Optimization of lipase-catalyzed synthesis of fructose stearate using response surface methodology

Abstract

In this study, immobilized lipase-catalyzed esterification reaction between fructose and stearic acid was examined for the synthesis of a useful compound, fructose stearate, using response surface methodology. The increase of water content in the reaction medium was the negative effect while the increase in initial stearic acid/fructose molar ratio was the greatest positive effect on the yield. The highest fructose stearate yield was obtained as 65% in tert-butanol. The product yield was enhanced in 1-butyl-3-methylimidazolium trifluoromethanesulfonate obtained as 74% under the optimized conditions. The spectroscopic and elemental analysis methods showed that the esterification reaction is regioselective and the product is fructose monostearate.

Keywords::

• Fructose stearate
• ionic liquids
• response surface methodology
• tert-butanol

Introduction

Sugar esters of fatty acids are used as biosurfactants in the food, cosmetic, detergent and pharmaceutical industries (Cao et al. 1996, Sabeder et al. 2006, Kennedy et al. 2006). These esters are completely biodegradable under aerobic and anaerobic conditions, non-toxic, non-skin irritants, odorless and tasteless (Coulon et al. 1999). In addition, their antitumoral and anti-bacterial activities, plant growth inhibition, antibiotic and insecticidal properties have also been reported (Sabeder et al. 2006, Kennedy et al. 2006, Ferrer et al. 1999, Reyes-Duarte et al. 2005). The synthesis of sugar esters of fatty acids can be performed either chemically or enzymatically. The chemical methodologies are mainly carried out at high temperatures using alkaline catalysts. Furthermore, browning of products, formation of by- products because of poor selectivity and high energy consumption are major disadvantages of these methods. Alternatively, enzyme-catalyzed regioselective acylation of sugars in organic solvents was studied to overcome the drawbacks of chemical methods (Tsuzuki et al. 1999, Sarney and Vulfson 2001, Yoo et al. 2007, Pyo and Hayes 2009). Lipases (EC 3.1.1.3; triacylglycerol acyl hydrolase) have been extensively used for a broad range of stereoselective and regioselective transformation reactions in anhydrous organic solvents (Pimentel et al. 2007, Sagiroglu 2008, Karadeniz et al. 2010). Lipases catalyze the synthesis of ester bonds instead of their hydrolysis at low water content and they have been extensively used to synthesize esters of sugars. A major question in the synthesis of esters of sugar fatty acids is the solubilization of sugars in anhydrous organic media. Several methods have been developed to overcome the solubility problem including the use of derivatized sugars such as alkyl glycosides or two-phase systems and the use of intermediate polarity solvents such as dimethyl formamide, pyridine, and co-solvents such as dimethyl sulfoxide. However, the enzyme stability and reaction rates are poor in these solvents (Sarney and Vulfson 2001) and the derivatization of sugars are unpractical because of requiring extra protecting and deprotecting steps (Ganske and Bornscheuer 2005).

Recently, the use of ionic liquids (ILs) as reaction media in lipase-catalyzed syntheses of fatty acid sugar esters have much attention since ILs exhibit unique physical characteristics to dissolve sugars and fatty acids. Moreover, their melting points, viscosities, densities and hydrophobicities can be easily altered by changing to the structure of the ions (Ganske and Bornscheuer 2005, Earle and Seddon 2000, Sheldon et al. 2002, Lee et al. 2007, Hara et al. 2009, Ha et al. 2010).

Response surface methodology (RSM) is a collection of statistical and mathematical techniques based on the multivariate non-linear model and useful for developing, improving, and optimizing processes. RSM provides better advantages than classical methods for optimization of parameters and includes three steps: (1) designing an experiment and executing of designed experiment, (2) calculating the coefficients of proposed mathematical model, and (3) testing the model adequacy and predicting the response (Myers et al. 2009, Ray 2006). In the literature, RSM has been employed for optimization of lipase-catalyzed synthesis of various enzymatic reactions (Kapucu et al. 2003, Jeong and Park 2007, Kahveci et al. 2010). Box–Behnken designs (BBD) are a class of rotatable or nearly rotatable second-order designs (Box and Behnken 1960) and necessitate fewer experiments than full factorial design (FFD) or central composite design (CCD) with the same number of factors (Myers et al. 2009, Bae and Shoda 2005).

The objectives of this study are (1) to optimize fructose stearate synthesis catalyzed by a commercial immobilized lipase, Novozym 435, in tert-butanol using a 4-factor and 3-level BBD, (2) to formulate fructose stearate synthesis, (3) to examine the synthesis of fructose stearate in ionic liquids such as 1-butyl-3-methylimidazolium trifluoromethanesulfonate ([Bmim][TfO]), 1-butyl-3-methylimidazolium tetrafluoroborate ([Bmim][BF₄]) and 1-butyl-3-methylimidazolium hexafluorophosphate ([Bmim][PF₆]) for the first time, (4) to determine thermal stability of immobilized lipase, and (5) to characterize fructose stearate by spectroscopic and elemental analysis methods.

Materials and methods

Materials

A commercial lipase, Novozym 435 (Lipase B, immobilized lipase from Candida antarctica onto macroporous acrylic resin, EC 3.1.1.3, 10000 PLU/g; propyl laurate units synthesized per gram of catalyst at 60°C) and D-Fructose were purchased from Sigma–Aldrich (St. Louis, MO, USA). Molecular sieve (3 Å) was supplied from Fluka (Steinheim, Italy). Stearic acid, tert-butanol, methanol, chloroform, [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆] were obtained from Merck (Darmstadt, Germany). All other chemicals used were analytical grade and used without further purification.

Methods

Experimental design and data analysis

RSM analysis was carried out by employing a 4-factor and 3-level BBD which includes a total of 27 experimental runs. Several preliminary experimental studies have been executed to determine independent parameters and the reaction temperature (X₁), water content of reaction mixture (X₂), initial stearic acid/fructose molar ratio (X₃) and immobilized lipase loading (X₄) were considered as important factors. The low, center, and high levels of each factor were donated as − 1, 0 and + 1, respectively, and are given in Table I. A second-order polynomial equation (Eq. 1) was used to establish a mathematical relationship between the variables and the response:

Where Y is the predicted response (fructose stearate yield %), N is the number of variables, Xi is the independent variable, b₀, bi, bii and b_ij are the intercept terms, the linear effect, the squared effect and the interaction effect, respectively. BBD and data analyses were achieved by using Design Expert statistical software (Design Expert 8.0.7).

Table I. The selected four variables and the coded levels of each variable.

Variables	Symbol	Variable level
		Low	Center	High
		−1	0	+1
Reaction temperature (°C)	X1	45	55	65
Water content (%)	X2	0.5	1.5	2.5
Initial stearic acid/fructose molar ratio	X3	1.0	3.0	5.0
Immobilized lipase loading (mg mL^−1)	X4	10	30	50

Lipase-catalyzed synthesis of fructose stearate in tert-butanol

In a typical experiment, 5.0 mL of tert-butanol was added onto 0.3 mmol fructose and then incubated at the related reaction temperature for 2 h. Subsequently, the appropriate amount of stearic acid and 100 mg molecular sieve (pre- activated at 250°C overnight under reduced pressure) were sequentially added onto it. After the dissolution of stearic acid, the reaction was started at different experimental conditions as given in Table II, at 100 rpm in a water bath (Figure 1). The immobilized lipase and molecular sieve were removed by filtration of the mixture after 48 h reaction time.

Figure 1. Lipase-catalyzed synthesis of fructose monostearate.

Table II. The observed and predicted results of BBD for immobilized C. antarctica lipase catalyzed synthesis of fructose stearate.

Run	Pattern	X1/Reaction temperature (°C)	X2/Water content (%)	X3/Stearic acid/fructose molar ratio	X4/Immobilized lipase loading (mg mL^−1)	Y/Yield (%) (observed)	Y/Yield (%) (predicted)
1	0 + 0+	55	2.5	3.0	50	18.42	17.00
2	+ 0 + 0	65	1.5	5.0	30	29.46	29.22
3	0000	55	1.5	3.0	30	23.09	22.94
4	0 − 0−	55	0.5	3.0	10	31.09	32.67
5	0++ 0	55	2.5	5.0	30	17.35	18.65
6	−− 00	45	0.5	3.0	30	26.88	27.66
7	+− 00	65	0.5	3.0	30	38.50	39.32
8	++ 00	65	2.5	3.0	30	10.71	10.15
9	− 00+	45	1.5	3.0	50	11.18	12.13
10	0+− 0	55	2.5	1.0	30	16.58	17.85
11	0 + 0−	55	2.5	3.0	10	13.38	13.39
12	0−+ 0	55	0.5	5.0	30	58.12	56.48
13	00−−	55	1.5	1.0	10	13.66	14.01
14	0000	55	1.5	3.0	30	22.87	22.94
15	+ 00−	65	1.5	3.0	10	11.62	10.30
16	− 0 + 0	45	1.5	5.0	30	14.45	14.36
17	+ 0 − 0	65	1.5	1.0	30	11.17	11.41
18	00++	55	1.5	5.0	50	37.10	36.97
19	− 00−	45	1.5	3.0	10	5.65	4.22
20	00−+	55	1.5	1.0	50	22.67	22.08
21	0000	55	1.5	3.0	30	22.86	22.94
22	0—0	55	0.5	1.0	30	34.87	33.20
23	+ 00+	65	1.5	3.0	50	23.16	24.22
24	0 − 0+	55	0.5	3.0	50	50.76	50.90
25	− 0 − 0	45	1.5	1.0	30	7.69	8.09
26	−+ 00	45	2.5	3.0	30	4.23	3.63
27	00+−	55	1.5	5.0	10	22.40	23.21

The solvent was evaporated under vacuum and the obtained solid was dissolved in a methanol/chloroform/water mixture (2/1/0.8, v/v/v). This final solution was quantitatively analyzed by high-performance liquid chromatography (HPLC) for stearic acid content. HPLC analyses were performed on a Nucleosil C18 column (5 μm, 4.6 × 250 mm). The mobile phase was methanol/chloroform/water mixture (2/1/0.8, v/v/v) at a flow rate of 0.5 mL min⁻¹. The column temperature was kept constant at 40°C during analysis. The detection of fructose, stearic acid and fructose stearate was performed with a refractive index detector (Shimadzu, RID-10A). Calibration curve was prepared using stearic acid and the quantitative data were obtained from peak areas. The formed ester amount was determined by subtracting the residual stearic acid amount in the reaction mixture from the initial stearic acid amount. A blank reaction without lipase was also run in parallel. Water activities of the reaction medium were determined using a Karl Fischer titrator.

Lipase-catalyzed synthesis of fructose stearate in ILs

Five mililiters of ILs containing 0.5% water was added onto 0.3 mmol fructose and the mixture was incubated at 90°C for 2 h. After the mixture was cooled to 59°C, 1.47 mmol stearic acid and 100 mg molecular sieve were added onto it. The reaction was started by adding 48 mg of immobilized lipase per mL of reaction mixture and allowed to continue by agitating at 100 rpm in a water bath for 48 h. Subsequently, the immobilized lipase and molecular sieve were removed by filtration of the mixture. The unreacted stearic acid and formed fructose stearate were extracted with acetone. The solvent was evaporated and the obtained solid was dissolved in the methanol/chloroform/water mixture then analyzed by HPLC for stearic acid content.

Solubility of fructose in ILs

Fructose (0.3 mmol) was weighted into a glass vial equipped with screw stopcock and then 5 mL of ILs containing 0.5% water was added onto it. The mixture was separately incubated at 59 and 90°C for 2 h. After centrifugation at 5000 rpm for 10 min, the supernatant was analyzed for fructose content using dinitrosalicylic acid (DNSA) method (Miller 1959).

Purification and structural analysis of fructose stearate

The purification of fructose stearate was accomplished by using a flash column (pore diameter 60 Å, particle size 45 μm) packed with octadecyl-modified silica. The column was equilibrated with the methanol/chloroform/water mixture. Fructose stearate was eluted with a methanol/ chloroform/water mixture (1/1/0.25, v/v/v) at a flow rate of 1.4 mL min⁻¹ using a peristaltic pump (Watson-Marlow Alitea, Cornwall TR 11 4RU, England). The collected fractions were analyzed by HPLC for fructose stearate content.

The presence of fructose stearate was confirmed by infrared spectra which was recorded on FTIR spectrophotometer (Spectrum RX I FT-IR System, Perkin Elmer) using a film of test surfactant between potassium bromide plates. ¹H NMR analysis was performed at 300 MHz employing a Bruker Ultrashield TM NMR spectrometer using DMSO (d-6) as solvent and tetramethylsilane (TMS) as internal standard. Elemental analysis was performed using an automated CHNS-932 (LECO) analyzer.

Results and discussion

In this study, the synthesis of fructose stearate was achieved by esterification of stearic acid with fructose in tert-butanol, catalyzed by immobilized C. antarctica lipase. BBD was performed to assess the effect of four variables on synthesis of fructose stearate and following analyses were achieved using RSM. In the preliminary studies, the effect of reaction time on the lipase-catalyzed synthesis of fructose stearate was investigated for 12, 24, 36, 48, and 72 h reaction times. The corresponding yields were determined as 9.2, 15.7, 18.8, 22.90, and 11.67% when water content of reaction medium, reaction temperature, stearic acid/fructose molar ratio and immobilized lipase loading were 1.5%, 55°C, 3 and 30 mg mL⁻¹, respectively. Therefore, the synthesis reaction was carried out at constant reaction time of 48 h for further investigations. The experiments were repeated three times and the results are given as average of three replicates in Table II. The quadratic model was used to determine the mathematical relationship between fructose stearate yield and variables. Based on the experimental results of BBD, the values of regression coefficients of Eq. (1) were calculated using Design-Expert software. The mathematical relationship between the response and the four variables was shown below:

Where Y is fructose stearate yield (%), X₁, X₂, X₃ and X₄ are reaction temperature (°C), water content of reaction mixture (%), stearic acid/fructose molar ratio and immobilized lipase loading (mg mL⁻¹), respectively. Equation (2) indicates that the reaction temperature (linear), stearic acid/fructose molar ratio (linear), immobilized lipase loading (linear), the interactions between temperature and stearic acid/fructose molar ratio (linear), temperature and immobilized lipase loading (linear) and stearic acid/fructose molar ratio and immobilized lipase loading (linear), temperature (quadratic) and stearic acid/fructose molar ratio (quadratic) have positive effect on the fructose stearate yield. The stearic acid/fructose molar ratio with a coefficient of 6.02 was the greatest positive linear effect whereas water content of reaction mixture with a coefficient of 13.29 was the greatest negative linear effect on the fructose stearate yield. As can be seen in Table II, the maximum and minimum fructose stearate yield were observed as 58.12 (run number 12) and 4.23% (run number 26), respectively.

The results of analysis of variance (ANOVA) test are shown in Table III. The calculated F-value (169.77) indicates the proposed model and is significant. The goodness of fit of the model can be tested by the coefficient of determination (R²). The values of R² and adjusted R² were calculated as 0.99 and 0.98, respectively. The actual and the predicted fructose stearate yields (%) are shown in Figure 2. It is clear that a very well linear correlation was obtained between the actual and the predicted results. This indicates that the proposed equation provides an appropriate approximation for the explanation of the relationship between the independent variables and fructose stearate yield. Values of Prob >F less than 0.0500 are accepted as model terms are significant. In this case X₁, X₂, X₃, X₄, X₁3X₃, X₁3X₄, X₂₃X₃, X₂3X₄, X²₁3X²₂, X²₂ and X²₃ are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. The F-value of Lack of Fit test are calculated as 133.12, indicating that the Lack of Fit is significant.

Figure 2. The observed and predicted fructose stearate yields (%).

Table III. ANOVA test for the proposed quadratic model.

Source	Sum of squares	Degree of freedom	Mean square	F value	p value Prob> F
Model	4462.71	14	318.77	169.77	< 0.0001	significant
X1	246.79	1	247.88	132.02	< 0.0001	significant
X2	2052.51	1	2121.35	72.21	< 0.0001	significant
X3	284.90	1	434.88	231.61	< 0.0001	significant
X4	374.87	1	357.41	190.35	< 0.0001	significant
X1X2	6.60	1	6.60	3.52	0.0853
X1X3	33.24	1	33.24	17.70	0.0012	significant
X1X4	9.03	1	9.03	4.81	0.0487	significant
X2X3	126.34	1	126.34	67.29	< 0.0001	significant
X2X4	53.51	1	53.51	28.50	0.0002	significant
X3X4	8.09	1	8.09	4.31	0.0600
X^21	457.49	1	457.49	243.65	< 0.0001	significant
X^22	226.17	1	226.17	120.46	< 0.0001	significant
X^23	23.32	1	23.32	12.42	0.0042	significant
X^24	4.95	1	4.95	2.63	0.1306
Residual	22.53	12	1.88
Lack of fit	22.50	10	2.25	133.12	0.0075	significant
Pure error	0.034	2	0.017

R² = 0.99, AdjR² = 0.98.

Water plays important role for lipase activity in organic media. A small amount of water molecules are necessary for enzyme activity. In this study, the synthesis reactions were carried out at different water content of reaction medium as 0.5% (a_w = 0.12), 1.5% (a_w = 0.23) and 2.5% (a_w = 0.35). Figure 3 shows that fructose stearate yield was drastically decreased with increasing water concentration of reaction medium. The fructose stearate yield was obtained as 26.88% when water concentration of reaction medium, reaction temperature, stearic acid/fructose molar ratio and immobilized lipase loading were 0.5%, 45°C, 3 and 30 mg mL⁻¹, respectively (run number 6). A 6.35-fold decrease in fructose stearate yield was observed when the water concentration of reaction medium was increased to 2.5% and the other reaction was kept constant (run number 26). Under the same conditions, no yield was obtained by increasing the water content of reaction mixture up to 5% (a_w = 0.75) (data not shown). The similar results were observed in the synthesis of lipase catalyzed sugar esters of fatty acids. Tsuzuki et al. (1999) reported that the presence of water (1%) in hexane inhibited the esterification of glucose catalyzed by Pseudomonas sp. lipase. Lee et al. (2007) noted that the increase of water content of reaction medium markedly decreased the esterification rate of glucose with vinyl laurate catalyzed Novozym 435 in ILs. As shown in Figure 3, the fructose stearate yield was significantly higher at 55 and 65°C than at 45°C. The predicted fructose stearate yield was calculated as 42.75% at 55°C when the reaction conditions were 0.5% of water, 3.0 of stearic acid/fructose molar ratio and 30 mg mL⁻¹ of immobilized lipase loading. At 65°C, the corresponding observed yields were 38.50% (run number 7) under the same conditions. The corresponding fructose stearate yield was 26.88% at 45°C (run number 6).

Figure 3. The estimated response surface plot of the effect of reaction temperature and water content of reaction medium on the fructose stearate yield at constant stearic acid/fructose molar ratio of 3.0 and and immobilized lipase loading of 30 mg mL⁻¹ after 48 h reaction time.

The response surface plot of the effect of initial stearic acid/fructose molar ratio and immobilized lipase loading on the fructose stearate yield is shown in Figure 4. When the equimolar fructose and stearic acid were used, the obtained fructose stearate yield were 13.66% at 55°C, 1.5% of water content and 10 mg mL⁻¹ of immobilized lipase loading (run number 13). As shown in Figure 4, the fructose stearate yield increased significantly by increasing of stearic acid/fructose molar ratio. The fructose stearate yield was observed as 22.40% when only the initial molar ratio of stearic acid to fructose was changed to 5.0 (run number 27). Further increase in stearic/fructose molar ratio (7.0) did not affect on the product yield (data not shown). Moreover, a 37.10% fructose stearate yield was obtained when increasing the immobilized lipase loading from 10 to 50 mg mL⁻¹ (run number 18).

Figure 4. The estimated response surface plot of the effect of stearic acid/fructose molar ratio and immobilized lipase loading on the fructose stearate yield at 55°C and water content of reaction medium of 1.5% after 48 h reaction time.

Enzyme loading is a crucial factor for product yield and its industrial application. As shown in Figure 5, the fructose stearate yield was obtained as 31.09% when the concentration of immobilized lipase was 10 mg mL⁻¹ (run number 4). When the immobilized lipase concentration was changed to 50 mg mL⁻¹, the fructose stearate yield markedly increased and was determined as 50.76% (run number 24). However, when the enzyme loading was increased up to 100 mg mL⁻¹, a significant increase was not observed in the fructose stearate yield. (data not shown).

Figure 5. The estimated response surface plot of the effect of immobilized lipase loading and reaction temperature on the fructose stearate yield at constant stearic acid/fructose molar ratio of 3.0 and water content of reaction medium of 1.5% after 48 h reaction time.

To determine the optimum conditions for fructose stearate synthesis, the desired fructose stearate yield was chosen as “maximize” and reaction temperature, water content of reaction mixture, initial stearic acid/fructose molar ratio and immobilized lipase loading were selected as “within the range on the desirability section of software menu. The best conditions were determined as 59°C, 0.5% of water content, 4.9 of stearic acid/fructose molar ratio and 48 mg mL⁻¹ of immobilized lipase loading for this way.

The selection of solvent for fatty acid sugar esters synthesis is very important because product yield strictly depends on the dissolution of reactants in reaction medium. Therefore, the solvents used in the lipase- catalyzed synthesis of fatty acid sugar esters should be appropriate for dissolution of sugars and fatty acids. It was reported that ILs might be good solvents for the lipase-catalyzed synthesis of fatty acid sugar esters because of their ability to dissolve both sugars and fatty acids (Lee et al. 2007). Therefore, the synthesis of fructose stearate was examined in three ionic liquids under the optimized conditions. [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆] were selected since these ILs were mostly used in the lipase-catalyzed synthesis of various reactions (Itoh et al. 2001, Dhake et al. 2009, Sunitha et al. 2007). The dissolution of stearic acid was completely achieved for all the used ILs whereas 41.2, 24 and 20.3% of initial amounts of fructose were dissolved in [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆]. To enhance the solubility of fructose, supersaturated fructose solutions were prepared. For this purpose, 0.3 mmol fructose was separately added in [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆] and heated to 90°C. After incubation for 2h, it was then slowly cooled to 59°C. In this way, fructose solubility was enhanced 2.1, 2.5 and 2.8 folds, respectively in [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆] compared to the solubility at 59°C. Figure 6 shows the fructose stearate yields obtained in [Bmim][TfO], [Bmim][BF₄] and [Bmim][PF₆] under the determined optimal conditions. The fructose stearate yield was increased in the order of [Bmim][PF₆] <tert-butanol < [Bmim][BF₄] < [Bmim][TfO] when all the other reaction parameters were kept constant. The highest fructose stearate yield was obtained as 74% in [Bmim][TfO]. In our study, the lowest fructose stearate yield (62%) was obtained for [Bmim][PF₆]. Lee and Lee (2005) reported that [Bmim][PF₆] was high viscosity than [Bmim][TfO]. The high viscosity of [Bmim][PF₆] probably limited the mass transfer of the substrates to the active site of enzyme leading to a decrease in activity. Lee et al. (2007) reached 90% yield of glucose laurate in [Bmim][TfO] and reported that glucose solubility was higher in [Bmim][TfO] than [Bmim][BF₄] and [Bmim][PF₆]. They also reported that fructose solubility was about 5.5-fold higher in [Bmim][TfO] than [Bmim][BF₄] at 60°C. Ha et al. (2010) obtained 84.4% yield for the synthesis of fructose palmitate catalyzed by Novozym 435 in the mixture of [Bmim][TfO] and 1-methyl-3-octylimidazolium bis-(trifluoromethyl sulfonyl) imide ([Omim][Tf₂N]) with a 1/1 volume ratio. Ganske and Bornscheuer (2005) reported that C. antarctica lipase catalyzed synthesis of glucose laurate was achieved with 62% yield using lauric acid vinyl ester as acyl donors in [Bmim][PF₆] in the presence of 40% tert-butanol at 60°C.

Figure 6. The effect of ILs on fructose stearate yield at the optimized conditions.

The thermal stability of immobilized lipase was investigated at the optimum conditions for 45, 55 and 65°C. The relative activity of immobilized lipase decreased depending on the increasing of incubation time for all investigated temperatures (Figure 7). At the end of 48 h incubation time, the relative activities of immobilized lipase were determined as 67, 65 and 57%, respectively for 45, 55 and 65°C. Half-life times of immobilized lipase for 45, 55 and 65°C were calculated from their first-order inactivation constants which were estimated the slope of plot of In(V₀/Vt) versus incubation time (t), where V₀ and Vt are the initial activity and the activity after t time, respectively (data not shown). The half-life times of immobilized lipase were calculated as 73.7, 64.8 and 53.7 h, respectively for 45, 55 and 65°C.

Figure 7. Thermal stability of immobilized lipase at 45, 55 and 65°C up to 48 h.

The purification of fructose stearate was performed by using flash chromatography. The elution of fructose stearate was achieved with a mixture of methanol/chloroform/water. The characterization of fructose stearate was accomplished by Fourier transform infrared spectroscopy (FTIR), nuclear magnetic resonance spectroscopy (¹H NMR) and elemental analysis methods. FTIR analysis of fructose stearate showed that the product had absorption bands at wave numbers 3422 cm⁻¹ (for the OH bond), 2850–2918 cm⁻¹ (for the C-H bond in –CH₂ or –CH₃), 1733 cm⁻¹ (for the C = O, ester bond), 1467 cm⁻¹ (for –CH₂, − CH₃), 1275 cm⁻¹ (for C-O-C bond), 1073 cm⁻¹ (for the C-O, ester bond) and 726 cm⁻¹ [for the (CH)₂ bond]. Rahman and Herawan analyzed fructose palmitate synthesized with Lipozyme IM by FTIR spectrophotometer and they reported that the product had absorption bands at wavenumbers 3381–3407 cm⁻¹ (for the OH bond), 2851–2919 cm⁻¹ (for the C-H bond in –CH or –CH₃), 1735 cm⁻¹ (for the C = O, ester bond), 1468 cm⁻¹ (for –CH₂, -CH₃), 1055–1183 cm⁻¹ (for the C-O, ester bond), and 723 cm⁻¹ (for the (CH)₂ bond) (Rahman and Herawan 2000).

The position of stearic acid in the fructose ester and monoacylation were supported by ¹H NMR spectroscopy. ¹H NMR data are as follows: 0.88 (3H, t, J = 7.1 Hz), 1.24 (28H, m), 1.49 (2H, m), 2.28–2.15 (2H, t, J = 7.0 Hz), 3.35–3.57 (1H, dd, J = 12 Hz), 3.80 (1H, ddd, J = 3.1 Hz), 3.96 (1H, d, J = 11.5), 4.09 (1H, d), 4.15 (1H, d, J = 11 Hz). Our ¹H NMR spectral data matched with literature data given for 1^’-O-stearyl fructose by Schlotterbeck et al. (1993) The elemental composition of fructose stearate was also determined and C and H percentages of purified fructose stearate were found to be 65 and 10.4, respectively. These results indicate that the acylation of fructose is occurred only at one position and that the product is fructose monostearate.

Conclusion

In this study, 4-factor and 3-level BBD was applied to establish a mathematical model for estimation of fructose stearate yield catalyzed by immobilized C. antarctica lipase in tert-butanol. The results of ANOVA analyses indicated that the proposed equation estimated the experimental results with 99% probability. The experimental results showed that the increase in temperature, initial stearic acid/fructose molar ratio and immobilized lipase loading were positive linear effect whereas the increase of water content in the reaction medium was negative linear effect on fructose stearate yield. The optimum reaction conditions for fructose stearate synthesis were determined as 59°C, 0.5% of water content, 4.9 of stearic acid/fructose molar ratio and 48 mg mL⁻¹ of immobilized lipase loading. Under these conditions, the fructose stearate yield was obtained as 65% in tert-butanol. [Bmim][TfO] and [Bmim][BF₄] ILs provided better reaction media than tert-butanol and the highest fructose stearate yield was determined as 74% in [Bmim][TfO]. The results of ¹H NMR and elemental analyses show that the C. antarctica lipase-catalyzed esterification of fructose with stearic acid occurs only at one position and the product is fructose monostearate.

Acknowledgments

The authors thank the University of Cukurova for financial support.

Declaration of interest

The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper. This work was supported by Research Grants FEF 2008 YL 4 from University of Cukurova.