Heterogeneous Catalysis in Esterification Reactions: Preparation of


Heterogeneous Catalysis in Esterification Reactions: Preparation of...

0 downloads 267 Views 1MB Size

Ind. Eng. Chem. Res. 1994,33, 2198-2208

2198

Heterogeneous Catalysis in Esterification Reactions: Preparation of Phenethyl Acetate and Cyclohexyl Acetate by Using a Variety of Solid Acidic Catalysts Ganapati D. Yadav* and Pranav H. Mehta Department of Chemical Technology, University of Bombay, Matunga, Bombay 400 019, India

Esterification is an important class of reactions in the preparation of perfumery and flavor chemicals, wherein homogeneous acid catalysts are normally used. The application of solid acidic and superacidic catalysts can prove t o be very effective from the viewpoint of activity, selectivity, reusability, and economy in the manufacture of perfumery esters, and thus, this paper delineates such studies in the preparation of phenethyl acetate and cyclohexyl acetate with a variety of solid acids including a complete theoretical and experimental analysis. The list of catalysts employed for this study is Filtrol-24, Amberlyst-15, sulfated zirconia, heteropolyacids (supported on silica and carbon and also unsupported), and concentrated sulfuric acid.

Introduction Organic esters are a very important class of chemicals having applications in a variety of areas such as perfumery, flavors, pharmaceuticals, plasticizers, solvents, and intermediates. Obviously different approaches have been employed on both laboratory and commercial scales to prepare esters, and the traditional homogeneous catalyzed reactions are being less favored owing to the attendant problems of separation and reuse. This paper is specifically written to highlight the importance of heterogeneous catalysis in the manufacture of fragrance and flavor chemicals, with theoretical and experimental analysis of two esterification reactions. The paper takes a brief overview of the present state of the art and then embraces our own research findings. There are several routes by which esters can be synthesized as briefly mentioned below (Ogliaruso and Wolfe, 1991). 1. Esters by Solvolytic Reactions. Direct conversion of carboxylic acids and acid derivatives into esters by reactions with hydroxylic compounds are the so-called solvolytic reactions,which include the following: (a) direct esterification of acids; (b) alkylation of carboxylic salts; (c) alcoholyses of acyl halides; (d) alcoholyses of anhydrides; (e) alcoholysesof nitriles and amides; (0alcoholyses of ketenes; (g) transesterification. 2. Esters by Condensation Reactions. Reactions involving carbanion intermediates can be used to prepare carboxylic acid esters, and several of above condensation reactions can be adapted to the preparation of esters. Some well-known reactions of this category are the following: (a) Knovengel reaction; (b) Darzens reactions; (c) Wittigtype reactions; (d) Reformatsky reaction; (e) acetoacetic ester synthesis; (0 malonic ester synthesis; (g) derivatives of a-anions of esters; (h) Michael reactions and related conjugate additions; (i) Claisen condensations. 3. Esters by Free-Radical Processes. (a) radical additions and substitution reactions; (b) acyloxylation reactions; (c) anodic dimerization. 4. Miscellaneous Ester Synthesis (a) from organoboranes, (b) from acetylenes, (c) from diazo esters, (d) by carbonylation of alcohol, and (e) by phase transfer catalyzed reaction of alkyl or aryl halide with inorganic salt of the carboxylic acid (both liquid-liquid and solidliquid).

* Author

to whom all correspondence should be addressed.

For the preparation of perfumery and flavor grade esters, only a few of the above-mentioned routes can be considered due to stringent specifications of the final product. The most widely employed and supposedlycleaner production technique for such esters involves the reaction of the appropriate carboxylicacid with an alcohol in the presence of a mineral acid catalyst or a heterogeneous catalyst at reflux. H+

R'COOH

+ R20H F' R'COOR2 + H20

Such reactions are equilibrium processes and must be displaced toward the desired ester by the use of an excess of one of the free reactants or by continuous removal of water by azeotropic distillation. The typical catalysts used for direct esterification as mentioned in the literature include sulfuric acid alone, sulfuric acid in conjunction with molecular sieves, hydrogen chloride, arylsulfonic acids, acidic ion exchange resins coupled with calcium sulfate, polymer-boundaluminum chloride (both as Lewis acid and a dehydrating agent), intercalated graphite bisulfite, boron trifluoride, and trifluoroacetic anhydride. Esters can also be made under neutral conditions a t room temperature by the reaction of carboxylic acids with alcohols in the presence of molar amounts of triphenylphosphine and diethyl azodicarboxylate. Phenethyl Acetate and Cyclohexyl Acetate. Phenethyl acetate and cyclohexyl acetate are two of the basic esters having widespread applications in the perfumery and flavor industries. Both these esters attracted our attention for several reasons including the possible commercial use of different solid acidic catalysts. We have been investigating in our laboratory several solid acidic and superacidic catalysts, with regards to their structural characterization and applications, and this reaction could be embraced. Phenethyl acetate is widely employed in perfume compositions, from everyday soap and detergent perfumes to fine-cosmeticfragrances, room sprays, deodorants, etc. Its sweetness, versatility, and very low cost renders it attractive and almost universally applicable in several fragrance and flavor formulations. For rose, jasmine, hyacinth, reseda, freesia, peony, magnolia, oriental, and even citrusy fragrance types, phenethyl acetate concentration from 1%to 10% or perhaps much higher is recommended. Phenethyl acetate is also commonly used in flavor compositions, for imitation butter, apple, apricot,

0888-5885/94/2633-219~~0~.50/ 0 0 1994 American Chemical Society

Ind. Eng. Chem. Res., Vol. 33, No. 9,1994 2199 Table 1. Various Catalysts Used in Esterification of Phenethyl Alcohol with Acetic Acid Leading to Phenethyl Acetate ~

catalyst toluenesulfonyl chloride graphite bisulfate

2-chloro-1-methylpyridiniumsalts dialkyl acylphosphonates 1-acyl-3-benzylimidazoliumhalide diethyl azodicarboxylate

3-methylbenzothiazole-2-selone hydrated sulfate salts

references Brewster and Ciotti (1956); Parish et al. (1965) Bertin et al. (1974); Kogan et al. (1977) Saigo et al. Sekino et al. (1981) Kamijo et al. (1983) Mitaunobo et 01. (1984) Mitaunobo et al. (1984) Huang et al. (1988)

Table 2. Various Catalysts Used in Esterification of Cyclohexanol with Acetic Acid Leading to Cyclohexyl Acetate catalyst references sulfuric acid McCracken and Dickeon (1967) ferric, stannic, and zinc chloride Selwitz (1972) Besenhard (1977) graphite hydrogen sulfate Olah et al. (1978) Nafion H 2-chloro-1,3,5-trinitrobenzene Takimoto et al. (1981) synthetic zeolites Ou and Chen (1985) 4,6-dimethyl-2-pyrimidinyl Sunggak and Sung (1986) dithiocarbonate aluminum phosphate supported Fernandez et a!. (1988) on polymer beads ~

~~

caramel, honey, passion-fruit, peach, strawberry, vanilla, and beer. The concentration in finished product is normally as low as 1-6 ppm (Arctander, 1988). Cyclohexyl acetate has a sweet fruity smell and is used in masking odors for industrial purposes. It also has applications in flavor compositions, mainly for imitation apple, blackberry, banana, and raspberry and in tuttifrutti flavors (Arctander, 1988). The esterification of phenethyl alcohol, as well as of cyclohexanol, with acetic acid has been studied by different researchers by using various homogeneous and heterogeneous catalysts as given in Tables 1and 2, respectively. The literature survey indicated that there exists scope for conducting detailed kinetic investigations by employing various heterogeneous catalysts other than the ones reported earlier. In addition, heterogeneous acidic and superacidic catalysts have proved to be very commercially viable from the viewpoint of activity, selectivity, reusability, noncorrosivity, and virtual absence of effluent treatment which is associated with homogeneous catalysts. In this paper, we would like to report our studies with such solid catalysts as Filtrol-24, an acidic clay, sulfated zirconia, a superacid, heteropolyacids (HPA, both supported and unsupported), and Amberlyst-15, an ion exchange resin in the preparation of bothphenethylacetate and cyclohexyl acetate. For the sake of comparison, aqueous sulfuric acid was also used as a homogeneous catalyst. Furthermore, the mathematical modeling and determination of kinetic parameters and activity of these catalysts are fully discussed.

Experimental Section Chemicals. Phenethyl alcohol, cyclohexanoland acetic acid were obtained from films of repute and were distilled or used directly if they were of analytical laboratory reagent grade. Catalysts. Filtrol-24 and Amberlyst-15 were obtained from Engelhard Corporation and Rohm and Haas, USA, respectively. The dodecatungstophosphoric acid (H3POq12WO~nHz0,DTPA, molecular weight 2880.17; BDH

Analab) supported on the silica was prepared by the wet incipient method described by Thorat et al. (1992) from our laboratory. Silica manufactured by Degussa was calcined at 600 "C for 6 h and then used for impregnation. A calculated amount of DTPA was dissolved in deionized water and then impregnated on the support followed by drying either by rotary evaporator or in an oven for 24 h at about 120 "C. The catalyst was calcined at 650 "C. There was no leaching of DTPA from the support as our earlier studies have shown (Thorat et al., 1992). The superacidic zirconia was prepared by the method described by Kumbhar and Yadav (1989) and Hino and Arata (1980). A calculated amount of ZrOClz-8HzOwas dissolved in deionized water followed by precipitation to zirconium hydroxide by using aqueous ammonia. The precipitated zirconium hydroxide was washed with deionized water until a neutral filtrate showing no chloride was detected (phenolphthalein and AgN03 tests) and subsequently dried at 110 "C for 24 h. I t was then crushed to obtain the desired average particle size (160 pm). The impregnation was performed by immersing the dried zirconium hydroxide in 1 N HzS04 in the ratio of 1 g of zirconium hydroxide to 15cm3of HzS04 followedby drying for 1.5 h and calcination at 650 "C in a Pyrex tube for 3 h. Table 3 presents some of the physical properties of the catalysts used in this study. The characterization of these catalyst has already been reported in a separate paper from this laboratory (Thorat et al., 1992).

Experimental Procedure

A slurry reactor was used for the experiments, which consisted of a 5-cm4.d. fully baffled mechanically agitated reactor of 250-cm3capacity, made of glass and equipped with a reflux condenser. A 1.5-cm-diameter six-bladed glass disk turbine impeller was used for agitation. The reactor was kept in a constant temperature bath with a temperature control of f0.5 "C. A thermowell with a thermometer was used to monitor the reaction temperature. In all experiments, acetic acid, always in stoichiometric excess over the reactant alcohol, and the required amount of catalyst were introduced into the reactor and brought to the reaction temperature while the reactor contents were stirred. The alcohol (phenethyl alcohol or cyclohexanol) was then added into the reactor at the same temperature as the reaction mixture and the moment of addition of alcohol was taken as the starting time of the reaction. Samples were withdrawn from the reactor at regular intervals, quenched, h d analyzed by gas chromatography (Perkin Elmer Model 8320). Method of Analysis. An stainless steel column 1/8in. X 2 m packed with 5% OV-17 on Chromosorb W/HP was used for quantitative analysis of the reaction mixture. Synthetic mixtures were used for quantification, and the formation of phenethyl acetate and cyclohexyl acetate was also confirmed by FTIR and NMR. Results and Discussion Reaction Model and Analysis of Kinetic Data. A general model developed by Kumbhar and Yadav (1989) for solid catalyzed liquid reactions was used to study the intrinsic kinetics of the reaction. A general equation is acid

A+zB

i-

catalyst

E+W

(1)

(where A = alcohol, B = acetic acid, E = acetate, W =

2200 Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 Table 3. Physical Properties of Catalysts physical property

Filtrol-24

Amberlyst-15

shape size, fim apparent bulk density, g/cm3 surface area, m2/g porosity, vol % temperature stability, OC hydrogen ion capacity, mequiv/g

granular 660 0.75 358 37 >250 0.3

beads 500 0.98 55 36 120 4.5

catalyst sulfated zirconia

DTPA/Sa

DTPA/Ca

91.4

193.77

921.87

650

650

650

160

0 DTPA/S = DTPA/silica; DTPA/C = DTPA/carbon. Wt percent of DTPA = 30% w/w. The pore structural properties of the support are included in the table with an assumption that the pore structure remains unaltered.

[E01

CATALYST

SURFACE

[A01

[as1 [AS1

BULK

LIQUID

1 LIQUID I FILM

S U R R OUN D I N G CATALYST SURFACE

E and W from catalyst surface to the exterior of the catalyst; (7) transport of desorbed E and W to the bulk of the liquid phase. The overall rate of reaction could be controlled by one or some of these resistances, depending on their relative magnitudes. Since the preliminary experiments had shown that the reactions were quantitative, the resistances associated with steps 6 and 7, i.e., desorption of the products from the active sites and their transport to the bulk liquid phase, were negligible. Some pertinent equations are given below (for details of derivation, etc., see Kumbhar and Yadav (1989)).

At steady state,the rate of mass transfer per unit volume of the liquid phase is given by

( A )

RA

= ksL.Aap{[&] -

(2)

= rate of diffusion of A from bulk liquid phase to catalyst surface (g-mol/(cm3.s))

RA = Zks~.~a,([Bo]- [B,])

(3)

= rate of diffusion of B from bulk liquid phase to catalyst surface

= rate of surface reaction per unit volume

(D)

(E)

Figure 1. Concentration profilesfor the reaction between two liquidphase reactants reacting on the catalyst surfaces: (A) generalized case; (B-E) various cases. [Ad = concentration of A in bulk liquid phase, gmol/cm3;[A,] = concentration of A at solid (catalyst) surface, gmol/cm3;[Bo1 = concentration of B in bulk liquid phase, gmol/crns; [B,] = concentration of B at solid-liquid interface, gmol/cm3.

water, and z = stoichiometric coefficient of reaction which is 1 for the present cases). Figure 1showsthe concentration profiles for the reaction between two liquids reacting on catalyst surface. Although A and B are liquid-phase reactants, they need to diffuse to the interior surface of the catalyst. The different steps are the following: (1)diffusion of A from bulk liquid phase to the exterior of catalyst surface (reflected in KSLA,solidliquid mass transfer coefficient for A); (2) diffusion of B from bulk liquid phase to the exterior of the catalyst surface (reflected in kSL-B, solid-liquid mass transfer coefficient for B); (3) intraparticle diffusion of A and B within the catalyst pores (reflected in effectiveness factor, 7); (4) adsorption of A and B on the catalyst surface (reflected in adsorption equilibrium constant); (5) reaction between adsorbed A and B to produce E and W; (6) desorption of

where kR2 is the second-order rate constant, (cm3/g-mol)(cm3/g)W s ) . Eliminating the unknown surface concentrations of the reactants, a quadratic equation in RAis obtained as given by Kumbhar and Yadav (1989). Equation 4 is a special case of the power-law model. When the adsorption and/ or desorption steps are likely to be important, eq 4 can be replaced by the Langmuir-Hinshelwood types of the model. Our preliminary experiments had indicated that the analysis based on the treatment mentioned in this paper explained the observed rates. Depending on the relative magnitude of the diffusional and reaction rates, various controlling mechanisms prevail. When the surface reaction is very rapid with respect to the diffusion of A and B or of either A alone or B alone, several interesting cases emerge depending on the concentration of A and B in the liquid phase (see Kumbhar and Yadav (1989) for detailed analysis of this aspect). When the surface reaction is much slower than the diffusional steps, the following relationships holds: 1

1

1

qkRZ~[AO1 [Bo] ” ksL~ap[&1’ZkSL-Bap[BOl

(5)

Then the surface reaction is the controlling mechanism and the overall rate of reaction will be the same as that

Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 2201 given by the surface reaction; i.e.,

= ki[AoI

RA = V k ~ z W [ & l [Bo1

(6)

In order to discern the controlling mechanisms, the effect of various parameters on the rate of reaction was studied as discussed in what follows. In our case, acetic acid was always taken in far excess in order to drive the reaction to the right (mole ratio A:B = 1:4 and 1:3, for the phenethyl acetate and cyclohexyl acetate, respectively, for all the catalysts employed in this study). Effect of Intraparticle Diffusion. Since the alcohol was used as a limiting reactant in most of the cases, its concentration may not be uniform within the catalyst particle due to intraparticle diffusional resistance, whereas the concentration of acetic acid inside the catalyst is almost uniform. To account for this resistance, differential equations for intraparticle diffusion with reaction on the surface of the catalyst need to be solved and this type of problem is well-known to chemical engineers (for instance, see Satterfield (1970) and Doraiswamy and Sharma (1984)). For a first-order reaction occurring in a spherical particle of radius R, one can relate the concentration of the alcohol at any radial distance r from the center of the catalyst to its concentration on the surface by the following equation: [AI --

[A,]

-

sinh(34rlR) [dRI sinh(34)

(7)

4 = Thiele modulus:

k ~ =1 pseudo-first-order rate constant, (cm3/g)(l/s), kR1=

k,Z[Bol

(9)

and pp = particle density, g/cm3. It is desirable to express the effect of intraparticle diffusion in terms of the effectiveness factor, q: q = [1/41[coth(34) - 1/34]

(10)

Equation (4) now needs to be rewritten as RA = ?I~R~W[A,]

(11)

Once again the concentration [A,] can be eliminated to get RA as RA = [&I(Ppdd6ks~~W + l/&W)-'

(12)

because a, = 6w/p,dP

(13)

RA = [AolW(Ppdd6k~~-~ + l/Vk~i)-'

(14)

Depending on the relative magnitudes of ksL-A and tlkR1, the controlling resistance can be found out. Intrinsic Kinetics. In the absence of both external and internal resistance to mass transfer, it is possible to determine the intrinsic kinetics. For a pseudo-first-order reaction, with excess of acetic acid,

(16)

where

Integrating eq 16, we get

or -ln(l - XA)= klt

(19)

where XA = fractional conversion of A,

Thus, a plot of -ln(l - XA)against t will yield a slope 1 be determined. of 121, from which k ~ can Furthermore, the effects of solid loading w and also particle diameter d, on the rate of reaction can be studied. A plot of kl against w will be a straight line passing through the origin with a slope of k ~ 1 . As mentioned earlier, acetic acid was always taken in far excess of the alcohol and the preliminary experiments were conducted under otherwise similar conditions of reactant concentration, catalyst loading, speed of agitation, and temperature except the type of catalyst to study the efficacy of the catalyst. Figures 2 and 3 give plots of the conversion of phenethyl alcohol and cyclohexanol, respectively, against time for the various catalysts. It was also thought desirable to use a homogeneous catalyst for the sake of comparison. The following catalysts were used (1)Filtrol-24; (2) Amberlyst-15;(3) sulfated zirconia; (4) DTPA supported on silica; (5) DTPA supported on carbon; (6) DTPA as a homogeneous catalyst; (7) concentrated sulfuric acid as a homogeneous catalyst. The effect of various parameters on the rate of reaction of the alcohol was studied by choosing Filtrol-24 as a catalyst wherein only one parameter was changed at a time under otherwise similar reaction conditions. Effect of Speed of Agitation. To ascertain the influence of external resistance to mass transfer of the reactants to the catalyst surface, the speed of agitation was varied over a range of 500-1800 and 250-1800 rpm for the phenethyl alcohol and cyclohexanol systems, respectively, under otherwise similar conditions (see Figure 4). A catalyst loading of 5 7% w/w (based on the alcohol) and temperature of 100 "C were employed. The percent conversion of the alcohol was measured as a function of time. It was observed that the speed of agitation had no effect on the conversion and hence rate of reaction beyond 1000 and 1200 rpm for the esterification reaction of phenethyl alcohol and cyclohexanol, respectively. Thus, there was no limitation of external mass transfer beyond the said speeds. Further experiments were conducted at or above the speeds for both cases. The values of the rates of mass transfer were theoretically calculated as per eqs 2 and 3. According to the model presented here both kSL-A and ksL-B need to be calculated a t the operating conditions. The diffusivities of A in bulk B,DAB,and B in bulk A, DBA,were estimated by the WilkeChang group contribution method given by Reid et al. (1977) for both systems. A typical method is described below:

2202 Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 ""I

-

100

X

0

50

100

150

Reaction

200

250

300

350

Time,min

Figure 2. Effect of various catalysts on conversion of phenethyl alcohol: mole ratio, phenethyl alcohokacetic acid (1:4); temperature, 100O C ; catalyst loading, 5 % (w/w) based on alcohol;speed of agitation, 1200 rpm. (-X-) Amberlyst-15; (-o-)Filtrol-24; (+-) sulfated zirconia; (-+-) DTPA/silica; (-*-) DTPA/carbon; (-0-) DTPA (homogeneous). 100,

0

500

1000

1500

2000

S p e e d of A g i t a t i o n , rpm

Figure 4. Effect of speed of agitation on conversionof alcohol: mole ratio, phenethyl alcohokacetic acid (1:4), cyclohexanokacetic acid (1:3); catalyst, Filtrol-24; temperature, 100 O C ; catalyst loading, 5% (w/w) based on alcohol; reaction time, 300 min. (*-) Phenethyl alcohol; (-+-) cyclohexanol.

solid-liquid external mass transfer is very much greater than the rate of reaction on the surface of the catalyst and hence the mechanism is surface reaction controlled. The values of k~zw[Bolwere determined from the initial rate of reaction, that is, initial rate of reaction = rate of surface reaction The values of initial rate of reaction were calculated for the phenethyl alcohol and cyclohexanol system, and compared with the rate of mass transfer. According to eq (5), the left-hand side (LHS) and right-hand side (RHS) for these two systems for a typical case are as follows:

Reaction

Time, m i n

Figure 3. Effect of various catalysts on conversion of cyclohexanol: mole ratio, cyclohexanokacetic acid (1:3); temperature, 100 O C ; catalyst loading, 5% (w/w) based on alcohol; speed of agitation, 1200 rpm. (-*-) Amberlyst-15; (-+-) Filtrol-24; (-04 sulfated zirconia; (+-) DTPAhilica; (-O-) DTPA/carbon; (-X-) DTPA (homogeneous).

For the phenethylalcohol (A)and acetic acid (B)system, the diffusivity values at 100 "C are Dm = 1.6342 X 10-5 cm2/s. cm2/s and DBA= 4.2768 X The solid-liquid mass transfer coefficients were calculated by assuming the Sherwood number, Sh = (ksLd,)/D = 2 as ~ S L - A= 4.95 X lo4 cm/s and ~ S L - B= 1.296 X cm/s. For Filtrol-24, d, = 0.066 cm, pp = 0.75 g/cm3, w = 0.017 53 g/cm3liquid volume, and up = (6w)/p,d, = 2.1247 cm2/cm3. For the cyclohexanol (A) and acetic acid (B) system, then diffusivity values at 100 "C are D m = 1.313 X cm2/s and DBA= 2.846 X lo3 cm2/s. Thus, the corresponding ~ S L Aand ~ S L . Bvalues are 3.98 X 10" and 8.63 X 104cm/s, respectively. For this system, w = 0.018 18 g/cm3liquid volume, d, = 0.066 cm, pp = 0.75 g/cm3, and up = 2.20 cmZ/cm3. Putting these values in eq 5, it is found that the rate of

first quantity, second quantity, system LHS RHS RHS phenethyl alcohol 3.9760 X 108 3.3245 X 106 3.1840 X 104 cyclohexanol 7.4680 X 108 3.1420 X 106 4.8361 X 104

From these values it is obvious that the LHS is greater than both quantities on the RHS and hence the condition given by eq 5 is satisfied indicating the absence of resistance to solid-liquid mass transfer at a reaction temperature of 100 "C. For other reaction conditions of temperature and concentration the same observation was found to hold. Effect of Intraparticle Resistance (Pore Diffusional Resistance). For the typical case of Filtrol-24 as catalyst with an average particle size of 660 pm, the value of Thiele modulus (4) as given by eq 8 was calculated as 0.077 and the effectiveness factor (7) as given by eq 10 was 1for the phenethyl alcohol system. For the cyclohexanol system, 4 and 7 were found to be 0.0425 and 1,respectively. Thus, in both systems there was no pore or intraparticle diffusional resistance. It was further checked that the intraparticle diffusional resistance was absent for all other catalysts covered in this study. Effect of Catalyst,Loading. The catalyst loading was varied over a range of 1-10% w/w on the basis of the amount of alcohol. Figures 5 and 6 show the effect of catalyst loading on the conversion of the alcohol. The rate of reaction is directly proportional to the catalyst loading because the surface area and hence the total number of active sites increase linearly with it.

Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 2203

0 R e a c t i o n T i m e , min

0

0.01

0.02

0.03

0.04

w , g / c m 3 0 f l i q u i d volume

Figure 7. Effect of mole ratio of phenethyl alcohol to acetic acid on conversion of phenethyl alcohol: temperature, 100 "C;catalyst, Filtrol-24; catalyst loading, 5 % (w/w)based on alcohol; speed of agitation, 1200 rpm. (+-) 1:l. (-+-) 1:3; (-*-) 1:4; (-m) 1:s.

Figure 5. Effect of catalyst loading on pseudo-first-order constant: mole ratio, phenethyl alcohokacetic acid (1:4), catalyst, Filtrol-24; temperature, 100 O C ; speed of agitation, 1200 rpm. (e) Filtrol-24.

0 Reaction T i m e , min

Figure 8. Effect of mole ratio of cyclohexanol to acetic acid on conversion of cyclohexanol: temperature, 100 "C;catalyst, Filtrol24; catalyst loading, 5 % (w/w)based on alcohol; speed of agitation, 1200 rpm. (-0-) 1:l; (-+-) 1:3; (-*-) 1:4. I

0

0.01

I

0.02

I

0.03

0.06

w , g / c m 3 0 f l i q u i d volume

Figure 6. Effect of catalyst loading on pseudo-first-order constant: mole ratio, cyclohexanokacetic acid (1:4), catalyst, Filtrol-24; temperature, 100 OC;speed of agitation, 1200 rpm. (+-) Filtrol-24.

Effect of Mole Ratio. The mole ratio of acetic acid to the alcohol was varied from 1:l to 5:l to assess its effect on the rate of reaction and to find the order of reaction at 100 "C, with a catalyst loading of 5% w/w based on the alcohol (Figures 7 and 8). It was found that as the mole ratio is increased from 1:l to 3:1, the conversion of the alcohol increases and the reaction becomes overall second order, first order in acetic acid, and first order in the alcohol. Beyond a mole ratio of 4:l and 3:l for the phenethyl alcohol and cyclohexanol systems, respectively, a pseudo-firstorder kinetics is obtained and conversion becomes independent of the mole ratio. This observation stems from the fact that

x, = 1- exp(-k,t)

= 1- eXp(-k~,WM[&]t)

(21)

Thus, by putting appropriate values in eq (21), the fractional conversion X Abecomes almost equal when M 2 4 and 3 for the two systems. Effect of Temperature. The effect of temperature on the rate of reaction was studied by conducting the reactions at 80, 100, and 115 "C for both cases under otherwise similar conditions. In addition, an experiment was also done at 66 "C for the cyclohexanol system. The ester conversion was found to increase with temperature (see Figures 9 and 10). Arrhenius plots (In k ~ vs 2 1/27 were made as shown in Figures 11and 12 to find out the activation energy and the preexponential factor which are presented in Table 4. The high values of activation energies further lend support to the earlier comments that there was a total absence of both external and internal mass transfer resistance and reaction was kinetically controlled in both esterifications. The activation energy for phenethyl alcohol with Filtrol-24 is 8.86 kcal/g-mol,and thus, the influence of mass transfer is likely to be significant for coarse particles. The calculation showed that the effectiveness factor under operating conditions was unity. Since Filtrol-24 was a commercial

2204 Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 '""I

0

50

100

200

150

250

300

350

Reaction T i m e , m i n

Figure 9. Effect of temperature on conversion of phenethyl alcohol: mole ratio, phenethyl alcohokacetic acid (1:4); catalyst, Filtrol-24; catalyst loading, 5% (w/w) based on alcohol; speed of agitation, 1200 rpm. (-O-) 80 O C ; (-+-) 100 O C ; (-*-) 115 O C .

250

270

2 60

280

2 90

300

1 x 1 0 5 , K-' 1 Figure 11. Arrhenius plot: mole ratio, phenethyl alcohokaceticacid (1:4); catalyst, Filtrol-24;catalyst loading, 5% (w/w) based on alcohol; speed of agitation, 1200 rpm. (*-) Filtrol-24.

0

50

100

150

200

Reaction T i m e , m i n

Figure 10. Effect of temperature on conversion of cyclohexanok mole ratio, cyclohexanokaceticacid (1:3);catalyst, Filtrol-24;catalyst based on alcohol; speed of agitation, 1200 rpm. loading, 5% (w/w) (-0-) 66 "C;(-+-) 80 O C ; (-*-) 100 "C;(-O-) 115 O C .

sample, the effect of particle size was not studied. In order to check that, for the best catalyst Amberlyst-15, there was no effect of intraparticle diffusion, the reaction was conducted at two different temperatures to get an activation energy of 12.2 kcal/g-mol. Kinetics of the Reaction. Since both external and internal resistance to mass transfer were absent for both reactions, the kinetics of the reaction was determined as explained earlier by eq 16. Thus, plots of -ln(l - XA)vs t were made to determine kl and hence kR1, for the pseudo-first-order case. Since, the overall rate of reaction follows a second-order kinetics,

-6.01 250

I

I

I

I

2 60

210

2 80

2 90

f

300

x ~ ~ K-'5 ,

Figure 12. Arrhenius plot: mole ratio, cyclohexanokaceticacid (1: 3); catalyst, Filtrol-24; catalyst loading, 5% (w/w) based on alcohol; Filtrol-24. speed of agitation, 1200 rpm. (e) Table 4. Values of Activation Energy and Preexponential Factor for Phenethyl Alcohol and Cyclohexanol preexponential factor, activation (cm3/g-mol) energy, system catalyst (cm3/g)(l/s) kcal/g-mol phenethyl alcohol Filtrol-24 65 512.74 8.86 cyclohexanol Filtrol-24 94 112 365.00 14.87

fractional conversion of the alcohol.

--X A - [&)liKR2wt For a mole ratio of M = [Bo]i/[&]i = 1, eq 22 can be integrated to give the following equation in terms of the

1 - x A

(23)

For the mole ratio of M # 1, the integration of eq 22

Ind. Eng. Chem. Res., Vol. 33, No. 9,1994 2205 Table 5. Value of the Rate Constant Catalyst at M = 1 system phenethyl alcohol cyclohexanol

b for Filtrol-24

~ R Z (cm3/g-mol)(cm3/g)(l/s) ,

0.600 0.233

Table 6. Values of the Rate Constant for the Various Catalysts for the Phenethyl Alcohol System Amberlyst-15 Filtrol-24 sulfated zirconia DTPA/silica DTPA/carbon

1.0420 0.4400 0.3919 0.2877 0.1339

Table 7. Values of the Rate Constant Catalysts for the cyclohexanol System

0

50

100

150

200

250

300

350

Reaction T i m e , min

Figure 13. Kinetic plots for the reaction between alcohols and acetic acid mole ratio, phenethyl alcohokacetic acid (1:4), cyclohexanol: acetic acid (1:3);catalyst, Filtrol-24;catalyst loading, 5% (w/w)based on alcohol;temperature, 100 O C ; speed of agitation, 1200rpm. (-O-) Phenethyl alcohol; (-+-) cyclohexanol.

catalyst Amberlyst-15 Filtrol-24 sulfated zirconia DTPA/silica DTPA/ carbon

kR21

for the Various

(cm3/g-mol)(cm3/g)(l/s) 0.4306 0.1863 0.1570 0.1059 0.0859

Activity and Reusability of Different Catalysts. In the present study, in all six different catalysts were employed under otherwise similar conditions of reaction temperature, concentrations, and catalyst loading and are Filtrol-24, Amberlyst-15, sulfated zirconia, DTPA supported on silica, and DTPA supported carbon, which are all solid catalysts, and DTPA as homogeneous catalyst. Figures 2 and 3 show the percent conversion vs time plots for phenethyl alcohol and cyclohexanol esterifications, respectively. The activity of the catalyst on the basis of unit weight of the catalyst shows the following trends. Phenethyl Alcohol: Amberlyst-15 = DTPA (homogeneous) > Filtrol-24 > sulfated zirconia > DTPA/silica (heterogeneous) > DTPA/carbon (heterogeneous) Cyclohexanol:

0

/+

/+-+ 0

50

100

150

200

250

300

350

Reaction T i m e , min

Figure 14. Kinetic plots for the reaction between alcoholsand acetic acid mole ratio, phenethyl alcohokacetic acid (1:4), cyclohexanol: acetic acid (1:3); catalyst, Filtrol-24; catalyst loading, 5 % (w/w)based on alcohol; temperature, 100 "C;speed of agitation, 1200rpm. (-03 Phenethyl alcohol; (-+-) cyclohexanol.

leads to the following:

Thus, plots were made of the LHS vs time for both eqs 23 and 24 to determine the slope of the line from which 2 calculated. See Figures 13 and 14 the values of k ~ were

and Tables 5-7. k R 2 is based on the weight of the catalyst having units (cm3/g-mol)(cm3/gof catalyst) (l/s). Corresponding values on surface area basis are also given in Tables 8 and 9.

Amberlyst-15 = DTPA (homogeneous) > Filtrol-24 > sulfated zirconia > DTPA/silica (heterogeneous) > DTPA/carbon (heterogeneous) Some comments are in order as far as the phenethyl alcohol esterification is concerned. It was observed that the Amberlyst-15and DTPA as homogeneous catalyst have almost similar activity but the latter is difficult to separate from the reaction mixture and reuse. However, with Amberlyst-15 it takes 5 h for 95.5% conversion of phenethyl alcohol. DTPA gives a conversion of 100% in 3 h. Reusability: The reusability of Filtrol-24 was tested by conducting three runs as shown in Figures 15and 16.There was a small decline in the catalytic activity; this may be because there was no treatment given t o the catalyst after the completion of reaction. Catalyst Structure. Our earlier work (Thorat et al., 1992)deals with the structure of heteropolyacids (HPAs) and the acid-treated zirconia in relation to their high catalytic activity. However, it is pertinent to brief about these catalysts. HPAs and some of their salts have a unique blend of properties which lend them high catalytic activity. HPAs are polynuclear complexes of hexavalent molybdenum, tungsten, and vanadium which can also incorporate

2206 Ind. Eng. Chem. Res., Vol. 33, No. 9, 1994 Table 8. Values of & for Various Heterogeneous Catalysts for the Phenethyl Alcohol System

km X lo', catalyst Amberlyst-15 Filtrol-24 sulfated zirconia DTPA/silica DTPA/carbon

(cmYg-mol)(crn~/cm*) (Us) 18.9400 1.2290 4.2990 1.4840 0.1452

Table 9. Values of h for Various Heterogeneous Catalysts for the Cyclohexanol System catalyst Amberlvst-15 Filtrol-i4 sulfated zirconia DTPA/silica DTPA/ carbon 0

50

100

150

200

250

300

350

Reaction Time, m i n

Figure 15. Reusability of Filtrol-24 for phenethyl alcohol system: mole ratio, phenethyl alcohokacetic acid (1:4);temperature, 100 O C ; catalyst loading, 5% (w/w) baaed on alcohol; speed of agitation, 1200 rpm. (-e-) Fresh catalyst; (-+-) first reuse; (-*-) second reuse. ""I

I

kw X lo', (cmS/g-mol)(cmS/cm2)(l/s) 7.82900 0.52039 1.72200 0.54650 0.09310

Table 10. Values of the Rate Constant 4 for the Phenethyl Alcohol and Cyclohexanol Systems Using DTPA Homogeneous Catalyst system phenethyl alcohol cyclohexanol

k2, (cm~/g-mol)(l/s)

0.043 86 0.013 59

responsible for the generation of strong acidity. The central metal Zr4+ cation acts as a Lewis acid, and H+ acts as a Bronsted acid (Thorat, 1994). The generation of highly acidic properties of the above structure is due to its dynamic transformation in which the bond character of the SO is altered by the adsorption and desorption of the reactant molecule. Bronsted acidity is also likely to be generated by adsorption of water molecule. Comparison of Acidity. Since the support plays an important role in lending high dispersion of the active species, the various rate constants kR2 can be expressed in terms of the internal surface area of the support.

k,

=

kR2

(cm3/g-mol)(cm3/g)(l/s) ai (cm2/g)

I/

O* 0

I

I

I

I

I

I

50

100

150

200

250

300

350

Reacfion Time, min

Figure 16. Reusability of Filtrol-24 for cyclohexanol system: mole ratio, cyclohexanokacetic acid (1:4); temperature, 100 O C ; catalyst loading, 5% (w/w)baaed on alcohol; speed of agitation, 1200 rpm. (-W Fresh catalyst; (-+-) first reuse; (-*-) second reuse.

with units of (cm3/gmol)(cm3/cm2)( l/s). Now arational comparison can be made of the km values (see Tables 8 and 9) to give the following order: Phenethyl alcohol: Amberlyst-15 > sulfated zirconia > DTPA/silica > Filtrol-24 > DTPA/carbon

other elements as central atoms of ligands, and are Cyclohexanol: multielectron oxidants as well as strong Bronsted acids. HPAs can be employed as both homogeneous and hetAmberlyst-15 > sulfated zirconia > DTPA/silica > erogeneous bifunctional oxidative and acidic catalysts. The Filtrol-24 > DTPA/carbon 12-HPA molecule, H%,XnMI2040 or H ~ , + , X n M ~ 2 V , 0 ~ have the Keggin structure, where X = central atom, SiIV, HomogeneousCatalysis. DTPA and sulfuric acid were GeN, Pv, AsV,etc., n = oxidation state of X', M = metal used as homogeneous catalysts as well. In that case, the atom, Mo*, Wvr,V = vanadium, and m = oxidation state rate of reaction is given by of vanadium. When protons are present as counterions with the heteropolyanion, the protons are highly mobile, R, = k3[H+l[AI [Bl lending high acidity. = k,[AI [Bl The sulfated zirconia has the following structure,

There are two covalent SO double bonds which are

where kz = k3[H+l for a specific catalyst concentration. The k2 values are given in Table 10. The efficacy of Filtrol-24 was compared with aqueous sulfuric acid by using the same milliequivalents at a Filtrol-24 loading of 5 % w/w based on the alcohol and under otherwise similar conditions. Figure 17gives the plots of percent conversion

Ind. Eng. Chem. Res., Vol. 33, No. 9,1994 2207 kR2 = second-order rate constant, (~m3/g-mol)(cm~/g)(l/s) k s ~ksLB ~ , = solid-liquid mass transfer coefficientfor A and B, respectively, cm/s kw = rate constant based on internal surface area of the support, (cm3/g-mol)(cm3/cm2)(Us) M = mole ratio of acetic acid to alcohol RA = rate of reaction for A, (g-mol/cm3s) R = catalyst radius, cm r = any radial distance from the center of the catalyst, cm w = catalyst loading, g/cm3 X A = fractional conversion of A z = stoichiometric coefficient Greek Letters 7 = effectiveness factor pp = density of catalyst particle, g/cm3 4 = Thiele modulus 50

100

150

200

250

300

350

R e a c t i o n T i m e , min

Figure 17. Comparisonof Filtrol-24and sulfuric acid for both alcohol systems: mole ratio, phenethyl alcohokaceticacid (1:4), cyclohexanol: acetic acid (1:3); temperature, 115 OC; speed of agitation, 1200 rpm. (e) Filtrol-24 (phenethyl alcohol); (-+-) sulfuric acid (phenethyl alcohol); (-*-) Filtrol-24 (cyclohexanol); (-W) sulfuric acid (cyclohexanol).

against time for both alcohols. It is obvious that the solid catalyst Filtrol-24 is more reactive than aqueous sulfuric acid at the same milliequivalents.

Conclusions Solid acid catalysis in the preparation of perfumery esters is very effective from the viewpoint of activity, selectivity, reuse, and noncorrosiveness in comparison with the homogeneous catalysts. For both the esterification of phenethyl alcohol and of cyclohexanol, following is the order of catalytic activity based on unit weight of the catalyst for the catalysts studied: Amberlyst-15 > Filtrol-24 > sulfated zirconia > DTPA/silica > DTPA/carbon

Acknowledgment P.H.M. thanks the University Grants Commissions for an award of Senior Research Fellowship which made this work possible.

Nomenclature [AI = concentration of A in catalyst pores, g-mol/cm3 [&I = concentration of A in bulk liquid phase, g-mol/cm3 [&]i = initial concentration of A in bulk liquid phase, g-mol/ cm3 [A,] = concentration of A at solid (catalyst) surface, g-mol/ cm3 ai = surface area of catalyst, cm2/g a, = solid-liquid interfacial area, cm2/cm3 [Bo] = concentration of B in bulk liquid phase, g-mol/cm3 [B& = initial concentration of B in bulk liquid phase, g-mol/ cm3 [B,] = concentration of B at solid-liquid interface, g-mol/ cm3 D m = diffusion coefficient of A in B, cm2/s DBA= diffusion coefficient of B in A, cm2/s De = effective diffusivity, cm2/s d, = diameter of catalyst, cm k2 = rate constant for homogeneous catalyst system, (cm3/ g-mol)(l/s) kR1 = pseudo-first-order rate constant, (cm3/g)(l/s)

Literature Cited Arctander, S. Perfume and Flavor Chemicals; Steffen Arctander Publications: Elizabeth, NJ, 1988. Bertin, J.; Kogan, H. B.; Luche, J.; Setton, R. Graphite Electrolytic Lamellar Reagents in Organic Chemistry. Esterification in the Presence of Graphite Bisulfate. J. Am. Chem. SOC.1974,96 (26), 8113. Besenhard, J. 0. Mechanism of the Esterification Catalysis by Graphite Hydrogen Sulfate. 2.Naturforsch.,Anorg. Chem., Org. Chem. 1977,32B (lo), 1210. Brewster, J. H.; Ciotti, C. J. Dehydration with Aromatic Sulfonyl Halides in Pyridine. A Convenient Method for the Preparation of Esters. J. Am. Chem. SOC.1956, 77, 6214. Doraiswamy, L. K.; Sharma, M. M. Heterogeneous Reactions: Analysis,Examples and Reactor Design; John Wiley: New York, 1984; Vol. 2. Fernandez, C.; Marinas, J. M.; Sinisterra, J. V. Aluminium Phosphate Supported on Polymer beads as Catalyst in the Esterification of Carboxylic Acids. Bull. SOC.Chim. Belg. 1988,97 (l),17. Hino, A.; Arata, K. Synthesis of Solid Superacid Catalyst with Acid Strength of Ho < -16.04. J. Chem. Soc., Chem. Commun. 1980, 24,1148. Huang, H.; Liu, F.; Zhang, H. Esterification Reactions Catalysed by Hydrated Sulfated Salts. Chem. Abstr. 1988, 109,148688~. Kamijo, T.; Yamamoto, R.; Harada, H.; Izuka, K. 1-Acyl-3-Substituted Imidazolium salts as Highly Reactive Acylating Agents. Chem. Pharm. Bull. 1982,30 ( l l ) , 4242. Kogan, H.; Bertin, J.; Luche, J.; Setton, R. Esterification of Carboxylic Acids in Presence of Intercalary Compounds of Acid and Graphite Bisulfate. Chem. Abstr. 1977,86, P88454k. Kumbhar, P. S.; Yadav, G. D. Catalysis by Sulfur-Promoted Superacidic Zirconia. Condensation reactions of Hydroquinone with Aniline and substituted Anilines. Chem. Eng. Sci. 1989,44 (ll),2535. McCracken, D. J.; Dickson, P. F. Reaction Kinetics of Cyclohexanol - Acetic Acid Esterification. Ind. Eng. Chem. Process Des. Dev. 1967,6 (3), 286. Mitsunobo, 0.;Takemasa, A.; Ends, R. Intermolecular Dehydration between Carboxylic Acids and Alcohols by the Use of Diethyl azodicarboxylate and 3-methyl benzothiazole-2-selone. Chem. Lett. 1984, 6, 855. Ogliaruso, M. A.; Wolfe, J. F. Synthesis of Carboxylic Acids, Esters, and Their Derivatiues; John Wiley and Sons: Chichester, 1991. Olah, G. A.; Keumi, T.; Meidar, D. Synthetic Methods and Reactions. A Convenientand Improved Method for Esterification Over Nafon H, a Superacidic Perfluorinated Resinsulfonic acid Catalyst. Synthesis 1978, 12, 929. Ou, Z.; Chen, Z. Esterification Catalysed by Synthetic Zeolites. Chem. Abstr. 1985, 104, 87775~. Parish, R. C.; Stoch, L. M. A Method for the Esterification of Hindered Acids. J. Org. Chem. 1965, 30, 927. Reid, R. C.; Prausnitz, J.; Sherwood, T. Properties of Gases and Liquids, 3rd ed.; McGraw-Hik New York, 1977. Satterfield, C. N. Mass Transfer in Heterogeneous Catalysis;MIT Press: Cambridge, MA, 1970. Saino. K.: Usui. M.: Kikuchi. K.: Shimada. E.: Mokaivama. T. New Method for the Preparation of Carboxylic Esters: Bull. Chem. SOC.Jpn. 1977, 50 (7), 1863.

2208 Ind. Eng.Chem. Res., Vol. 33, No. 9, 1994 Sekino, M.; Kume, A.; Hata, T. Dialkylphosphonates: A New Acylating Agent of Alcohols. Tetrahedron Lett. 1981, 22 (37), 3617. Selwitz, C. M. Preparing Esters by Reacting a Monobasic Organic Acid with a Secondary Alkyl Halide in the Presence of Ferric, Stannic or Zinc Chloride or Bromides. Chern. Abstr. 1972, 76, P126407a. Sunggak, K.; Sung,S. K. S,S-Bis[4,6-dimethyl-2-pyrimidinyl]Dithiocarbonate. A New Reactive Coupling Agent for the Direct Esterification of Carboxylic Acids. Synthesis 1986, 12, 1017. Takimoto, S.; Jnanaya, J.; Kabuki, T.; Yamaguchi, M. Esterification of CarboxylicAcidsby Alcohols with 2-chloro-1,3,5-trinitrobenzene as Coupling Agent. Bull. Chem. SOC.Jpn. 1981,54 (5), 1470.

Thorat, T. S. Heterogeneous Catalytic Reactions. Ph.D. Thesis, University of Bombay, Bombay, 1994. Thorat, T. S.; Yadav, V. M.; Yadav, G. D. Esterification of Phthalic anhydride with 2-ethylhexanol by Solid Superacidic Catalysts. Appl. Catal. A Gen. 1992,90, 73. Received for review December 7, 1993 Revised manuscript received May 9, 1994 Accepted May 31, 1994@

* Abstractpublished in Advance ACSAbstracts, July 15,1994.