Chemical

Engineering

and Processing,

33 i 1994) 413-418

Kinetics of the esterification of phthalic anhydride 2-ethylhexanol. I. Sulfuric acid as a catalyst

Jerzy Skrzypek,*, Jan Zbigniew Sadlowski, Maria Lachowska,

with

Marian Turzariskih

~ Institute of Chemicnf f%ginerring, Polish Academy oj Sciences, ul. Bultycka 5. 44-100 Gliwice, Poland bInstitute qfMathematics, Silesian University, Katon!ice, Poland Received 30 June 1993; accepted 3 March 1994

Abstract
The kinetics of the esterification of mono-2-ethylhexyl phthdlate in the presence of sulfuric acid as a catalyst has been investigated in an isothermal semibatch reactor. The reaction appears to be first order with respect to mono-2-ethylhexyl phthalate alone and the reaction rate does not depend on the concentration of 2-ethylhexanol. The kinetic parameters have been

determined. Keywords: Kinetics; Esterification; Phthalic anhydrldc

Introduction Di-2-ethylhexyl phthalate is an important industrial chemical. Its main use is as an additive to the wellknown plastic PVC. The synthesis proceeds in two stages:

has been well developed industrially and presents no problems. The mass production of di-2-ethylhexyl phthalate is usually carried out in a separate plant from that used for the preparation and storage of the mono2-ethylhexyl phthalate.

0 + CH1(CH&CHCH,OH-

C-OCH,CH(CH&CH,

(1)

C-OCH,CH(CH,),CH, + CH,(CH&CHCH,OH w

(2)

The first step is the fast non-catalytic and complete reaction of phthalic anhydride with one molecule of 2-ethylhexanol, the monocarboxylic acid generated being readily obtained on dissolution of phthalic anhydride in 2-ethylhexyl alcohol at 50-90 C. This stage

*Author

to whom correspondence

should

be addressed.

The second stage in the preparation is a slow equilibrium reaction which industrially is taken to completion by removing the water formed. Both acid catalysts (sulfuric acid or p-toluenesulfonic acid) and non-acidic alkyl titanates are used for this stage in industrial practice. Few industrial installations exist in which no catalyst is employed or, strictly speaking, where the $ 1994 ~ Elsrvier Science S.A. All rights reserved ,

0X5-2701 /94/!$7.00 SSDI 0255-2701(94)00517-L

414

.l. Skrqpk

et al / Chemicrrl Engimwing

and Proctming

33 (1994) 413-418

phthalate monoester produced during the first stage plays the role of a weak acid. The most recent generation of catalysts for the esterification of phthalic anhydride. which have yet to be used commercialized, includes heteropolyacids which exhibit superior activity at low temperatures and possess good selectivity [ 1,2]. The strong acid catalyst which has been employed for the greater length of time is sulfuric acid, but it has the disadvantage of generating relatively high quantities of byproducts. In constrast, p-toluenesulfonic acid is relatively expensive and toxic, and there are serious problems in recycling this catalyst. Very low levels of byproduct are attained using titanates as catalysts; however, their activity is much less than that of sulfuric acid. A suitable degree of removal of water is always important from the standpoint of eliminating the retarding effect of the reverse reaction of hydrolysis. This can be achieved either by using (i) vacuum conditions, (ii) co-solvents which form a low-boiling heteroazeotrope with water (e.g. toluene), or (iii) a flow of inert gas (e.g. nitrogen).

Kinetics of the esterification of mono-Z-ethylhexyl phthalate with 2-ethylhexanol in the presence of sulfuric acid as a catalyst - a literature survey

making it necessary to introduce several empirical coefficients into the resulting rate equation. These workers also pointed out that during the first few minutes of the reaction a distinct anomalous acceleration of the process occurred. This phenomenon was subsequently explained in terms of the transformation of the sulfuric acid into alkyl sulfuric acid during the first few minutes, and this compound then acts as a catalyst under the reaction conditions employed [S-7]. A kinetic equation frequently encountered in the literature is the second order one proposed by Leyes and Othmer [8] for the esterification of phthalic anhydride with butyl alcohol. Drimus et al. [9] employed this equation for the study of the kinetics of esterification of phthalic anhydride with 2-ethylhexyl alcohol in the presence of sulfuric acid and found a satisfactory agreement with the experimental data. The object of the series of articles of which this is the first is to develop rigorous kinetic equations for the esterification of mono-2-ethylhexyl phthalate with 2ethylhexanol in the presence of various catalysts, allowing the possible application of the results to the modelling and optimization of industrial reactors. This first part in the series describes studies in which sulfuric acid has been used as the catalyst.

Literature reports dealing with this subject are relatively scarce and very often contradictory. Some of the kinetic studies described have been carried out with continuous removal of the water formed in the reaction, whereas others have been performed in closed isolated systems in which the effect of the reverse hydrolysis of the esters produced must have been substantial. The absence of stable and suitable conditions for the reaction from the standpoint of its reversibility or irreversibility seems to be the reason for the discrepancies between the various kinetic equations given in the litersecond-order and empirical ature. Both first-order, equations have been proposed. In describing a continuous industrial cascade process developed by BASF for the production of di-2-ethylhexyl phthalate, Suter [3] has suggested that the reaction is first order (but only with resqect to the monoester). Unfortunately, he did not provide any description of the experiments which might have substantiated such a conclusion. According to Nosovski et ul. [4], the reaction may be considered as close to first order with respect to monoester and sulfuric acid provided that a number of empirical coefficients are assumed. Nosovski et al. [4] carried out their studies in a semibatch reactor, removing water from the system by the use of vacuum. However, they did recycle 2-ethylhexanol which in our opinion was contaminated with water. Hence, a certain amount of water was re-introduced into the system and this may have affected the unidirectional course of the reaction,

Experimental

The equipment employed in this study allowed the immediate and complete removal of the water from the system. The most important component of the apparatus consisted of a four-necked glass flask of 1 dm3 capacity which was placed in a thermostat. This flask was equipped with a thermometer, a nitrogen and 2-ethylhexyl alcohol inlet, a high-speed mixer, a head for collecting samples and an azeotropic head with a cooler. Losses of 2-ethylhexyl alcohol during the reaction were compensated for by the addition of equivalent amounts of the pure anhydrous reactant. The removal of water from the system was effected by the use of a suitably adjusted nitrogen flow. A sufficiently high flow of N, (of the order 60-80 dm3 h- was used to ensure the efficient removal ) of water, and consequently elimination of the reverse hydrolysis reaction. The level of the nitrogen flow was determined experimentally by using progressively increasing N, flow rates. A maximum limiting value of the flow was thus established above which further increase in the flow rate had no effect. Under these circumstances the removal of water was virtually instantaneous and chromatographic analysis of a sample of the reacting liquid did not reveal the presence of water. The initial reaction mixture of a given composition was prepared by dissolving a suitable amount of phthalic anhydride in 2-ethylhexanol at temperatures up to 8.5 (358 K). A suitable excess of mono-2-ethylC

.I. Skrzypek

et al. 1 Chemical

Engineering

and Processing 33 (1994) 413-418

415

hexylphthalate was thus formed in the solution of 2ethylhexyl alcohol. The following initial molar ratios of 2-ethylhexanol/mono-2-ethylhexyl phthalate were used: l:l, 1.4:1, 3:1, 6:1, 11~1. Following dissolution of the phthalic anhydride and formation of the monoester, the solution was cooled to c. 20 C (293 K) and its acid number determined. The acid number obtained did not differ by more than 1% from the theoretical value for a given excess of 2-ethylhexanol. The monoester solution (0.75 dm3) and 2-ethylhexanol in a suitable amount were introduced into a heated reaction flask. The liquid was heated to a given temperature and a weighted portion of sulfuric acid (96.7%), dissolved in 0.01 dm3 of the reacting mixture, introduced. The flow of nitrogen was then commenced and samples were collected after the temperature had stabilized. Samples were taken every IO, 15, 30 or 60 min, depending on the rate of the reaction. The acid number of each sample was determined, corrected for the small amount of H,SO, present and recalculated to obtain the degree of conversion of the monoester. Chromatographic analyses were also randomly performed which indicated that no substantial side reactions occurred during the kinetic experiments. Virtually the only reaction taking place in the system was the single irreversible reaction expressed in eqn. (2) above. The concentrations of sulfuric acid employed were 0.0078, 0.0156, 0.0235 and 0.0313 mol dm- (at 20 C), corresponding to 0.6, 1.2, 1.8 and 2.4g of H,SO,, respectively, in 0.75 dm of the reacting solution. These concentrations were selected on the basis of industrial practice where the amount of H2S04 used is approximately equal to 0.5 wt.% relative to the amount of phthalic anhydride used. This corresponds to a concentration of H,SO, in the monocstcr/alcohol solution of c. 0.0156 mol dmm3. Excessive amounts of H,SO, lead to a visible darkening of the final product and an excessive consumption of basic solutions for neutrahzing the crude diester formed. The degree of conversion, of the mono-2-ethylI, hexyl phthalate was calculated from:

where AN0 is the initial acid number of the reaction system, AN that after a time 7, 56 000 is an analytical factor (the molar mass of KOH expressed in mg mol- ) and 18 is the molar mass of water. Equations (3) or (3a) take account the decrease in the total reaction mass due to removal of water from the system during the course of the reaction. For high acid numbers, i.e. for r + 0, a good approximation is provided by the formula: X0-X rl=-= x0 which does AN -AN AN0 not take the change in mass (4) into account.

Fig. I. Exactness of fit for first-order alcohol/monoester (A/M) molar ratios: IX= 3.0; *, T = 6.0; and T>$, = 11.0. z

kinetics r = k*C,. Initial LI, z = 1: *, z : 1.4; 0,

on j

X-0 where x0 is the hexyl phthalate, the molar mass mono-2-ethylhexyi Equation (3) cal form as: AN - AN rl=

(3)

initial concentration of mono-2-ethylX the concentration after a time r, M, of water and A the molar mass of , phthalate. may be written in an equivalent practi-

56 000 - 18AN 56 000 - 18AN > AN (3a)

Fig. 2. Exactness alcohol/monoester of fit for first-order kinetics r = k*C,. Initial

z=

(A/M) 3.0; and a, r = 11.0.

molar

ratios:

a.

a = 1; *, a =

1.4, 0,

416

.I. Skrzypk cf ul. Chemical 1

Engineering

and Processing

33 (1994) 413-418

Table 1 Calculated Run No.

Results
values of first-order rate constant k* A/M molar ratio, a

k (min-)

C,., (mol dm-))

The reaction appeared to be first order with respect to monoester and sulfuric acid as a catalyst, i.e.
r = k*C,

At 393 K 2 9 14 22 28 34 38 45 50 57 At 403 K 3 10 I5 19 23 29 39 51

0.0118 0.0062 0.0114 0.0241 0.0 126 0.0228 0.0105 0.0236 0.0114 0.0227

0.014 0.007 0.014 0.028 0.014 0.028 0.014 0.028 0.014 0.028

1.0 1.4 1.4 1.4 3.0 3.0 6.0 6.0 11.0 11.0

where k* = kc,,,

(6)

Calculated values of the first-order reaction constant k * for each run are listed in Table 1, while Figs. l-5 illustrate the exactness of fit. Arrhenius plots for several C,,, values are presented in Fig. 6. Table 2 lists the

0.0161 0.0088 0.0200 0.0243 0.0321 0.0168 0.0142 0.0168

0.014 0.007 0.014 0.021 0.028 0.014 0.014 0.014

1.0

1.4
1.4 1.4 1.4 3.0 6.0 11.0

At
5 12 17 25 31 41 53

413 K 0.0337 0.0185 0.0348 0.0704 0.0329 0.0347 0.0336 0.014 0.007 0.014 0.028 0.014 0.014 0.014 1.0 1.4 1.4 1.4 3.0 6.0 11.0

At 423 K 5 12 17 25 31 41 53 At 433 K 1 6 7 8 13 18 21 26 27 32 33 35 37 42 44 47 49 54 56 59

0.0337 0.0185 0.0348 0.0704 0.0329 0.0347 0.0336

0.014 0.007 0.014 0.028 0.014 0.014 0.014

1.0 1.4 1.4 1.4 3.0 6.0 11.0

Fig. 3. Exactness of fit for first-order alcohol/monoester (A/M) molar ratios: a = 3.0; Sr, c( = 6.0; and +T. bl = 11 .O.

kinetics I = k*C,. Initial 0, a = I; *, 1 = 1.4; 17,

0.0209 0.043 I 0.0648 0.1060 0.0245 0.0465 0.0676 0.0902 0.0199 0.0588 0.0641 0.0932 0.0214 0.0426 0.0654 0.0518 0.0216 0.0459 0.0731 0.0887

0.007 0.014 0.021 0.028 0.007 0.014 0.021 0.028 0.007 0.014 0.021 0.028 0.007 0.014 0.021 0.028 0.007 0.014 0.021 0.028

1.0 1.0 I .o 1.0 1.4 1.4 1.4 1.4 3.0 3.0 3.0 3.0 6.0 6.0 6.0 6.0 11.0 11.0 11.0 11.0

min

Fig. 4. Exactness alcohol/monoester

a = 3.U; *. a = 6.0;

of fit (A/M)
and

for first-order molar ratios:

P, z = 1 I .O

kinetics r =k*C,. initial LI, 1 = 1; *, x = 1.4; 0,

.I. Skrzypek

et al. / Chemical Engineering and Processing

Table Mean Run Run 2 9 14 22 28 34 38 45 50 51 2 values No. No.

33 (1994)

413-418

417

of catalyst

concentration-independent

rate constant

k = k*lC,,,
(dm3mol 0.823 0.860 0.795 0.841 0.882 0.797 0.731 0.824 0.799 0.792 min )

Mean value of k (dm3 mol I min-)

At 403K
3 10 15 19 23 29 39 51 1.13 1.24 1.40 1.14 I.12 1.18 I .oo 1.18

I
0.814 1.17 1.70 } 2.41

Fig. 5. Exactness of fit for first-order alcohol/monoester (A/M) molar ratios: a=3.O;t,a=6.0;andA.a=ll.O.

kinetics r = k *CM. Initial n, c( = 1; *, a = 1.4; 0,

At 413 K
4 11 16 24 30 40 43 46 48 52 55 58

1.62 1.95
1.97 1.68 1.68 1.57 I .47 1.58 1.72 1.63 1.69 1.75

At 423 K
5 12 17 25 31 41 53 Fig. 6. Arrhenius plots for several values of C,,,. 2.40 2.64 2.48 2.51 2.36 2.48 2.41

At 433 K
I 6 7 13 18 21 26 27 33 35 37 42 44 49 54 56 59 3.02 3.11 3.12 3.54 3.36 3.25 3.26 2.87 3.08 3.31 3.10 3.07 3.15 3.12 3.32 3.52 3.32

mean values of the catalyst concentration-independent rate constant k, whilst an Arrhenius plot for k is shown in Fig. 7. From the latter plot the activation may be calculated enabling the final as I1 300 cal mol- (47 300 J mol-I), expression for k to be written as: k = 1.66 x 106exp(-F) [dm3molm min- 1 (7) It should be noted that in eqn. (5) r = kC,C,,,,, with C H2SObeing constant during the course of the reaction. l

3.21

418

J. Skrzypek

et al. / Chemical Engineering and Processing 33 (1994) 413-418

Nomenclature

-0.50

acid number concentration, mol dm- first-order reaction rate constant, min- catalyst concentration-independent reaction constant, dm3 mol- min- total mass of reaction mixture, g molar mass of water, g mol - number of moles reaction rate, mol dme3 min- concentration, mol g-

rate

initial alcohol/monoester (A/M) molar ratio degree of conversion of mono-2-ethylhexyl phthalate reaction time, min-
Fig. 7. Arrhenius plot for k.

Subscripts

M cat
Mechanism of the reaction

mono-2-ethylhexyl catalyst

phthalate

Superscripts 0 initial T reaction References

temperature,

Although the object of the present study was not to elucidate the reaction mechanism, the kinetic results require at least cursory comment in relation to other mechanisms. In the presence of H,SO,, the reaction proceeds as a first-order process. Basically, such an order excludes some other mechanisms notably the &..,2 mechanism for nucleophilic substitution frequently encountered during esterification and hydrolysis processes since, in this case, the reaction should be second order in nature. On the basis of a first-order reaction, the nature of the process could be plausibly explained by assuming the existence of an intermediate carbonium ion which might be stabilized by a non-polar solvent (2-ethylhexyl alcohol) or alternatively stabilized by the structure of the molecule. However, such an assumption is risky. It should be noted that no detailed studies are described in the literature on the mechanism of esterification employing o-phthalic anhydride and higher alcohols, despite the fact that the formation and hydrolysis of esters is one of the reactions most often studied. On the basis of the literature, it seems that the first order observed in our studies could be satisfactorily explained by the action of an adjoining group [IO] or by the formation of a non-classic carbonium ion [ll]. Esterification of a neighbouring group could produce an anchimeric acceleration of those groups undergoing binary esterification. In that case the slowest reaction would be intramolecular rearrangement following protonation of the monoester. This would lead to the observation of first-order reaction kinetics despite the fact that, in reality, the reaction is bimolecular.

as Ill M. A. Schwegler and H. van Bekkum, Heteropolyacids catalysts for the production of phthalate diesrers, Appl. Cum/., 74 (1991) 191. r21 T. S. Thorat, V. M. Yadav and G. D. Yadav, Esterification of phthalic anhydride with 2-ethylhexanol by solid superacidic catalysts, Appl. Catal., 90 (1992) 13. [31 H. Suter, Phthalsiiurranhydrid und seine Verwendung, Mechanismus und Kinetic Verwendung, Dietrich Steinkopffverlag Edition, Darmstadt, 1972. [41 J. Nosovskii, S. Osinceva, A. Kucenko and A. Alejeva, Nekotoryje kineticzeskije zakonomiernosti etierifikacji ftalewowo angidrida 2-etylheksanolom w prisutswii kislych katalizatorow, Chim. Prom., 2 (1974) 108 (in Russian). Potencjometryczne oznaczanie zwiazk6w r51 H. Trzmielewska, kwaSnych we ftalanach dwualkilowych, Chem. Anal. (Warsaw), 19 (1974) 649 (in Polish). Potencjometryczne oznaczanie zwiqzkbw 161 H. Trzmielewska, kwaSnych we ftalanach dwualkilowych, II Kwas ftalowy, Chem. Anal. (Warsaw), 22 (1977) 357 (in Polish). 171 V. Dhanuka, V. Malske and S. Chandalia, Kinetics of the liquid phase esteritication of carboxylic acids with alcohols in the presence of acid catalyst: R&interpretation of published data, Chem. Eng. Sci., 32 ( 1977) 551. [gl Ch. Leyes and D. Othmer, Continuous esterification of butanol and acetic acid, kinetic and distillation considerations, Trans. Am. Inst. Chem. Eng., 41 (1945) 151. r91 I.Drimus, D. Morath and H. Glatt, Cinetica reactiei anhidridei ftalice cu 2-etilhexanol, Rev. Chim. Acad. Repub. Pop. Roum.. 28 (1977) 325 (in Roumanian). group participation, Q. Rev. Chem. [lOI A. Capon. Neighbouring sot., 18 (19643 76. IllI S. Cram, Phenonium ions as discrete intermediates in certain Wagner-Meerwein rearrangements, J. Am. Chem. Sot., 86 (1964) 3767.