Ind. Eng. Chem. Res. 1987,26, 1513-1518
1513
Carbonylation of 1,4-Butanediol Diacetate Using Rhodium Complex Catalyst: A Kinetic Study? Subodh B. Dake, Raghuraj V. Gholap, and Raghunath V. Chaudhari* Chemical Engineering Division, National Chemical Laboratory, Pune 41 1 008, India
Kinetics of carbonylation of 1,4-butanediol diacetate (BDDA) in the presence of Rh complex catalyst and HI promotor has been reported. T h e effect of catalyst, HI, and BDDA concentration, carbon monoxide partial pressure, and temperature on the concentration vs. time data in a batch reactor was studied. A batch reactor model has been developed, using the rate parameters of the two consecutive reaction steps that were evaluated. The reaction was found to be zero order with respect t o CO partial pressure and nearly first order with respect to catalyst and HI concentrations for both of the stew. T h e activation energies for the two steps were found t o be 91.36 X lo3 and 88.1 X lo3 k J / kmo1,-respectively. I
It is well-known that carbonylation of alcohols and esters gives corresponding carboxylic acids as products. This route can also be extended for the carbonylation of diols and diesters to produce the corresponding dicarboxylic acids. Carbonylation of 1,4-butanediol diacetate is an important reaction since it can be useful in the synthesis of adipic acid, which is an important intermediate for several end products. Adipic acid is a key intermediate in the manufacture of Nylon 66, and with the increasing demand of Nylon 66, several new routes for its manufacture are being investigated. The existing process for the manufacture of adipic acid involves two-step oxidation of cyclohexane (Danly and Campbell, 1978). In the fiist step, cyclohexane is oxidized to a mixture of cyclohexanone and cyclohexanol (K-A mixture), while in the second step, this mixture is oxidized to adipic acid by using nitric acid as an oxidizing agent. A new route via carbonylation of l,&butadiene has also been reported (Weiss, 1979; Kesling and Zehner, 1980; Kutepow, 1975). In these reports, it is mentioned that carboalkoxylation or alkoxycarbonylation of l,&butadiene using CO and waterJalcoho1 gives adipic acid/ester as a product. Cobalt, palladium, and rhodium are reported to be effective catalysts for this process. Recently, carbonylation of 1,4-butanediolor its esters using homogeneous catalysts has also been reported (Monsanto Co., 1972; Nakamura, 1979; Nakamura and Sado, 1978; Yoneda et al., 1978; Sad0 and Tajima, 1980). The aim of this work was t o study the kinetics of carbonylation of 1,Cbutanediol diacetate (BDDA) using homogeneous rhodium complex catalyst. The reaction of BDDA wit+ CO, catalyzed by Rh complex and HI as a promoter, can be described by the following stoichiometry:
!!
!!
I1
11
catalyst
CH~CO(CHZLOCCH~ CO. H z 0 A
9
-
II
catalyst
CH~CO(CHZ)~COOH
co.
H20
0
COOH(CHz)4COOH
+
CH3COOH ( 2 )
This is a typical case of a consecutive gas-liquid catalytic reaction and therefore an interesting example for kinetic
* To whom correspondence should be addressed. +NCL Communication No. 4119.
0888-5885/87/2626-1513$01.50/0
study. Effect of catalyst (RhC13.3H20),promoter (aqueous HI) and BDDA concentrations, partial pressure of CO, and temperature on the rate or carbonylation was investigated. The products formed during the carbonylation reaction were characterized in a few experiments. From the observed data on concentration of BDDA and CO consumed as a function of time, the kinetic parameters were evaluated by using a batch reactor model.
Experimental Section Materials. 1,4-Butanediol diacetate (BDDA) was synthesized from LR grade 1,4-butanediol (Fluka make, Switzerland) and acetic acid (BDH Laboratory, Bombay) by esterification. The purity of BDDA obtained was found to be >99% by GC. Other materials were of the same specifications as described elsewhere (Dake et al., 1984). Analysis. 1,4-Butanediol diacetate (BDDA) and other products were analyzed by using a H P 5840 A gas chromatograph equipped with a glass column 2.0 m long and packed with Chromosorb material impregnated with 1.0% H,PO,. Carrier gas flow rate = 3.3 x io-’ m3/s, oven temperature = 433 K, injection temperature = 523 K, and FID temperature = 573 K were the other GC conditions. Products B and C are high boiling products and hence cannot be analyzed by GC. For characterization of these products, the following procedure was followed. A sample of liquid (5 mL) was taken into a round-bottom flask, 20 mL of ethanol was added to it, and the contents were refluxed for about 2 h. Ethyl acetate and other volatile low boilers were then removed by conventional distillation, until about 5 mL of the liquid was left. The same procedure was repeated twice, each time adding a fresh lot of 20 mL of ethanol and treating it similarly. The sample thus obtained was then kept under vacuum for about 2 h, diluted with ethanol to about 5 mL, and then analyzed by GC. The same sample was then analyzed by GC / MS technique. Procedure for Kinetic Study. In a typical run for kinetic study, known quantities of catalyst (RhC1,.3H20), HI, and BDDA were charged to the reactor along with 5 mL of water and acetic acid as the solvent. The total volume of solution was 100 cm3 in all experiments. The reactor used was a Parr autoclave of 0.3-L capacity, provided with automatic temperature control, four-bladed propeller-type stirrer, a cooling coil, and liquid sampling arrangement. The reactor was then flushed with CO and heated to a desired temperature. A liquid sample was taken at this stage for analysis of BDDA. Reador was then pressurized with CO up to a desired pressure, and reaction started by switching the stirrer on. The reaction was carried out at a constant pressure by supplying CO through 0 1987 American Chemical Society
1514 Ind. Eng. Chem. Res., Vol. 26, No. 8, 1987 Table I. Range of Operating Conditions parameter conditions 1.91 X 1OW3-7.64X catalyst concn, kmol/m3 HI concn, kmol/m3 0.076-0.228 water concn, kmol/m3 3-3.5 1.5 X 103-3.5 X lo3 partial pressure of CO, kPa temp, K 463-488
z w I-
I
a reservior vessel using a constant pressure regulator. The progress of the reaction was followed by observation of the amount of CO consumed and change in the concentration of BDDA in the liquid at different time intervals. For this purpose, pressure drop in the reservior vessel was continuously recorded, and also liquid samples were withdrawn at different time intervals for analysis of BDDA. By use of this procedure, several experiments were carried out a t different catalyst and HI concentrations, partial pressures of CO, and temperatures. The ranges of conditions investigated are given in Table I.
Figure 1. Qualitative picture of the product distribution at two conversion levels (GC analysis).
1
Results and Discussion Product Identification. The various side reactions and intermediate steps likely to occur during the carbonylation of BDDA in the presence of rhodium complex catalyst can be described as CO BDDA
+ H20
+
HI
+
2HI
catalyst
- I + -
COz
+
Hz
(3)
+
C H ~ C O O ( C H Z ) ~ I CH3COOH
A
BDDA
-
I(CH2l4I
+
2CH3COOH
(4)
m/e
Figure 2. Mass spectrum of GC peak i.
(5)
1
1
3
7
101
73
I01
I1
CH~COO(CHZ)~COOH H I
I(CH2)4COOH
+
CH3COOH (6)
+
CH3COOH ( 7 )
111 U
-
CH~COO(CHZ)~COOH
II
c-
I
HzC,
0 \
YHz C-CH2 Hz
From the above list, water gas shift reaction (eq 3) was not found to occur under the conditions used, as no COz formation was noticed in the analysis of the gas phase, in some cases. Similarly, no trace of the 6-valerolactone (eq 7) was detected in the liquid sample, thus leaving mainly B and C, the mono- and diacids, respectively, as the major products. The reaction obviously proceeds through intermediate iodo compounds (I, 11, and 111);however, these were formed in very small quantities proportional to the concentration of the promoter (HI). Reaction 5 is likely only a t higher HI concentrations. In general, the carbonylation of BDDA was found to proceed at 463 K and 1.5 X lo3kPa pressure in the presence of Rh complex catalyst. In a few experiments, the liquid samples were withdrawn a t different conversion levels of BDDA for identification of products. These samples were analyzed by the procedure described earlier. It was observed that a t intermediate level of conversion (15-20%), the products were found to contain products of esterification of both monoand diacids. On the other hand, at higher conversion level (90%),of BDDA, most of the product consisted of the ester of adipic acid. The qualitative picture of the product distribution at two different conversions is shown in Figure 1.
Fragmentation pattern of GC peak i is presented in Figure 2. The molecular ion peak at m l e 100, a ( m + 1) peak at m l e 101, and another splitting pattern confirm that GC peak i corresponds to 3-valerolactone. Formation
2
1
,
A i ' l 'SO I20
,
11
110
- . 73
101
111,
' ,
I "
100
€0
90
14' 70
111
€0
50
m /e
Figure 3. Mass spectrum of GC peak ii.
of 6-valerolactonejustifies the formation of intermediate acid, CH3COO(CHJ4COOH,which on self-esterification gives 6-valerolactone by the reaction U
II
I
H2/0C-
c\H2
CH,-Ct ,CHz C2HsOH H-0%~- C H ~
I
-
II
0
H~F-CHZ \
FHz
+ CH3COOC2H5
rcHz
(8)
0
The MS pattern of GC peak ii is presented in Figure 3. The molecular ion peak ( m l e 202), which is insignificant, undergoes consecutive loss of two ethanol molecules, thereby producing significant peaks at m l e 156 and 110, respectively. The cleavage to the carbonyl function produces a strong peak a t m l e 73. Another splitting pattern is also found to be consistent with diethyl ester of adipic acid. GC peak iii was found to correspond to 1,4-butanediol. Thus, it was established that the only products of carbonylation reaction are CH3COO(CH2)4COOH and COOH(CH2)&OOH in the range of conditions studied. The overall reaction can thus be simplified as A+CO-B
catalyst catalyst
B+CO-C
(10)
Ind. Eng. Chem. Res., Vol. 26, No. 8, 1987 1515
08t
li
.
I
i
Kinetic Study. In the present work, the batch reactor data were directly used for evaluation of the kinetic parameters. Effect of catalyst, HI, and BDDA concentration and CO partial pressure and temperature on the concentration of BDDA and CO consumed VI. time was observed. All experiments were carried out in such a way that CO pressure was kept constant throughout the experiment. The only variables in a given run were the concentrations of BDDA and the products. For the purpose of deriving rate equations, the following reaction scheme was considered:
oV 2000
1000
1500
WAVE
N U M B E R IO)
Figure 4. Infrared spectrum of the reaction product. Catalyst concentration-7.64
"16' kmol/m3
H I c o n c e n t r a t i o n - 0 , 0 7 6 kmol/m3
,#'
Temperature-488 K P a r t i a l pressure of CO - 3 , 4 4 5 XlO' kPa
,/' I
,A'
1.0
Y
40.5
!-
2
I
4 TIME ,hrs
I
6
8
cataIySt.H20
HOOC(CH2)4COOH
(14)
C
0
"
4.
I
R,, ¶ l o w
' I 1 co +
/
0 0
0
0
500
0
Figure 5. Typical plot of BDDA and CO consumed as a function of time.
A general observation made in most experiments was that if experiments were carried out until CO absorption stopped, it was possible to completely convert BDDA as well as the intermediate, B, to adipic acid. In some experiments carried out until complete conversion, a solid product was isolated, which on purification and analysis was found to be adipic acid. For example, in a typical experiment carried out a t 488 K, catalyst concentration of 3.82 X kmol/m3, HI concentration of 0.228 kmol/m3, BDDA concentration of 1.1kmol/m3, and partial pressure of CO of 3445 P a , the amount of CO absorbed was 0.208 mol, with 0.107 mol of BDDA consumed. Adipic acid formed was 0.105 mol. This indicates that all BDDA is converted to adipic acid. The elemental analysis and melting point are C = 49.1%, H = 6.9570, and mp = 151 OC, which agree well with the theoretical values (C = 49.3%, H = 6.85%, mp = 152 "C). Also, IR spectrum of the solid product is shown in Figure 4 which shows IR frequencies a t v = 1695 cm-l, confirming the presence of the carboxylic acid group. Thus, all the analysis supports the formation of adipic acid. A typical plot of concentration of BDDA and CO consumed as a function of time is shown in Figure 5. In most experiments, it was observed that a t a stage where CO absorption stopped, the amount of CO consumed was twice that of BDDA consumed, indicating complete conversion to adipic acid. In the initial stages (