In situ monitoring of turbid immobilized lipase-catalyzed esterification of oleic acid using fiber-optic Raman spectroscopy

Article abstract of DOI:10.1016/j.cattod.2009.09.001

Raman spectroscopy with a fiber-optic probe was used to monitor the immobilized Candida antarctica lipase B (Novozym 435)-catalyzed esterification of oleic acid with ethanol in iso-octane as an organic solvent. The effects of reaction temperature and molar ratio of substrates were studied. An optimal reaction rate was found at 60 °C, beyond which deactivation due to denaturing of the lipase was observed. The variation in molar ratio of substrates suggests that the esterification of oleic acid and ethanol proceeds via a Ping-Pong Bi-Bi mechanism with ethanol exhibiting an inhibitory effect. A good fit was obtained between the experimental results and the best-fit Ping-Pong Bi-Bi mechanism. This current work shows that fiber-optic Raman spectroscopy is indeed a suitable instrument to monitor immobilized lipase-catalyzed reaction in turbid organic systems in situ. Since this approach is reliable, simple to use, allows automatic data acquisition, and accurate, it should be applicable to the detailed kinetic analysis on other immobilized enzymatic reactions as well.

Full text of DOI:10.1016/j.cattod.2009.09.001

Catalysis Today 155 (2010) 223–226
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
In situ monitoring of turbid immobilized lipase-catalyzed esterification of oleic
acid using fiber-optic Raman spectroscopy
Erick Elfanso a, Marc Garland b, Kai Chee Loh a, M.M. Rahman Talukder b, , Effendi Widjaja
*
b,**
a Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, Singapore 117576, Singapore
b Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (A*STAR), 1 Pesek Road, Jurong Island, Singapore 627833, Singapore
A R T I C L E I N F O
A B S T R A C T
Article history:
Raman spectroscopy with a fiber-optic probe was used to monitor the immobilized Candida antarctica
lipase B (Novozym 435)-catalyzed esterification of oleic acid with ethanol in iso-octane as an organic
solvent. The effects of reaction temperature and molar ratio of substrates were studied. An optimal
reaction rate was found at 60 8C, beyond which deactivation due to denaturing of the lipase was
observed. The variation in molar ratio of substrates suggests that the esterification of oleic acid and
ethanol proceeds via a Ping-Pong Bi-Bi mechanism with ethanol exhibiting an inhibitory effect. A good fit
was obtained between the experimental results and the best-fit Ping-Pong Bi-Bi mechanism. This current
Available online 9 October 2009
Keywords:
Esterification
Lipase
Novozym 435
In situ monitoring
Fiber-optic Raman spectroscopy
Turbid systems
work shows that fiber-optic Raman spectroscopy is indeed
a suitable instrument to monitor
immobilized lipase-catalyzed reaction in turbid organic systems in situ. Since this approach is reliable,
simple to use, allows automatic data acquisition, and accurate, it should be applicable to the detailed
kinetic analysis on other immobilized enzymatic reactions as well.
ß 2009 Elsevier B.V. All rights reserved.
1. Introduction
organic functional groups are pronounced. Consequently, Raman
spectroscopy readily allows detection of other constituents present
Owing to the advantages offered by enzymes as catalysts in
organic solvent, the interest in lipase-catalyzed reactions, such as
esterification, inter-esterification, and trans-esterification, has
increased significantly [1]. It has also prompted chemical
industries to exploit the catalytic characteristics of lipases isolated
from various microbial sources. Off-line chromatography is one of
the most common analytical techniques employed to investigate
the kinetics of enzymatic reactions. However, this conventional
technique generally requires more sample preparation, longer
analysis time, and it is also invasive.
Alternative analytical techniques, such as in situ spectroscopy are
certainly preferred, since the progress of the reaction can be
monitored in real time. In recent years, various in situ spectroscopies
have been used to monitor enzymatic reactions. These included
Fourier transform infrared spectroscopy (FTIR) [2], ultraviolet (UV)
spectroscopy[3,4], nuclearmagneticresonance(NMR)spectroscopy
[5,6], and Raman spectroscopy [7,8]. The use of Raman spectroscopy
to monitor reactions in the aqueous phase is advantageous, since
water is a poor Raman scatterer, and hence the vibrations of the
intheaqueoussystem[9]. Thisisincontrasttoinfraredspectroscopy
which suffers from the intense infrared bands of water, and thus
identification and quantification of other constituents observed in
aqueous system can be rather difficult.
In the present study, we investigate the use of fiber-optic
Raman spectroscopy to monitor in situ the lipase-catalyzed
synthesis of ethyl oleate by the esterification of oleic acid and
ethanol in iso-octane. Since water is being formed as one of the
products from this esterification reaction, Raman spectroscopy is
probably the preferable choice. Since this is a multi-phase system,
involving suspended immobilized enzymes, a flow-through Raman
cell may not be appropriate. Instead, a Raman fiber-optic probe
immersed into the reaction system may provide a better approach.
The present contribution attempts to demonstrate the usefulness
and the reliability of this approach to monitor turbid enzymatic
systems and obtain quantitative information. The effects of
reaction temperature and molar ratio of substrates are also
studied and reported.
2. Materials and methods
2.1. Materials
* Corresponding author. Tel.: +65 6796 3826; fax: +65 6316 6182.
** Corresponding author. Tel.: +65 6796 3943; fax: +65 6316 6185.
E-mail addresses: rahman_talukder@ices.a-star.edu.sg
All substrates (oleic acid and ethanol), solvent (iso-octane), and
(M.M. Rahman Talukder), effendi_widjaja@ices.a-star.edu.sg (E. Widjaja).
immobilized Candida antarctica lipase B (Novozym 435) were
0920-5861/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2009.09.001

224
E. Elfanso et al. / Catalysis Today 155 (2010) 223–226
purchased from Sigma–Aldrich. All chemicals were of analytical
grade and used as received.
2.2. Reaction temperature study
The standard reaction mixture consisted of 10 g of oleic acid,
1.95 g of ethanol and 0.4 g of Novozym 435 in 10 ml of iso-octane
as an organic solvent. The molar ratio (mol ethanol/mol oleic acid)
was fixed at 1.2 for all runs. The reaction mixtures were incubated
at various different temperatures (25, 30, 35, 40, 50, 60, and 65 8C).
The mixture was then agitated continuously at the speed of
200 rpm throughout the reaction course.
2.3. Molar ratio study
Fig. 1. Pure component Raman reference spectra of pure substances (dashed double
dotted curve refers to iso-octane; dotted curve refers to ethanol, solid black curve
refers to oleic acid, and dashed curve refers to ethyl oleate).
The standard reaction mixture consisted of 0.4 g C. antarctica
lipase in 10 ml of iso-octane as the organic solvent at 40 8C. The
reaction mixtures were reacted under different molar ratio of
substrates. Two sets of molar ratio experiments were performed.
The first set (A) was carried out by varying the amount of ethanol
while keeping the amount of oleic acid fixed at 10 g (mol ethanol/
mol oleic acid = 1, 1.2, 1.5, 2, 2.5 and 3). The second set (B) was
carried out by varying amount of oleic acid while keeping the
amount of ethanol fixed at 1.95 g (mol oleic acid/mol ethanol = 1,
1.5, 1.7, 2, and 2.5).
standard initial reaction conditions (see Sections 2.2 and 2.3
experiment A). A similar curve was generated for the other series of
experiments (Section 2.3 experiment B).
2.6. Effect of turbidity on spectral quality
The presence of suspended immobilized enzyme results in
turbidity of the reaction mixture. The Raman measurements of the
reaction mixture are also effected by this increased light scattering.
This results in a slightly reduced intensity for the constituents
present, but more importantly, to a decrease in the signal-to-noise
of the signals. Fig. 3 shows a Raman measurement of one of the
calibration solutions and a representative measurement of a
reaction mixture. This figure confirms that rather good quality
Raman probe measurements can be made of the turbid reaction
solution, however, at a reduced signal-to-noise ratio.
2.4. Experimental set up
The reactions were carried out in a 3-neck 50-ml jacketed and
thermostated glass reactor. Stirring was achieved with a magnetic
stirrer controlled at 200 rpm. The enzyme-catalyzed reaction was
then monitored in situ throughout the reaction course, using a
dispersive fiber-optic probe attached to an inVia Reflex Raman
microscope, Renishaw, UK. The laser source employed for Raman
scattering was a near infrared laser (785 nm) with a power setting
of 100%. The spectral range was 1300–1800 cmꢀ1 with a resolution
of circa 1.1 cmꢀ1. Each Raman spectrum was measured with a 10 s
acquisition time and the interval between each scan was 60 s.
3. Results and discussion
3.1. Effect of reaction temperature
2.5. Spectral analysis
The time-dependent concentration profiles for ethyl oleate at
different reaction temperatures are shown in Figs. 4 and 5. The
results are separated in order to more clearly show the effect of
temperature. The maximum concentrations of ethyl oleate shown
in Figs. 3 and 4 correspond to circa 85–90% yield. The ‘‘noisy’’
concentration profiles are due in part to the turbidity of the
reaction mixtures and the resulting decrease in signal-to-noise
ratio in the spectra.
The collected in situ Raman spectra were processed and
analyzed to determine the percentage conversion of substrates
to ethyl oleate ester. Since the raw spectra acquired from the
reaction mixture were a combination of Raman scattering signals,
spikes due to cosmic rays, and some autofluorescence background,
some spectral preprocessing was needed to generate the Raman
spectral alone. This spectral preprocessing includes noise removal
by adjacent 3-point averaging, spikes removal due to cosmic rays,
and baseline correction using third order modified polynomial
fitting.
The time-dependent concentration profiles of ethyl oleate were
then used to calculate the initial reaction velocity. Fig. 6 shows the
Fig. 1 shows the pure component Raman reference spectra of
iso-octane, ethanol, oleic acid and ethyl oleate.
As shown in Fig. 1, there is considerable similarity between the
Raman spectra of pure oleic acid and pure ethyl oleate, although
the latter has a band around 1743 cmꢀ1 due to the ester group (see
small circle in Fig. 1). The spectral peaks enclosed by the large
ellipse indicate the presence of carbonyl groups in both ethyl
oleate and oleic acid. Therefore, in order to evaluate the reaction
progress, the ratio of the integrated ester group band area to the
total integrated carbonyl band area was used.
A calibration was performed in order to obtain the relationship
between ratio of integrated areas and conversion. The calibration
data were generated with known amounts of the substrates
(ethanol and oleic acid) and product (ethyl oleate). Fig. 2 provides
the calibration curve used for evaluating conversions with the
Fig. 2. A calibration curve for experiment A.

E. Elfanso et al. / Catalysis Today 155 (2010) 223–226
225
Fig. 5. Concentration of ethyl oleate at various reaction temperatures. Arrow
indicates increasing temperatures (50, 60 and 65 8C).
Fig. 3. A Raman spectrum of one the calibration solutions (top) and a representative
Raman spectrum from one of the reaction mixtures (bottom) with lipase added.
plot of initial reaction velocity obtained at various reaction
temperatures. The initial velocity is observed to increase steadily
with respect to increasing temperature till around 60 8C. This effect
of reaction temperature suggests that the optimal reaction
temperature is circa 60 8C. Beyond this temperature, the initial
reaction velocity starts to decrease due to thermal denaturing of
the lipase.
3.2. Effect of molar ratio
Fig. 6. Effect of reaction temperature on initial reaction velocity.
Similar experiments were conducted by varying the molar
ratios of reactants (see Section 2.3). The initial velocity of the
reactions for different molar ratio of substrates (mol ethanol/mol
oleic acid) is shown in Fig. 7.
with the result found by Chulalaksananukul et al. [10] that ethanol
acts as substrate inhibitor to lipase enzyme activity at higher
concentration.
Fig. 7 indicates that the initial reaction velocity reaches a
maximum at a ratio of 1.2 (mol ethanol/mol oleic acid), after which
it decreases. The decrease in reaction velocity at higher ratios
indicates that increasing ethanol indeed inhibits Novozym 435
enzyme activity. This was further verified by plotting the reciprocal
of initial reaction rate against reciprocal of ethanol concentration
(shown in Fig. 8). Since the plot shown in Fig. 8 is not a straight line,
but instead hyperbolic, the result indicates substrate inhibition.
This plot is indeed quite similar to the Lineweaver-Burk plot in the
presence of substrate inhibitor. This result is also in agreement
The effect of the excess of oleic acid on the initial velocity of the
esterification reaction is shown in Fig. 9. As seen, an increase in the
ratio of oleic acid to ethanol increases the initial reaction velocity.
This result indicates that oleic acid does not exhibit any inhibitory
behavior on the present reaction.
Since the inhibition effect of ethanol was only found at low
concentrations of oleic acid, i.e. higher molar ratio (mol ethanol/
mol oleic acid) of 2, 2.5 and 3, as shown in Fig. 7, it appears that a
Ping-Pong Bi-Bi mechanism is underlying the kinetics. No ternary
complex is expected to be formed. In this manner, lipase reacts
Fig. 7. The effect of molar ratio (mol ethanol/mol oleic acid) on the initial reaction
Fig. 4. Concentration of ethyl oleate at various reaction temperatures. Arrow
indicates increasing temperatures 25, 30, 35, 40 and 50 8C.
velocity (varying amount of ethanol at fixed amount of oleic acid).

226
E. Elfanso et al. / Catalysis Today 155 (2010) 223–226
velocity yo obtained from equation 1 and from the spectroscopic
measurements. Corana’s simulated annealing [13] was chosen as
the global optimizer for present problem, since it has been proven
to be quite robust in obtaining global solutions for highly nonlinear
optimization cases. The major advantage of this optimization
approach over the conservative graphical method is that less
experimental data is generally needed, but the parameters
modeled prove to be quite accurate. Using the experimental
results obtained from 10 different experiment runs, the values of
KmA
, KmB, KsiB, and V were calculated. These values are
2.51E5 ꢁ 4.07E4 M, 1.53Eꢀ2 ꢁ 3.04Eꢀ2 M, 0.13 ꢁ 0.03 M, and
224 ꢁ 111 mol gꢀ1 minꢀ1, respectively. The sum of squared error
between predicted and actual experimental data is circa 7.68Eꢀ9.
Fig. 8. The reciprocal initial velocity of the reaction versus reciprocal ethanol
4. Conclusion
concentration.
In the present study, it has been demonstrated that fiber-optic
Raman spectroscopy can be used to facilitate the kinetic
investigation of immobilized enzymatic reaction in turbid systems.
Although the signal-to-noise ratio of the Raman measurements
were reduced somewhat due to the turbidity, spectral measure-
ments in the range of 1300–1800 cmꢀ1 permitted quantitative
assessment of the carbonyl and ester vibrational bands associated
with oleic acid and ethyl oleate. This information was subse-
quently used to obtain the concentration profiles of ethyl oleate as
a function of time and the values of initial reaction velocity. The
experimental runs conducted at various reaction temperatures
suggested an optimal reaction temperature of 60 8C. The experi-
mental runs with varying substrates ratios suggest that this
reaction undergoes a Ping-Pong Bi-Bi mechanism with an ethanol
exhibiting inhibitory effect. The present analytical approach to
monitor turbid immobilized enzymatic reaction is rather general
and not restricted to the present reaction alone. It appears
applicable to a wide range of enzymatic reactions conducted in
either aqueous or organic solvents.
Fig. 9. The effect of molar ratio (mol ethanol/mol oleic acid) on the initial reaction
velocity (varying amount of oleic acid at fixed amount of ethanol).
with oleic acid to form a lipase-oleic acid complex. The lipase–oleic
acid complex then transforms to a carboxylic-lipase intermediate
releasing water. This is then followed by the binding of carboxylic-
lipase intermediate to ethanol to form another binary complex,
and subsequently yielding ethyl oleate and free lipase. A similar
deduction on this reaction sequence has been suggested by
Chulalaksananukul et al. [10].
Acknowledgement
This work was supported by Agency for Science, Technology
and Research (A*STAR), Singapore.
References
Since the present esterification reaction follows a Ping-Pong Bi-
Bi mechanism with the dead-end inhibition by ethanol, the
equation for initial velocity of reaction can be written as follows:
[1] S.H. Krishna, N.G. Karanth, Catal. Rev. 44 (2002) 499.
[2] M. Haberkorn, P. Hinsmann, B. Lendl, Analyst 127 (2002) 109.
[3] A. Kandelbauer, O. Maute, R.W. Kessler, A. Erlacher, G.M. Guebitz, Biotechnol.
Bioeng. 87 (2004) 552.
[4] M.P.A. Ribeiro, T.F. Padua, O.D. Leite, R.L.C. Giordano, R.C. Giordano, Chemom.
Intell. Lab. Syst. 90 (2008) 169.
Vab
yo
¼
(1)
K
mBa þ KmAbð1 þ b=KsiBÞ þ ab
where b refers to the inhibitory substrate (ethanol) concentration,
a refers to oleic acid concentration, V is the maximum possible of
reaction rate when A (oleic acid) and B (ethanol) are both
saturating; yo is the initial reaction velocity; KmA, known as the
limiting Michaelis constant for A, is the concentration of A which
gives 1/2 V when B is saturating; KmB, known as the limiting
Michaelis constant for B, is the concentration of B which gives 1/2 V
when A is saturating; KsiB is the dissociation constant of EB, with E
referring to enzyme. Further details of the ping-pong bi-bi
mechanism can be found in the detailed article by Chulalaksana-
nukul et al. [10], as well as [11,12].
[5] M.E. Amato, G. Ansanelli, S. Fisichella, R. Lamanna, G. Scarlata, A.P. Sobolev, A.
Segre, J. Agric. Food Chem. 52 (2004) 823.
[6] Z.Y. Wang, T. Nozawa, in: J.-N. Barbotin, J.-C. Portais (Eds.), NMR in Microbiology:
Theory and Applications, Horizon Scientific Press, Amiens, 2000, p. 221 (Chapter 10).
[7] E. Horvath, J. Gajari, J. Kristof, A. Redey, L. Kocsis, Anal. Chim. Acta 370 (1998) 191.
[8] M. Lopez-Pastor, A. Dominguez-Vidal, M.J. Ayora-Canada, B. Lendl, M. Valcarcel,
Microchem. J. 87 (2007) 93.
[9] T.I. Williams, T.W. Collette, in: I.R. Lewis, H.G.M. Edwards (Eds.), Handbook of
Raman Spectroscopy: From The Research Laboratory to the Process Line, CRC
Press, New York, 2001, p. 683.
[10] W. Chulalaksananukul, J.S. Condoret, P. Delorme, R.M. Willemot, FEBS Lett. 276
(1990) 181.
[11] A. Marty, W. Chulalaksananukul, R.M. Willemot, J.S. Condoret, Biotechnol. Bioeng.
39 (1992) 273.
[12] I.H. Segel, Enzyme Kinetics. Behavior and Analysis of Rapid Equilibrium and
Steady-State Enzyme Systems, Wiley–Interscience, USA, 1975.
[13] A. Corana, M. Marchesi, C. Martini, S. Ridella, ACM Trans. Math. Softw. 13 (1987)
262.
In order to determine the kinetic parameters (KmA, KmB, KsiB, and
V) of the above equation, a global optimization approach was
employed to minimize the differences between the initial reaction