Fuel 140 (2017) 724–730

Contents lists available at ScienceDirect

Fuel
j o u r n a l h o m e p a g e : w w w . e l s ev i e r . c o m / l o c a t e / f u e l

Determination of the kinetics of biodiesel saponification in alcoholic

hydroxide solutions

Valentine C. Eze, Adam P. Harvey, Anh N. Phan ⇑

School of Chemical Engineering & Advanced Materials (CEAM), Newcastle University, NE1 7RU, UK

highlights

Quantitatively investigated kinetics of fatty acid alkyl esters saponification in alcoholic hydroxides.
FAME saponification occurred 5–7 times more rapidly in ethanol than in methanol.
Found actual saponification rate constants for the methanol–KOH and ethanol–KOH.
Found activation energies for the FAME saponification in the alcoholic hydroxides.
Provided useful data in understanding/modelling base-catalysed transesterification reactions.

article info abstract

Article history: The research focused mainly on kinetics of the side reactions occurring during homogeneous base-catalysed transesterification
Received 8 July 2016 processes, which were not studied previously. The rates of fatty acid methyl esters (FAME) saponification with sodium and
Received in revised form 18 September 2016 potassium hydroxides were investigated in methanol and ethanol (commonly used in biodiesel production) at temperatures of
40, 50 and 60 LC. The effect of water on the rate of FAME saponification in the alcoholic hydroxide solutions was also
Accepted 1 October 2017 Available studied (at 60 LC). The apparent rate constants of FAME saponification were strongly influenced by the types of alcohol, but
online 18 October 2017 little affected by variation of the metal hydroxides. The apparent rate constants for FAME saponification using the ethanol
hydroxides were 5–7 times higher than for methanol hydroxides. This was due to the lower acidity of ethanol, leading to a
Keywords: higher concentration of hydroxide ions in the ethanol–hydroxide solu-tions. The rate constant of the FAME saponification
Kinetics
increased with temperature. For example, it was 1.31 L mol 1 min 1 at 40 LC, but 3.08 and 5.63 L mol 1 min 1 at 50 and 60 LC
Transesterification
FAME Saponification in the methanol–NaOH solu-tion. Arrhenius activation energies for the FAME saponification in the alcoholic hydroxides were
Rate constants in the range of 60.3–64.0 kJ mol 1. It was found that the FAME saponification rate in ethanol–hydroxides was 3.5 times higher
Hydroxide-alkoxide equillibrium than that in methanol–hydroxides in the presence of water. This indicates that use of methanol should be preferred over
ethanol in alkali-catalysed transesterification, as the amount of soap produced by saponification of triglycerides and alkyl
esters will be lower.

Crown Copyright 2017 Published by Elsevier Ltd. All rights reserved.

1. Introduction MG + ROH -> AEþGL (3)

Transesterification of triglycerides with alcohols with/without catalyst is
one method for producing biodiesel, a renewable alter-native to diesel. It
proceeds via three reversible stepwise reactions [1,2] as follows
TA + ROH -> AEþDG
DG + ROH -> AEþMG
⇑ Corresponding author. Tel.: +44 0191 208 5212. E-mail
address: anh.phan@ncl.ac.uk (A.N. Phan).
Overall reaction:
TG + 3ROH -> 3AE + GL (4)
(TG: triglyceride; DG: diglyceride; MG: monoglyceride; GL:
glycerol; ROH: alcohol; AE: alkyl esters).
(1) In alkali-catalysed homogeneous transesterifications, hydrox-ides and
methoxides of sodium and potassium are commonly used. In a hydroxide-
(2) catalysed transesterification process, alongside three main transesterification
steps (Eqs. (1)–(3)), there can also be saponification of the oil (TG) and the
alkyl esters (AE) (Eqs. (6) and (7)) to form soap. This occurs alongside the
main reactions due to the existence of the hydroxide-alkoxide equilibrium
(Eq.

http://dx.doi.org/10.1016/j.fuel.2014.10.001
0016-2361/Crown Copyright 2017 Published by Elsevier Ltd. All rights reserved.
 V.C. Eze et al. / Fuel 140 (2017) 724–730
(5)). This equilibrium shifts towards the formation of hydroxide
when the water content is increased. The formation of soap leads
to emulsification, which renders the downstream separation of
glycerol very difficult. The deleterious effect of water means that
in practice transesterification is limited to very dry (<0.5 wt% mois-
ture) vegetable oils.
Keq
ROHþOH $RO þH2O ð5Þ
TG þ 3OH ! 3Soap þ GL ð6Þ
AE þ OH ! Soap þ ROH ð7Þ
Saponification of alkyl esters can occur in aqueous hydroxide
solutions [3], mixtures of water and soluble organic solvents, such
as dimethyl sulphoxide (DMSO) [4], ethanol [5,6] and mixtures of
protic solvents, e.g. methanol and dichloromethane [7]. It has been
reported that saponification of carboxylic acid esters in hydroxide
solutions is a bimolecular reaction, which is first order for ester and
hydroxide ions [3,8]. Hydroxide ions attack the carbonyl to form a
tetrahedral intermediate, which then decomposes to the products
(Fig. 1a). However, recent studies [7,9] found that the reactions
did not occur via simple bimolecular collisions, but rather required
a molecule of water to form the tetrahedral intermediate. This indi-
cates that a water molecule stabilises the transition-state complex
through hydrogen bonding (Fig. 1b). Such hydrogen bonding could
also be provided by other protic solvents such as simple alcohols.
The rate of saponification of alkyl esters was strongly depen-
dent on steric and electronic interactions of the esters and metal
hydroxides with the solvent [7]. Such interactions determine the
rate of saponification in various solvent environments. For
instance, the hydroxide anions are highly solvated because of the
electronic charge in a predominantly aqueous and protic solvent
environment. The solvation of the OH anions reduces the rate of
direct collisions between the anions and the alkyl ester molecules
in the solution [10]. This phenomenon was used to explain the
higher rate of saponification of esters in 0.5 M KOH prepared in
80–90% (v/v) DMSO in water than that of 0.5 M KOH prepared in
80–90% (v/v) ethanol in water [4]. 63.5% of the ester was saponified
in a 0.5 M KOH solution prepared in 83.5% (v/v) aqueous DMSO,
whereas only 9.2% was saponified in a 0.5 M KOH prepared in
83.5% (v/v) aqueous ethanol after 3 min at the same conditions.
However, the differences in the rates of saponification of the esters
in 0.5 M KOH prepared in 83.5% (v/v) aqueous DMSO and 0.5 M
KOH prepared in 83.5% (v/v) aqueous ethanol were not entirely
due to the reduction in the desolvation energy of the OH in the
aqueous DMSO solution. The authors did not take into account
the reactions of the KOH with ethanol to form potassium ethoxide
(Fig. 2), which reduces the amount of KOH in the solutions and
consequently lowering the rate of saponification.
(a) OH
- -
R C
OR' + OH R C
-
O O
(b) OR' OR'
- -
R C
+ O H R C O
-
O O O O
H H
Fig. 1. Saponification of alkyl esters (a) reaction mechanism for alkyl esters saponification; (b) bimolecular collision by solvated hydroxide ions to form H-bond stabilised tetrahedral intermediate.
725
CH3CH2OH + KOH ↔ CH3CH2OK + H2O
Fig. 2. Ethanol–KOH reaction.
The depletion of metal hydroxide in a saponification mixture
containing alcohol where carboxylic acid esters were saponified
using 0.1–0.5 M NaOH prepared in approximately 9:1 (v/v) of CH 2-
Cl2/methanol was reported [7]. The presence of sodium methoxide
formed by the reaction of NaOH with methanol lead to the transe-
sterification of the original esters to form ‘‘methyl ester’’ [7]. How-
ever, the authors explained these results due to ‘‘hydroxide ions
assisted’’ bimolecular collisions of methanol and the original
esters.
Furthermore, the magnitudes of activation energies for alkyl
esters saponification indicate that the differences in the desolva-
tion energies of the OH in aqueous/protic and aprotic solvents
was not entirely responsible for the higher rates of alkyl esters
saponification in dipolar aprotic co-solvents. For instance, the
Arrhenius activation energies of 41.40–63.18 kJ mol 1 [3,5,6,11]
for alkyl esters saponification in aqueous/protic medium were
too small to account for the desolvation energy of 423.4 kJ mol 1
[9] required for the desolvation of OH in the aqueous medium.
The values of the activation energies for alky esters saponification
suggest that the direct collision of desolvated OH ions is
implausible.
Although extensive research on the kinetics of alkaline hydroly-
sis of esters has been done, no study has been carried out on the
kinetics of fatty acid alkyl esters saponification in alcohol hydrox-
ides. Such investigation is essential to understanding the compet-
itive saponification of biodiesel produced during alkali-catalysed
homogeneous transesterification. As methanol and ethanol are
commonly used for the biodiesel production, kinetic studies of
saponification of fatty acid alkyl esters (biodiesel) in these alco-
holic media provide fundamental in choosing alcohol and/or oper-
ating conditions to achieve high yield and quality product. In this
study, the kinetics of fatty acid methyl esters (FAME) saponifica-
tion in both methanol–hydroxide and ethanol–hydroxide solutions
of NaOH and KOH at a temperature range of 40–60 LC were inves-
tigated. These conditions are similar to those used in commercial
homogeneously catalysed biodiesel processes. The effect of water
on the rate of the FAME saponification in the alcoholic NaOH and
KOH solutions was also studied.
2. Materials and methods
2.1. Materials
Refined RSO (Henry Colbeck, UK) used in production of the
FAME contained approximately 98.7 wt% triglycerides, 0.03 wt%
OR' R C O + R'OH
O
H R C
O + R'OH + H2O
O
H H
726 V.C. Eze et al. / Fuel 140 (2017) 724–730
moisture and 0.05 wt% free fatty acids. The measured oil’s viscosity and
density at 60 LC were 25.5 mPas and 890 kg/m3 [12].
Fatty acid methyl esters (97.85 wt% purity), known as FAME, were
obtained via a continuous alkaline catalysed transesterifica-tion process at a
molar ratio of methanol to rapeseed oil of 6:1, reaction temperature of 60 LC
as described in previous studied [12]. FAME consisted mainly of esters of
oleic acid (63.2%), linoleic acid (19.8%), linolenic acid (8.6%) and less than
10 wt% of others (palmitic acid: 5.3%, stearic acid: 1.7% and icosenoic acid:
1.4%).
Sodium hydroxide (97% purity), potassium hydroxide (90% purity),
methanol (99.8% purity), hydrochloric acid (36.5–38.0% purity), potassium
hydrogen phthalate (99.5% purity) and other analytical standard chemicals
such as methyl heptadecanoate (99.0% purity), methyl oleate (99.0% purity)
were purchased from Sigma–Aldrich. Ethanol (99.99% purity) was obtained
from Fisher Scientific.
2.2. Experimental methods
FAME was dried in oven for 12 h at 105 LC to remove all water. The
dried FAME was then saponified with 0.5 M NaOH (or KOH) in alcoholic
solutions using a 100 ml three-neck jacketed glass reac-tor connecting to a
water condenser and a thermocouple for mon-itoring reaction temperature as
shown in Fig. 3.
Approximately 75 ml of 0.5 M alcoholic hydroxide solution was heated in
the reactor to the desired temperature of 40, 50 or 60 LC by circulating hot
water inside the jacket. As soon as the desired temperature reached, 6 g of
pre-heated FAME was transferred into the reactor and mixed vigorously using
a magnetic stirrer. The ratio of FAME and hydroxide alcohol solution (w/v) in
this study was adapted from the saponification procedure of fats and oils [13].
The same procedure was also used to obtain the weight percentage of FAME
in alcohol solution in absence of hydroxide – without saponification (Wo).
About 0.5 ml of the reaction sample was collected using a 1000 lm
micropipette at various time intervals for 60 min. The sample was transferred
into a pre-weighed 2 ml vial containing 0.5 ml HCl (0.5 M) to quench the
reaction immediately. All the samples were stored in a freezer for further
analysis.
2.3. Analysis
FAME content in the sample collected during the experiments were
analysed using a 5890 Hewlett Packard Series II gas chroma-tography (GC)
equipped with a 30 m length, 0.25 lm thickness and
Condenser
Sampling port
3-neck glass
jacketed reactor
Magnetic stirrer
Fig. 3. Experimental set-up for FAME saponification.
0.25 mm diameter BPX70 CP wax capillary column. The oven tem-perature
was set at 210 LC (isothermal) while the injector and FID detector
temperatures were set at 250 and 260 LC, respectively. The amounts of
FAME in the samples were quantified using the British standard for
quantification of FAME content (BS EN 14103: 2003). Errors in FAME
analysis were ±1.3%. The FAME saponification experiments at 60 LC using
methanol–KOH were repeated three times for reproducibility (errors less than
±1.6%). Conversion of the FAME (±0.012) to soap was calculated as follows:
Wo Wt
FAME conversion to soap ðXÞ: X ¼ Wo ð8Þ
(wt: weight fraction of FAME in the sample; wo: weight fraction of FAME at
zero saponification).
3. Results and discussion
3.1. Effect of alcohols and alkali on the rates of the FAME
saponification
Fatty acid alkyl esters saponification is an undesired reaction during
alkali-catalysed transesterification, as it hinders the separa-tion process,
reduces yield and reduces the quality of the raw bio-diesel product. As
presented in Table 1, KOH has a stronger effect on the saponification of
FAME than NaOH in methanol solution. For instance, at 60 LC reaction
temperature, the FAME conversion was 0.06 for KOH, but was only 0.03 for
NaOH after 2 min. The finding in this study was similar to others [14], who
reported that saponification was much faster for KOH than for NaOH. The
differ-ence became smaller when the reaction time was above 30 min.
Generally, the effect of temperature and reaction time on the saponification of
FAME was similar for both metal hydroxides. Increasing the reaction
temperature increased the rate of conver-sion of FAME into soap, e.g. the
conversion was 0.14 at 40 LC but 0.34 at 60 LC for methanol–NaOH at 30
min reaction time.
Because of the similarity of KOH and NaOH as shown in the Table 1, the
following discussion only focuses on the effect of the alcoholic hydroxides of
the potassium solutions. As shown in Fig. 4, the conversion of the FAME to
soap at the reaction times of 0–30 min was 4 times faster for ethanol–
hydroxide solutions than for methanol–hydroxide solutions for all tested
reaction tem-peratures. For example, the FAME conversion was 0.1 for
metha-nol–KOH, but 0.43 for ethanol–KOH at 50 LC and 10 min reaction
time (Fig. 4(b)). At 60 LC, the conversion was 0.36 for methanol– KOH but
was 0.86 for ethanol–KOH after 30 min reaction time (Fig. 4(c)). This agreed
well with other findings which the amount of soap formation in both ester-
rich phase and glycerine-rich phase was 3.5 times higher when using ethanol
than when using metha-nol for alkaline catalysed transesterification [14].
Menegheti et al.
[17] obtained higher fatty acid alkyl esters conversion and lower soap
formation when using methanol than ethanol. The higher rates of
saponification in the ethanol–hydroxide systems accounts for the lower ester
contents in the ethanolysis.
The higher conversions of the FAME to soap obtained for etha-nol–KOH
in the Fig. 4 can be explained by the existence of more hydroxide ions (OH )
in the ethanol–KOH than in the methanol– KOH due to the stronger acidity of
methanol than ethanol (4.4 times higher), which will be discussed later in
detail (Section 3.3). As a result, the formation of methoxide ions (CH3O ) is
more favourable in a methanol–KOH system, resulting in less free hydroxide
ions (OH ) for the FAME saponification, than that of ethoxide ions (C 2H5O )
in an ethanol–KOH system. Therefore, the higher rates of FAME
saponification in the ethanol–KOH solution was as a result of higher amount
of available hydroxide ions concentration in the solution. This study provided
a full explanation about the reduction
 V.C. Eze et al. / Fuel 140 (2017) 724–730 727

Table 1
Effect of sodium hydroxide and potassium hydroxide in methanol on the conversion of FAME into soap at temperatures of 40 LC, 50 LC and 60 LC.

Temperature, LC Conversion of FAME into soap (X)

0.5 M NaOH solution prepared in 99.8% methanol 0.5 M KOH solution prepared in 99.8% methanol

Reaction time (min) 60 LC 50 LC 40 LC 60 LC 50 LC 40 LC

0 0.00 0.00 0.00 0.00 0.00 0.00
0.5 0.01 0.00 0.00 0.01 0.01 0.00
2 0.03 0.04 0.02 0.06 0.04 0.01
5 0.10 0.04 0.04 0.14 0.06 0.05
10 0.16 0.12 0.07 0.20 0.14 0.08
20 0.26 0.19 0.11 0.30 0.20 0.10
30 0.34 0.26 0.14 0.36 0.27 0.16
40 0.40 0.30 0.17 0.42 0.32 0.19
50 0.47 0.34 0.20 0.50 0.38 0.22
60 0.53 0.38 0.23 0.53 0.40 0.24

1.0 1.0
(a) 40oC Methanol solution (b) 50oC Methanol solution
Ethanol solution Ethanol solution
0.8 0.8
Soap formation

Soap formation
0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Reaction time (mins) Reaction time (mins)

1.0
(c) 60oC Methanol solution
Ethanol solution
0.8
Soap formation

0.6

0.4

0.2

0.0
0 10 20 30 40 50 60
Reaction time (mins)

Fig. 4. FAME saponification in 0.5 M alcohol–KOH solutions at a molar ratio of FAME to KOH of 1:1.85 and reaction temperatures of (a) 40 LC, (b) 50 LC and (c) 60 LC.

of fatty acid alkyl esters in alkali-catalysed biodiesel production with
methanol [12,15,16] and with ethanol [14,17] at greater reaction times,
particularly at high catalyst concentrations. Previously this has been either
explained qualitatively [12,14,16,17] or wrongly attributed to increased
reverse transeste-rification [15].
Menegheti et al. [17] and Mendow et al. [14] also observed that more soap
formation and consequently less yield of fatty acid alkyl esters was obtained
in the ethanolysis of sunflower oil than the methanolysis. An explanation for
the increase in saponification rate was due to higher solubility of the
hydroxide catalyst in the oil-ethyl esters phase than in the oil-methyl esters
phase [14]. However, the results in this study indicate that the higher rate of
saponification in ethanol than that in methanol is due to higher amount of
hydroxide ions (OH ). This suggests that the choice of alcohol for
homogeneous alkali-catalysed biodiesel process must
be restricted to short chain alcohols. The longer chain (higher molecular
weight) alcohols such as propanol and butanol contain even more mole
fractions of hydroxide ions in the alcoholic hydroxide solution, which
accelerate saponification reactions of the biodiesel. Therefore, in order to
minimise saponification reac-tions, methanol must be chosen in homogeneous
alkali-catalysed transesterification. In addition, fatty acid methyl esters have
lower viscosity than other alkyl esters produced from ethanol and longer
chain alcohols [18].
3.2. Kinetics of fatty acid methyl esters (FAME) saponification in the
alcoholic bases
Kinetic studies of the fatty acid alkyl esters saponification in association
with biodiesel production have not been reported in literature. The aim of this
section is to explain quantitatively the
728 V.C. Eze et al. / Fuel 140 (2017) 724–730

reduction in fatty acid alkyl esters in homogeneously alkali-catalysed R0OH þ OH R0O þ H2O ð4Þ
biodiesel production at various reaction times and catalyst concentrations.
The amount of hydroxide ions available for the saponification of FAME is
a fraction of the original hydroxide dissolved in the alco-hol. Caldin and Long
The saponification rate of FAME in an alcoholic hydroxide solu-tion was
[19] reported that the hydroxide ion fraction available in the 0.5 M ethanol–
analysed based on bimolecular collisions of the metal hydroxide and FAME,
sodium hydroxide at 25 LC was 0.037. This hydroxide ion fraction was
with FAME as the limiting reactant [3,8,9].
assumed to be same for the 0.5 M ethanol–potassium hydroxide at the same
The saponification rate of FAME in alcoholic hydroxide solution can be
condition. In this study, the hydroxide ion fraction of 0.037 was used to
written as:
estimate the actual rate constants for the FAME saponification in the etha-
dXME nol–hydroxides. The actual rate constants of the FAME saponifica-tion in the
methanol–hydroxides were calculated based on the relative acidity of ethanol
ð1 XMEÞðhB XMEÞ ¼ kbdt ð9Þ and methanol, e.g. methanol was 4.4 times more acidic than ethanol at 26 LC
Integrating Eq. (9): [21]. The estimated rela-tive apparent saponification rate constants of
1 X ethanol–hydroxides to methanol–hydroxides were 5.396, 5.423 and 6.938 at
h

¼ kbt ¼ k0t 40, 50 and 60 LC, respectively. These corresponded to 0.0069, 0.0068 and
ð B Þ
ME

1 ME ln 1 X
ðhB Þ½ o h Bð MEÞ
Eq. (10) could be simplified as:
A lnB ¼ k0t
where k = kb is the apparent saponification rate constant (L mol 1 min 1), k is
actual saponification rate constant (L mol 1 min 1), b is mole fraction of
unreacted metal hydroxide in the alcoholic hydroxide solutions, ME is Fatty
acid methyl esters, [ME]o is initial concentration of FAME, (mol/L), XME is
conversion of FAME to soap, hB is the ratio of the initial concentration of
alcoholic hydroxide solutions to FAME.
The apparent rate constants for the saponification of FAME at
temperatures of 40, 50 and 60 LC were determined using Eq. (10) and the
experimental data obtained from Table 1 and Fig. 4. The correlation
coefficients (R2) of all the plots used to estimate the rate constants were more
than 0.99. There was a linear relationship between the saponification rate
constants and temperature for all the tested temperatures. As shown in Table
2, the apparent rate constants (k0) increased approximately a factor of 2 for
every 10 LC increase in the temperature, e.g. it increased from 0.0536 L mol 1
min 1 at 40 LC to 0.114 L mol 1 min 1 at 50 LC and
ð10Þ
0.0053, mole fractions hydroxide ions in the methanol– hydroxide solutions at
40, 50 and 60 LC. Table 3 shows the actual rate constants of the FAME
saponification in the alcoholic hydrox-ide solutions using these hydroxide
ions fractions.
The actual FAME saponification rate constants (k) were inde-pendent of
both alcohol and metal hydroxide but depended on temperatures as shown in
Table 3. The Arrhenius activation ener-gies of the FAME saponification
ranged from 60 to 64 kJ mol 1, within the range reported in the literature for
alkyl esters saponi-fication [3,5,6].
The actual saponification rate constants for the ethyl acetate, ethyl
propanoate and ethyl laurate esters in the Table 4 were cal-culated using the
data from Caldin and Long [19] for sodium hydroxide–ethoxide–water
solution and the apparent rate con-stant for the saponification in 85 wt%
aqueous ethanol [5,6]. As listed in Table 4, the alkyl ester saponification rate
constants are strongly dependent upon the carboxylic acid chain length. The
saponification rates constants at 50 LC (55 LC) were 8.07 L mol 1 min 1 for
acetic acid ester, 4.10 L mol 1 min 1 for propanoic acid ester and 3.17 L mol 1
min 1 for lauric acid ester.

0.2152 L mol 1 min 1 at 60 LC for the ethanol–potassium hydrox-ide solution.

This indicates that the saponification of the fatty acid alkyl esters is an
endothermic reaction. The difference in the appar-ent FAME saponification
rate constants between NaOH and KOH was small, e.g. the apparent rate
constant was 0.105 L mol 1 min 1 for the ethanol–NaOH and 0.114 L mol 1
min 1 for ethanol–KOH at 50 LC reaction temperature (8% difference), but
significant between methanol and ethanol. The apparent rate constant for
FAME sapon-ification was 5–7 times higher for the ethanol–hydroxide than
for the methanol–hydroxide. This was due to more hydroxide ions available
in the ethanol–hydroxide solution than in the metha-nol–hydroxide solution as
explained above.
3.4. Effect of water in the alcoholic hydroxide solutions on FAME
saponification
Anhydrous ethanol (99.99% purity) and methanol (99.8% purity) were
used in order to determine the effect of alkaline catalysts on the saponification
of FAME during transesterification, and to eval-uate the effect of water on the
saponification. The results for the FAME saponification at a reaction
temperature of 60 LC using alco-holic hydroxide solutions containing 0–10%
(v/v) of water in the alcohols, are presented in the Figs. 5 and 6.

3.3. The hydroxide–alkoxide equilibrium for the alcoholic solutions of the
NaOH and KOH
When a metal hydroxide is dissolved in an alcoholic solution, there exists
an equilibrium between hydroxide and ethoxide ions as shown below:
Figs. 5 and 6 clearly show that water accelerates the rates of FAME
saponification. For instance, at 2 min reaction time, the FAME conversions to
soap in the methanol–hydroxides increased rapidly from 0.03 to 0.10, 0.17
and then to 0.21 for the metha-nol–NaOH; and from 0.06 to 0.13, 0.19 and to
0.26 for the metha-nol–KOH when the water content rose from 0% to 2, 4 and
10 vol% (Fig. 5). The rapid increase in the FAME conversion to soap in the

Table 3
Table 2
Actual rate constants (k) and activation energy (Ea) for the saponification of FAME in 0.5 M
Apparent rate constants (k0 ) for the methyl esters saponification in 0.5 M alcoholic hydroxide
alcoholic hydroxide solutions at various temperatures (L mol 1 min 1).
solutions at various temperatures (L mol 1 min 1).
1
Solution k (L mol min 1) Activation energy (kJ mol 1)
Solution k0 (L mol 1 min 1)
40 LC 50 LC 60 LC
40 LC 50 LC 60 LC
Methanol–NaOH 1.313 3.078 5.626 61.73
Methanol–NaOH 0.0090 0.0208 0.0300
Methanol–KOH [20] 1.444 2.921 5.926 61.16
Methanol–KOH [20] 0.0099 0.0195 0.0316
Ethanol–NaOH 1.308 2.838 5.723 64.01
Ethanol–NaOH 0.0484 0.1050 0.2120
Ethanol–KOH 1.449 3.081 5.816 60.26
Ethanol–KOH 0.0536 0.1140 0.2152
 V.C. Eze et al. / Fuel 140 (2017) 724–730 729

Table 4
Comparison of the kinetic parameters for the FAME saponification with that reported in the literature for other alkyl esters.

Alkyl esters and the saponification media k (L mol 1 min 1) Ea (kJ mol 1) References
40 LC 50 LC
Ethyl acetate and 0.05 M NaOH in aqueous ethanol 3.88 8.07 61.50 Smith and Levenson [6]
Ethyl propanoate and 0.05 M NaOH in aqueous ethanol 1.98 4.10 61.50 Smith and Levenson [6]
Ethyl laurate and 0.05 M NaOH in aqueous ethanol 1.57 @ 45 LC 3.17 @ 55 LC 63.18 Levenson and Smith [5]
FAME and 0.5 M methanol–KOH 1.44 2.92 61.16 Eze et al. [20]
FAME and 0.5 M methanol–NaOH 1.31 3.08 61.73 This study
FAME and 0.5 M ethanol–NaOH 1.31 2.84 64.01 This study
FAME and 0.5 M ethanol–KOH 1.45 3.08 60.26 This study

presence of water is due to the reaction of the water with alkoxide ions in the 1.0
hydroxide–alkoxide equilibrium (Eq. (4)), producing more hydroxide ions, by
Le Chatelier’s principle. The rate of the FAME saponification in methanol–
hydroxide (Fig. 5) was again below that of the ethanol–hydroxide shown in
0.8
Fig. 6, because methanol has a higher acidity than ethanol [21], resulting in

to soap
fewer hydroxide ions being available for saponification. As shown in Table 5,
the apparent rate constants of the FAME saponification increased with
0.6
increasing water content, e.g. from 0.03 to 0.399 L mol 1 min 1 for methanol–

E conversion
NaOH and from 0.212 to 1.494 L mol 1 min 1 for the ethanol–NaOH, when
the water con-tent rose from 0 to 10 vol%. This corresponds to an increase in 0.4
the apparent rate constants by factors of 3.7% and 12.7% per unit increase in
the vol% water in the alcohol for the methanol–NaOH and ethanol–NaOH, FAM Ethanol-NaOH solution
respectively. The deleterious effects of water on alkali-catalysed 0% (v/v)
homogeneous transesterification can be fully inferred from this study. The 0.2 2% (v/v)
increased hydroxide ions concentra-tion in the presence of water leads to the 4% (v/v)
10% (v/v)
saponification of both glycerides and fatty acid alkyl esters (biodiesel). Soap
formation due to the saponification of glyceride and biodiesel reduces the 0.0
biodiesel yield, and causes emulsification, which adversely affects glycerol 0 5 10 15 20 25 30 35
separation unit operations in the downstream processing [22,23]. To minimise Time (mins)
such reactions, anhydrous reactants must be used for the reactions. Some
studies suggest that reactants that contain some moisture (less than 0.3 wt%) 1.0
can be used in the base-catalysed homogeneous transesterification [24], such
process achieves less biodiesel conversions due to the increased triglycer-ide
and biodiesel saponification in the presence of water.
0.8
FAME conversion soap
to

0.6
The actual rate constants (k) for the FAME saponification were expected
to be similar at the fixed reaction temperature of 60 LC. The increases in the
apparent rate constants (k0) when increased water content in Table 5 are 0.4 Ethanol-KOH solution
plainly due to the increase in the mole

0% (v/v)
1.0 0.2 2% (v/v)
4% (v/v)
conversion tosoap

10% (v/v)

0.8 0.0
0 5 10 15 20 25 30 35
Time (mins)
0.6
Fig. 6. Effect of water level on FAME saponification in 0.5 M solutions of ethanol– NaOH and
ethanol–KOH at a molar ratio of FAME to hydroxide of 1:1.85, 60 LC.
0.4
FAME

0.2 2% (v/v) fractions (b) of the hydroxides (k0 = kb). The actual rate constants for the
0% (v/v) fatty acids alkyl esters saponification at the various water contents, calculated
4% (v/v) using the hydroxide–alkoxide equilibrium in the Eq. (4), were consistent with
10% (v/v) the values in the Table 3. For instance, the actual rate constants for the FAME
0.0 saponification were 5.808 L mol 1 min 1 at 2% (v/v) water, 5.942 L mol 1 min
0 10 20 30 40 50 60 70 1 at 4% (v/v) water and 5.598 L mol 1 min 1 at 10% (v/v) water, for the
Time (mins)
saponification of FAME in 0.5 M NaOH–methanol solutions containing
Fig. 5. Effect of water level on FAME saponification in 0.5 M solutions of methanol– NaOH at
various levels of water.
a molar ratio FAME to hydroxide of 1:1.85, 60 LC.
730 V.C. Eze et al. / Fuel 140 (2017) 724–730
Table 5
Apparent rate constants (k0 ) for fatty acid alkyl saponification at 60 LC using alcoholic hydroxide solutions with various vol% water concentrations in the alcohols (L mol 1 min 1).
Alkyl esters in alcoholic solutions k0 (L mol 1 min
0% (v/v)
FAME in 0.5 M NaOH–methanol 0.0300
FAME in 0.5 M KOH–methanol [20] 0.0316
FAME in 0.5 M NaOH–ethanol 0.2120
FAME in 0.5 M KOH–ethanol 0.2152
Ethyl laurate in 0.05 M NaOH–ethanol [5] – –
a
Calculated for 12 vol% water and 55 LC reaction temperature.
4. Conclusion
The kinetics of fatty acid methyl esters (FAME) saponification with
sodium and potassium hydroxides (NaOH and KOH) in meth-anol and
ethanol solutions were studied over the range 40–60 LC. This is similar to the
range used in commercial biodiesel produc-tion. This study provides for a
clear understanding of the saponifi-cation side reactions occurring in
homogeneous base-catalysed transesterification processes.
FAME saponification occurred 5–7 times more rapidly in etha-nol–
hydroxide than in methanol–hydroxide solutions. This was attributed to the
greater availability hydroxide ions in the etha-nol–hydroxide solution than in
the methanol–hydroxide solution, as a result of the lower acidity of the
ethanol. However, the actual rate constants for the FAME saponification in
the 2 alcoholic hydroxides were similar. For instance, the apparent FAME
saponi-fication rate constants at 60 LC were 0.032 L mol 1 min 1 for the
methanol–KOH and 0.215 L mol 1 min 1 for the ethanol–KOH; whereas the
estimated actual saponification rate constants were 5.93 L mol 1 min 1 and
5.82 L mol 1 min 1, for the methanol– KOH and ethanol–KOH, respectively.
Arrhenius activation energies for the FAME saponification in the alcoholic
hydroxides were in the range 60.26–64.01 kJ mol 1. This is the first time that
thorough and quantitative investigation has been carried out on the kinetics of
fatty acid alkyl esters saponification in alcoholic hydroxides. Hence, these
kinetic studies provide useful data in understanding and modelling
homogeneous base-catalysed transesterification reactions. Previously, only
qualitative explanations were given to account for the saponification side
reactions during the base-catalysed transesterification processes. Based on the
findings in this study, it is recommended that methanol should be chosen in
preference to ethanol in homogeneous base-catalysed transesteri-fications, to
increase the yield and purity of biodiesel produced, as its use results in
significantly lower rates of saponification.
In base-catalysed transesterification processes, triglyceride and biodiesel
saponification occurs in the presence of water, which may lead to
emulsification. Anhydrous reactants are, therefore, rec-ommended for
homogenous alkali-catalysed biodiesel production.
The increased rates of FAME saponification with increasing water content
of the alcohol can be used for rapid quantitative determinations of
saponification values of esters. Furthermore, the trend of increasing hydroxide
ion concentration, and therefore saponification rate, with increasing alcohol
chain length suggests that longer chain alcohols, such as propanol, may be
used for rapid, quantitative determinations of the saponification values of
esters and vegetable oils.
1 )
2% (v/v) 4% (v/v) 10% (v/v)
0.116 0.198 0.399
0.104 0.175 0.344
0.517 0.768 1.494
0.464 0.647 1.340
– 1.122a
References
[1] Bambase ME, Nakamura N, Tanaka J, Matsumura M. Kinetics of hydroxide-catalyzed
methanolysis of crude sunflower oil for the production of fuel-grade methyl esters. J
Chem Technol Biotechnol 2007;82:273–80.
[2] Darnoko D, Cheryan M. Kinetics of palm oil transesterification in a batch reactor. J Am
Oil Chem Soc 2000;77:1263–7.
[3] Das K, Sahoo P, Sai Baba M, Murali N, Swaminathan P. Kinetic studies on saponification
of ethyl acetate using an innovative conductivity-monitoring instrument with a pulsating
sensor. Int J of Chem Kinet 2011;43:648–56.
[4] Vinson JA, Fritz JS, Kingsbury CA. Quantitative determination of esters by saponification
in dimethyl sulphoxide. Talanta 1966;13:1673–7.
[5] Levenson HS, Smith HA. Kinetics of the saponification of the ethyl esters of several
phenyl substituted aliphatic acids. J Am Chem Soc 1940:2324–7.
[6] Smith HA, Levenson HS. Kinetics of the saponification of the ethyl esters of normal
aliphatic acids. J Am Chem Soc 1939;61:1172–5.
[7] Theodorou V, Skobridis K, Tzakos AG, Ragoussis V. A simple method for the alkaline
hydrolysis of esters. Tetrahedron Lett 2007;48:8230–3.
[8] Price CC, Michel RH. Rates of saponification of substituted ethyl 2-naphthoates. J Am
Chem Soc 1952;74:3652–7.
[9] Mata-Segreda JF. Hydroxide as general base in the saponification of ethyl acetate. J Am
Chem Soc 2002;124:2259–62.
[10] Matta MS, Toenjes AA. Solvation effects on the alkaline hydrolysis of some p-
nitrophenyl esters. J Am Chem Soc 1985;107:7591–6.
[11] Schneider MA, Stoessel F. Determination of the kinetic parameters of fast exothermal
reactions using a novel microreactor-based calorimeter. Chem Eng J 2005;115:73–83.
[12] Phan AN, Harvey AP, Eze V. Rapid production of biodiesel in mesoscale oscillatory
baffled reactors. Chem Eng Technol 2012;35:1214–20.
[13] Firestone D. Official methods and recommended practices of the American Oil Chemists’
Society. 5th ed. Champaign, Illinois: AOCS Press; 1998.
[14] Mendow G, Veizaga NS, Querini CA. Ethyl ester production by homogeneous alkaline
transesterification: influence of the catalyst. Bioresource Technol 2011;102:6385–91.
[15] Darnoko D, Cheryan M. Continuous production of palm methyl esters. J Am Oil Chem
Soc 2000;77:1269–72.
[16] Keera ST, El Sabagh SM, Taman AR. Transesterification of vegetable oil to biodiesel fuel
using alkaline catalyst. Fuel 2011;90:42–7.
[17] Meneghetti SMP, Meneghetti MR, Wolf CR, Silva EC, Lima GES, de Lira Silva L, et al.
Biodiesel from castor oil: a comparison of ethanolysis versus methanolysis. Energ Fuels
2006;20:2262–5.
[18] Knothe G. Analyzing biodiesel: standards and other methods. J Am Oil Chem Soc
2006;83:823–33.
[19] Caldin EF, Long G. The equilibrium between ethoxide and hydroxide ions in ethanol and
in ethanol–water mixtures. J Chem Soc 1954:3737–42.
[20] Eze VC, Phan AN, Harvey AP. A more robust model of the biodiesel reaction, allowing
identification of process conditions for significantly enhanced rate and water tolerance.
Bioresource Technol 2014;156:222–31.
[21] Reeve W, Erikson CM, Aluotto PF. A new method for the determination of the relative
acidities of alcohols in alcoholic solutions: the nucleophilicities and competitive
reactivities of alkoxides and phenoxides. Can J Chem 1979;57:2747–54.
[22] Demirbas A. Biodiesel production from vegetable oils via catalytic and non-catalytic
supercritical methanol transesterification methods. Prog Energ Combust 2005;31:466–87.
[23] Demirbas A. Progress and recent trends in biodiesel fuels. Energ Convers Manage
2009;50:14–34.
[24] Freedman B, Pryde EH, Mounts TL. Variables affecting the yields of fatty esters from
transesterified vegetable oils. J Am Oil ChemSoc 1984;61:1638–43.