This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
To define a simple kinetic model, the ultrasound-assisted sunflower oil methanolysis catalyzed by KOH was studied under a set of operational variables, namely methanol-to-oil molar ratios of 4.5:1, 6:1 and 7.5:1, KOH loadings of 0.3, 0.5 and 0.7% (based on the oil weight) and reaction temperatures of 20, 30 and 40ËšC. The methanolysis process rate was observed to be controlled by the mass transfer limitation in the initial heterogeneous regime and by the chemical reaction in the pseudo-homogeneous regime. The model involving the irreversible second-order kinetics was established and used for simulation of the triglyceride conversion and the FAME formation in the latter regime. The model assumed that the neutralization of free acids and the saponification reaction were negligible. A good agreement between the proposed model and the experimental data in the chemically controlled regime, where the reaction mixture could be considered as a pseudo-homogeneous system, was found.
Keywords: Biodiesel, kinetics, methanolysis, modeling, sunflower oil, ultrasound.
Fatty acid alkyl esters, known as biodiesel, are a renewable, biodegradable and non-toxic fuel, which is a good alternative for fossil fuels.1 The use of biodiesel leads to many environmental benefits such as less pollution of air, water and soil and minimal impact on human health.2 The most often used process for the biodiesel production is methanolysis, a reaction between triglycerides (TG) from a variety of plant oils and animal fats, and methanol in the presence of a catalyst. Base catalysts such as KOH, NaOH and methoxides in homogeneous media are usually applied for the industrial biodiesel production.
The operational variables that affect fatty acid methyl ester (FAME) formation are: the type and the catalyst loading, the molar ratio of methanol to vegetable oil, the reaction temperature, the agitation intensity and the purity of the reactants. Methanol and TG are immiscible3 and the reaction between the reactants occurs on the interfacial surface.4 Intensive mixing should be applied to enhance the contact surface area between the two immiscible reactants, causing the increase of the methanolysis reaction rate.4,5 Based on the known effects of ultrasonic cavitation such as mixing, heating and disruption of the interface, some researchers have sped up the reaction rate by carrying out methanolysis in the presence of ultrasound.6-13
The kinetics of the base-catalyzed methanolysis reaction, fundamental to the reactor design, has been widely studied by several researchers.5,11,12,14-19 In these studies, either a complex kinetic model based on the mechanism of the methanolysis reaction consisting of three reversible consecutive-competitive reactions5,15,16,18,19 or the simplified kinetic models involving irreversible reaction11,12,14,17 have been used. Also, an initial heterogeneous stage controlled by mass transfer rate and a chemically controlled pseudo-homogeneous stage are well-recognized from the beginning to the completion of the methanolysis reaction. The former stage is usually observed at lower reaction temperatures and in the case of non-adequate mixing. Under these reaction conditions, the methanolysis kinetic model should include the mass transfer limitations.17 In modeling the chemically controlled stage forward and reverse second-order reactions have been most often used.5,16,18 Exceptionally, Freedman et al.15 used a combination of second-order consecutive and fourth-order shunt reactions to describe the kinetics of methanolysis. Darnoko and Cheryan20 suggested a second-order kinetic for the initial stages of the reaction, but they considered the forward reactions only. To simplify the reactor design, simple kinetic models, which do not require complex computer calculations, are favored. StamenkoviÄ‡ et al.17 reported a simple kinetic model which consisted of an irreversible second-order reaction followed by a reversible second-order reaction close to the completion of the methanolysis reaction.
The base-catalyzed methanolysis kinetics in the presence of ultrasound has hardly been studied and contradictory results have been reported. For a 6:1 methanol/soybean oil molar ratio and a 25 to 60oC temperature range in the presence of low frequency ultrasound (20 kHz), Colucci et al.14 established a pseudo second-order kinetic model with respect to TG for the base-catalyzed methanolysis of soybean oil. On the other hand, Georgogianni et al.11,12 reported the first and second-order reaction with respect to TG for the base-catalyzed methanolysis of sunflower and cottonseed oils using both low frequency ultrasonication and mechanical agitation, although a better fit was obtained for the first order reaction kinetics. Deshmane et al.21 verified a kinetic model based on the first-order reaction followed by the second order with respect to fatty acids for ultrasound-assisted acid-catalyzed esterification of palm fatty acid distillate.
In the present work, the kinetics of the ultrasound-assisted sunflower oil methanolysis catalyzed by potassium hydroxide was studied in ranges of three operational variables: methanol-to-oil molar ratio (4.5:1, 6:1 and 7.5:1), catalyst loading (0.3, 0.5 and 0.7% based on the oil weight) and reaction temperature (20, 30 and 40ËšC). These ranges were based on the reports dealing with methanolysis optimization. The FAME formation in the ultrasound-assisted methanolysis of soybean oil was not substantially increased with increasing the methanol-to-oil molar ratio from 6.0:1 to 12:1 and the catalyst loading from 0.6 to 1.0%.22 For the same reaction driven by low-frequency ultrasound, Santos et al.23 optimized the methanol-to-oil molar ratio and the catalyst loading in the ranges of 3:1 to 12:1 and 0.2 to 0.6%, respectively. By increasing catalyst loading, FAME yield and reaction rate also increase, 23 but the excess catalyst negatively affects FAME yield.9,12,16,17 The reaction in the presence of NaOH at the concentration higher than 1.0%6,11,13,24,25 was followed by the strong soap formation, decreasing the FAME yield. Since the reaction temperature should be relatively low to reduce the energy consumption and the use of ultrasound at boiling temperature is futile, the temperature range below 40oC was chosen. The main goals were to develop a simple kinetic model of the base-catalyzed methanolysis reaction in the presence of ultrasound and to simulate the TG conversion and the FAME formation in the chemically-controlled stage of the process.
2. Theoretical background
The overall vegetable oil methanolysis reaction can be presented by the following stoichiometric equation:
where A is TG, B is methanol, R is FAME and S is glycerol.
For modeling the ultrasound-assisted sunflower oil methanolysis, the following assumptions are introduced:
The methanolysis process occurs via the initial heterogeneous regime, followed by the pseudo-homogeneous regime, where the mass transfer and the chemical reaction control the overall process kinetics, respectively. These regimes are well-recognized in the studies of the methanolysis reaction kinetics in the absence of ultrasound.17
The methanolysis of TG is the irreversible second-order reaction with respect to TG in the chemically-controlled regime. Due to the excess of methanol and the low product concentration, one can expect the reverse reactions to be negligible. The irreversible second-order reaction kinetic model with respect to TG has already been employed for base-catalyzed methanolysis of palm20 and sunflower17 oil without ultrasonic irradiation as well as for the ultrasound-assisted base-catalyzed methanolysis of soybean oil20.
The dispersion and the composition of each phase of the reaction mixture are uniform due to perfect mixing.
The content of free fatty acid in the oil is negligible, so the neutralization reaction can be ignored. From the oil and initial catalyst amounts as well as the acid value of the oil it was estimated that only less than 3% of the catalyst amount could participate in the neutralization of free fatty acids.
The saponification reaction is negligible because low catalyst amounts are used to enhance methanolysis.6
Mass transfer controlled regime
In the modeling of the methanolysis kinetics, only StamenkoviÄ‡ et al.17 have included the mass transfer limitation in the initial heterogeneous regime of the reaction carried out in a batch stirred reactor. According to the assumption (a), the rate of TG conversion, (), must be equal to the rate of TG mass transfer from the oil phase towards the interfacial area methanol-to-oil. Since the TG concentration on the interfacial area is equal to zero (cA,s → 0), it follows: 17
where kc is the TG mass transfer coefficient, a is the specific interfacial area, cA is the TG concentration in the oil phase and t is the time.
The TG concentration is related to the conversion degree of TG, :
so equation (1) is transformed into the following:
where cA and cAo are the actual and initial TG concentrations.
The instantaneous overall volumetric TG mass transfer coefficient, , for the mass transfer controlled regime can be estimated from equation (3) as follows:
Chemical reaction controlled regime
According to the assumption (b), the rate of the TG conversion in the chemically controlled regime is:
The apparent reaction rate constant is expected to depend on the reaction temperature, the initial methanol concentration and the catalyst concentration, as follows
where is the reaction rate constant, is the initial methanol concentration and is the catalyst concentration.
Equation (6) can be easily transformed into the following
For the boundary values: , and , (the beginning and any moment of the chemically controlled regime, respectively), then the integration of equation (7) gives:
where the integration constant is defined as follows:
The apparent reaction rate constant,, and the integration constant, , can be estimated from the slope and intercept of the linear dependence of versus time, respectively. With known values of the apparent reaction rate and integration constant, the TG conversion degree can be calculated for any moment of the reaction during the chemically controlled regime using equation (9).
3. Experimental Section
Refined, edible sunflower oil (Vital, Vrbas, Serbia) was used. The acid, saponification and iodine values of the oil were 0.09 mg KOH/g, 190 mg KOH/g and 129 g J2/100 g, respectively, determined by AOCS official methods.26 The content of water was 0.07% as determined by the volumetric Karl Fisher titration. Absolute methanol of 99.5% purity was purchased from Zorka Pharma (Šabac, Srbija). Potasium hydroxide pellets of min 95% purity were purchased from Mos-Lab (Belgrade, Srbija). Hydrochloric acid, conc. was obtained from Lach-Ner (Neratovce, Czech Republic). Methanol, 2-propanol and n-hexane, all HPLC grade, were obtained from Lab-Scan (Dublin, Irland). The HPLC standards for methyl esters of palmitic, stearic, oleic and linoleic acids, triolein, diolein and monoolein were purchased from Sigma Aldrich.
The reaction was carried out in a 250 mL three-neck round-bottom glass flask equipped with a condenser which was immersed in an ultrasonic cleaning bath (Sonic, Niš, Serbia; total power: 3 - 50 W; and internal dimensions: 30 - 15 - 20 cm) operating at 40 kHz frequency. The ultrasound generator had to be switched off after 60 minutes, and therefore, the ultrasound-assisted reaction was being followed for 60 minutes. The bath was filled with distilled water up to 1/3 of its volume (about 2.5 L). The temperature was controlled and maintained at the desired level (±0.1oC) by water circulating from a thermostated bath by a pump.
3.3. Reaction conditions
The methanolysis of sunflower oil was carried out at methanol-to-oil molar ratios of 4.5:1, 6:1 and 7.5:1, KOH loadings of 0.3, 0.5 and 0.7% (based on the oil weight) and reaction temperatures of 20, 30 and 40ËšC under the atmospheric pressure. All experiments were carried out in duplicate (total of 54 runs).
3.4. Experimental procedure
Methanol and KOH were fed into a three neck glass vessel. The vessel was placed into the ultrasonic bath and submitted to ultrasound at the desired temperature until all the catalyst was dissolved. The sunflower oil (45.96 g) was thermostated separately and added to the vessel. As soon as the oil was added, the reaction was timed. During the reaction, the samples (1 mL) were removed from the reaction mixture at certain time intervals, immediately quenched by adding a required amount of the aqueous hydrochloric acid solution (4.5%) to neutralize KOH and centrifuged at 3500 rpm (average 700g) for 15 min using a laboratory centrifuge (LC 320, Tehtnica, Zelezniki, Slovenia). The upper layer was withdrawn, dissolved in a 2-propanol/n-hexane (5:4 v/v) in an appropriate ratio (1:200) and filtered through a 0.45 μm Millipore filter. The resulting filtrate was used for HPLC analysis.
3.5. Composition of the reaction mixture
The composition of the reaction mixture samples was determined by a HPLC chromatograph (Agilent 1100 Series) using the somewhat modified method of HolÄapek et al.,27 as described elsewhere.4 A Zorbax Eclipse XDB-C18 column (4.6x150 mm with 5 mm particle size) held at a constant temperature (40oC) was used for separation. The mobile phase was methanol (reservoir A) and 2-propanol/n-hexane (5:4 v/v; reservoir B). A linear gradient from 100% A to 40% A + 60% B in 15 min was employed. The mobile phase flow rate was 1 mL/min. The injection volume was 20 mL and the components were detected at 205 nm. The calibration curves were prepared by using the standard mixture of FAME and the standard glycerides and used for the quantification of the FAME and the glycerides present in the samples analyzed.
The conversion degree of TG was calculated from the content of TG in the FAME/oil fraction of the reaction mixture, (in %), by the following equation:
The TG conversion degree was fitted to give the sigmoidal curve:
known as the Boltzman function, which describes the variation of TG conversion degree from its lowest ( = 0) to the highest () value; is the time at which TG conversion degree is halfway between the lowest and the highest value and is the steepness of the curve. Values of , , and were calculated by the non-linear regression method using a computer program (SIGMA PLOT 11.0 Trial).
4. Results and discussion
4.1. Methanolysis reaction analysis
Typical variations of the reaction mixture composition with the progress of the ultrasound-assisted sunflower oil methanolysis are shown in Figure 1. The shape of the curves representing the variations of FAME fraction with time was sigmoidal, showing the change of the reaction rate with time. The initial rate was slow due to mass transfer limitation caused by a very low interfacial area available for the reaction. As the reaction proceeded, the drops of methanol were disintegrated by an ultrasound action and additionally stabilized by monoglycerides (MG), diglycerides (DG) and soaps that acted as emulsifiers.28 In this way, the interfacial area available for mass transfer increased. It was surprising that ultrasound was not efficient to emulsify the reactants at the beginning of the reaction. Recently, Stavarache et al.8 have reported that the major part of the ultrasound-assisted base-catalyzed alcoholisys of vegetable oils takes place in the first 3 to 10 minutes of the reaction. The observed disagreement between this and the present study is probably due to different reactor geometries and sonication conditions (power and frequency) applied in the two studies. The decrease of the TG concentration was followed by the increase of FAME concentration. It should be mentioned that the FAME yield after the 60 minute reaction varied depending on the reaction conditions from 30% to 95%. The low FAME yield was due to the uncompleted methanolysis reaction under mild conditions. The concentrations of intermediate products, MG and DG, increased at the beginning of the reaction achieving the maximum, then decreased and finally remained nearly constant.
Mahamuni and Adewuyi22 first reported the sigmoidal shape of curves representing the variation of FAME yield during the base-catalyzed methanolysis of soybean oil in the presence of high frequency ultrasound at certain reaction conditions. They showed that power and frequency of ultrasound played significant roles in the methanolysis reaction. The former affects the drop breakage process and the latter influences the critical size of the cavitation bubble. The initial induction period shortened and the formation rate increased with increasing the power of ultrasound, while an optimum frequency for maximum FAME formation for a set of other reaction conditions was observed. For an optimum combination of power and frequency of ultrasound, the induction period shortened and the FAME formation rate increased with increasing the catalyst loading from 0.1% to 1%. The shortest induction period and the highest FAME formation rate were observed for the methanol-to-oil molar ratio of 6:1.
The sigmoidal shape of curves representing the variation of FAME yield meant that the methanolysis reaction occurred via the initial heterogeneous regime, where TG mass transfer from the bulk of the oil phase to the interfacial area limits the overall process rate (assumption a). Thus, it would be useful to analyze the variation of the overall volumetric TG mass transfer coefficient with the reaction time at different temperatures. For this purpose, Figure 2 shows the variations of the overall volumetric TG mass transfer coefficient with the reaction time at 20, 30 and 40oC in the initial mass transfer controlled regime of the methanolysis process performed at the methanol-to-oil molar ratio of 7.5:1 and the catalyst loading of 0.7%. The overall volumetric TG mass transfer coefficient were calculated using equation (4) and sigmoidal-fitted values of (equation 11).
At the higher reaction temperature, the overall volumetric TG mass transfer coefficient was higher at any moment of the reaction and increased rapidly with the methanolysis progress. This was explained by a faster MG and DG formation at higher temperatures, which was favorable to the stabilization of the emulsion formed. The increase of the reaction temperature decreased the reaction mixture viscosity which favored the agitation action on the drop breakage process.4 The specific interfacial area was increased because of the drop breakage, causing the increase of the overall volumetric TG mass transfer coefficient. On the other hand, the TG mass transfer coefficient decreased with the drop size reduction, due to lower internal circulation inside small drops. Decreasing of TG mass transfer coefficient with drop size reduction was predicted by the simulation model applied for alkaline transesterification reaction of soybean oil with alcohols catalyzed by potassium hydroxide within an ultrasonic reactor. 29 Therefore, the increase of the specific interfacial area had a primary effect on the increase of the TG mass transfer rate and the FAME formation rate both with the increase in the reaction temperature and with the methanolysis progress.
Figure 3, 4 and 5 illustrate the effects of the catalyst loading, the methanol-to-oil molar ratio and the reaction temperature on the FAME formation, respectively. At the methanol-to-oil molar ratio of 7.5:1 and the reaction temperature of 30oC, the reaction rate increased with increasing the catalyst loading in the range from 0.3 to 0.7% (based on the oil weight), due to the increase pronounced nucleophile methoxide anions that actually catalyzed the reaction (Figure 3). The effect of the catalyst amount on the reaction rate reduced at higher temperatures. The variation of the amount of FAME (moles) with time is shown in Figure 4. Independent of the catalyst amount or the reaction temperature, the reaction rate increased with increasing the initial methanol-to-oil molar ratio from 4.5:1 to 7.5:1 (Figure 4) because the forward reaction and the cavitation bubble formation were favored22 in the excess of methanol. FAME moles were analyzed because the FAME concentration was reduced with increasing the methanol-to-oil molar ratio at a constant oil mass, due to the increase of the total reaction mixture volume. The reaction rate increased with increasing the reaction temperature from 20oC to 40oC (Figure 5), due to the increase of the methanolysis reaction rate constant. Also, viscosity of sunflower oil reduced with increasing the temperature which resulted in the increase in cavitation events and the rate of the emulsion formation.22 In other words, a decrease in viscosity reduces the cavitation threshold, which leads to increase how much cavitation occurs.30 The temperature effect on the reaction rate reduced as the catalyst loading and the methanol-to-oil molar ratio increased.
Figure 3, 4 and 5
4.2. Reaction kinetic model
According to the assumption (a), the methanolysis process occurs via the initial heterogeneous regime followed by the pseudo-homogenous regime, where the mass transfer and the chemical reaction control the overall process kinetics. Since the alcoholic drop size was not measured during the methanolysis process due to the ultrasonic bath construction, the present kinetic model included only the latter regime. Variations of with time are shown in Figure 6. It was obvious that equation (7) corresponding to the irreversible pseudo second-order reaction (assumptions b) agreed well with the experimental data during the chemical reaction controlled regime. This agreed with the result of Colluchi et al.14 and Georgogianni et al.11,12, who established the irreversible pseudo-second-order kinetic model with respect to TG for the base-catalyzed methanolysis of soybean, sunflower and cotton seed oils in the presence of ultrasound. The irreversible pseudo-second-order kinetics has been also reported for the pseudo-homogeneous stage of vegetable oil methanolysis in the absence of ultrasound such as for sunflower17 and palm20 oil. The proposed kinetic model satisfactorily fitted the changes of the FAME concentration with the reaction time in the chemical reaction controlled regime as it could be concluded from the linear correlation coefficient and the relative deviations of the calculated and the experimental TG conversion degrees. Values of are higher than 0.91, except one which is 0.88, showing that the kinetic model is valid. The relative deviations of the calculated and the experimental TG conversion degrees (in the range where they fitted) were between 12.6% in all experiments.
To check the order of the base-catalyzed methanolysis of vegetable oils different from sunflower oil under ultrasound irradiation, we analyzed the kinetic data of Stavarache et al.7 on the TG conversion. Figure 7 shows the pseudo second-order reaction with respect to TG for commercial, corn and grape seed oils during the whole reaction period and only in the initial period of the reaction for canola and palm oils. Since different vegetable oils contain various types of fatty acids in different concentrations, the kinetic model and the value of the reaction rate constant was generally affected by the type of raw material. It appears that base-catalyzed methanolysis of vegetable oils with high percentage of linoleic acid follows the second-order reaction kinetics such as for the commercial, corn, grape seed (52.5, 66,5 and 82.2%, respectively)7, sunflower (68.7%)12, cotton seed (57.4%)11 and soybean (typically 55,5%)31 oil. However, for the oils with higher contents of palmitic and oleic acids, such as canola (60.0 and 26.4% of oleic and linoleic acid esters, respectively) and palm (44.0% of palmitic acid esters and 39.4% of oleic acid esters) oil,7 the base-catalyzed methanolysis is the second-order reaction in the initial period and the kinetic model changes in the later period. Darnoko and Cheryan20 also observed that palm oil methanolysis with 1% of KOH as catalyst and a molar ratio of 6:1 in the absence of ultrasound, was pseudo second order through the first 30 min of reaction.
When the irreversible pseudo-second order apparent reaction rate constant was preliminary correlated with the catalyst concentration, the experimental data obtained at a methanol-to-oil molar ratio and a reaction temperature gathered around a straight line passing through the origin point. Since separate lines were obtained for the possible combinations of methanol-to-oil molar ratio and reaction temperature, we concluded that the irreversible pseudo-second order apparent reaction rate constant was affected, as expected, not only by the reaction temperature and the catalyst concentration but also by the initial methanol concentration. In the next step, we correlated with at different reaction temperatures and linear curve were obtained as can be seen in Figure 8. The slope of the linear curves represented the reaction rate constant . Now, equation (6) can be rewritten as follows:
When the was correlated with , then a straight line was obtained, as can be seen in Figure 9. Thus, the Arrhenius equation can be applied for determining the activation energy for the methanolysis reactions:
where is the pre-exponential factor, is the activation energy, and is the gas constant. The activation energy and the pre-exponential factor for the irreversible pseudo-second-order reaction were calculated to be 46,2 kJ/mol and L3 mol-3.min-1, respectively ( = 0.999). A somewhat higher value of the activation energy (53.13 kJ/mol) has been recently published for ultrasound-assisted KOH-catalyzed methanolysis of soybean oil.14 Generally, the activation energy for the methanolysis reaction in the presence of ultrasonic irradiation is of the same order as that in its absence (53.3 kJ/mol).17 This suggests that there is no change in the reaction mechanism in the presence of ultrasonic irradiation, which can be explained by purely physical nature of ultrasound effect on methanolysis reaction.32
The kinetic model and the experimental data were also compared based on the variations of the molar concentrations of TG and FAME. Figure 9 shows that the kinetic model agrees well with the experimental data in the chemical reaction controlled regime. The molar TG concentration was calculated by equation (2) and the molar FAME concentration was calculated using the TG conversion degree corrected for the MG and DG formation.
Ultrasound-assisted base-catalyzed sunflower oil methanolysis was studied under various reaction conditions to determine the effects of the reaction temperature, methanol-to-oil molar ratio and catalyst loading on the reaction rate, and to model the methanolysis reaction kinetics. Generally, the reaction rate increased with increasing methanol-to-oil molar ratio, catalyst loading and the reaction temperature. A simple model which consisted of the initial mass transfer controlled regime followed by the chemical reaction controlled regime was shown to exist for the sunflower oil methanolysis performed under the applied reaction conditions. The former regime was not included in the modeling of the overall ultrasound-assisted methanolysis process kinetics because the alcohol phase drop size was not measured. In the second regime, the ultrasound-assisted methanolysis was the irreversible pseudo second-order reaction with respect to TG. This kinetic model is actually the same as that for the base-catalyzed methanolysis in the absence of ultrasound, showing that the same phenomena control the process rate independently of the type of mixing applied.
- Specific interfacial area,
- Pre-exponential factor, equation (13),
, - Parameters in equation (11), 1
- Concentration of TG in the oil phase,
- Initial concentration of TG,
- Concentration of TG on the interfacial area,
- Initial concentration of methanol,
- Concentration of catalyst,
- Integration constant, 1
- Concentration of FAME,
- Activation energy of the reaction,
- TG mass transfer coefficient,
- instantaneous overall volumetric TG mass transfer coefficient,
- Reaction rate constant,
- Apparent reaction rate constant for the irreversible pseudo second-order reaction,
- Rate of TG disappearance,
- Gas constant,
- Time, min
, - Parameters in equation (11),
- Content of TG in the FAME/oil fraction of the reaction mixture, %
- TG conversion degree, 1
DG - Diglycerides
FAME - Fatty acid methyl esters
MG - Monoglycerides
TG - Triglycerides