Académique Documents
Professionnel Documents
Culture Documents
www.elsevier.com/locate/procbio
Abstract
Experimental determination of the separate effects of palm oil and methanol concentrations on the rate of their enzymatic transesterification
was used to propose suitable mechanismic steps and to test the generated kinetic model. The reaction took place in n-hexane organic medium and
the lipase used was from Mucor miehei. At a constant methanol concentration of 300 mol m3, it was found that, initially as the palm oil
concentration increased, the initial reaction rate increased. However, the initial rate dropped sharply at substrate concentrations larger than
1250 mol m3. Similar behaviour was observed for methanol concentration effect, where at a constant substrate concentration of 1000 mol m3,
the initial rate of reaction dropped at methanol concentrations larger than 3000 mol m3. Ping Pong Bi Bi mechanism with inhibition by both
reactants was adopted as it best explains the experimental findings. A mathematical model was developed from a proposed kinetic mechanism and
was used to identify the regions where the effect of inhibition by both substrates arised. The proposed model equation is essential for predicting the
rate of methanolysis of palm oil in a batch or a continuous reactor and for determining the optimal conditions for biodiesel production.
# 2007 Elsevier Ltd. All rights reserved.
and Shimada et al. [7] have both reported that fatty acids methyl practical experimental set-up. A standard 1.5 l agitated
esters (FAME) production increased when methanol concen- fermentor having a working volume of 1 l to determine the
tration increased up to an oil to methanol ratio of 3:1 separate effects of substrate and alcohol concentrations was
equivalents and then decreased when methanol concentration used, which constitute a promising industrial expansion. The
was further increased. This was also found by Noureddini et al. work also included proposing a suitable kinetic model and
[8] although the ratio was much lower (7.5:1, equivalents). numerical determination of the constants found in this model,
Lipases efficiently catalysed reactions when the substrates using the collected experimental data.
dissolved each other. It was believed that when methanol
amount exceeded its solubility limits, lipase was deactivated by 2. Materials and methods
the insoluble methanol that existed as drops in the oil, as
proteins in general are unstable in short-chain alcohols. 2.1. Materials
In addition to substrates concentrations, the temperature was
Palm oil was obtained from Yee Lee Edible Oils Private Limited, Malaysia.
also reported to affect the rate of reaction considerably. It was The substrate considered was the ester bond on the triglyceride. Analytical
found that the reaction rate initially increased as the grade n-hexane and methanol were obtained from Fisher Scientific, UK. Liquid
temperature increased. That was mainly due to the increase lipase (EC 3.1.1.3) from M. miehei (claimed activity 100 kLU ml1) was
in rate constants with temperature and partly due to decrease in obtained from Novo Nordisk, Denmark. The enzyme chosen was the one that
mass transfer resistances. However, the reaction rate decreased maintains its activity under the highest water content, in view to use it for waste
oil in the future [30,31]. The reuse of lipase is essential from the economic point
sharply at the onset of denaturation of the enzyme [9]. This of view, which can be achieved by using the lipase in immobilised form.
trend was observed in all the studies that investigated the effect However, soluble enzyme and not immobilised was purposely used in this study
of temperature on the production of biodiesel by lipase. The to avoid the mass transfer limitations or clogging problems, which could
optimum operating temperature were of lipases from different complicate the kinetic model. Using soluble lipase has the advantage of having
sources, including the one used in this work (Mucor miehei) the interaction with the substrate be at molecular level, with no mass transfer
limitations or clogging problems. Therefore, it was justified to assume negli-
were determined to be close to 40 8C [2,3,8]. gible external mass transfer resistance in the subsequent analysis. The model
Although the application of lipase in the production of derived in this study could be used to determine a mass transfer resistance of
biodiesel from vegetable oils has been thoroughly addressed in immobilised enzyme of the same strain, however, this is beyond the scope of
the literature [2,4,6,1015], these studies were purely experi- this paper.
mental. On the other hand, all the kinetic studies found in the
literature considered the esterification of free fatty acids rather 2.2. Experimental set-up
than the transesterification of vegetable oil [1623]. The main
The reactor used was a 1.5 l batch stirred fermentor with four baffle plates
industrial interest, however, is on the production of biodiesel provided to reduce vertex effects and enhance the mixing (Miniforce,
from the triacylglyceride (oil), not the free fatty acids. INFORCE, Switzerland). Two rushton impellers attached to the same shaft
Therefore, it was desired to develop a reliable kinetic model were used for agitation. A schematic diagram of the experimental set-up is
for this important reaction. An earlier attempt to model shown in Fig. 1. A stirring of 150 rpm was kept constant throughout the
progress of all experiments. The temperature of the reaction system was kept at
vegetable oil transesterification was done by the author,
40 8C using a controlled heating metal jacket. The temperature used was that
assuming that the reaction took place in two consecutive steps presented in the literature to be the optimum [2,3] and the agitation speed was
[24,25]. In the first step, triglycerides were hydrolysed to chosen to provide suitable mixing without affecting the activity of the enzyme.
produce free fatty acids, and in the second step, the free fatty The working volume at the beginning of each run was 1 l, consisting of
acids produced in the first step were esterified to produce fatty different methanol concentrations in the range of 3005000 mol m3 and
substrate (i.e., the ester bond on the triglyceride palm oil) concentrations in the
acids methyl esters. However, in the current work it was shown
that it was more accurate to assume that transesterification takes
place by direct alcoholysis of the triglycerides.
The main difference between esterification of free fatty acids
and transesterification of triglycerides (oils) is that in the first
OH bonds are broken, whereas in the second ester bonds are
the ones that are broken. In addition, the by-product of
esterification is water whereas it is glycerol in transesterifica-
tion. In this study, experimental determination of the separate
effects of palm oil and methanol concentrations on the rate of
enzymatic transesterification were used to understand the
reaction behaviour and propose suitable mechanismic steps.
On the other hand, most of the experimental data reported in
literature were found using reactors of small volumes in the
range of 2 ml [2] to 50 ml [7,12,13,1518,2629]. The mixing
was achieved by orbital shakers or magnetic stirrers. It was
obvious that these experimental results could not represent the
large scale industrial reactors, and therefore it was desired to
generate more reliable experimental results using a more Fig. 1. Experimental set-up.
S. Al-Zuhair et al. / Process Biochemistry 42 (2007) 951960 953
range of 1001250 mol m3 in n-hexane. The reactor was covered tightly
throughout the progress of the experiments to prevent evaporation. After the
desired conditions were reached in the bioreactor, enzyme solution with an
activity of 20 LU was added to initiate the reaction. The enzyme activity was
chosen low enough to ensure a slow reaction rate to obtain more accurate
determination of the initial rate of reaction. If the rate was fast, the substrate
concentration would drop considerably within short times and result in
inaccurate estimation of the initial rate of reaction. In addition, if the enzyme
concentration was high, the effect of substrate concentrations would not be
clear, especially at enzyme concentrations close to the critical saturation
concentration [9].
At regular intervals, 5 ml sample was withdrawn from the reactor by a
special sampling port, using a syringe, for subsequent analysis. Directly after
being withdrawn, the sample was placed in a screw-capped bottle and heated to
100 8C for at least 5 min to denature the enzyme and stop the reaction. The
sample was then analysed for fatty acids methyl esters content. In the range of
substrate and alcohol concentrations considered in this study, the reaction
Fig. 2. Total fatty acids methyl esters concentration vs. time (initial substrate
mixture was completely homogeneous and was well agitated in the reactor.
and methanol concentrations of 250 and 300 mol m3, respectively).
4. Kinetic model
k2
Fig. 5. Initial rate of FAME production vs. initial palm oil ester bond con- E:Ac:G $ E:Ac G (3)
centration at constant methanol concentration of 300 mol m3. k2
S. Al-Zuhair et al. / Process Biochemistry 42 (2007) 951960 955
where, k1 and k1, k2 and k2 and k3 and k3 are the rate
constants for the reversible formation of enzymesubstrate
complex, E.S, acylated enzymeglycerol moiety complex,
E.Ac.G, acylated enzymealcohol complex, E.Ac.A, respec-
tively and k4 and k4 are the rate constants for the product
formation and enzyme regeneration.
To account for the inhibition by alcohol, competitive
inhibition was assumed when an alcohol molecule reacts with
Fig. 7. Comparison between the experimental values of the initial rate of lipase
the enzyme directly to produce a dead-end enzymealcohol methanolysis of palm oil and those predicted by Eq. (11) at constant alcohol
complex (E.A). And to account for the inhibition by the concentration of 300 mol m3 and different substrate concentrations.
substrate, competitive inhibition was also assumed when a
substrate molecule reacts with the acylated enzyme to produce
an equation to be used when the water [W], taken as one of the
another dead-end complex, namely, acylated enzymesubstrate
products, was assumed to inhibit the reaction. This modification
complex (E-Ac.S). The two competitive inhibition reactions
was applicable when free fatty acids were considered as the
mentioned above were presented by the following equations,
substrate. However, when the substrate was the triglyceride,
respectively:
water was replaced with monoglyceride, diglyceride or glycerol.
k5 And unlike water which is usually present in the reaction medium
E A $ E:A (7)
k5 at time zero, these products were not. Therefore, the product
inhibition was neglected, especially when considering the initial
k6
E:Ac S $ E:Ac:S (8) rate of reaction, as done in this study.
k6
Eq. (9) describes the rate of transesterification reaction of
where, k5 and k5 and k6 and k6 are the rate constants for the triglycerides using lipase, which was derived from the
reversible formation of dead-end complexes, enzymealcohol, mechanismic steps of the reaction (Eqs. (1)(8)). The equation
E.S, and acylated enzymesubstrate, respectively. is applicable for oil, alcohol and enzyme concentrations in the
With the above mechanism and assumptions, the reaction region of complete homogeneity and could be used to predict
rate presented in the following equation was derived: the optimal operating conditions. The kinetic model obtained
(Eq. (9)) is non-monotonic which have important implications
V max
y on optimal design. Under specific condition, the non-
1 K IA =A1 S=K S K IS =S1 A=K A monotonic kinetics might give rise to complex bifurcation
(9) behaviour when applied in a continuous processes; the simplest
where y is the initial reaction rate, Vmax is the maximum reaction of it is multiplicity of the steady states. This bifurcation occurs
rate, KS and KA are the dissociation constants for the substrate (S) when a small smooth change made to the parameter values
and the alcohol (A), respectively, and KIS and KIA are the
inhibition constants for the substrate and the alcohol, respec-
tively. Eq. (9) describes the initial reaction rate in the absence of
any product inhibition, which is similar to the one proposed by
Krishna and Karanth [12] for the esterification of short chain fatty
acids with alcohol. On the other hand, Janssen et al. [17] derived
Table 1
Comparison between the values of Vmax, KS, KA, KIS and KIA, from different
references
Reference
[12] [17] Current work
3 1
Vmax (mol m min ) 11.72 6.0 0.041
KS (mol m3) 3.03 10 6 44 430
KA (mol m3) 3.06 10 6 61 350
KIS (mol m3) 1050 N/A 4.45 104 Fig. 8. Comparison between the experimental values of the initial rate of lipase
KIA (mol m3) 6550 107 3.3 104 methanolysis of palm oil and those predicted by Eq. (11) at constant substrate
concentration of 1000 mol m3 and different alcohol concentrations.
S. Al-Zuhair et al. / Process Biochemistry 42 (2007) 951960 957
(e.g., substrate concentrations) of a system causes a sudden significance of their effects. The results shown in Figs. 3 and 4
qualitative or topological change in the systems long-term were used to determine the numerical values of the parameters
dynamical behaviour. The development of reaction kinetic found in Eq. (9). The kinetic parameters found for esterification
model, such as Eq. (9) would sturdily assist in predicting the of free fatty acids using lipase from similar strain [12], were
dynamic behaviour of the system and avoid any unexpected first assumed. The values were then optimised using Excel
responses. solver to find the minimum objective function (Eq. (10)) that
compares the measured rate of reaction with those predicted by
5. Determination of kinetic parameters the proposed kinetic equation (Eq. (9)):
X
In this study, it was proposed to consider the changes of both OF ypred yexpt 2 (10)
methanol and substrate concentrations to demonstrate the
Fig. 9. Effect of kinetic parameters on the initial rate of reaction predicted by Eq. (11) at different substrate concentrations.
958 S. Al-Zuhair et al. / Process Biochemistry 42 (2007) 951960
The estimated kinetic parameters for the transesterification of It was found that the dissociation constants for the substrate,
triglycerides compared to those for the esterification of free KS, and alcohol, KA, found in this study, were in the same
fatty acids (Table 1) [12,17]. The proposed model equation order of magnitude, which agreed with the previous results
(Eq. (9)) with the estimated constants is presented in the [12,17]. However, the inhibition constant for the substrate,
following equation: KIS, was found to be slightly higher than that for the alcohol,
KIA, which disagreed with that found earlier [12]. It is
0:041
y 4
2:8 105 important to notice that the comparison was done with
1 3:3 10 =A1 S=430 values extracted from experiments done on free fatty acids,
4:45 105 =S1 A=350 namely butyric acid, rather than the triglyceride, since the
(11) kinetic studies on triglycerides transesterification is lacking.
Fig. 10. Effect of kinetic parameters on the initial rate of reaction predicted by Eq. (11) at different alcohol concentrations.
S. Al-Zuhair et al. / Process Biochemistry 42 (2007) 951960 959
[S] substrate concentration (i.e., the ester bond on the [16] Watanabe Y, Shimada Y, Sugihara A, Tominaga Y. Conversion of
degummed soybean oil to biodiesel fuel with immobilized Candida
triglyceride) (mol m3)
antarctica lipase. J Mol Catal B Enzym 2002;17:1515.
Vmax maximum reaction rate (mol m3 min1) [17] Janssen AEM, Sjursnes BJ, Vakurov AV, Halling PJ. Kinetics of lipase-
catalyzed esterification in organic media: correct model and solvent
Greek letter effects on parameters. Enzyme Microbiol Technol 1999;24:46370.
y initial reaction rate (m3 mol1 min1) [18] Janssen AEM, Vaidya AM, Halling PJ. Substrate specificity and kinetics
of Candida rugosa lipase in organic media. Enzyme Microbiol Technol
1996;18:3406.
[19] Marty A, Chulalaksananukul W, Willemot RM, Condoret JS. Kinetics of
References lipase-catalyzed esterification in supercritical CO2. Biotechnol Bioeng
1992;39:27380.
[1] Tickell J. From the fryer to the fuel tank: the complete guide to using [20] Rizzi M, Stylos P, Riek A, Reuss M. A kinetic study of immobilized lipase
vegetable oil as an alternative fuel, 3rd ed., Tallahassee, FL: Tickell catalysing the synthesis of isoamyl acetate by transesterification in n-
Energy Consulting; 2000. hexane. Enzyme Microbiol Technol 1992;14:70914.
[2] Soumanou MM, Bornscheuer UT. Improvement in lipase-catalysed synth- [21] Van Tol JBA, Jongejan JA, Duine JA, Kierkels HGT, Gelade EFT, Mosterd
esis of fatty acid methyl esters from sunflower oil. Enzyme Microbiol F, et al. Thermodynamic and kinetic parameters of lipase-catalyzed ester
Technol 2003;33:97103. hydrolysis in biphasic systems with varying organic solvents. Biotechnol
[3] Fukuda H, Kondo A, Noda H. Biodiesel fuel production by transester- Bioeng 1995;48:17989.
ification of oils. J Biosci Bioeng 2001;9(25):40516. [22] Van Tol JBA, Kraayveld DE, Jongejan JA, Duine JA. The catalytic
[4] Kaieda M, Samukawa T, Matsumoto T, Ban K, Kondo A, Shimada Y, et al. performance of pig pancreas lipase in enantioselective transesterification
Biodiesel fuel production from plant oil catalyzed by Rhizopus oryzae in organic solvents. Biocatal Biotrans 1995;12:11936.
lipase in a water-containing system without an organic solvent. J Biosci [23] Van Tol JBA, Stevens RMM, Veldhuizen WJ, Jongejan JA, Duine JA. Do
Bioeng 1999;88(6):62731. organic solvents affect the catalytic properties of lipase? Intrinsic kinetic
[5] Lai CC, Zulaikah S, Vali SR, Ju Y. Lipase-catalyzed production of parameters of lipases in ester hydrolysis and formation in various organic
biodiesel from rice bran oil. J Chem Technol Biotechnol 2005;80: solvents. Biotechnol Bioeng 1995;47:7181.
3317. [24] Al-Zuhair S. The effect of substrate concentrations on the production of
[6] Samukawa T, Kaieda M, Matsumoto T, Ban K, Kondo A, Shimada Y, et al. biodiesel by lipase-catalysed transesterification of vegetable oils. J Chem
Pretreatment of immobilized Candida antarctica lipase for biodiesel fuel Technol Biotechnol 2006;81:299305.
production from plant oil. J Biosci Bioeng 2000;90(2):1803. [25] Al-Zuhair S. Production of biodiesel by lipase-catalysed transesterification
[7] Shimada Y, Watanabe Y, Sugihara A, Tominaga Y. Review: enzymatic of vegetable oils: a kinetics study. Biotechnol Prog 2005;21(5):14428.
alcoholysis for biodiesel fuel production and application of the reaction to [26] Yadav GD, Lathi PS. Synthesis of citronellol laurate in organic media
oil processing. J Mol Catal B Enzym 2002;17:13342. catalysed by immobilised lipases: kinetic study. J Mol Catal B Enzym
[8] Noureddini H, Gao X, Philkana RS. Immobilized Pseudomonas cepacia 2004;27:1139.
lipase for biodiesel fuel production from soybean oil. Bioresour Technol [27] Shah S, Sharma S, Gupta MN. Biodiesel preparation by lipase-catalysed
2005;96:76977. transesterification of jatrropha oil. Energy Fuels 2004;18:1549.
[9] Al-Zuhair S, Hasan M, Ramachandran KB. Kinetic hydrolysis of palm oil [28] Tweddell RJ, Kermasha S, Combes D, Marty A. Esterification and
using lipase. Process Biochem 2003;38:115563. interesterification activities of lipase from Rhizopus niveus and M. miehei
[10] Bjorkling F, Godtfredson SE, Kirk O. The future impact of industrial in three different types of organic media: a comparitive study. Enzyme
lipases. Trends Biotechnol 1991;9:3603. Microb Technol 1998;22:43945.
[11] Mukherjee KD. Lipase-catalyzed reactions for modification of fats and [29] Kontogianni A, Skouridou V, Sereti V, Stamatis H, Kolisis FN. Lipase-
other lipids. Biocatalysis 1990;3:27793. catalysed esterification of rutin and naringin with fatty acids of medium
[12] Krishna SH, Karanth NG. Lipase-catalyzed synthesis of isoamyl butyrate carbon chain. J Mol Catal B Enzym 2003;21:5962.
a kinetic study. Biochim Biophys Acta 2001;1547:2627. [30] Nelson LA, Foglia A, Marmer WN. Lipase-catalyzed production of
[13] Du W, Xu Y, Liu D, Zeng J. Comparative study on lipase-catalysed biodiesel. JAOCS 1996;73:11915.
transformation of soybean oil for biodiesel production with different acyl [31] Al-Zuhair S, Jayaraman KV, Krishnan S, Chan WH. The effect of fatty
acceptors. J Mol Catal B Enzym 2004;30:1259. acid concentration and water content on the production of biodiesel by
[14] Kaieda M, Samukawa T, Kondo A, Fukuda H. Effect of methanol lipase. Biochem Eng J 2006;30(2):2127.
and water contents on production of biodiesel fuel from plant oil [32] George S, Arumughan C. Positional distribution of fatty acids in the
catalyzed by various lipases in a solvent-free system. J Biosci Bioeng triacylglycerols of developing oil palm fruit. JAOCS 1993;70(12):12558.
2001;91:125. [33] Bailey JE, Ollis DF. Biochemical engineering fundamentals, 2nd ed.,
[15] Iso M, Chen B, Eguchi M, Kudo T, Shrestha S. Production of biodiesel fuel McGraw Hill; 1986.
from triglycerides and alcohol using immobilized lipase. J Mol Catal B [34] Panalotov I, Verger R. Physical chemistry of biological interfaces. New
Enzym 2001;16:538. York: Marcel Dekker Inc.; 2000.