Scholarly article on topic 'Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol'

Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol Academic research paper on "Chemical sciences"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{Triglyceride / "Kinetic modelling" / "Dual sites" / Transesterification / Deactivation}

Abstract of research paper on Chemical sciences, author of scientific article — M.A. Olutoye, B.H. Hameed

Abstract In this work, a dual-site heterogeneous catalyst was used to study the kinetics of the transesterification of crude jatropha oil with methanol in batch experiments. Experimental data were obtained between 150 and 182°C at a constant molar ratio of alcohol to oil (11:1) and at a set catalyst concentration (3.32wt% based on weight of oil). Kinetic modelling was performed with an assumed pseudo-first order to describe the rate and catalyst deactivation, which suitably fit the experimental data. The model parameters were determined for both the mass transfer controlled and reaction regimes. The triglyceride (TG) conversion, methyl ester (ME) formation, and system thermodynamics were also evaluated. Based on the calculated values, within an acceptable range, the equilibrium constant, K e =1.42; activation energy, E a =161kJ/mol; and free energy, ΔG =−1286J/mol indicate that the developed model adequately described the methanolysis of the oil, which may be useful in reactor design and process simulation. The values that were calculated from the kinetic equations agree well with the experimental values, and the results help to understand and predict the behaviour of dual-site catalysts in practical applications for the production of biodiesel.

Academic research paper on topic "Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol"

Accepted Manuscript

Title: Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol

Author: M.A. Olutoye B.H. Hameed

PII: DOI:

Reference:

S1658-3655(15)00144-2

http://dx.doi.org/doi:10.1016/j.jtusci.2015.10.001 JTUSCI 231

To appear in:

Received date: Revised date: Accepted date:

13-7-2015

14-10-2015 28-10-2015

Please cite this article as: M.A. Olutoye, B.H. Hameed, Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol, Journal of Taibah University for Science (2015), http://dx.doi.org/10.1016/j.jtusci.2015.10.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Kinetics and deactivation of a dual sites heterogeneous oxide catalyst during transesterification of crude Jatropha oil with methanol.

M. A. Olutoye1*, B. H. Hameed2

'Department of Chemical Engineering, Federal University of Technology, P.M.B. 65, Minna. Nigeria

2School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, '4300 Nibong Tebal, Penang, Malaysia

*Corresponding author: m.olutoye@futminna.edu.ng

Abstract

In this work, dual sites heterogeneous catalyst was used to study the kinetics of transesterification of crude jatropha oil with methanol in batch experiments. Experimental data were obtained between 150 and 182 °C at constant molar ratio of alcohol to oil (11:1) and at concentration of catalyst (3.32 wt% based on weight of oil). Kinetic modeling was performed with assumed pseudo-first order to describe the rate and catalyst deactivation. This was found to suitably fit the experimental data. Model parameters were determined for both the mass transfer controlled and the reaction regimes. The triglyceride (TG) conversion, methyl ester (ME) formation, and system thermodynamics were also evaluated. Based on the calculated values, within acceptable range, equilibrium constant, Ke =1.42, activation energy, Ea = 161 kJ/mol and free energy, AG= -1286 J/mol indicate that the model developed adequately described the methanolysis of the oil which may find useful application in reactor design and process simulation. The calculated values from the kinetic equations are in good agreement with experimental values and the results help to understand and predict the behavior of dual sites catalyst in practical applications for production of biodiesel.

Keywords: Triglyceride; Kinetic modeling; Dual sites; Transesterification, Deactivation

1. Introduction

Over the last few years, synthesis of fatty acid methyl esters (FAME) has assumed importance as much research has been intensified on the utilization of vegetable oils and animal fats derivatives for liquid fuels (biodiesel) production [1-5]. FAME can be obtained from renewable feedstock, such as vegetable oils or animal fats by the process of transesterification with alcohols in the presence of an acid or an alkaline catalyst. It is a clean burning fuel, which is nontoxic, biodegradable, and considered as the fuel of the future [6]. Biodiesel, in its pure form, is referred to as neat fuel (B100), or in the blended form (B20 or B80) when mixed with petroleum diesel. It can be utilized in compression ignition engines under a variety of operating conditions. The B100 contains no petroleum fuels and emits virtually no sulfur, aromatics, particulates, or carcinogenic compounds and is thus a safer alternative to petroleum diesel. Biodiesel can be used in all conventional diesel engines and its performance and engine durability closely match that of petroleum diesel, and requires no modifications in fuel handling and delivery systems [7, 8]. Three consecutive and reversible reactions are believed to occur during the transesterification of vegetable oils and fats. They are the conversion to intermediate products formed during the reaction which are diglycerides (DGs), monoglycerides (MGs), and the final methyl ester product. The reaction scheme is as shown in Scheme 1. H2C-O—CO-R H2C-O—CO-R1

hc-0-c0-r2 + ch3oh ^^ ch3-o~co-r3 + ch-o-co-r2

H2^ O— CO-R3 h2c 0H

Triglyceride Methanol Methyl ester Diglyceride

H2C-OH

H2C-O—CO-R I

cL-CCMR, + CH3OH ^ ^ -o-cc«, +

I H2C— OH

Diglyceride Methanol Methyl ester Monoglyceride

H2C-OH

H2C~OH

CH3OH CH3 -O~CO-R2 + CH_ OH

H2C—OH

CH- O —CO-R2 + - k6

H2C—OH

Monoglyceride Methanol Methyl ester Glycerol

Scheme 1: 3-step reversible reactions of triglyceride where R1, R2, and R3 are long chains of carbons and hydrogen atoms (fatty acid chains)

The various parameters which influenced the synthesis of FAME have been investigated by many authors due to increasing industrial application of methyl esters [7]. These include the molar ratio of alcohol used to vegetable oil, catalyst type and loading (wt % of oil), reaction temperature, agitation intensity and presence of impurities such as free fatty acids and moisture content. There are a number of kinetic studies in the literature on transesterification of esters with alcohol catalyzed by homogeneous base catalysts; but very few information is available on the kinetics of transesterification reactions by solid-base-catalysts [9]. Most especially, in this regard, only few studies dealt with transesterification reactions of vegetable oils and FFAs with the application of a dual site composite catalyst. Hattori et al. [10] proposed a mechanism for the transesterification of ethyl acetate with different alcohols using variety of solid base catalysts, particularly alkaline-earth-metal oxides. However, this mechanistic study neither provided values of activation energies nor rate constants. Recently, a kinetic model was proposed to describe the

transesterification of ethyl acetate with methanol catalyzed by a heterogeneous magnesium oxide catalyst [11, 12]. This model was based on a three-step 'Eley-Rideal' type of mechanism applied in liquid phase where methanol adsorption is assumed to be rate determining. The value of activation energy determined was 20 kJ/mol. with methanol adsorption equilibrium coefficient of 3.13*10-3 m3/mol. The objective of this study is to evaluate the use of a dual site heterogeneously catalyzed transesterification reaction in batch reactors at predetermined conditions using the kinetic model based on the three-step 'Eley-Rideal' type mechanism assuming methanol adsorption as rate-determining step. The kinetics and deactivation studies of a dual site catalyst in the transesterification of CJO (non-edible stock) are of importance to enable reactor design parameters to be determined and advance the stage of efficient production of biodiesel.

In addition, kinetic model for the reaction of triglyceride over solid catalyst would be developed and the deactivation characteristic during the transesterification and esterification reactions with methanol would be observed. The influence of the various reaction parameters on the catalyst activity and kinetic rate will be evaluated. Since the kinetics of transesterification and esterification reactions are increased by using basic and acidic catalysts, respectively, it thus make the research interesting with the use of a dual site catalyst for reactions of both transesterification of vegetable oils and esterification of FFAs because of its wide industrial use and also fundamental to reactor design.

2. Experimental

2.1 Materials

Crude Jatropha (Jatropha curcas Linnaeus) oil supplied by Telegamadu Sdn. Bdh.,

Butterworth, Penang, Malaysia was used for the study. The acid value, moisture content and

saponification values of the oil were 14.47 %, 3.28 % and 193 mg KOH/g of oil respectively. Methanol (HPLC, analytical reagent grade 99.9%) purchased from Merck (Malaysia) was used. Al(NO3)3.9 H2O ( 99%), Mg(NO3)2.6H2O (> 99%), Zn(NO3)2.6H2O (> 98%) and analytical grade NH4OH (> 85%) used to synthesize the catalysts were purchased from Sigma-Aldrich Pty Ltd., Malaysia. Other chemicals used in the experiment such as acetonitrile (> 99 %), and acetone (> 99.9 %) for the Ultra-Fast Liquid Chromatography (UFLC) analysis were purchased from Merck (Malaysia). These reagents were used without further purification for catalyst synthesis and the kinetics of trans-esterification of crude jatropha oil.

2.2 Catalyst preparation

The dual sites catalyst (MgZnAlO) was prepared by co-precipitation using 25% ammonia water solution as precipitating medium. The mixture was stirred continuously for 6 h under constant heat at 80 oC until when a completely homogenized milky solution was obtained. Basic strength was determined by the indicator method which gives a pH value range of 8-9 and was maintained throughout the catalyst preparation. The mixture was filtered, dried, and calcined at 461 oC for 4 h 25min. The catalyst powder was stored in a well closed glass desiccator to keep it moisture free before it was used for the kinetic and trans-esterification experiments. The detailed procedure for the preparation and characterization has been previously reported [13]

2.2.1 Characterization

Characterization of the catalyst by N2 adsorption-desorption isotherm at 77 K were obtained using Micromeritics ASAP 2020 gas adsorption analyzer. The surface area was calculated using the Brunauer-Emmett-Teller (BET) equation and mean pore diameter was obtained by applying the Barret-Joyner-Halenda (BJH) method on the desorption branch. Thermogravimetry measurement (TG/DTG) was carried out with thermo balance SETARAM (accuracy ±0.04^g) thermal analyzer from 298 K to 1073 K at a heating rate of 293 K min-1 under N2 atmosphere. The structure of the catalyst was checked by X-ray diffractometer (Philips PW 1710) with Cu Ka (Ni-filtered) radiation. The particle microstructures were studied on a Philips XL30S model Scanning Electron Microscopy (SEM). The element composition was analyzed by using an energy dispersive X-ray detector (EDX) mounted on the microscope. Fourier transform infrared (FT-IR) spectra were recorded for the active surface functional groups. The spectra were recorded in the range 4000-400

2.3 Determination of the surface acidity and basicity of the catalysts

The surface basicity and acidity are important parameters of solid catalyst properties which influences its activity. The surface properties were evaluated by the method of titration after a displacement reaction, in which the adhered free OH- and H+ on solid catalyst were treated with acidic and basic species to protonate and displace from the surface through neutralization reaction. Briefly, a catalyst sample (1 g) each was contacted with 25 mL of 0.1 M solution of NaOH and HCl in two separate batch experiments. The mixture in 100 mL glass beakers were mixed thoroughly for 24 h. The contents at the end of reaction were allowed to settle, filtered and the filtrate titrated each with 0.1 M solution of NaOH and HCl using

phenolthalein indicator. The measurements and the evaluation of the results were determined from the acidic or basic indicator molecules on the solid catalyst surface in order to count acid or base sites, establish the dual sites component of either Lewis or Bronsted characteristics, and determine the site strength in relation to known properties of the substance. The surface acidity was determined from the expression in Eqs. 1 and 2

r T , .. „ ftimole\ nXaQH{iniiiBij-n.\-aQHifi>iBO

Acid concentration, C. -I =-;----[11

\ g J mass of catalyst

^ /fimole\ nnciUnitiaiy-nHCl {final")

Base concentration, LB -| =-:------(2)

\ g / mass of catalyst

2.4 Equipment and reaction conditions

The transesterification reaction and kinetic experiments were carried out in an automated high pressure, high-temperature batch reactor (PARR Instruments, 4843) equipment. It has a maximum capacity of 300 mL and is equipped with PID and a thermocouple which regulates the temperature inside the reactor, a heating chamber, water cooling and pressure devices. A long shaft rod with impeller blade mounted at the tip provides efficient stirring of the mixture.

The molar ratio of methanol to crude jatropha oil used in all the kinetic experiments was 11:1 which represents an optimum condition for the process as previously determined [13]. Other reaction conditions include a catalyst loading of 3.32 wt % of oil. The temperature and time of reaction were varied at 150, 160, 170, and 182 oC over a time range of 0- 360 min respectively. All experiments were run at the methanol vapour pressure that was built-up in the autoclave

reactor (no external pressure was supplied). The impeller speed of 300 rpm (max.) was applied throughout the reaction to produce uniform dispersion and to avoid mass transfer limitations.

2.5 Experimental procedure for the kinetic study

The reactor was charged with a mixture of 38 mL methanol, 82 mL CJO and 2.424 g of solid catalyst for each experimental batch. The experiment was timed immediately the stirrer was turned on and it continued for the batch process over different time intervals of 0, 5, 10, 15, 30, 60, 180, and 360 min. At the completion of each batch process, the samples were immediately quenched in a cold water bath to stop further reaction between participating species in the reactor and thereafter discharged. The product (a mixture of fatty acid methyl esters and glycerol) was centrifuged and then separated before being sent for HPLC analysis. In all the kinetic experiment, 1 h was established for the equipment to reach the target set-temperature. Thus, the experiments were initialized for of all the batch runs with TG, methanol and the catalyst charged into the reactor and run for 1 h till the target temperature was reached. At this point, the reaction was stopped and quenched with cooling water and the content analyzed for FAME. The TG conversions obtained at other experimental points were determined based on the initialization and the reaction was allowed to run at the target temperature over a specified period of time. The experiments were planned to enable the determination of reaction rate constants and to study the effect of molar concentration of alcohol to oil ratio, time and temperature on conversion.

2.6 Separation and analysis of fatty acid, FAME, MG, DG and TG in crude jatropha oil.

CJO was analyzed by the high-performance Ultra-Fast Liquid Chromatography (UFLC model LC20AC series, Schimadzu, Japan). The same procedures were adopted to analyze all the product samples. The LC system is equipped with a split/splitless injection system with a column dimension 250 x 4.6 mm, Hypersil Gold. The method adopted in the quantitation of products is based on isocratic (fix gradient) pump mode with mobile phase: A (Acetone); B (Acetonitrile) at 36.5 % and 63.5 % respectively. The pump was set to a flow rate of 1 mL/min and injector volume is 60 |iL. The total analysis time is 60 min with column C18 at oven temperature of 30 oC. The UFLC method used in the present study had been successfully validated in-house with repeated (n=10; df approximately 1: 200) analysis of CJO sample used in transesterification. For each sample determination, 20 |iL was mixed with 3820 |iL of acetone to give a dilution factor (df) of approximately 1: 200. The calibration curve (Figure not shown) was obtained for the CJO using different concentrations, based on the weight of oil, prepared in the range 0.01-1.5g/10 mL of acetone and the component peak area. Because of the difference in the partition coefficient between solid- and liquid- phase, four eluates containing mixture of fatty acid and fatty acid methyl ester, the partial glycerides (mono- and di-glycerides) in addition to triglycerides are separated in this order using the solvent system acetone/acetonitrile ratio. The TG concentration was then calculated based on the regression equation as expressed in Eq. 3 Concentration rCTG (^/¿j = Si 1Q~7A -f 2.1727 (3)

where A is the peak area of TG component as determined by UFLC and the coefficient of regression (R2) given as 0.9625. The fractions obtained were characterized as FA+FAME, MG, DG, and TG, but the sample were further characterized for FA in each to determine the actual ester content. Based on the change in the percentage composition of TG before and after the

reaction, the conversion values were obtained. The contents of other minor components in CJO were neglected in the calculation of conversion.

3. Results and discussion

3.1 Reaction kinetic model and analysis of kinetic data

Batch kinetic experiments were carried out to study the kinetics of transesterification at various reaction temperature and time while other parameters, initial reactant molar ratio and catalyst concentration were kept constant. The simplified reaction for the conversion of triglycerides to products is presented in Scheme 2.

H2C-O-CO-R H2C-OH

hc-o-co-r +3CH3OH^— 3ch3-o-co-r + ch_oh

o— CO—R H2C— OH

Triglyceride Methanol Methyl esters Glycerol

Scheme 2: Triglyceride conversion to methyl ester

The reactants and products in Scheme 2 are presented in coded form and written as: TG + MeOH = ME + GLY (4)

or written as

A + 3B 3P + D (5)

(where A=triglyceride, B=alcohol, P=FAME, D=glycerol)

Although more than one species is involved in the rate-determining step of this reaction as shown in Scheme 1, a pseudo-first order reaction is proposed. Initially, CP=0 and CD=0, and concentration dependent rate function, -rA= fn(Cj) is defined; where j represents the reacting species A and B. Thus, a general rate can be written for Eq. 5 as—rA = kAC£C^. But since B is present in excess, its concentration virtually remains constant during the course of the reaction. Thus, the rate of reaction practically depends on the concentration of A in the system (the limiting reactant). In the absence of both external and internal resistance to mass transfer, the intrinsic kinetics can be determine using the pseudo-first-order reaction, that is

km, am, CA,b =CA, and CA,s are the mass transfer coefficient, external surface area, concentration in the bulk liquid and concentration on the catalyst surface respectively. If the transport process is much smaller than the surface reaction process which implies that it is rate determining then CAfS = 0 Also, take ki =kmam, we have

and by integration gives

where XA is the fractional conversion of A. The plot of — ln( 1 — XA} against t was used to determine k], from which km was also obtained. To obtain a first approximation of the kinetic model, it was assumed that the trans-esterification reaction behaved like a first order irreversible

reaction. Therefore, if this model is valid, a plot of —

versus t will be a straight line,

and the value of the slope will be k1 according to Eq. 7. The reaction orders with respect to the triglyceride and the alcohol are found to obey zero and pseudo-first orders with well fitted data which indicate a shift from lower order to higher order as the reactant concentration drops in the system as illustrated in Fig.1A-D. As shown, the experimental data followed a linear trend with a correlation factor, R2, of 0.984 and with a slope value of 0.0075 (min-1) with respect to Fig. 1A. The irreversibility of this reaction is favoured with an excess of methanol that induces low concentrations of triglycerides in the equilibrium composition. This determination suggests that surface reaction is the rate limiting step (RLS). Excess of methanol (> 11:1 methanol: TG) was used to compensate for the slow reaction rate and it would be expected that under this condition, the reaction order with respect to methanol approaches zero. The above observation is supported by the works of Chantrasa et al. [14] and Casas et al. [15].

The chemical reaction rate, -rA for reversible process of the type in Eq. 5 is given by

For the system under consideration, the initial concentration of FAME and glycerol is taken to be zero. If the TG concentration (A) is related to the conversion degree of TG, XA, by the relation C. = C.. ■ _ - ; the initial concentration of methanol is denoted by CBo. Thus,

CB = CBo - 3CAoXA, CP = 3CAoXA and CD = CAoXA, it follows that:

C-r ') = = k C C3 - k C^C

f AS *-JJ

CAa = kfCAJl - xA){cBo - 3 CAXA) - 3K(CAXA)

At equilibrium, r4=0 therefore Eq. 9 becomes

And after re-arrangement becomes

where Xdg represent equilibrium fractional conversion. Table 1 shows the kinetics and thermodynamics parameters obtained for the process at the reaction conditions for all of the experiments performed in this work.

3.2 The thermodynamics of process

For each step of the reversible reaction, a dynamic equilibrium can be established in accordance with the law of mass action such that at equilibrium, the rate of forward reaction equals the rate of the reverse reaction. The process thermodynamics based on the experimental data was used to quantify heat effects in the reacting systems and calculate the equilibrium conversion; the maximum conversion that could be obtained during the reaction. Thus, re-writing Eq. 8 we have

where kf and kr are the reaction rate constants for the forward and reverse reactions, respectively and CB, CP and CD are the actual concentrations of methanol, FAME and glycerol, respectively. At equilibrium —rA= 0,

where Ke is the dimensionless equilibrium constant in terms of concentration and Eqs. 12 and 13 combine to give

fcjrCjCn i

' f j.4 B

= .: = c (14)

For the reversible reaction as stated in stoichiometric equation of the reaction being considered, the equilibrium constant, Ke at temperature T were evaluated from the change in the Gibb's free energy expressed as a function of forward and backward rate in terms of

■:: " :=■= -FT-./.kK 7 :=■ ; (15)

* = RT) (16)

= - -'z - - -'.-.. (17)

A Gf is the standard Gibb's free energy of formation of a given species Gf . The relationship between the change in Gibb's free energy and enthalpy, H, and entropy, S of this system is thus defined by

^ = ^ - T^S (18)

The summary of all the calculation for the kinetic and thermodynamic parameters is presented in Table 1. It was observed that the value of equilibrium constant for the TG methanolysis was Ke =1.42. Thus, species A and B are dominant in the reaction mixture and Ke >1 indicates FAME production is favoured.

3.3 Catalyst deactivation model and mechanism

Eley-Rideal and Langmuir-Hinshelwood kinetics were fitted to the batch reaction data for

the heterogeneous catalyst. The kinetics is assumed to involve the adsorption of one reactant, in

this case the alcohol, which is then attacked by the other reactant from the liquid phase. The

reaction which occurs through the Eley-Rideal mechanism is between a protonated fatty acid

and the methanol coming from the liquid phase. Scheme 3 shows the transesterification reaction of triglycerides over a solid base catalyst. The scheme shows an initiation step of an exchange between fatty acid and a protonated methanol which is followed by the Eley-Rideal surface reaction step that involves the protonated fatty acid and methanol to produce protonated methyl ester. Water is also formed in this step due to large moisture content present in the feedstock.

(/ I //

R— C'7 + -OR'' ^^ R—C

/ + -OR'' R—C— OR'' ^^^^ R— C^ + OR'

\0R' OR' ^ OR''

Scheme 3: Alcohol exchange scheme with FA moiety of TG for basic site catalytic reaction There are partial glycerides (MG and DG) complex that are formed which have both alcohol and ester functionality, so for the back reaction, it was assumed that the partial glycerides are adsorbed on catalyst surface and their alcohol functionality reacts with that of ester product. The surface coverage of adsorbed species is represented by 6Ads and the relationship between the adsorbed and bulk species is derived through use of the Langmuir isotherm. The Langmuir isotherm relates the adsorption and desorption of the species from the surface as given in Eq.19 which applies for all species being adsorbed. This is based on Langmuir-Hinshelwood kinetics which requires the adsorption of both reactants onto adjacent sites on the catalyst surface. The reaction takes place on the catalyst surface and both products then desorb.

n _ Ads A ( iq)

Ads 1 v r* V '

1+KAdCA

where KAds is the adsorption equilibrium constant and CA is the concentration of triglyceride. The value of the adsorption coefficient was calculated from experimental data for the

transesterification of CJO with methanol as presented in Table 1. This coefficient was assumed to be constant for all species.

Transesterification of high molecular TG with methanol in a batch reactor with the use of dual site solid catalyst is a highly complex chemical reaction and yielded various conversions to products. The results of these conversions are presented in Fig. 2 where it was observed generally that from the start of the reaction, the conversion to FAME increases as the DG, MG and TG values drop in the reaction system. There are various assumptions underlying models for reaction taking place in closed reactors, and the model development for the process will include both steady and unsteady state predictions. The kinetic equation was developed by investigating the general behaviour of the system in terms of one variable influencing another based on theoretical knowledge of the system with relevant physical information in the form of conservation laws and rate equations. Assumptions for the development of the kinetic equation based on the Langmuir approach are: (a) all the surface of the catalyst has the same activity for adsorption, that is, the surface is energetically uniform, (b) There is no interaction between adsorbed molecules, (c) All adsorption occur by the same mechanism and each adsorbed complex has the same structure, (d) The extent of adsorption is less than one complete monomolecular layer on the surface, (e) The TG molecule is in contact with the catalyst and undergo a three-stage process of adsorption (rAds= rA), surface reaction (rS), desorption (rD= -rAds=rP) and that near equilibrium, all the rates rA, rS, rD are equal.

If TG is adsorbed on catalyst surface for the conversion of triglycerides to products (neglecting the other reactions in the series) then, ki

TG(liquid) + X ^ TG.X(ads)

k 1 (20)

TG represents specie A and the mechanisms from which these equations were obtained are summarized in five consecutive steps 1-5. From Langmuir-Hinshelwood kinetics with surface reaction being the RLS, the observed reaction would be

1 A + X -»A*

rA = k1CA Cv — k1C A (21)

2. B + X^B.X

rA = kfBCv — k Cb (22)

The net rate of adsorption, rA gives

rA = k1CA (Cm — C) — k1CA

rA = k1 [CA Cv — -¡^C A ]

Kads (23)

where Cv is the concentration of vacant sites. If simultaneous reaction is considered and species A and B are adjacent to each other and adsorption process and surface reaction of the type given below occurred.

3. A.X+ B^P.X

rs = k2 cacb — k2 cp (24)

If reaction takes place between adsorbed A and adsorbed B, then

4 A.X+B.X^P.X+X (25)

From the reactions above, it is assumed that only the adsorbed A immediately adjacent to adsorbed B will react to yield the products.

If C m = molar concentration of total sites

and 9B = Fraction of total surface occupied by B (methanol)

9v = Fraction of total surface that is vacant Net rate expression for the reaction gives

i /—r C B 1 ' ^ C V

rS _ k3 Ca=--k3 C p =—

C m Cm

rS _ -¿^[CaCb —— CpCv] S c Tf

m (26)

If the resistance is negligible at the surface reaction with respect to others, the process will occur at equilibrium, rs=0

^ CpCv ^ v CaCB !

If the resistance is high such that it is rate controlling then, rate = rS in Eq. 26 above Desorption.

For the process being considered, let the desorption process take the form 5. P.X -+P + X

rp _ k4Cp - k4CPCv

rP _ -k4[CPCv —— Cp]

K4 (28)

Eq. 28 shows that it is the reverse of adsorption process. X represents the vacant sites, P denotes

the methyl ester, A.X, B.X and P.X are the adsorbed species. The concentrations of the adsorbed

species can be obtained considering that the reaction rate constant for the formation of the

intermediates is large compared to the reaction rate. Thus,

rA - JjL - 0,

kads kB (2Q)

which represents a pseudo-equilibrium, therefore,

CAe _ KAC

CBe _ KbCBCV

With K4=Kd, at equilibrium, rP=0 then the equation of desorption gives the expression for C i

Cp _ K4Cp Cv (32)

It must be noted that

Cm _ Cv + Ca + Cb + Cp (33)

Therefore, combining equations 26, 30, 31 and 32 and assuming that the surface reaction controls, then

Cv _j-—-, (34)

[1 + KaCA + KBCB + KdCp ]

Substituting for Cv, surface reaction rate will be given by:

rate _ rS

C K C ^

K K C c dp

v KS J

In this study, the Cv was based on the summation of surface acidity and basicity as determined by Eqs.1 and 2.

Although A and B are in the liquid phase, diffusion into the interior surface of the catalyst is required according to the postulated mechanism above through the process of adsorption, surface reaction and desorption in and out of the catalyst to the bulk of the liquid. The overall rate of reaction could be controlled by one or some of these processes, depending on their relative magnitude. In order to be able to evaluate which is the controlling mechanism, the kinetics and equilibrium constants in Table 1 were determined from the experimental data. It will

be observed that temperature in this case was not used to establish the controlling mechanism because of the dependence and variation of equilibrium constant with temperature.

3.4 External mass transfer effects on the kinetics

The rate determining step in this heterogeneous kinetic study involved species being transported from bulk solution to the catalyst particle, or occurred in form of transport within the particle to the active sites or in the reaction itself, as depicted by the diagram in Fig. 3.

If component A enters the pore of the catalyst in the system through constant agitation, and resulted in the production or disappearance by chemical reaction, the amount of A in moles in the pore is given by the product of cross sectional area of contact and the number of moles if the catalyst surface is assumed spherical. Thus, NA in moles, NA (4nr2) is the amount of material on a spherical surface radius r, If the disappearance of A takes place at the Ar surface by chemical reaction, then the amount of materials which disappeared at this surface is given by

— rA (4nr 2)Ar where rA is the rate. Also, if effective diffusivity for specie A in the porous medium is defined by

concentration, temperature and catalyst pore structure. Neglecting any volume change with

CA= concentration of A contained within the pores,

effective diffusivity, is a function of

reaction (liquid phase), the conversion of CJO, XA is given by, XA=1- CA/CAo where CA and CAo are the CJO molar concentration at a particular time and initially, respectively.

Since A was used as the limiting reactant, its concentration may not be uniform within the catalyst particle due to intra-particle diffusional resistance and the alcohol concentration is almost uniform. To account for this resistance, differential equations for intra-particle diffusion with reaction on the surface of the catalyst were solved. The different steps in which these were achieved are presented in Eqs. 20-35. The occurrence of active sites on catalyst surface is supported by the work of Trionfetti et al. [16] where the characterization and mechanism of regeneration was presented. Assuming a first-order reaction occurring in a spherical particle of radius rs, the concentration of A at any radial distance r from the center of the catalyst to its

concentration on the surface will be expressed by Eq. 37

= -; (37)

CAs r Si-nh

-"■-= Thiele-type modulus for spherical pellet, k1 is the pseudo-first-order rate constant, /?.p=particle density, g/cm3. If

- = (39)

rp=reaction rate for the pellet, rs= diffusion rate of reactant into the pellet and n= effectiveness factor. Then rp= r| rs or rp= r| k,

= ---L: (40)

from the above, the rp can be obtained from rp= r| k, C^ as

depending on the relative magnitudes of mass transfer coefficient km as obtained from Eqn. 6 and n k1, the controlling resistance was evaluated. These values and other important determinations are presented in Table 1

The influence of intra-particle mass transfer resistances on dual sites catalyzed transesterification of CJO was evaluated by first calculating the value of Thiele modulus for the catalyst (assumed spherical pellet) =0.373 which explain the significance of diffusion and chemical reaction. Clearly <p3< 1 and effectiveness factor r| —> 1 with a calculated value of 0.93 indicating that intra-particle mass transport has no effect on the rate hence chemical reaction controls. Internal diffusion is negligible. Although limited mass transfer could be assumed for the reaction at the temperature range investigated, the limitations were accounted for in the effectiveness factor included in the developed kinetic model. If the initial rate is taken as maximum rate for each experiment, the effective diffusivity of TG in methanol liquid phase is 6.83*10-10 m2/s obtained from the Wilke-Chang equation. Also, transesterification reactions are bimolecular and reversible, the calculated parameters in which a pseudo first order, irreversible, and isothermal reactions assumed was justifiable since methanol was always in considerable excess (>11:1 molar ratio with CJO) which tends to benefit the forward reaction and the equilibrium conversion in all experiments was > 67 %.

3.5 The mechanism of dual site catalyst reaction 3.5.1 Acidic site catalyzed process

The long FA moieties in TG can be esterified by alcohols in the presence of a suitable acidic catalyst as illustrated in Scheme 4. Fatty acids can occur in nature in the free (unesterified) state and they are often found as esters linked to glycerol. The dual site nature of the catalyst employed for this study is proved from the plot of rate versus time as shown in Fig. 1 where the profile gave two rate constants with different R2. The higher R2 being the rate determining step. Briefly, it is believe that the initial step is the protonation of the TG (the R1,R2 or R3-chain) to give an oxonium ion which thereafter undergo an exchange reaction with an alcohol to give the intermediate dihydric complex and this in turn can lose a proton to produce an ester. It is proposed that the acid site of this dual catalyst will exchange hydrogen ion with other components involved in the reaction which are adsorbed on the catalyst surface. This will enhance absorption equilibrium and subsequently give rise to successive protonic exchange equilibrium reaction. Since methanol is in excess, it is preferably adsorbed to the TG and the most part of the active sites of the catalyst is occupied by the acid protonated methanol.

R—C- OH

+ORi H

R—C" oh

R—C^+OH

R— C — ORi

R—C—ORi

Scheme 4: Alcohol exchange scheme with FA moiety of TG for acid site catalytic reaction

The process is characterized by reversible steps and the equilibrium point of the reaction is displaced to the right in the presence excess alcohol. However, in the presence of water (a stronger electron donor than the aliphatic alcohols), the forward reaction in esterification of the FA is not favoured and may hinder formation of intermediate. Ester exchange or transesterification occurs under similar conditions. In the reaction under consideration, initial protonation of the ester is followed by addition of the exchanging alcohol to give a partially distorted intermediate complex that dissociates easily via transition states to give the ester product. Side reaction, dissociation of the intermediate by hydrolysis, due to the presence of water could take place and was excluded in the reaction system. Water content of the TG used

for this study is quite high; hence hydrolysis cannot be completely eliminated as reported by Suwannakarn et al. [17]. The kinetic study gave good conversion to FAME of > 97 % with the water content. Fig. 4 shows the plot of FAME yield obtained and other converted products. They have shown that addition of water in the system may be useful to increase the polarity. The overall reaction scheme showed how the acidic end of the dual site catalyst could have influenced both the esterification and transesterification steps to produce methyl esters from crude jatropha oil. The 3-stepwise reversible processes occurred over the dual site catalyst with excess methanol via a three-phase medium. Also, as reported by Berrios et al. [18], evaporation of water through the oil when TG are broken down will lead to the formation of partial glycerides and FFAs in the reaction medium.

3.5.2 Basic site catalyzed process

As shown in Scheme 3 the reaction mechanism is proposed for the base-catalyzed transesterification of oils on the basic site. The alcohol used for the reaction produced an anionic -OR" intermediate which can dissociate back to the original ester or form a new ester product if excess of the alcohol is used. The excess of -OR" anion will shift the equilibrium point of the reaction in the forward direction. In a similar reaction, an unesterified fatty acid will be converted to carboxylate ion, RCOO-, in basic medium, and this is not subject to nucleophilic attack by alcohols because of its negative charge. Transesterification will then proceed by the proposed mechanism with the basic catalyst (basic site) but not with esterification. As in the case of acid-catalyzed reactions, the presence of water will cause dissociation of the intermediate component irreversibly to the free acid. Thus, a base-catalyzed transesterification of esters

require an excess of alcohol in the absence of water from the reaction medium to obtain high yield of products. In this study, FAME, DG, MG and TG yields as analyzed using the HPLC are presented in Table 2. The results showed high conversion for FAME, however, a nearly complete conversion of the TG as the temperature increases was observed an indicative of higher rate order and that the chemical reaction predominates.

3.6 Effect of temperature

The evaluation of the effect of temperature is very important in kinetics studies, since it is useful to calculate the activation energy of the reaction. In this regard, temperature effect on the rate of reaction was studied by conducting the reaction at 150, 160,170, and 182 oC having previously determined optimum conditions for the CJO methanolysis. Typical results are shown in Fig. 4. It can be seen that the ester conversion increases with increasing reaction temperature. However, the equilibrium conversion was nearly equal in the range of temperatures considered in this work. Similar observation was reported by Delgado et al. [19] with a conversion value of above 70 %. Generally, in most transesterification reactions, the equilibrium constant is a weak function of the temperature because of the small value of the heat of reaction. This fact is corroborated by the work of Sanz et al. [20]. The Arrhenius plots were constructed (figure not shown) and was used to establish the activation energy and the pre-exponential factor as presented in Table 1. The high values of activation energy support the fact that there was no external and internal mass transfer resistance and reaction was kinetically controlled for the transesterification. Two distinctive slopes were calculated from the Arrhenius plot corresponding to activation energies of Ea, of the first order and zero order kinetics. The change in activation energy in relation to the order as the temperature varies suggest a change in the reaction

mechanism or the rate limiting step. This could be due to the dual site having involved in the reaction of both esterification and transesterification. Based on the activation energy results, these reactions did not appear pore diffusion-limited at the reaction conditions employed. Stamenkovic and co-workers [21] observed similar trend to the results presented in this study. In addition, they found that the rate constants increased when the reaction temperature increased.

3.7 Catalyst reusability and deactivation studies

Heterogeneous catalytic reactions are usually affected by deactivation although they have the potential to be recovered, regenerated and reused. In the present work, the catalyst used was employed under similar conditions of reaction temperature, concentration, and catalyst loading to generate the plot of conversion against time. Deactivation studies were carried out for the catalyst employed in this study. Figure 5 shows the results of triglyceride conversion for five successive 6 h reaction cycles. The catalyst activity reactions were carried out in the reactor as described earlier in this article except that the conditions were fixed at optimum values. Batch runs were prepared for the catalyst and the reaction quenched in order to be consistent with the kinetic study experiments. The reaction mixture samples were centrifuged prior to analysis, to separate the catalyst from the reactant mixture and to stop the reaction. For the catalyst deactivation studies, reaction was started as previously described [13]. The observed slight drop in the conversion for the reusability runs could be due to deposition of carbonaceous residues from reactants, products or intermediates (coking), chemisorption of impurities of the feed stream (poisoning) and losses encountered during the handling process. The regeneration and

reusability test on the catalyst for five cycles showed that there was no appreciable decline in the conversion and confirms the catalyst activity with minimal surface deactivation.

4. Conclusions

The kinetics of transesterification of CJO with methanol using a dual site (Alumina-Mg-Zn) catalyst has been studied. It was shown that reaction follows both zero and first order rate models in the formation of ester. Further observations from both the concentration-time and ratetime plots revealed that there was a shift in the order of the reaction from low (zero) to high (first) order and the reaction was found to occur between an adsorbed methanol molecule and a TG molecule in the bulk liquid phase. As a result, it was concluded that the dual sites catalyst offers a viable and suitable alternative to the existing heterogeneous catalysts for the transesterification of CJO with methanol at the stated conditions with equilibrium conversion, Ke above 67 %, equilibrium constant of TG adsorption, KA of 0.121 (cm3/mol) and diffusivity of the

TG-MeOH system, e of 6.83 x10-10 (m2/s). The catalyst is simple in preparation and separates

easily from product mixture and thus recommended for the industrial production of biodiesel.

Acknowledgement

The authors acknowledge the research grant provided by the Ministry of Higher Education, Malaysia under the Fundamental Research Grant Scheme (FRGS) with the Account No: 203/PJKIMIA/6071206 that resulted in this article.

References

[1] B. O. Aderemi, B. H. Hameed, Alum as a heterogeneous catalyst for the transesterification of

palm oil, Appl. Catal. A: General. 370 (2009) 54-58.

[2] W. M. Antunes, C. O. Veloso, C. A. Henriques, Transesterification of soybean oil with

methanol catalyzed by basic solids, Catal.Today. 133-135, (2008) 548-554.

[3] A. A. Apostolakou, I. K. Kookos, C. Marazioti, K.C. Angelopoulo, Techno-economic

analysis of a biodiesel production process from vegetable oils, Fuel Process. Technol. 90 (2009) 1023-103.

[4] R. Sarin, M. Sharma, A. A., Terminalia belerica Roxb. seed oil: A potential biodiesel

resource, Bioresour. Technol.. 101 (2010) 1380-1384.

[5] S. Pinzi, L. M. Gandía, G. Arzamendi, J. J. Ruiz, M.P. Dorado, Influence of vegetable oils

fatty acid composition on reaction temperature and glycerides conversion to biodiesel during transesterification, Bioresour. Technol.. 102 (2011) 1044-1050.

[6] X. Yu, Z. Wen, H. Li, S. Tu, J. Yan, Transesterification of Pistacia chinensis oil for biodiesel

catalyzed by CaO-CeO2 mixed oxides, Fuel. 90 (2011) 1868-1874.

[7] M. A. Olutoye, B. H. Hameed, KyMg1_xZn1+xO3 as a heterogeneous catalyst in the

transesterification of palm oil to fatty acid methyl esters Appl. Catal. A: General. 371 (2009) 191-198.

[8] Z. Wan, B. H. Hameed, Transesterification of palm oil to methyl ester on activated carbon

supported calcium oxide catalyst, Bioresour. Technol. 102 (2011) 2659-2664.

[9] T. F. Dossin, M. Reyniers, R. J. Berger, G. B. Marin, Simulation of heterogeneously MgO-

catalyzed transesterification for fine-chemical and biodiesel industrial production, Appl. Catal. B: Environmental. 67 (2006) 136-148.

[10] H. Hattori, M. Shima, H. Kabashima, Alcoholysis of ester and epoxide catalyzed by solid

bases Stud. Surf. Sci. Catal. 130 (2000) 3507-3512.

[11] T. F. Dossin, M. Reyniers, G. B. Marin, Kinetics of heterogeneously MgO-catalyzed

transesterification, Appl. Catal. B: Environmental. 62 (2006) 35-45.

[12] P. Patil, V. G. Gude, S. Pinappu, S. Deng, Transesterification kinetics of Camelina sativa

oil on metal oxide catalysts under conventional and microwave heating conditions, Chem. Eng. J. 168 (2011) 12Q6-1300.

[13] M. A. Olutoye, B.H. Hameed, Synthesis of fatty acid methyl ester from crude jatropha

(Jatropha curcas Linnaeus) oil using aluminium oxide modified Mg-Zn heterogeneous catalyst, Bioresour. Technol. 102 (2011) 63Q2-63Q8.

[14] A. Chantrasa, N. Phlernjai, J. G. Goodwin, Kinetics of hydrotalcite catalyzed

transesterification of tricaprylin and methanol for biodiesel synthesis, Chem. Eng. J. 168 (2011) 333-340.

[15] A. Casas, et al., Kinetics of chemical interesterification of sunflower oil with methyl acetate

for biodiesel and triacetin production, Chem. Eng. J. (2011), doi:10.1016/j.cej.2011.05.037.

[16] C. Trionfetti, S. Crapanzan, I. V. Babich, K. Seshan, L. Lefferts, Lithium ions incorporation

in MgO for oxidative dehydrogenation/cracking of propane: Active site characterization and mechanism of regeneration, Catal. Today. 145 (2009) 19-26.

[17] K. Suwannakarn, E. Lotero, K. Ngaosuwan, J. G. Goodwin, Simultaneous Free Fatty

Acid Esterification and Triglyceride Transesterification Using a Solid Acid Catalyst with in Situ Removal of Water and Unreacted Methanol Ind. Eng. Chem. Res. 48 (2009) 2810-2818.

[18] M. Berrios, M. A. Martin, A.F. Chica, A. Martin, Study of esterification and

transesterification in biodiesel production from used frying oils in a closed system, Chem. Eng. J. 160 (2010) 473-479.

[19] P. Delgado, M.T. Sanz, S. Beltran, Kinetic study for esterification of lactic acid with ethanol

and hydrolysis of ethyl lactate using ion-exchange resin catalyst, Bioresour. Technol. 126 (2007) 111-118.

[20] M. T. Sanz, S. M. Beltran, J. L. Cabezas, J. Coca, Autocatalyzed and ion-exchange-resin-

catalyzed esterification kinetics of lactic acid with methanol, Ind. Eng. Chem. Res. 41 (2002) 512-517.

[21] O. S. Stamenkovic, Z. B. Todorovic, M. L. Lazic, V. B. Veljkovic, D. U. Skala, Kinetics of

sunflower oil methanolysis at low temperatures, Bioresour. Technol. 99 (2008) 11311140.

Figure Captions

Fig.1: Ca vs time plot and the LnCA/CAo vs time plot, A= 150 oC; B=160 oC; C=170 oC; and D=182 oC.

Fig. 2: FAME and partial glycerides yield during transesterification of CJO, A= % FAME

content at T=170 oC; B= % MG content at T=170 oC; C= % DG content at T=170 oC and D= % TG converted at T=170 oC.

Fig. 3: Schematic process of solid-liquid transport showing catalyst pores (enlarged)

Fig. 4: Plot of % FAME yield versus time

Fig. 5: Conversion of TG as a function of time using the dual site catalyst at various temperatures, a=150 oC; b=160oC; c=170 oC; and d=182 oC, e and f=182 oC (regeneration and reuse runs)

Table Captions

Table 1: Values of experimental and kinetic model parameters

Table 2: Products and yields of transesterification reactions using the dual site catalyst

Table 1

Parameters Notation and units Calculated value

Pre-exponential factor A (L mol-1 min-1) 1.33 x 105, 9.9 x 103

Activation energy Ea (103 Jmol-1) 26.103, 161.42

Equilibrium constant of TG adsorption Ka (cm3mol-1) 0.121

Equilibrium constant of the overall reaction Ke 1.42

Diffusivity TG-MeOH system ie (m2s-1) 6.83 x10-10

Thiele modulus <Ps 0.373

Effectiveness factor n 0.93

Gibb's free energy AG, (Jmol-1) -1285.61

Adsorption coefficient ^0Ads 0.211

Table 2

Temperature, Time, min Monoglyceride, Diglyceride, Triglyceride, FAME, %

oC % % %

150 5.0 15.09 1.53 13.25 67.50

15.0 13.20 1.36 9.77 73.17

30.0 12.18 1.04 7.56 76.74

60.0 8.22 0.68 2.55 86.23

180.0 1.86 0.04 0.27 95.62

360.0 1.53 0.06 0.14 96.10

160 5.0 11.20 0.35 2.68 83.36

15.0 8.76 0.097 2.15 86.81

30.0 6.99 0.26 1.15 89.59

60.0 3.07 0.13 0.51 94.47

180.0 1.49 0.07 0.11 96.65

360.0 1.74 0.23 0.12 96.56

170 5.0 5.58 0.30 0.98 87.87

15.0 4.62 0.18 0.59 91.85

30.0 2.61 0.06 0.33 93.49

60.0 2.03 0.05 0.18 96.10

180.0 1.60 0.03 0.03 97.01

360.0 1.20 0.24 0.12 97.73

Table 2 continue

Temperature, Time, min Monoglyceride, Diglyceride, Triglyceride, FAME, %

oC % % %

182 5.0 3.82 0.16 0.66 94.46

15.0 2.90 0.15 0.38 95.73

30.0 1.53 0.18 0.25 97.37

60.0 1.22 0.20 0.17 97.85

180.0 1.61 0.31 0.73 96.91

360.0 1.91 0.60 0.02 97.14

Figures

200 300

Time (mill)

200 300

Time (miu)

200 300

Time (mill)

100 200 300

Time (miu)

y = 3E-05X -t- 0.4-9 7 £

♦ First orde r ■ Zeroorder

200 300

Time (miu)

oi -0.09 -o.oa -

0 07 -0.06 -0.05 -0.04- -003 -002 -

001 -0

y = -2E-06x+0.0951 R2 - 1

y= 0.0015x RJ=0.B31

♦ First ordi r ■ Zeroordiir

200 300 400 Time (miu)

♦ First order HI Zero order

♦ First order ■ Zeroorder

100 200 300 400 Time (miu)

Fig.1 Ca vs time plot and the LnCA/CAo vs time plot, A= 150 oC; B=160 oC; C=170 oC; and D=182 oC.

100 90 SO | 7C 1 SO M 5C

3 % 40

5 3c 20 1C

§ 0 25 v

.o 0.15 C.l C.C5 O

75 90 12C Time (tilin)

240 42C

Time (min)

12C 240

SSS ■:■:■:

45 40 35 3C 25 2C 15 1C 5 C

90 120

Time (min)

75 90 12C

Time (min)

240 42C

Fig. 2 FAME and partial glycerides yield during transesterification of CJO, A= % FAME content at T=170 oC; B= % MG content at T=170 oC; C= % DG content at T=170 oC and D= % TG converted at T=170 oC.

0 50 100 150 200 250 300 350 400 450

Time (min)

Fig. 4 Plot of FAME yield, % versus time (min)

O 60 65 75 90 120 240 420

Time imiu)

Fig. 5 Conversion of TG as a function of time using the dual site catalyst at various temperatures, a=150 oC; b=160oC; c=170 oC; and d=182 oC, e and f=182 oC