Environ Sci Pollut Res DOI 10.1007/s11356-015-4583-7
POLLUTION CONTROL TECHNOLOGIES AND ALTERNATE ENERGY OPTIONS
Production and optimization of biodiesel using mixed immobilized biocatalysts in packed bed reactor S. Bakkiyaraj 1 & Mahin Basha Syed 1,2 & M. G. Devanesan 1 & Viruthagiri Thangavelu 1
Received: 15 March 2015 / Accepted: 21 April 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Vegetable oils are used as raw materials for biodiesel production using transesterification reaction. Several methods for the production of biodiesel were developed using chemical (alkali and acidic compounds) and biological catalysts (lipases). Biodiesel production catalyzed by lipases is energy and cost-saving processes and is carried out at normal temperature and pressure. The need for an efficient method for screening larger number of variables has led to the adoption of statistical experimental design. In the present study, packed bed reactor was designed to study with mixed immobilized biocatalysts to have higher productivity under optimum conditions. Contrary to the single-step acyl migration mechanism, a two-step stepwise reaction mechanism involving immobilized Candida rugosa lipase and immobilized Rhizopus oryzae cells was employed for the present work. This method was chosen because enzymatic hydrolysis followed by esterification can tolerate high free fatty acid containing oils. The effects of flow rate and bed height on biodiesel yield were studied using two factors five-level central composite design (CCD) and response surface methodology (RSM). Maximum biodiesel yield of 85 and 81 % was obtained for jatropha oil and karanja oil with the optimum bed height and optimum flow rate of 32.6 cm and 1.35 L/h, and 32.6 cm and 1.36 L/h, respectively. Responsible editor: Bingcai Pan * Mahin Basha Syed [email protected] 1
Biochemical Engineering Lab, Chemical Engineering Department, Annamalai University, Chidambaram 608002, Tamilnadu, India
2
Environmental Engineering Lab, Nawab Shah Alam Khan College of Engineering and Technology, Hyderabad 500024, Telangana, India
Keywords Biodiesel . Response surface methodology . Central composite design . Packed bed reactor . Rhizopus oryzae . Candida rugosa lipase
Introduction Recently, the increased crude oil prices and environmental awareness have renewed the focus on the use of vegetable oils and animal fats as alternative fuels. The raw materials used for production of biodiesel can be either crude refined or waste such as frying oils/fats (Marchetti et al. 2008). The use of alternative fuel for diesel engines dates back almost a century, when Rudolf Diesel first demonstrated the diesel engine at the 1900 world exhibition in Paris with peanut oil. Given the potential of vegetable oils as an alternative fuel, different nations, depending on their unique climate and soil conditions, are exploring a variety of vegetable oils for diesel fuel applications. The vegetable oils therefore need to be chemically modified to make them suitable as engine fuels (Harwood 1984). Vegetable oils comprise 90–98 % triglycerides and small amounts of monoglycerides, diglycerides, free fatty acids (FFA), phospholipids, phosphatides, carotenes, tocophenols, sulfur compounds, and traces of water. Triglycerides are esters of three fatty acids and one glycerol molecule and contains substantial amount of oxygen in its structure. Fatty acids of vegetable oils vary in their carbon chain length and in the number of double bonds (Barnwal and Sharma 2005). An important factor in biodiesel production is the fatty acid composition of the source oil or fat. Oils containing higher levels of saturated fatty acids than unsaturated fatty acids (have one or more double bonds) may solidify and clog the fuel lines during the winter condition (Pinto et al. 2005; Akoh et al. 2007; Demirbas 2008).
Environ Sci Pollut Res
Transesterification consists of a sequence of three consecutive reversible reactions where triglycerides are converted to diglycerides, monoglycerides, and free fatty acids with formation of an ester molecule and glycerol at each step; thus, three ester molecules are produced from one molecule of triglyceride. Since the reactions are reversible, excess alcohol is used to shift the equilibrium towards the product side. However, the yield of biodiesel is independent of the type of the alcohol used, and the selection of alcohol is based on cost and performance of it. Methanol and ethanol are the most preferred, especially methanol because of its low cost and its physical and chemical advantages (polar- and short-chained alcohol) over ethanol. Pressure shockwaves produced by underwater high-voltage discharges on underground and inadequately soaked seeds of Jatropha curcas L. The depth of penetration of excess methanol solvent and the number of shockwaves seem to be very important and increase the oil yield up to 94 % (Maroušek et al. 2012). It was found that grinding, followed by maceration that may enhance the effect of following pressure shockwaves (50 to 60 MPa), will increase oil extraction over 94 % (Maroušek et al. 2013a). Enzymatic hydrolysis enhanced by pressure shockwaves opens new possibilities in J. curcas L. (Maroušek et al. 2013b). Microemulsion also causes heavy carbon deposit, injector needle deposit, and incomplete combustion during long-term endurance tests (Pryde 1983). Thermal decomposition of triglycerides produces a mixture of compounds such as alkanes, alkenes, alkadienes, aromatics, and carboxylic acid (Billaud et al. 1995; Ma and Hanna 1999). The process is however economically unviable due to high cost of equipments. This paper evaluates production of biodiesel by mixed immobilized biocatalysts in packed bed reactor using karanja and jatropha oils. The effects of flow rate and bed height on biodiesel yield were studied using two-factor five-level central composite design (CCD) and response surface methodology (RSM). RSM is a powerful and efficient mathematical approach widely applied in the optimization of fermentation process. It can give information about the interaction between process parameters, provide information necessary for design and process optimization, and give multiple responses at the same time. The factors that have significant effects were optimized using a CCD and response surface analysis.
Materials and methods Medium preparation Rhizopus oryzae, MTCC (262), was obtained from microbial type culture collection Chandigarh (India). R. oryzae was maintained on 7 % potato dextrose agar medium. After 3 days of incubation at 25 °C, the agar slants were stored at 4 °C. The growth medium used contains the following: olive oil 30 g/L,
polypeptone (50 wt% pepton, 50 wt% trypton) 70 g/L, KH2PO4 (99.0 %) 1.0 g/L, MgSO4·7 H2O (99.0 %) 0.5 g/L, and NaNO3 (98.0 %) 1.0 g/L (Ban et al. 2002). R. oryzae was grown aerobically in 250-mL Erlenmeyer flasks containing the growth medium. Lipase preparation Lipase powder (Sigma-Aldrich) (1 g) was suspended in 100 mL of 0.25 M sodium phosphate buffer (pH 7); lipase solution was centrifuged (Remi centrifuge) at 4000 rpm for 15 min at 4 °C. The supernatant was stored at the same temperature prior to immobilization. Normalization of the activity was done by diluting the enzyme solution when the transesterification ability of the enzymes was compared. Immobilization by entrapment Sodium alginate solution was prepared by dissolving 5 % (w/v) sodium alginate in boiling water and autoclaved at 121 °C for 15 min. Both alginate slurry and 10 g of cells were mixed and stirred for 10 min to get a uniform alginate/cell mixture. The mixture was extruded drop by drop into the cold sterile CaCl2 (94.0 %) (0.2 M) solution through a sterile burette and was cured at 4 °C for 1 h. The beads were hardened by suspending it again in a fresh CaCl2 solution for 24 h at 4 °C with gentle agitation. Finally, these beads were washed with distilled water to remove excess calcium ions and un-entrapped cells. The beads were preserved in 0.9 % sodium chloride (99 %) solution in the refrigerator. The same procedure was adopted for enzyme immobilization; instead of cells, the enzyme supernatant was used for preparing the alginate/enzyme mixture. All other steps were similar for both the cases. Degumming of crude oils Crude karanja and jatropha oils were first degummed to remove the phospholipids prior to subjecting it to enzymatic transesterification. Degumming was done by heating the oils to 75 °C and adding 0.7 % (w/v) of NaCl at the same temperature followed by stirring at a constant speed of 400 rpm. The stirring rate was reduced gradually to 30 rpm based on the formation of colloidal particles, and the stirring was completely stopped when there was no suspended particles in the oils. The colloidal particles formed were removed by centrifugation at 2012g for 10 min. The resultant oils was dried by vacuum drying for 30 min and used for transesterification (Yuan et al. 2008). Batch transesterification Methanolysis reaction was conducted with pre-designed values of molar ratio of oil to alcohol, reaction temperature, reaction time, added water content, and immobilized beads
Environ Sci Pollut Res
concentration in a 50-mL Erlenmeyer flask with constant shaking (120 rpm) using incubated orbital shaker. Solvent (n-hexane) (99.9 %) is added to the reaction mixture to enhance the solubility of the reactants. After pre-designed time, the biocatalysts were removed from the reaction mixture by filtration. The produced ester and by-product glycerol were separated using a separating funnel. The produced esters are quantified using high-performance liquid chromatography equipment (HPLC). Transesterification reactions were carried out in triplicate, and the mean results were recorded.
Packed bed reactor Packed bed reactor designed consists of a cylindrical perspex column of 2.5-cm diameter and 25 cm, 40 cm in height to study the effect of bed height on biodiesel production as shown in Fig. 1. The reactor was used for the continuous biodiesel production studies from karanja and jatropha oils using mixed immobilized biocatalysts. Packed bed column contains the immobilized beads of R. oryzae cells and Candida rugosa lipase in fixed ratio. In the columns, wire mesh was placed at the bottom and top of the column to retain the beads inside the column. The raw oil and methanol (reactant mixture) at a known flow rate based on CCD is pumped in to the column in an upward direction through the column using a peristaltic pump at fixed flow rate. The outflow from the column is again mixed with the inflow in the reactant storage vessel. Samples from outflow were collected at regular intervals and were analyzed for biodiesel production.
Fig 1 Production of biodiesel from nonedible oils using mixed immobilized biocatalysts
High-performance liquid chromatographic analysis A High-Performance Liquid Chromatographic system (HPLC Model- LC 20 AT Prominence, Shimadzu, Japan) fitted with Refractive index detector (RID-10A, Shimadzu, Japan) and millennium 32 system software was used to quantify the fatty acid methyl esters produced during reaction. Separations were carried out on a 238 nm in Luna C18 column of particle size 5 μm, and (250×4.6 mm) I.D. Methanol (99.8 %) mobile phase was filtered through a 0.45-μm membrane filter (Millipore) and then degassed ultrasonically prior to use. The flow rate was 1 mL/min, the injection volume was 20 μL, and the column oven temperature was maintained at 40 °C. Each component in the samples analyzed was identified by comparing its retention time with that of the respective standards. Quantification was carried out by integration of the peaks using external standards followed by calculating the % yield as weight of methyl esters produced to weight of oil initially taken. Central composite experimental design and response surface methodology Statistical methods provide an efficient alternative methodology for traditional one factor at a time approach to optimize a particular process by considering the mutual interactions among the variables and to give an estimate of the combined effects of these variables (Myers and Montgomery 1995). The graphical representation of the polynomial model resulted in the response surface plots that characterize the individual, squared, and interactive effects of test variables on the response. The fractional factorial design consisted of 16 factorial points, 10 axial points, and six center points. The variables and their levels selected for the study were the following: molar ratio of alcohol to oil (1:1–1:5), reaction temperature (30–50 °C), reaction time (varied depending on type of biocatalyst used, 72 h as the center point for cells, 24 h for enzymes and 40 h mixed immobilized biocatalysts respectively), water content (0–20 % w/w of oil), and concentration of immobilized beads (1–5 % w/w of oil). The independent factors (xi), levels, and experimental design in terms of coded and uncoded values are shown in the BResults^ section. For statistical calculation, the test factors were coded according to Eq. 2.1: . xi ¼ X i –X o ΔX i i ¼ 1; 2; 3; ………:: k ð2:1Þ where xi is the coded value of an independent variable, Xi is the real value of the independent variable, Xo is the real value of the independent variable at the center point, and ΔXi is the step change value. The Design Expert 6.0 software was used for regression and graphical analysis of the data obtained. The maximum
Environ Sci Pollut Res
values of the percentage yield of biodiesel (or) methyl esters were taken as the responses of the design experiment. Once the experiments were performed, the response variable was fitted as second-order model in order to correlate the response variable to the independent variable. The general form of the second-order polynomial equation for the yield of fatty acid methyl ester is given below in Eq. 2.2 Y ¼ β0 þ
X
β i xi þ
X
2
βii x1 þ
X
β i j xi x j
ð2:2Þ
where methyl esters yield (or) biodiesel yield (Y) is the predicted response variable, i and j are the linear and quadratic coefficients respectively, β is the regression coefficient of the model, and xi, xj ( i=1,3; j=1,3, i=j) represent the independent variables (reaction conditions) in the form of coded values. The MINITAB 16 software statistical package was used for regression analysis of the experimental data obtained and to estimate the coefficients of the regression equation. The accuracy and general predictive ability of the polynomial model was evaluated by the multiple correlation coefficient (R) and by the coefficient of determination (R2). The statistical significance of the model was determined by F-test. The mean squares are obtained by dividing the sum of squares of each of the two sources of variation, the model and the error variance, by the respective degrees of freedom. The Fisher’s variance ratio—F value is the ratio of the mean square due to regression to the mean square due to the error which is a statistically valid measure of how well the factors describe the variations in the data about its mean—can be calculated from ANOVA. Since coding of the variable enables direct comparison of the partial regression coefficients, their significance was determined by Student’s t test and the associated probabilities. The P values are used as a tool to check the Table 2 Biodiesel production from Jatropha oil in a packed bed reactor using RSM
Run no.
1 2 3 4 5 6 7 8 9 10 11 12 13
Table 1 Experimental range and levels of the independent variables used in CCD and RSM for biodiesel production from jatropha and karanja oils using mixed biocatalyst in packed bed reactor Parameters with designate
X1: Packed bed height (cm) X2: Substrate flow rate (L/h)
Coded levels −1.414
−1
0
1
1.414
22.93 0.793
25 1
30 1.5
35 2
37.07 2.207
significance of each of the coefficients, which in turn may indicate the pattern of the interaction between the variables. The smaller the value of P, the more significant is the corresponding coefficient. The response surface plots were used to describe the individual and cumulative effects of the independent variables as well as the mutual interactions between the variables on the biodiesel yield. The second degree polynomial equation was maximized by the constraint search procedure using MATL AB software (Version 7.0, The MathWorks, Inc. Natick, USA) to obtain the optimal levels of the independent variables and the predicted maximum biodiesel yield. Optimization of packed bed height and substrate flow rate using CCD and RSM for transesterification of jatropha and karanja oils in packed bed reactor The experiments were conducted under the optimal conditions of substrate molar ratio (1:3), hexane (solvent) 25 % (w/w of oil), bead size (0.25 cm), reaction time (42 h), and added water content 11 % (w/w of oil ) and were maintained constant for all the experiments conducted in Packed bed reactor. The mixed immobilized biocatalysts consisting of immobilized R. oryzae
CCD experimental design matrix
Biodiesel yield % (w/w of oil)
X1
X2
Experimental
RSM predicted
25 35 25 35 22.93 37.07 30 30 30 30 30 30 30
1 1 2 2 1.5 1.5 0.793 2.207 1.5 1.5 1.5 1.5 1.5
74.500 80.600 74.100 75.800 75.500 82.900 78.300 70.200 83.400 83.800 83.500 84.100 84.200
75.186 81.953 73.222 75.589 75.734 82.191 76.957 71.068 83.800 83.800 83.800 83.800 83.800
X1 packed bed height, X2 substrate flow rate
Environ Sci Pollut Res Table 3 Biodiesel production from karanja oil in a packed bed reactor using RSM
Run no.
CCD experimental design matrix
Biodiesel yield % (w/w of oil)
X1
X2
Experimental
RSM predicted
1
25
1
71.30
71.99
2 3 4 5 6 7 8 9 10 11 12 13
35 25 35 22.93 37.07 30 30 30 30 30 30 30
1 2 2 1.5 1.5 0.793 2.207 1.5 1.5 1.5 1.5 1.5
76.50 71.20 72.20 71.90 78.80 75.10 65.80 79.20 79.80 80.20 79.60 79.00
78.08 69.71 71.60 72.48 78.12 73.50 67.29 79.56 79.56 79.56 79.56 79.56
X1 packed bed height, X2 substrate flow rate
cells and immobilized C. rugosa enzyme in the ratio (4:1) were used as catalyst. The effects of substrate flow rate and packed bed height were studied for transesterification of jatropha and karanja oils. The reactions were carried out at room temperature, and the samples were collected after 42 h of reaction time.
Results and discussion Shimada et al. (1999) studied the production of fatty acid methyl esters from vegetable oils using immobilized Candida antarctica lipase. Shaw et al. (1991) studied lipase catalyzed ethanolysis and isopropanolysis of triacyl glycerols with longchain fatty acids using celite-immobilized Pseudomonas lipase. Selmi and Thomas (1998) studied immobilized lipase catalyzed ethanolysis of sunflower oil in a solvent-free medium using lipozyme. Kaieda et al. (1999) produced fatty acid methyl esters from plant oils in a solvent-free system catalyzed by various
Table 4 Regression results from the data of CCD experiments for biodiesel production using Jatropha oil
lipases including Rhizopus oryzae lipase, Pseudomonas cepacia lipase, C. rugosa lipase, Pseudomonas fluorescens lipase, Penicillium roqueforti lipase, and Candida lipolytica lipase. Enzymatic transesterification has several drawbacks: longer reaction time—higher catalyst concentration is required to completion of reaction—and high cost of production—repeated use of lipase becomes possible after immobilization of lipase on carrier and it loses its activity in 100 days of application (Jeong and Park 2008; Fjerbaek et al. 2009). A five-level two-factor CCD was employed in this optimization study. Packed bed height (X1) and substrate flow rate (X2) were optimized for the transesterification of jatropha and karanja oils. Biodiesel yield (Y) was taken as the response of the designed experiments. The uncoded (actual) levels of the independent variables are given in Table 1. Thirteen experiments were conducted with five replications at the center points to evaluate the pure error as shown in Tables 2 and 3. Once the experiments are performed, the response variable (biodiesel yield) was
Model terms
Coefficient
Standard error of coefficient
t value
P value
Constant x1 x2 x1x1 x2x2 x1x2
83.800 2.283 −2.082 −2.419 −4.894 −1.100
0.4395 0.3467 0.3467 0.3718 0.3718 0.4903
191.105 6.556 −6.005 −6.506 −13.164 −2.244
0.000 0.000 0.000 0.000 0.000 0.060
R2 =97.6 % R2 (adj)=95.8 %
Environ Sci Pollut Res Table 5 Regression results from the data of CCD experiments for biodiesel production using Karanja oil
Model terms
Coefficient
Standard error of coefficient
t value
P value
Constant x1 x2 x1x1 x2x2
79.56 1.995 −2.194 −2.130 −4.580
0.5577 0.4645 0.4645 0.4983 0.4983
135.37 4.293 −4.723 −4.275 −9.192
0.000 0.004 0.000 0.004 0.000
x1x2
−1.050
0.6571
−1.593
0.154
R2 =95.2 % R2 (adj)=91.2 %
fitted with a second-order polynomial model in order to correlate the response variable to the independent variable. Multiple regression analysis of the experimental data
on biodiesel yield gave the following second-order polynomial Eqs. 1 and 2 for jatropha and karanja oils, respectively.
Y 1 ðBiodiesel yieldÞ ¼ 83:800 þ 2:283x1 −2:082x2 – 1:100x1 x2 −2:419x1 2 – 4:894x2 2 Eq
ð1Þ
Y 2 ðBiodiesel yieldÞ ¼ 79:560 þ 1:995x1 −2:194x2 – 1:050x1 x2 −2:130x1 2 – 4:580x2 2 Eq
ð2Þ
The Student’s t test distribution and the corresponding P values, along with parameter estimates for the biodiesel yield were evaluated using MINITAB software and are given in Tables 4 and 5 for jatropha and karanja oils, respectively. If the value of P is less than 0.05, the corresponding coefficient is significant. The P values indicate that the linear effects of packed bed height (X1), substrate flow rate (X2), quadratic effects of packed bed height (X1), and substrate flow rate (X2) were found to be significant at 5 % confidence level as shown in the Table 4 for maximum biodiesel yield from jatropha oil. The P values from Table 5 indicate that linear effects of packed bed height (X1), substrate flow rate (X2), quadratic effects of packed bed height (X1), and substrate flow rate (X2) were found to be significant at 5 % confidence level for maximum biodiesel yield from karanja oil. The values of the determination coefficient (R2) for jatropha and karanja oils are found to be 97.6 and Table 6 Analysis of variance (ANOVA) for the fitted quadratic polynomial model for optimization of transesterification conditions using Jatropha oil
95.2 % indicating the better goodness of the fit, and only 2.4 and 4.8 % of the total variations are not explained by the model for jatropha and karanja oils, respectively. The adjusted determination coefficient Adj. (R2) for jatropha and karanja oils was found to be 95.8 and 91.2 %, respectively. The ANOVA results for the quadratic regression model for biodiesel yield from jatropha and karanja oils are given in Tables 6 and 7, respectively. The ANOVA of the regression model demonstrates that the model is highly significant, as evident from the F values (Fmodel =56.22 and Fmodel =27.48) and low probability values (Pmodel > F=0.000 and Pmodel >F=0.000) for jatropha and karanja oils, respectively. The linear effects of packed bed height and substrate flow rate, and the quadratic effects of packed bed height and substrate flow rate were highly significant at 5 % level for jatropha oil. This is evident from the P values in Table 4
Source
Degree of freedom
Sum of squares
Mean square
F value
P value
Regression Linear Square Interaction Residual error Lack of fit Pure error Total
5 2 2 1 7 3 4 12
270.55 75.376 189.034 4.840 6.730 6.230 0.500 270.980
54.050 38.1881 94.5169 4.8400 0.9614 2.0767 0.1250
56.22 39.72 98.31 5.03
0.000 0.000 0.000 0.060
16.61
0.100
Environ Sci Pollut Res Table 7 Analysis of variance (ANOVA) for the fitted quadratic polynomial model for optimization of transesterification conditions using karanja oil
Source Regression Linear Square Interaction Residual error Lack of fit Pure error Total
Degree of freedom
Sum of squares
Mean square
F value
P value
5 2 2 1 7
239.298 70.342 162.546 4.410 12.090
42.596 35.1712 81.2728 4.4100 1.7271
27.48 20.36 47.06 2.55
0.000 0.001 0.000 0.154
3 4 12
11.178 0.912 249.388
3.7259 0.2280
16.34
0.010
(x1 =0.000), (x2 =0.000), (x22 =0.000), and (x12 =0.000). The linear effects of packed bed height and substrate flow rate, and the quadratic effects of packed bed height and substrate flow rate were found to be highly significant at 5 % level for karanja oil. This is evident from the P values in Table 5 (x1 = 0.004), (x2 =0.0004), (x12 =0.004), and (x22 =0.000). The biodiesel yield was found to increase initially when there was an increase in substrate flow rate and packed bed height. The maximum biodiesel yield of 84.2 and 80.2 % was achieved at 30-cm packed bed height with substrate flow rate around 1.5 L/h. Further increase in substrate flow rate (>1.5 L/h) gave significant decrease in biodiesel yields. At low flow rates, low yields of biodiesel were obtained due to mass transfer resistance at liquid film layer. An increase in substrate flow rate caused reduction in mass transfer limitations causing higher reaction rates and hence resulting in increase in the biodiesel yield (Tepe and Dursun 2008). However, further increase in substrate flow rate (
POLLUTION CONTROL TECHNOLOGIES AND ALTERNATE ENERGY OPTIONS
Production and optimization of biodiesel using mixed immobilized biocatalysts in packed bed reactor S. Bakkiyaraj 1 & Mahin Basha Syed 1,2 & M. G. Devanesan 1 & Viruthagiri Thangavelu 1
Received: 15 March 2015 / Accepted: 21 April 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Vegetable oils are used as raw materials for biodiesel production using transesterification reaction. Several methods for the production of biodiesel were developed using chemical (alkali and acidic compounds) and biological catalysts (lipases). Biodiesel production catalyzed by lipases is energy and cost-saving processes and is carried out at normal temperature and pressure. The need for an efficient method for screening larger number of variables has led to the adoption of statistical experimental design. In the present study, packed bed reactor was designed to study with mixed immobilized biocatalysts to have higher productivity under optimum conditions. Contrary to the single-step acyl migration mechanism, a two-step stepwise reaction mechanism involving immobilized Candida rugosa lipase and immobilized Rhizopus oryzae cells was employed for the present work. This method was chosen because enzymatic hydrolysis followed by esterification can tolerate high free fatty acid containing oils. The effects of flow rate and bed height on biodiesel yield were studied using two factors five-level central composite design (CCD) and response surface methodology (RSM). Maximum biodiesel yield of 85 and 81 % was obtained for jatropha oil and karanja oil with the optimum bed height and optimum flow rate of 32.6 cm and 1.35 L/h, and 32.6 cm and 1.36 L/h, respectively. Responsible editor: Bingcai Pan * Mahin Basha Syed [email protected] 1
Biochemical Engineering Lab, Chemical Engineering Department, Annamalai University, Chidambaram 608002, Tamilnadu, India
2
Environmental Engineering Lab, Nawab Shah Alam Khan College of Engineering and Technology, Hyderabad 500024, Telangana, India
Keywords Biodiesel . Response surface methodology . Central composite design . Packed bed reactor . Rhizopus oryzae . Candida rugosa lipase
Introduction Recently, the increased crude oil prices and environmental awareness have renewed the focus on the use of vegetable oils and animal fats as alternative fuels. The raw materials used for production of biodiesel can be either crude refined or waste such as frying oils/fats (Marchetti et al. 2008). The use of alternative fuel for diesel engines dates back almost a century, when Rudolf Diesel first demonstrated the diesel engine at the 1900 world exhibition in Paris with peanut oil. Given the potential of vegetable oils as an alternative fuel, different nations, depending on their unique climate and soil conditions, are exploring a variety of vegetable oils for diesel fuel applications. The vegetable oils therefore need to be chemically modified to make them suitable as engine fuels (Harwood 1984). Vegetable oils comprise 90–98 % triglycerides and small amounts of monoglycerides, diglycerides, free fatty acids (FFA), phospholipids, phosphatides, carotenes, tocophenols, sulfur compounds, and traces of water. Triglycerides are esters of three fatty acids and one glycerol molecule and contains substantial amount of oxygen in its structure. Fatty acids of vegetable oils vary in their carbon chain length and in the number of double bonds (Barnwal and Sharma 2005). An important factor in biodiesel production is the fatty acid composition of the source oil or fat. Oils containing higher levels of saturated fatty acids than unsaturated fatty acids (have one or more double bonds) may solidify and clog the fuel lines during the winter condition (Pinto et al. 2005; Akoh et al. 2007; Demirbas 2008).
Environ Sci Pollut Res
Transesterification consists of a sequence of three consecutive reversible reactions where triglycerides are converted to diglycerides, monoglycerides, and free fatty acids with formation of an ester molecule and glycerol at each step; thus, three ester molecules are produced from one molecule of triglyceride. Since the reactions are reversible, excess alcohol is used to shift the equilibrium towards the product side. However, the yield of biodiesel is independent of the type of the alcohol used, and the selection of alcohol is based on cost and performance of it. Methanol and ethanol are the most preferred, especially methanol because of its low cost and its physical and chemical advantages (polar- and short-chained alcohol) over ethanol. Pressure shockwaves produced by underwater high-voltage discharges on underground and inadequately soaked seeds of Jatropha curcas L. The depth of penetration of excess methanol solvent and the number of shockwaves seem to be very important and increase the oil yield up to 94 % (Maroušek et al. 2012). It was found that grinding, followed by maceration that may enhance the effect of following pressure shockwaves (50 to 60 MPa), will increase oil extraction over 94 % (Maroušek et al. 2013a). Enzymatic hydrolysis enhanced by pressure shockwaves opens new possibilities in J. curcas L. (Maroušek et al. 2013b). Microemulsion also causes heavy carbon deposit, injector needle deposit, and incomplete combustion during long-term endurance tests (Pryde 1983). Thermal decomposition of triglycerides produces a mixture of compounds such as alkanes, alkenes, alkadienes, aromatics, and carboxylic acid (Billaud et al. 1995; Ma and Hanna 1999). The process is however economically unviable due to high cost of equipments. This paper evaluates production of biodiesel by mixed immobilized biocatalysts in packed bed reactor using karanja and jatropha oils. The effects of flow rate and bed height on biodiesel yield were studied using two-factor five-level central composite design (CCD) and response surface methodology (RSM). RSM is a powerful and efficient mathematical approach widely applied in the optimization of fermentation process. It can give information about the interaction between process parameters, provide information necessary for design and process optimization, and give multiple responses at the same time. The factors that have significant effects were optimized using a CCD and response surface analysis.
Materials and methods Medium preparation Rhizopus oryzae, MTCC (262), was obtained from microbial type culture collection Chandigarh (India). R. oryzae was maintained on 7 % potato dextrose agar medium. After 3 days of incubation at 25 °C, the agar slants were stored at 4 °C. The growth medium used contains the following: olive oil 30 g/L,
polypeptone (50 wt% pepton, 50 wt% trypton) 70 g/L, KH2PO4 (99.0 %) 1.0 g/L, MgSO4·7 H2O (99.0 %) 0.5 g/L, and NaNO3 (98.0 %) 1.0 g/L (Ban et al. 2002). R. oryzae was grown aerobically in 250-mL Erlenmeyer flasks containing the growth medium. Lipase preparation Lipase powder (Sigma-Aldrich) (1 g) was suspended in 100 mL of 0.25 M sodium phosphate buffer (pH 7); lipase solution was centrifuged (Remi centrifuge) at 4000 rpm for 15 min at 4 °C. The supernatant was stored at the same temperature prior to immobilization. Normalization of the activity was done by diluting the enzyme solution when the transesterification ability of the enzymes was compared. Immobilization by entrapment Sodium alginate solution was prepared by dissolving 5 % (w/v) sodium alginate in boiling water and autoclaved at 121 °C for 15 min. Both alginate slurry and 10 g of cells were mixed and stirred for 10 min to get a uniform alginate/cell mixture. The mixture was extruded drop by drop into the cold sterile CaCl2 (94.0 %) (0.2 M) solution through a sterile burette and was cured at 4 °C for 1 h. The beads were hardened by suspending it again in a fresh CaCl2 solution for 24 h at 4 °C with gentle agitation. Finally, these beads were washed with distilled water to remove excess calcium ions and un-entrapped cells. The beads were preserved in 0.9 % sodium chloride (99 %) solution in the refrigerator. The same procedure was adopted for enzyme immobilization; instead of cells, the enzyme supernatant was used for preparing the alginate/enzyme mixture. All other steps were similar for both the cases. Degumming of crude oils Crude karanja and jatropha oils were first degummed to remove the phospholipids prior to subjecting it to enzymatic transesterification. Degumming was done by heating the oils to 75 °C and adding 0.7 % (w/v) of NaCl at the same temperature followed by stirring at a constant speed of 400 rpm. The stirring rate was reduced gradually to 30 rpm based on the formation of colloidal particles, and the stirring was completely stopped when there was no suspended particles in the oils. The colloidal particles formed were removed by centrifugation at 2012g for 10 min. The resultant oils was dried by vacuum drying for 30 min and used for transesterification (Yuan et al. 2008). Batch transesterification Methanolysis reaction was conducted with pre-designed values of molar ratio of oil to alcohol, reaction temperature, reaction time, added water content, and immobilized beads
Environ Sci Pollut Res
concentration in a 50-mL Erlenmeyer flask with constant shaking (120 rpm) using incubated orbital shaker. Solvent (n-hexane) (99.9 %) is added to the reaction mixture to enhance the solubility of the reactants. After pre-designed time, the biocatalysts were removed from the reaction mixture by filtration. The produced ester and by-product glycerol were separated using a separating funnel. The produced esters are quantified using high-performance liquid chromatography equipment (HPLC). Transesterification reactions were carried out in triplicate, and the mean results were recorded.
Packed bed reactor Packed bed reactor designed consists of a cylindrical perspex column of 2.5-cm diameter and 25 cm, 40 cm in height to study the effect of bed height on biodiesel production as shown in Fig. 1. The reactor was used for the continuous biodiesel production studies from karanja and jatropha oils using mixed immobilized biocatalysts. Packed bed column contains the immobilized beads of R. oryzae cells and Candida rugosa lipase in fixed ratio. In the columns, wire mesh was placed at the bottom and top of the column to retain the beads inside the column. The raw oil and methanol (reactant mixture) at a known flow rate based on CCD is pumped in to the column in an upward direction through the column using a peristaltic pump at fixed flow rate. The outflow from the column is again mixed with the inflow in the reactant storage vessel. Samples from outflow were collected at regular intervals and were analyzed for biodiesel production.
Fig 1 Production of biodiesel from nonedible oils using mixed immobilized biocatalysts
High-performance liquid chromatographic analysis A High-Performance Liquid Chromatographic system (HPLC Model- LC 20 AT Prominence, Shimadzu, Japan) fitted with Refractive index detector (RID-10A, Shimadzu, Japan) and millennium 32 system software was used to quantify the fatty acid methyl esters produced during reaction. Separations were carried out on a 238 nm in Luna C18 column of particle size 5 μm, and (250×4.6 mm) I.D. Methanol (99.8 %) mobile phase was filtered through a 0.45-μm membrane filter (Millipore) and then degassed ultrasonically prior to use. The flow rate was 1 mL/min, the injection volume was 20 μL, and the column oven temperature was maintained at 40 °C. Each component in the samples analyzed was identified by comparing its retention time with that of the respective standards. Quantification was carried out by integration of the peaks using external standards followed by calculating the % yield as weight of methyl esters produced to weight of oil initially taken. Central composite experimental design and response surface methodology Statistical methods provide an efficient alternative methodology for traditional one factor at a time approach to optimize a particular process by considering the mutual interactions among the variables and to give an estimate of the combined effects of these variables (Myers and Montgomery 1995). The graphical representation of the polynomial model resulted in the response surface plots that characterize the individual, squared, and interactive effects of test variables on the response. The fractional factorial design consisted of 16 factorial points, 10 axial points, and six center points. The variables and their levels selected for the study were the following: molar ratio of alcohol to oil (1:1–1:5), reaction temperature (30–50 °C), reaction time (varied depending on type of biocatalyst used, 72 h as the center point for cells, 24 h for enzymes and 40 h mixed immobilized biocatalysts respectively), water content (0–20 % w/w of oil), and concentration of immobilized beads (1–5 % w/w of oil). The independent factors (xi), levels, and experimental design in terms of coded and uncoded values are shown in the BResults^ section. For statistical calculation, the test factors were coded according to Eq. 2.1: . xi ¼ X i –X o ΔX i i ¼ 1; 2; 3; ………:: k ð2:1Þ where xi is the coded value of an independent variable, Xi is the real value of the independent variable, Xo is the real value of the independent variable at the center point, and ΔXi is the step change value. The Design Expert 6.0 software was used for regression and graphical analysis of the data obtained. The maximum
Environ Sci Pollut Res
values of the percentage yield of biodiesel (or) methyl esters were taken as the responses of the design experiment. Once the experiments were performed, the response variable was fitted as second-order model in order to correlate the response variable to the independent variable. The general form of the second-order polynomial equation for the yield of fatty acid methyl ester is given below in Eq. 2.2 Y ¼ β0 þ
X
β i xi þ
X
2
βii x1 þ
X
β i j xi x j
ð2:2Þ
where methyl esters yield (or) biodiesel yield (Y) is the predicted response variable, i and j are the linear and quadratic coefficients respectively, β is the regression coefficient of the model, and xi, xj ( i=1,3; j=1,3, i=j) represent the independent variables (reaction conditions) in the form of coded values. The MINITAB 16 software statistical package was used for regression analysis of the experimental data obtained and to estimate the coefficients of the regression equation. The accuracy and general predictive ability of the polynomial model was evaluated by the multiple correlation coefficient (R) and by the coefficient of determination (R2). The statistical significance of the model was determined by F-test. The mean squares are obtained by dividing the sum of squares of each of the two sources of variation, the model and the error variance, by the respective degrees of freedom. The Fisher’s variance ratio—F value is the ratio of the mean square due to regression to the mean square due to the error which is a statistically valid measure of how well the factors describe the variations in the data about its mean—can be calculated from ANOVA. Since coding of the variable enables direct comparison of the partial regression coefficients, their significance was determined by Student’s t test and the associated probabilities. The P values are used as a tool to check the Table 2 Biodiesel production from Jatropha oil in a packed bed reactor using RSM
Run no.
1 2 3 4 5 6 7 8 9 10 11 12 13
Table 1 Experimental range and levels of the independent variables used in CCD and RSM for biodiesel production from jatropha and karanja oils using mixed biocatalyst in packed bed reactor Parameters with designate
X1: Packed bed height (cm) X2: Substrate flow rate (L/h)
Coded levels −1.414
−1
0
1
1.414
22.93 0.793
25 1
30 1.5
35 2
37.07 2.207
significance of each of the coefficients, which in turn may indicate the pattern of the interaction between the variables. The smaller the value of P, the more significant is the corresponding coefficient. The response surface plots were used to describe the individual and cumulative effects of the independent variables as well as the mutual interactions between the variables on the biodiesel yield. The second degree polynomial equation was maximized by the constraint search procedure using MATL AB software (Version 7.0, The MathWorks, Inc. Natick, USA) to obtain the optimal levels of the independent variables and the predicted maximum biodiesel yield. Optimization of packed bed height and substrate flow rate using CCD and RSM for transesterification of jatropha and karanja oils in packed bed reactor The experiments were conducted under the optimal conditions of substrate molar ratio (1:3), hexane (solvent) 25 % (w/w of oil), bead size (0.25 cm), reaction time (42 h), and added water content 11 % (w/w of oil ) and were maintained constant for all the experiments conducted in Packed bed reactor. The mixed immobilized biocatalysts consisting of immobilized R. oryzae
CCD experimental design matrix
Biodiesel yield % (w/w of oil)
X1
X2
Experimental
RSM predicted
25 35 25 35 22.93 37.07 30 30 30 30 30 30 30
1 1 2 2 1.5 1.5 0.793 2.207 1.5 1.5 1.5 1.5 1.5
74.500 80.600 74.100 75.800 75.500 82.900 78.300 70.200 83.400 83.800 83.500 84.100 84.200
75.186 81.953 73.222 75.589 75.734 82.191 76.957 71.068 83.800 83.800 83.800 83.800 83.800
X1 packed bed height, X2 substrate flow rate
Environ Sci Pollut Res Table 3 Biodiesel production from karanja oil in a packed bed reactor using RSM
Run no.
CCD experimental design matrix
Biodiesel yield % (w/w of oil)
X1
X2
Experimental
RSM predicted
1
25
1
71.30
71.99
2 3 4 5 6 7 8 9 10 11 12 13
35 25 35 22.93 37.07 30 30 30 30 30 30 30
1 2 2 1.5 1.5 0.793 2.207 1.5 1.5 1.5 1.5 1.5
76.50 71.20 72.20 71.90 78.80 75.10 65.80 79.20 79.80 80.20 79.60 79.00
78.08 69.71 71.60 72.48 78.12 73.50 67.29 79.56 79.56 79.56 79.56 79.56
X1 packed bed height, X2 substrate flow rate
cells and immobilized C. rugosa enzyme in the ratio (4:1) were used as catalyst. The effects of substrate flow rate and packed bed height were studied for transesterification of jatropha and karanja oils. The reactions were carried out at room temperature, and the samples were collected after 42 h of reaction time.
Results and discussion Shimada et al. (1999) studied the production of fatty acid methyl esters from vegetable oils using immobilized Candida antarctica lipase. Shaw et al. (1991) studied lipase catalyzed ethanolysis and isopropanolysis of triacyl glycerols with longchain fatty acids using celite-immobilized Pseudomonas lipase. Selmi and Thomas (1998) studied immobilized lipase catalyzed ethanolysis of sunflower oil in a solvent-free medium using lipozyme. Kaieda et al. (1999) produced fatty acid methyl esters from plant oils in a solvent-free system catalyzed by various
Table 4 Regression results from the data of CCD experiments for biodiesel production using Jatropha oil
lipases including Rhizopus oryzae lipase, Pseudomonas cepacia lipase, C. rugosa lipase, Pseudomonas fluorescens lipase, Penicillium roqueforti lipase, and Candida lipolytica lipase. Enzymatic transesterification has several drawbacks: longer reaction time—higher catalyst concentration is required to completion of reaction—and high cost of production—repeated use of lipase becomes possible after immobilization of lipase on carrier and it loses its activity in 100 days of application (Jeong and Park 2008; Fjerbaek et al. 2009). A five-level two-factor CCD was employed in this optimization study. Packed bed height (X1) and substrate flow rate (X2) were optimized for the transesterification of jatropha and karanja oils. Biodiesel yield (Y) was taken as the response of the designed experiments. The uncoded (actual) levels of the independent variables are given in Table 1. Thirteen experiments were conducted with five replications at the center points to evaluate the pure error as shown in Tables 2 and 3. Once the experiments are performed, the response variable (biodiesel yield) was
Model terms
Coefficient
Standard error of coefficient
t value
P value
Constant x1 x2 x1x1 x2x2 x1x2
83.800 2.283 −2.082 −2.419 −4.894 −1.100
0.4395 0.3467 0.3467 0.3718 0.3718 0.4903
191.105 6.556 −6.005 −6.506 −13.164 −2.244
0.000 0.000 0.000 0.000 0.000 0.060
R2 =97.6 % R2 (adj)=95.8 %
Environ Sci Pollut Res Table 5 Regression results from the data of CCD experiments for biodiesel production using Karanja oil
Model terms
Coefficient
Standard error of coefficient
t value
P value
Constant x1 x2 x1x1 x2x2
79.56 1.995 −2.194 −2.130 −4.580
0.5577 0.4645 0.4645 0.4983 0.4983
135.37 4.293 −4.723 −4.275 −9.192
0.000 0.004 0.000 0.004 0.000
x1x2
−1.050
0.6571
−1.593
0.154
R2 =95.2 % R2 (adj)=91.2 %
fitted with a second-order polynomial model in order to correlate the response variable to the independent variable. Multiple regression analysis of the experimental data
on biodiesel yield gave the following second-order polynomial Eqs. 1 and 2 for jatropha and karanja oils, respectively.
Y 1 ðBiodiesel yieldÞ ¼ 83:800 þ 2:283x1 −2:082x2 – 1:100x1 x2 −2:419x1 2 – 4:894x2 2 Eq
ð1Þ
Y 2 ðBiodiesel yieldÞ ¼ 79:560 þ 1:995x1 −2:194x2 – 1:050x1 x2 −2:130x1 2 – 4:580x2 2 Eq
ð2Þ
The Student’s t test distribution and the corresponding P values, along with parameter estimates for the biodiesel yield were evaluated using MINITAB software and are given in Tables 4 and 5 for jatropha and karanja oils, respectively. If the value of P is less than 0.05, the corresponding coefficient is significant. The P values indicate that the linear effects of packed bed height (X1), substrate flow rate (X2), quadratic effects of packed bed height (X1), and substrate flow rate (X2) were found to be significant at 5 % confidence level as shown in the Table 4 for maximum biodiesel yield from jatropha oil. The P values from Table 5 indicate that linear effects of packed bed height (X1), substrate flow rate (X2), quadratic effects of packed bed height (X1), and substrate flow rate (X2) were found to be significant at 5 % confidence level for maximum biodiesel yield from karanja oil. The values of the determination coefficient (R2) for jatropha and karanja oils are found to be 97.6 and Table 6 Analysis of variance (ANOVA) for the fitted quadratic polynomial model for optimization of transesterification conditions using Jatropha oil
95.2 % indicating the better goodness of the fit, and only 2.4 and 4.8 % of the total variations are not explained by the model for jatropha and karanja oils, respectively. The adjusted determination coefficient Adj. (R2) for jatropha and karanja oils was found to be 95.8 and 91.2 %, respectively. The ANOVA results for the quadratic regression model for biodiesel yield from jatropha and karanja oils are given in Tables 6 and 7, respectively. The ANOVA of the regression model demonstrates that the model is highly significant, as evident from the F values (Fmodel =56.22 and Fmodel =27.48) and low probability values (Pmodel > F=0.000 and Pmodel >F=0.000) for jatropha and karanja oils, respectively. The linear effects of packed bed height and substrate flow rate, and the quadratic effects of packed bed height and substrate flow rate were highly significant at 5 % level for jatropha oil. This is evident from the P values in Table 4
Source
Degree of freedom
Sum of squares
Mean square
F value
P value
Regression Linear Square Interaction Residual error Lack of fit Pure error Total
5 2 2 1 7 3 4 12
270.55 75.376 189.034 4.840 6.730 6.230 0.500 270.980
54.050 38.1881 94.5169 4.8400 0.9614 2.0767 0.1250
56.22 39.72 98.31 5.03
0.000 0.000 0.000 0.060
16.61
0.100
Environ Sci Pollut Res Table 7 Analysis of variance (ANOVA) for the fitted quadratic polynomial model for optimization of transesterification conditions using karanja oil
Source Regression Linear Square Interaction Residual error Lack of fit Pure error Total
Degree of freedom
Sum of squares
Mean square
F value
P value
5 2 2 1 7
239.298 70.342 162.546 4.410 12.090
42.596 35.1712 81.2728 4.4100 1.7271
27.48 20.36 47.06 2.55
0.000 0.001 0.000 0.154
3 4 12
11.178 0.912 249.388
3.7259 0.2280
16.34
0.010
(x1 =0.000), (x2 =0.000), (x22 =0.000), and (x12 =0.000). The linear effects of packed bed height and substrate flow rate, and the quadratic effects of packed bed height and substrate flow rate were found to be highly significant at 5 % level for karanja oil. This is evident from the P values in Table 5 (x1 = 0.004), (x2 =0.0004), (x12 =0.004), and (x22 =0.000). The biodiesel yield was found to increase initially when there was an increase in substrate flow rate and packed bed height. The maximum biodiesel yield of 84.2 and 80.2 % was achieved at 30-cm packed bed height with substrate flow rate around 1.5 L/h. Further increase in substrate flow rate (>1.5 L/h) gave significant decrease in biodiesel yields. At low flow rates, low yields of biodiesel were obtained due to mass transfer resistance at liquid film layer. An increase in substrate flow rate caused reduction in mass transfer limitations causing higher reaction rates and hence resulting in increase in the biodiesel yield (Tepe and Dursun 2008). However, further increase in substrate flow rate (
