Optimized enzymatic synthesis of levulinate ester

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.


Ethyl levulinate, produced through esterification of levulinic acid, is a ketoester with various applications. Synthesis of ethyl levulinate was carried out in solvent-free system using immobilized Candida antarctica lipase B (Novozym 435) as the biocatalyst for the reaction. Response surface methodology (RSM) based on a four-factor-five-level central composite rotatable design (CCRD) was used to study and optimize the reaction conditions in the synthesis of levulinate ester. The effect of four main reaction parameters including temperature, time, ethanol/levulinic acid molar ratio and amount of enzyme on the synthesis of ester were analyzed. A quadratic polynomial model was fitted to the data with an R2 of 0.8993. Model validation experiments show good correspondence between actual and predicted values. A high conversion yield (96.2%) is obtained at the optimum conditions of 51.4oC, 41.9 min, 292.3 mg enzyme amount and 1.1:1 alcohol:acid molar ratio.

Keywords: Levulinic acid; Enzymatic synthesis; Esterification; Solvent-free system; Response surface methodology


Levulinic acid, also known as gamma-ketovaleric acid, is a platform chemical with numerous potential applications by having a ketone and a carboxylic group (Fang, 2002; Bozell et al., 2000). It is commercially produced from renewable biomass such as cane sugar, starch and lignocellulosic materials from agricultural wastes. Ethyl levulinate is an industrially important derivative of levulinic acid, made by esterifying its carboxylic group with fuel-grade ethanol (Wetzel et al., 2006). The esterification reaction is usually carried out at high temperature in the presence of an acid catalyst such as sulphuric, polyphosphoric or p-toluenesulfonic acids. Ethyl levulinate has an oxygen content of 33% and properties similar to the biodiesel fatty acid methyl esters (FAME), which make it suitable to be used as an oxygenate diesel additive. Adding ethyl levulinate to the diesel results in a cleaner burning fuel with high lubricity, flashpoint stability, reduced sulphur content and improved viscosity that can be used in regular diesel engines (Hayes, 2009). Ethyl levulinate has also been applied in the flavoring and fragrance industries. It is a substrate for a variety of condensation and addition reactions at the ester and keto groups (Olson, 2001).

Application of enzymes to the synthesis of esters has increased extensively in recent years. Compared to conventional chemical esterification, enzymatic synthesis has several advantages such as mild reaction conditions, high productivity due to low energy requirements and minimal waste disposal, ease of product isolation, and biocatalyst reusability (Petersson et al., 2005). Enzyme-catalyzed synthesis of levulinate ester has been reported previously by Yadav and Borkar (2008). Their work focused on the kinetics and mechanism study of lipase-catalyzed esterification of levulinic acid with n-butanol using tetrabutyl methyl ether as the solvent. However, so far there is no report on lipase-catalyzed synthesis of levulinate esters in a solvent-free system. Moreover, neither specific information on the interactive effect of reaction parameters nor any detailed optimal conditions have been reported for these esters.

Although organic solvents provide several advantages in enzymatic reactions, their use in industrial processes is not desirable. They are a source of volatile organic compounds (VOCs), their use requires costly post-treatment actions, larger and more expensive reactors and auxiliary equipments, and they also have some inhibition effects on the enzyme (Tufvesson et al., 2007). Performing the reactions under solvent-free conditions can help to overcome these drawbacks. Furthermore, higher selectivity and volumetric productivity, improved substrate and product concentrations and fewer purification steps are some other advantages of using solvent-free system (Otero et al., 2001). In this study, synthesis of levulinate ester has been carried out in a solvent-free system.

In order to optimize the conversion of ester, response surface methodology (RSM) with central composite rotatable design (CCRD) was applied. RSM is a fast and economical statistical technique useful for developing, improving and optimizing processes (Myers et al., 2009). It has been successfully applied for the study and optimization of the enzymatic synthesis of various esters (Jeong et al., 2009; Keng et al., 2005).

The objective of the present work is to investigate the possibility of the lipase-catalyzed synthesis of levulinate ester in a solvent-free system. This study also helps to better understand relationships between the main reaction parameters (temperature, time, enzyme amount, and substrate molar ratio (alcohol:acid)) and the response (reaction yield) and to determine the optimal conditions of ethyl levulinate synthesis using CCRD and RSM analyses.



Novozym® 435 (specific activity of 10000 PLU/g) was purchased from NOVO Nordisk A/S (Bagsvaerd, Denmark) and consists of Candida antarctica lipase B (triacylglycerol hydrolase, EC physically adsorbed within the macroporous acrylic resin. Levulinic acid and ethanol were purchased from Merck Co. (Darmstadt, Germany). All other chemicals used in this study were of analytical grade.

Lipase-catalyzed esterification reaction

Different molar ratios of ethanol and levulinic acid (corresponding to the different substrate molar ratios generated by RSM) were mixed in 30 mL closed vials. Different amounts of Novozym 435 were subsequently added. The reaction was carried out in a horizontal water bath at 150 rpm at different temperatures and for different time periods, as presented in Table 1. The selection of Novozym 435 as catalyst for the reaction was based on the previous study (Yadav and Borkar, 2008) in which several commercial lipases including Novozym 435, Lipozyme RM IM and Lipozyme TL IM were screened for activity via lipase catalyzed synthesis of levulinate ester. Novozym 435 was found to be the best with maximum initial rate and conversion.

Analysis and characterization

The reaction was terminated by dilution with ethanol: acetone (50:50 v/v) and the enzyme was removed by filtration. The remaining unreacted acid was determined by titration with 0.1M NaOH to an end point of pH 8.25 using phenolphthalein as the indicator. The moles of acid reacted were calculated from the values obtained for the control (without enzyme) and the test samples. The ester formed was expressed as equivalent to conversion of the acid (Abdul Rahman et al., 2008; Radzi et al., 2005). Product was also identified by thin-layer chromatography (TLC) using chloroform as the solvent and gas chromatography/mass spectroscopy (GC/MS) on a Shimadzu (model GC 14B; model MS QP5050A; Shimadzu Corp, Tokyo, Japan) instrument equipped with a FID and a semi polar column BP-10 (0.33mm - 50m, 0.25 ?m). The carrier gas was nitrogen and the total gas flow rate was 50mL min-1. The injector and the detector temperatures were set at 280 and 310oC, respectively. The oven temperature was maintained at 100oC for 3 min, elevated to 300oC at a rate of 10oCmin-1 and held for 7 min.

Experimental design, statistical analysis and optimization

A four-factor-five-level central composite rotatable design (CCRD) was employed in this study. Rotatability indicates that the variation in the predicted response is constant at a given distance from the center point of the design (Anderson and Whitcomb, 2005). The total number of required experiments was 30 obtained by the following equation:

N=2k + 2k + n0

where k is the number of independent variables and n0 is the number of replicated center points (Mohammad et al., 2006). The fractional factorial design consisted of 16 factorial points, 8 axial points and 6 center points. Center point is repeated six times to give a good estimate of the experimental error. The variable and their levels selected for the synthesis of levulinate ester in solvent-free system were: temperature (25-75oC); time (30-240 min); amount of enzyme (20-400 mg) and substrate molar ratio of ethanol to levulinic acid (1:1-5:1). High and low levels of each variable were coded as 2 and -2, respectively, and the mean value was coded as zero (Table 1). The experiments were produced in random order and triplicate measurements of esterification percentage were run on each experiment. The design of experiments employed is presented in Table 2.

A software package of Design Expert Version 6.0.6 (State-Ease Inc., Statistics Made Easy, Minneapolis, MN, USA) was used to fit the data obtained for the response to a second-order polynomial model using the following equation:

where y is the dependent variable (percentage of yield) to be modeled, xi and xj are the independent variables (factors), b0, bi, bii and bij are the regression coefficients of model and e is the error of model. An analysis of variance (ANOVA) and R2 statistic were used to determine whether the produced model was adequate to describe the observed data. By using an F-test, it was possible to test the variation of the data around the fitted model (lack of fit). The significance of each factor in the model was estimated by testing the null hypothesis. Small P-value results in rejection of the null hypothesis, which means that the factor is significant. The optimal conditions for the synthesis of ester were generated using the software's numerical optimization function.


Identification of ester product

GC-MS analysis of the reaction mixture shows the presence of ethyl levulinate at a retention time of 10.098 min. The mass spectrum of the product exhibits molecular ion at m/z 144 that corresponds to molecular formula of ethyl levulinate (C7H12O3). The base peak of the fragmentation of the ester is related to CH3CO (m/z = 43). The other two important ion peaks are due to the formation of ion acylium, RCO+ that gives the fragment ions at m/z 99 (because of the loss of alkoxy group from the ester, R-O) and m/z 129 because of the loss of methyl group from the ester. Other bonds cleavage occur through some pathways and gave fragments ions at m/z 41, 56, 74, 84 and 116.

Model fitting and analysis of variance (ANOVA)

Fitting of the data to various models and their subsequent ANOVA shows that the reaction of levulinic acid and ethanol in solvent-free system is most suitably described with a quadratic polynomial model. The final equation of the model (based on the coded values) is as follows:

Yield (%) = +56.94 - 4.05 A + 0.79 B - 6.04 C + 14.11 D + 2.08 A2 + 2.21 C2 - 1.54 AD + 2.58 BC - 3.44 CD

where A is the temperature; B is the time; C is the substrate molar ratio and D is the amount of enzyme.

The ANOVA for the model is presented in Table 3. The computed F-value of the model (19.84) is higher than the tabular value of F9,20 (= 2.39), implying the model is significant at the 5% confidence level. A very small P-value (< 0.0001) and a suitable coefficient of determination (R2 = 0.8993) shows that the model can satisfactorily represent the real relationship among the reaction parameters (Gunawan et al., 2005). Figure 1 shows the experimental versus predicted yields obtained from the Equation 3. A linear distribution is observed which is indicative of a well-fitting model. Normal probability plot is also presented in Figure 1. The plot indicates that the residuals (difference between actual and predicted values) follow a normal distribution and form an approximately straight line. Adequate precision shows signal to noise ratio. Ratios greater than 4 are suitable. The adequate precision of the developed model is 23.231 indicating that the model can be used to navigate the design space. The lack of fit F-value of 3.09is lower than the tabular F15,5 -value (4.62), implying that there is no lack of fit in the model at 95% level of significance. The coefficients of the response surface model are also presented in Table 3. A P-value less than 0.05 indicates that the model term is significant. In this case A, C, D, and CD are significant terms. Equation 3 was used then to study the effect of various parameters and their interactions on the esterification yield.

Effect of reaction parameters

The effect of the four independent variables on the synthesis of levulinate ester is shown in Figure 2. The reaction yield gradually decreases from 63.0 to 55% by increasing temperature at the center point of other variables (Figure 2(a)). Increasing temperature causes increase in the acid solubility and dissociation and decrease in the binding equilibrium, leading to unfavorable esterification conditions (Hari Krishna et al., 2001). According to the analysis of variance, time is not a significant parameter that can influence the yield. This also can be observed in Figure 2(b) in which increasing time from 30 to 240 min causes a small increase in the yield from 56.1 to 57.7% due to equilibrium of the esterification reaction. In solvent-free systems, one substrate is generally used in a large excess over another in order to act as a solvent for other reactants (Yamane, 2001). Yadav and Borkar (2008) found that by increasing the mole ratio of n-butanol from 1 to 3, the conversion and initial rates are increased in solvent-based synthesis of levulinate ester. In this study, the mole ratio of ethanol was increased from 1 to 5. However, the highest conversion of ester was obtained at alcohol:acid molar ratio of 1:1 (Figure 2(c)). According to Carta et al. (1992), ethanol inhibits the catalytic function of lipase at even low concentrations. An irreversible denaturation of the enzyme also occurs at high concentrations of ethanol.

As the amount of enzyme is increased, the ester production is also increased (Figure 2(d)). The presence of higher amount of enzyme provides more active sites for the formation of the acyl-enzyme complex and also increases the probability of substrate-enzyme collision and subsequent reaction (Soo et al., 2004).

The interaction effects of two parameters on the synthesis of levulinate ester were examined by three-dimensional response surface plots (Figures 3-5). Contour plots are also very helpful for interpreting the main effects of reaction parameters and their mutual interactions (Myers et al., 2009). Contour plots representing the effect of varying parameters on the synthesis of the ester are shown in Figures 3-5.

Figure 3 shows the effect of temperature and enzyme amount on the synthesis of ester at 135 min and alcohol:acid molar ratio of 3:1. The effect of temperature is more significant at higher amounts of enzyme. Maximum yield can be obtained using more enzyme quantity at lower temperatures. Figure 4 represents the effect of time and substrate molar ratio on the synthesis of ethyl levulinate. Temperature and enzyme amount were fixed at their center points. The reaction with low substrate molar ratio (1:1) and low incubation time (30 min) gives the highest yield of 86.1%. In alcohol:acid molar ratio of 2.7:1, the yield is constant at 58.9% during 4 hours of the reaction, indicating that the equilibrium condition is achieved. The effect of varying enzyme amount and substrate molar ratio at 50oC and 135 min is shown in Figure 5. As the amount of enzyme increased, the yield also increased at each molar ratio. According to the analysis of variance, the interaction of enzyme amount and substrate molar ratio is a significant term (P-value = 0.0355). Using substrate molar ratio 1:1 to 2:1 and enzyme amount 311 to 400 mg results in a predicted yield of 100%.

Model validation and optimum conditions

The ?2 goodness-of-fit test was used to examine the validity of the model (Table 4) (Mooney and Swift, 1999). The test shows that there is not a significant difference between the predicted and actual values since the ?2 value (0.56) is much smaller than the cutoff value of ?2.05 for 4 degrees of freedom (9.49). This indicates that the generated model is valid at 95% confidence level.

Table 4 presents the optimal combination of parameters that can be used to obtain high percentage of yields. The optimum conditions can be used for future upscale synthesis of the ester (Ismail et al., 1999). A short time (41.9 min) is required to attain the maximum conversion of ester (96.2%). From economic standpoint, it would be favorable to use the minimum time and enzyme amount to achieve maximum yield. Maximum yield (90.0%) is predicted using 50 mg enzyme at 50.0oC, 71.8 min and substrate molar ratio 1.1:1. The actual yield obtained is 89.5 with 0.5% deviation.


Immobilized Candida antarctica lipase B-catalyzed synthesis of ethyl levulinate in an organic solvent-free system is successfully performed. Central composite rotatable design and response surface methodology are effectively applied to the optimization of the reaction parameters. Temperature, enzyme amount and substrate molar ratio are the significant process variables that affected the synthesis of levulinate ester. A high percentage yield (96.0%) is obtained in a short reaction time (41.9 min) which matches well with the predicted value of 96.1%. The developed model and optimum conditions can be used for future process scale-up.


  • Anderson, M.J., Whitcomb, P.J., 2005. RSM Simplified: Optimizing Processes Using Response Surface Methods for Design of Experiments, 1st edn. Productivity Press, New York.
  • Bozell, J. J., Moens L., Elliott, D. C., Wang, Y., Neuenscwander, G. G., Fitzpatrick, S.W., Bilski, R. J., Jarnefeld, J. L., 2000. Production of levulinic acid and use as a platform chemical for derived products. Resour. Conserv. Recy. 28, 227-239.
  • Carta, G., Gainer, J.L., Gibson, M.E., 1992. Synthesis of esters using a nylon-immobilized lipase in batch and continuous reactors. Enzyme Microb. Tech. 14, 904-910.
  • Fang, Q., Hanna, M.A., 2002. Experimental studies for levulinic acid production from whole kernel grain sorghum. Bioresource Technol. 81, 187-192.
  • Gunawan, E.R., Basri, M., Abdul Rahman, M.B., Salleh, A.B., Rahman, R.N.Z.A., 2005. Study on response surface methodology of lipase-catalyzed synthesis of palm-based wax ester. Enzyme Microb. Tech. 37, 739-744.
  • Hari Krishna, S., Sattur, A.P., Karanth, N.G., 2001. Lipae-catalyzed synthesis of isoamyl isobutyrate- optimization using a central composite rotatable design. Process Biochem. 37, 9-16.
  • Ismail, A., Linder, M., Ghoul, M., 1999. Optimization of butylgalactoside synthesis by ?-galactosidase from Aspergillus oryzae. Enzyme Microb. Tech. 25, 208-213.
  • Jeong, G.T., Yang, H.S., Park, D.H., 2009. Optimization of transesterification of animal fat ester using response surface methodology. Bioresource Technol. 100, 25-30.
  • Keng, P.S., Basri, M., Abdul Rahman, M.B., Salleh, A.B., Rahman, R.N.Z.A., Ariff, A., 2005. Optimization of palm based wax esters production using statistical experimenal designs. J. Oleo Sci. 54, 519-528.
  • Mooney, D.D., Swift, R.J., 1999. A Course in Mathematical Modeling, 1st edn. The Mathematical Association of America.
  • Myers, R.H., Montgomery, D.C., Anderson-Cook, C.M., 2009. Response Surface Methodology: Process and Product Optimization Using Designed Experiments, 3rd edn. John Wiley & sons, New Jersey.
  • Olson, E.S., 2001. Conversion of Lignocellulosic Material to Chemicals and Fuels. Energy & Environmental Research Center. University of North Dakota.
  • Otero, C., Arcos, J.A., Garcia, H.S., Hill, C.G., 2001. Enzymatic synthesis and hydrolysis reactions of acylglycerols in solvent-free systems. In: Vulfson, E.N., Halling, P.J., Holland, H.L. (Eds.), Enzymes in Nonaqueous Solvents: Methods and Protocols. Humana Press Inc., Totowa, NJ, pp. 479-496.
  • Soo, E.L., Salleh, A.B., Basri, M., Rahman, R.N.Z.A., Kamaruddin, K., 2004. Response surface methodological study on lipase-catalyzed synthesis of amino acid surfactants. Process Biochem. 39, 1511-1518.
  • Tufvesson, P., Annerling, A., Hatti-Kaul, R., Adlercreutz, D., 2007. Solvent-free enzymatic synthesis of fatty alkanolamides. Biotechnol. Bioeng. 97, 447-453.
  • Wetzel, S., Duchesne, L.C., Laporte, M.F., 2006. Bioproducts from Canada's Forests: New Partnerships in the Bioeconomy, 1st edn. Springer, the Netherlands.
  • Yadav, G.D., Borkar, I.V., 2008. Kinetic modeling of immobilized lipase catalysis in synthesis of n-butyl levulinate. Ind. Eng. Chem. Res. 47, 3358-3363.
  • Yamane, T., 2001. Solvent-free biotransformations of lipids. In: Vulfson, E.N., Halling, P.J., Holland, H.L. (Eds.), Enzymes in Nonaqueous Solvents: Methods and Protocols. Humana Press Inc., Totowa, NJ, pp. 509-516.