
ature using a thermostat and stirred gently. The reac- compounds were analyzed using a Shimadzu HPLC sys-
tion was started by adding the desired amount of tem (Japan) consisting of a liquid pump LC-10 AD, an
KOH/MeOH into the mixture. The mixture was stirred auto-injector SIL-10AF, a column oven CTO-10 AD, and
for 1 hour. The reaction mixture obtained was trans- a refractive index detector RID-10A. The HPLC column
ferred to a separatory funnel and allowed to separate used was Asahipak GF 310 HQ (5 μm particle size,
for 30 minutes. IPA in the upper FAME phase was put 300  7.5 mm, Shodex Co., Ltd., Tokyo, Japan). The col-
into an eggplant flask and recovered using a rotary umn temperature was 30
C. The mobile phase was a mix-
evaporator N1000 (EYELA, Tokyo, Japan) controlled at ture of acetone/acetonitrile (70:30 vol/vol) at a flow rate
80
C in a water bath under reduced pressure of of 0.5 mL min1. The injection volume was 20 μL.
52 mmHg. The solvent vapor was condensed by cooling
water at 11
C. The amount of solvent recovered was
measured by a difference of weight of the flask before 2.4 | Validation of quantification
and after evaporation. After recovering IPA, the crude
FAME was washed with purified water until the pH of Data are represented herein as the mean ± SD, based on
the washed water was 7 and finally heated to 90
C for triplicate analyses of each sample. Analysis of variance
1 hour under reduced pressure for removal of water. (ANOVA) and Tukey's comparison tests were performed
using the software developed by Assaad et al42 and
Microsoft Office 2010 to identify the significant difference
between the two values. A difference was considered sta-
tistically significant if P < .05.

2.5 | Determination of FAME yield

Transesterification proceeds by methoxyl anions (CH3

O) produced from the reaction of MeOH with KOH
under the anhydrous condition. The transesterification
reactions are shown in the following equations
(Equation 1-3)43:

KOH
TG þ CH3 OH ⇄ DG þ R1  CðOÞOCH3 ð1Þ

KOH
DG þ CH3 OH ⇄ MG þ R2  CðOÞOCH3 ð2Þ

KOH
MG þ CH3 OH ⇄ GL þ R3  CðOÞOCH3 ð3Þ

And the equations are summarized as follows:

F I G U R E 1 Flow diagram of BDF production from canola oil KOH

using IPA amount of 5 to 20 wt% TG þ 3 CH3 OH ⇄ GL þ 3 R  CðOÞOCH3 ð4Þ
4 PHAM THI PHUONG ET AL.

The FAME yields of transesterification are calculated FAME, where the MeOH/oil ratio was 4.5.38 Accordingly,
using the following formula12,15,39,40,44: the effect of FAIE on the physiological property of BDF
might be insignificant, while the FAIE has better cold-
FAME yeild ð%Þ ¼ ðW FAME =M FAME Þ  100=ð3  W oil =M oil Þ, flow property than FAME because of its branched
chain.45,46
where WFAME and Woil are the weight of the FAME in
the FAME phase and the weight of oil, respectively.
MFAME and Moil are the average molecular weights of 3.2 | Back reaction of transesterification
FAME and canola oil, respectively.
It has been reported that there has been the possibility of
the back reaction as the transesterification is a reversible
3 | R ES U L T S A N D D I S C U S S I O N reaction, as shown in Equations (1-3) in Section 2.5.43
The backward reaction causes the deterioration of BDF
3.1 | Transesterification of canola oil quality. The back reaction between GL and FAME
with various kinds of alkyl alcohol (Equation 3) was examined as follows: 10 g of GL and
96.5 g of MO (as FAME) that is equivalent to a GL/MO
The transesterification of canola oil with five types of molar ratio of 1:3 were mixed with 1 wt% KOH catalyst
alkyl alcohol such as MeOH, ethanol, IPA, 1-propanol, (to the weight of GL) and stirred continuously at a reac-
and 1-butanol was examined using KOH as a catalyst. tion temperature of 40
C for 24 hours. The TG, DG, MG,
The procedure was as same as described in Section 2.2. and FAME contents were monitored for 24 hours at cer-
The results are shown in Figure 2. Fatty acid ester (FAE) tain intervals using HPLC. During 24 hours, there were
yields of five alcohols except IPA increased to 42.7 ± 0.8% no changes in their contents. The results indicated that
in 5 minutes and attained 84.9%, 83.5%, 80.5%, and 74.3% the back reaction hardly occurred at the moderate condi-
in 1 hour corresponding to the reagent of MeOH, etha- tion.38
nol, 1-propanol, and 1-butanol, respectively. These results
indicated that the FAE yield decreased with the molecu-
lar size, that is, molecular weight. As for the IPA, the 3.3 | Transesterification of canola oil
FAE yield was 7% in 5 minutes; then, the reaction using IPA as a co-solvent
reached equilibrium. These results indicated that the
transesterification with the secondary alcohol of IPA 3.3.1 | The effect of IPA amount on
could not easily proceed at the modulate condition.38 transesterification of the oil
Using 25 wt% IPA co-solvent, the content of fatty acid
isopropyl ester (FAIE) was approximately 1 wt% of The effect of IPA on FAME yield was examined by
varying IPA amount to oil in the range of 0 to 30 wt%.
The results are shown in Figure 3. The formation of
FAME proceeded rapidly with 5 to 30 wt% IPA in com-
parison to 0 wt% IPA. The FAME yield after 1 hour was
84.9% with 0 wt% IPA, and attained 96.0% with 5 wt%
IPA, and 99.0 ± 0.3% with 10 to 30 wt% IPA. The result
indicated that the yield was attained ≥98% with
10 wt% IPA.
The time variations of FAME yields during 0 to
6 minutes are shown in Figure 3. The FAME yields were
increased exponentially with the reaction time; on the
other hand, the oil (TG) was decreased exponentially. To
evaluate the transesterification kinetics, we estimated the
rate constants according to the method reported by
Kongmany et al47 and Hoa et al,48 and plotted the corre-
lation between the ln (C/C0) and reaction values, where
F I G U R E 2 Transesterification of canola oil with various kinds C0 and C are the concentrations of canola oil (TG) at
of alcohol. The error bars are expressed as means ± SD. The beginning and duration of the reaction, respectively. The
reaction conditions were as follows: alcohol/oil molar ratio of 6:1, value of C decreased exponentially with reaction time,
1.0 wt% KOH, and reaction temperature of 40
C resulting in a linear relationship between the two
PHAM THI PHUONG ET AL. 5

F I G U R E 3 Effect of the IPA

amount on the FAME yield. The error
bars are expressed as means ± SD. The
reaction conditions were as follows:
MeOH/oil molar ratio of 6:1, 1.0 wt%
KOH, and reaction temperature of 40
C

T A B L E 2 The rate constants of transesterification and

correlation coefficients (r) of the linear regression in 0 to 30 wt% of 3.3.2 | Phase separation after reaction
IPA (P < 0.05)
After transesterification of 1 hour, the photograph of the
Rate constant Correlation
bottles of reaction mixture settled in 20 minutes in 0 to
IPA (wt%) (k) (min1) coefficient (r)
30 wt% IPA is shown in Figure 5. The lower GL phase
0 0.93  101 0.9688
occurred and gradually and slightly increased in 0 to 15 wt
1
5 3.47  10 0.9821 % IPA, but decreased in 20 to 30 wt% IPA, and dis-
1
10 4.58  10 0.9863 appeared completely with 30 wt% IPA. On the other hand,
15 4.91  101 0.9841 the FAME layer increased with the increase of 0 to 30 wt%
20 5.53  10 1
0.9530 IPA and became a single FAME phase with 30 wt% IPA.
1 These results indicated that the transesterification
25 5.85  10 0.9122
1
proceeded in heterogeneous (two FAME and GL) phases
30 6.12  10 0.8651
in 0 to 25 wt% IPA and in a homogeneous (a single FAME
phase) phase with 30 wt% IPA. The phase separation of
the reaction mixtures in 5 to 25 wt% IPA after reaction
variables that indicated a pseudo first-order reaction1,49 was completed within 20 minutes, as shown in Figure 5.
as the excess amount of MeOH exists. Based on the slope IPA dissolves the related materials for trans-
of the linear regression, we estimated the rate constant esterification of oil, namely TG, FAME, GL, MeOH, and
(k) for the transesterification of the oil. The k and the cor- KOH catalyst. In the beginning, the transesterification
relation coefficients (r) of the linear regression in the IPA was initiated in the mixture of IPA and oil by adding the
amount of 0 to 30 wt% are summarized in Table 2. The reagent of MeOH/KOH. GL and FAME were rapidly pro-
linear correlation (r: 0.87-0.99) between the two values of duced, and both GL and FAME phases were separated in
the ln (C/C0) and reaction values were observed in the ≲25 wt% IPA, that is, the transesterification was
IPA of 0 to 30 wt% (P < 0.05). proceeded in the heterogeneous conditions (FAME and
The relationship between the k of transesterification GL phases), where the reaction was retarded by the GL
and the IPA amount of wt% is shown in Figure 4. The phase.38 The IPA was distributed mostly in the FAME
correlation (r: 0.97) between the two values was observed phase, of which the volume increased in the IPA amount
except for 0 wt% IPA (P < 0.05), that is, the trans- of 5 to 30 wt%, as shown in Figure 5, which is discussed
esterification rate was correlated with wt% IPA and the in Section 3.3.1. In other words, the effect of the GL
reaction was accelerated by the IPA amount. These phase on the reaction rate decreased with the increase of
results suggested that the transesterification of the oil IPA amount. With the IPA of ≥30 wt%, the trans-
with 30 wt% IPA co-solvent was the most optimal condi- esterification proceeded rapidly in a homogenous condi-
tion for BDF production, because of the homogeneous tion (a single FAME phase). However, the large amount
reaction mixture of a single FAME phase. of waste corresponding to around half the volume of
6 PHAM THI PHUONG ET AL.

F I G U R E 4 Correlation between the

rate constant (k) of transesterification and
the wt% IPA

F I G U R E 5 Photograph of phase separation in IPA amount of 0 to 30 wt% after 20 minutes settling of the reaction products. The
reaction conditions were as follows: MeOH/oil molar ratio of 6:1, 1.0 wt% KOH, reaction temperature of 40
C, and reaction time of 1 hour

FAME might be caused by the purification process of were 91.0%, 99.1%, and 99.5%, corresponding to a KOH
FAME, as shown in Figure 5. amount of 0.5, 1.0, and 1.5 wt%, that is, the FAME yield
We examined the transesterification of canola oil attained >99% with 1.0 wt% KOH. Although a large amount
using IPA amount of 10 wt%, that is, the reaction of KOH catalyst could accelerate the reaction rate, the FAME
proceeded in the heterogeneous condition and compared yield satisfied the criteria of the BDF standards (>96.5%) with
the optimum conditions for transesterification with those KOH amount of 1.0 wt%. High-quality BDFs were produced
in homogeneous condition (a single FAME phase) using from Vernicia montana oil39 and waste cooking oil40 using
IPA amount of 25 wt%.38 acetone as a co-solvent with 1.0 wt% KOH catalyst. These
results were coincident with the present results.

3.3.3 | Effect of the catalyst amount on

FAME yield 3.3.4 | Effect of the molar ratio of MeOH/oil
on FAME yield
The effect of the catalyst amount on FAME yield was exam-
ined by varying KOH in the range of 0.5 to 1.5 wt% to oil. For transesterification of oil (TG), the stoichiometric
The results are shown in Figure 6. The FAME yields in 1 hour molar ratio of alcohol to oil is 3; thus, at least 3-fold
PHAM THI PHUONG ET AL. 7

F I G U R E 6 Effect of KOH catalyst on the FAME yield. The F I G U R E 8 Effect of the reaction temperature on the FAME
error bars are expressed as means ± SD. The reaction conditions yield. The error bars are expressed as means ± SD. The reaction
were as follows: 10 wt% IPA, MeOH/oil molar ratio of 6:1, and conditions were as follows: 10 wt% IPA, MeOH/oil molar ratio of
reaction temperature of 40
C 6:1, and 1.0 wt% KOH

3.3.5 | Effect of reaction temperature on

FAME yield

The reaction temperature is a critical factor for controlling

the transesterification rate. The effect of the temperature on
the FAME yield was examined in the range of 10
C to 40
C.
The results are shown in Figure 8. The FAME yields were
89.7%, 95.8%, 98.7%, and 99.1%, corresponding to the temper-
ature of 10
C, 20
C, 30
C, and 40
C in 1 hour. These results
indicate that temperature had a positive effect on the reaction
rate, and the FAME yield increased with increasing tempera-
ture. The FAME yields were >96.5% at a reaction tempera-
ture of ≥30
C in 1 hour, which meets the EN 14214 and JIS
K2390 standards. The reaction temperature was set at 30
C.
The transesterification using co-solvent method has been
F I G U R E 7 Effect of molar ratio of MeOH/oil on the FAME
conducted at a relatively lower temperature of ≤40
C, and
yield. The error bars are expressed as means ± SD. The reaction
conditions were as follows: 10 wt% IPA, 1.0 wt% KOH, and reaction
the high quality of BDF was obtained.13,15,29,38–41
temperature of 40
C The optimum conditions of transesterification using IPA
amount of 10 wt% and of 25 wt% are summarized in Table 3.
In order to attain >98% of FAME yield using 10 wt% IPA,
that is, a heterogeneous condition (FAME and GL phases),
molar of MeOH to oil is required. The effect of the more amount of reagents (MeOH and KOH) are required,
molar ratio of MeOH/oil on the FAME yield was exam- and the reaction should be conducted at slightly higher tem-
ined in the molar ratio of 3:1 to 6:1. The results are perature, but the less amount of waste occurs in the process
shown in Figure 7. The FAME yield in 1 hour was of FAME purification in comparison with using 25 wt% IPA,
86.9%, 95.6%, 97.8%, and 99.1%, corresponding to the that is, a homogeneous condition (single FAME phase).
molar ratios of 3:1, 4:1, 5:1, and 6:1. It is necessary to
add the reagent of MeOH to a raw material of oil with
their molar ratios of ≥5:1 to attain the BDF quality 3.4 | Physicochemical properties of BDF
standard (96.5%). A high FAME yield (>98%) of BDF produced from canola oil using IPA as co-
was obtained from Vernicia montana oil using MeOH/ solvent
oil molar ratio of 6:1 using acetone co-solvent.39 The
subsequent transesterification was conducted with the The BDF was produced from canola oil using 10 wt%
molar ratio of MeOH/oil 6:1. IPA and purified by the procedure as shown in
8 PHAM THI PHUONG ET AL.

Figure 1. The physicochemical properties of BDF using 3.5 | Recovery of IPA

10 wt% of IPA co-solvent and those using acetone co-
solvent29 are presented in Table 4, along with the qual- As discussed previously, the IPA was mostly distributed
ity criteria of BDF regulated in EN 14214 and JIS in the FAME phase: approximately 70% of it was con-
K2390. The transesterification using acetone co-solvent tained in the phase using 10 wt% IPA, as shown in
was proceeded in a heterogeneous condition (FAME Figure 5. The recoveries of IPA and acetone from reac-
and GL phases)14 as same as using 10 wt% of IPA co- tion mixtures were examined. A mixture of 25 mL IPA
solvent. The quality of BDF produced from canola oil and acetone with the same volume of FAME was pre-
using both co-solvents satisfied the JIS K2390 and EN pared in an eggplant flask, respectively. Each solvent was
14214 standards. recovered from the mixture as described in Section 2.2.
The FAME content produced from canola oil using The recovery of IPA was 97% and that of acetone was
acetone co-solvent (25 wt%) was 98.3%,29 and that using 78%. The recovery of IPA was higher than that of acetone
IPA co-solvent was 98.7%. The density of BDF produced because the vapor pressure of acetone (231 mmHg at
using both co-solvents was 875 kg m3. The acid value, 25
C) is higher in comparison with IPA (45.4 mmHg at
MG, TG, and total glycerol content obtained using IPA 25
C). These results suggest that IPA can be recovered
co-solvent were slightly higher in comparison with using from crude FAME under moderate conditions and thus
acetone co-solvent, but these two values were coincident reused for other various purposes.
with each other.

4 | CONCLUSION
T A B L E 3 The comparison of the optimum conditions between A co-solvent method using IPA amount of 0 to 30 wt%
with IPA amount of 10 wt% and with that of 25 wt%
was investigated for BDF production from canola oil. In
10 wt% 25 wt% the presence of IPA amount ≤25 wt%, the trans-
Parameters IPAa IPAb esterification proceeded in a heterogeneous condition
Molar ratio of MeOH to oil 6:1 4.5:1 (FAME and GL phases), and the reaction rate was dis-
KOH amount (wt%) 1.0 0.5
turbed by the formation of the GL phase. However, the


reaction rate increased with the increase of IPA co-sol-
Reaction temperature ( C) 30 20
vent. The optimum conditions for BDF production with
FAME yield in 1 h (%) 99.0 ± 0.3 99.0 ± 0.3
10 wt% IPA were as follows: 1.0 wt% KOH, molar ratio of
Phase separation after reaction (min) 20 min None MeOH/oil 6:1, and a reaction temperature of 30
C. In
a
This experiment: heterogeneous condition (FAME and GL). comparison with IPA amount of 25 wt%, i.e. in the homo-
b
Reference 38: homogeneous condition (a single FAME phase). geneous condition (a single FAME phase), the more

T A B L E 4 Physicochemical properties of biodiesel produced from canola oil by co-solvent method using IPA and acetone29 and EN
14214 and JIS K2390 standards

BDF produced from canola oil by co-solvent method

Property Unit IPA Acetone29 EN 14214/JIS K2390

Ester content wt% 98.7 ± 0.5 98.3 ≥96.5
Density (15
C) kg m3 875 ± 9 875 860-900
Water content mg kg1 470 ± 5 — ≤500
1
Acid value mg KOH g 0.21 ± 0.01 0.12 ≤0.5
MG content wt% 0.4 ± 0.02 0.32 ≤0.7
DG content wt% 0.10 ± 0.01 0.14 ≤0.2
TG content wt% 0.12 ± 0.01 0.10 ≤0.2
Free glycerol wt% ND <0.01 ≤0.02
Total glycerol wt% 0.13 ± 0.01 0.11 ≤0.25

Note: Total glycerol = free glycerol +0.255MG + 0.146DG + 0.103TG. 39

Abbreviations: DG; di-glyceride; MG, mono-glyceride; ND, not detected; TG, triglyceride.
PHAM THI PHUONG ET AL. 9

amounts of reagents (MeOH and KOH) and the slightly 11. Casas A, Fernandez CM, Ramos MJ, Pérez A,  Rodríguez JF.
higher reaction temperature were needed to attain ≥99% Optimization of the reaction parameters for fast pseudo single-
of FAME yield, but the amount of waste could be phase transesterification of sunflower oil. Fuel. 2010;89:650-
658. doi:10.1016/j.fuel.2009.08.004
reduced, and furthermore, IPA could be recovered from
12. Nguyen HPU, Pham TPT, Asada R, Furuta M, Maeda Y. A new
FAME phase. The physicochemical properties of BDF method for production of green biodiesel fuel using FAME as a
satisfied the JIS K2390 and EN 14214 standards. co-solvent. Eco Eng. 2020;32:61-67. doi:10.11450/seitaikogaku.
32.61
CONFLICT OF INTEREST 13. Luu PD, Truong HT, Van Luu B, et al. Production of biodiesel
The authors have none of any financial interests or per- from Vietnamese Jatropha curcas oil by a co-solvent method.
sonal relationships that could have influenced the work Bioresour Technol. 2014;173:309-316. doi:10.1016/j.biortech.
2014.09.114
in this paper.
14. Chang CC, Teng S, Yuan MH, et al. Esterification of jatropha
oil with isopropanol via ultrasonic irradiation. Energies. 2018;
A U T H O R C ON T R I B U T I O NS 11:1456. doi:10.3390/en11061456
Thao Pham Thi Phuong: Investigation, formal analysis, 15. Pham LN, Van Luu B, Phuoc HD, et al. Production of biodiesel
writing—original draft preparation, and visualization. from candlenut oil using a two-step co-solvent method and
Uyen Nguyen Huynh Phuong: Formal analysis. Kiyoshi evaluation of its gaseous emissions. J Oleo Sci. 2018;67:1-2. doi:
Imamura: Writing, review, and editing. Masakazu 10.5650/jos.ess17220
Furuta: Fund acquisition. Yasuaki Maeda: Conceptuali- 16. Bdulkadir BAA, Emura YU, Amli AR, et al. Production of Bio-
diesel from Rubber Seeds (Hevea Brasiliensis) by In situ Trans-
zation, resources, and fund acquisition.
esterification Method; 2015.763-768.
17. Montcho PS, Konfo T. R. C, Agbangnan D. C. P,
ORCID Sidouhounde A, Sohounhloue C. K. D. Comparative study of
Thao Pham Thi Phuong https://orcid.org/0000-0001- transesterification processes for Biodiesel production
5565-2477 (a review). Elixir Appl Chem. 2018;120:51235-51242.
18. Bin Mohiddin MN, Tan YH, Seow YX, et al. Evaluation on
feedstock, technologies, catalyst and reactor for sustainable bio-
R EF E RE N C E S diesel production: a review. J Ind Eng Chem. 2021;98:60-81.
1. Encinar JM, Pardal A, Sanchez N. An improvement to the doi:10.1016/j.jiec.2021.03.036
transesterification process by the use of co-solvents to produce 19. Chueluecha N, Kaewchada A, Jaree A. Enhancement of bio-
biodiesel. Fuel. 2016;166:51-58. doi:10.1016/j.fuel.2015.10.110 diesel synthesis using co-solvent in a packed-microchannel.
2. Singh D, Sharma D, Soni SL, Sharma S, Kumar Sharma P, J Ind Eng Chem. 2017;51:162-171. doi:10.1016/j.jiec.2017.
Jhalani A. A review on feedstocks, production processes, and 02.028
yield for different generations of biodiesel. Fuel. 2020;262: 20. Roschat W, Siritanon T, Kaewpuang T, Yoosuk B, Promarak V.
116553. doi:10.1016/j.fuel.2019.116553 Economical and green biodiesel production process using river
3. Unglert M, Bockey D, Bofinger C, et al. Action areas and the snail shells-derived heterogeneous catalyst and co-solvent
need for research in biofuels. Fuel. 2020;268. doi:10.1016/j.fuel. method. Bioresour Technol. 2016;209:343-350. doi:10.1016/j.
2020.117227 biortech.2016.03.038
4. Joshua CW, Matthew GD, Alexander MS, Hutchinson 21. Chinh Nguyen H, Pan J-W, Su C-H, et al. Sol-gel synthesized
Williamson IV G. A comparative analysis of Biodiesel and die- lithium orthosilicate as a reusable solid catalyst for biodiesel
sel emissions, Worcester Polytecnic Institute; 2010.1-128. production. Int J Energy Res. 2021;45:6239-6249. doi:10.1002/er.
5. Bajpai D, Tyagi VK. Biodiesel: source, production, composition, 6246
properties and its benefits. J Oleo Sci. 2006;55:487-502. doi:10. 22. Jamil F, Al-Riyami M, Al-Haj L, et al. Waste Balanites
5650/jos.55.487 aegyptiaca seed oil as a potential source for biodiesel pro-
6. Specification For Biodiesel (B100)-ASTM D6751-07b; 2007. duction in the presence of a novel mixed metallic oxide cata-
https://www.glycerintraders.com/ASTM%206751%20spec.pdf lyst. Int J Energy Res. 2021;45:17189-17202. doi:10.1002/er.
7. Thanh LT, Okitsu K, Van Boi L, Maeda Y. Catalytic technolo- 5609
gies for biodiesel fuel production and utilization of glycerol: a 23. Farooq M, Ramli A, Naeem A, et al. A green route for biodiesel
review. Catalysts. 2012;2:191-222. doi:10.3390/catal2010191 production from waste cooking oil over base heterogeneous
8. Singh D, Sharma D, Soni SL, Sharma S, Kumari D. Chemical catalyst. Int J Energy Res. 2019;43:5438-5446. doi:10.1002/er.
compositions, properties, and standards for different genera- 4646
tion biodiesels: a review. Fuel. 2019;253:60-71. doi:10.1016/j. 24. de Pietre MK, Almeida LCP, Landers R, Vinhas RCG, Luna FJ.
fuel.2019.04.174 H3PO4- and H2SO4-treated niobic acid as heterogeneous cata-
9. Knothe G. Dependence of biodiesel fuel properties on the lyst for methyl ester production. React Kinet Mech Catal. 2010;
structure of fatty acid alkyl esters. Fuel Process Technol. 2005; 99:269-280. doi:10.1007/s11144-009-0143-9
86:1059-1070. doi:10.1016/j.fuproc.2004.11.002 25. Garcia CM, Teixeira S, Marciniuk LL, Schuchardt U. Trans-
10. Fadhil AB, Al-Tikrity ETB, Albadree MA. Biodiesel production esterification of soybean oil catalyzed by sulfated zirconia. Bio-
from mixed non-edible oils, castor seed oil and waste fish oil. resour Technol. 2008;99:6608-6613. doi:10.1016/j.biortech.2007.
Fuel. 2017;210:721-728. doi:10.1016/j.fuel.2017.09.009 09.092
10 PHAM THI PHUONG ET AL.

26. Thangaraj B, Solomon PR, Muniyandi B, Ranganathan S, 40. Luu PD, Takenaka N, Van Luu B, Pham LN, Imamura K,
Lin L. Catalysis in biodiesel production - a review. Clean Maeda Y. Co-solvent method produce biodiesel form waste
Energy. 2019;3:2-23. doi:10.1093/ce/zky020 cooking oil with small pilot plant. Energy Procedia. 2014;61:
27. Cavalcante FTT, Neto FS, de Aguiar Falc~ao IR, et al. Opportu- 2822-2832. doi:10.1016/j.egypro.2014.12.303
nities for improving biodiesel production via lipase catalysis. 41. Le HNT, Imamura K, Watanabe N, et al. Biodiesel production
Fuel. 2021;288:119577. doi:10.1016/J.FUEL.2020.119577 from rubber seed oil by transesterification using a co-solvent of
28. Velusamy K, Devanand J, Senthil Kumar P, et al. A review on fatty acid methyl esters. Chem Eng Technol. 2018;41:1013-1018.
nano-catalysts and biochar-based catalysts for biofuel produc- doi:10.1002/ceat.201700575
tion. Fuel. 2021;306:121632. doi:10.1016/j.fuel.2021.121632 42. Assaad HI, Hou Y, Zhou L, Carroll RJ, Wu G. Rapid
29. Thanh LT, Okitsu K, Sadanaga Y, Takenaka N, Maeda Y, publication-ready MS-word tables for two-way ANOVA.
Bandow H. A new co-solvent method for the green production Springerplus. 2015;4:33. doi:10.1186/s40064-015-0795-z
of biodiesel fuel - optimization and practical application. Fuel. 43. Motasemi F, Ani FN. A review on microwave-assisted produc-
2013;103:742-748. doi:10.1016/j.fuel.2012.09.029 tion of biodiesel. Renewable Sustainable Energy Rev. 2012;16:
30. Sharma YC, Singh B, Korstad J. High yield and conversion of 4719-4733. doi:10.1016/j.rser.2012.03.069
Biodiesel from a nonedible feedstock (Pongamia pinnata). 44. Thanh LT, Okitsu K, Sadanaga Y, Takenaka N, Maeda Y,
J Agric Food Chem. 2010;58:242-247. doi:10.1021/jf903227e Bandow H. Ultrasound-assisted production of biodiesel fuel
31. Georgogianni KG, Kontominas MG, Pomonis PJ, Avlonitis D, from vegetable oils in a small scale circulation process. Bio-
Gergis V. Conventional and in situ transesterification of sun- resour Technol. 2010;101:639-645. doi:10.1016/j.biortech.2009.
flower seed oil for the production of biodiesel. Fuel Process 08.050
Technol. 2008;89:503-509. doi:10.1016/j.fuproc.2007.10.004 45. Deshmane VG, Gogate PR, Pandit AB. Ultrasound assisted syn-
32. Stavarache C, Vinatoru M, Maeda Y, Bandow H. Ultrasonically thesis of isopropyl esters from palm fatty acid distillate. Ultra-
driven continuous process for vegetable oil transesterification. son Sonochem. 2009;16:345-350. doi:10.1016/j.ultsonch.2008.
Ultrason Sonochem. 2007;14:413-417. doi:10.1016/j.ultsonch. 09.004
2006.09.014 46. Lee I, Johnson LA, Hammond EG. Use of branched-chain esters
33. Zhao Z, Xue Y, Xu G, et al. Reaction conditions of ultrasound- to reduce the crystallization temperature of biodiesel. J Am Oil
assisted production of biodiesel: a review. Int J Energy Res. Chem Soc. 1995;72:1155-1160. doi:10.1007/BF02540982
2017;41:1081-1095. doi:10.1002/er.3673 47. Kongmany S, Truong HT, Hanh LTN, et al. Photodegradation
34. Osmieri L, Esfahani RAM, Recasens F. Continuous biodiesel of 12-deoxy-13,16-dioxy phorbol esters in aqueous solution
production in supercritical two-step process: phase equilibrium under UV light and natural sunlight. J Environ Chem Eng.
and process design. J Supercrit Fluids. 2017;124:57-71. doi:10. 2017;5:6045-6052. doi:10.1016/j.jece.2017.11.006
1016/j.supflu.2017.01.010 48. Truong Thi H. Utilization of oil-rich biomass for the produc-
35. Encinar JM, Pardal A, Martínez G. Transesterification of rape- tion of biodiesel fuel and exploration of the multi-beneficial
seed oil in subcritical methanol conditions. Fuel Process components; 2018.
Technol. 2012;94:40-46. doi:10.1016/j.fuproc.2011.10.018 49. Wu L, Huang K, Wei T, Lin Z, Zou Y, Tong Z. Process intensifi-
36. Ma F, Clements LD, Hanna MA. Biodiesel fuel from animal cation of NaOH-catalyzed transesterification for biodiesel pro-
fat. Ancillary studies on transesterification of beef tallow. Ind duction by the use of bentonite and co-solvent (diethyl ether).
Eng Chem Res. 1998;37:3768-3771. doi:10.1021/ie980162s Fuel. 2016;186:597-604. doi:10.1016/j.fuel.2016.08.106
37. Simonelli G, Júnior JMF, Pires CADM, dos Santos LCL. Biodie-
sel production using co-solvents: a review. Res Soc Dev. 2020;9:
1-27. doi:10.33448/rsd-v9i1.1672
38. Maeda Y, Thanh LT, Imamura K, et al. New technology for the How to cite this article: Pham T Thi Phuong,
production of biodiesel fuel. Green Chem. 2011;13:1124-1128. Nguyen U Huynh Phuong, Imamura K, Furuta M,
doi:10.1039/c1gc15049a Maeda Y. Transesterification using isopropanol as
39. Le HNT, Imamura K, Furuta M, Van Boi L, Maeda Y. Produc-
a co-solvent for the production of green biodiesel
tion of biodiesel from Vernicia Montana Lour. Oil using a co-
fuel. Int J Energy Res. 2021;1-10.
solvent method and the subsequent evaluation of its stability
during storage. Green Process Synth. 2018;7:170-179. doi:10. doi:10.1002/er.7432
1515/gps-2016-0215