JOURNAL OF CAT.

4LYSIS 51,96-100 (1978)

Mechanistic Study of Methanol Carbonylation Catalyzed by

an Iridium Complex in the Presence of Methyl Iodide

TOSHIO MATSUMOTO, TSUTOMU MIZOROKI, AND ATSUMU OZAKI

Research Laboratory of Resources Utilization, Tokyo Institute of Technology,

Ohokayama, Meguro-ku, Tokyo 16.9 Japan

Received April 15, 1977; revised September 16, 1977

Kinetics of methanol carbonylation in acetophenone solvent catalyzed by an iridium com-

plex in the presence of methyl iodide was found to be first order with respect to methanol,
under reaction conditions of apparent zero order with respect to both carbon monoxide and
methyl iodide, where carbon monoxide pressures higher than 15 kg/cm2 and a CHJ/Ir ratio
higher than 20 were adopted at 125-175°C. Acetyl iridium(II1) complexes were isolated from
product solutions by stabilizing with triphenylphosphine. Both the observed kinetics and the
formation of acetyl iridium(II1) complexes are consistent with the rate-determining methanoly-
sis of an acetyl iridium(II1) species in the catalytic cycle.

INTRODUCTION elucidate the mechanism of carbonylation

Rhodium compounds have been reported with iridium catalyst.
to be highly active and selective, in the
EXPERIMENTAL
presence of methyl iodide, for the catalytic
carbonylation of methanol to acetic acid Reaction procedure. All runs were carried
(1, S). On the basis of reaction kinetics, it out using IrCb-Hz0 as the catalyst pre-
has been claimed that the rate-determining cursor in acetophenone solvent. Acetophe-
step is the oxidative addition of methyl none has been found to be effective in
iodide to a rhodium(I) complex (3, 4). preventing the formation of dimethyl ether
Recently Forster elaborately showed that in rhodium-catalyzed methanol carbonyla-
the active species is an anionic rhodium(I) tion (6). Kinetic measurements were per-
complex, [Rh (CO) JJ- (6). Iridium com- formed in a Ti-Mn alloy autoclave (100
pounds are reportedly active for the cata- ml) equipped with a magnetic stirrer by
lytic carbonylation of methanol (I), while following consumption of carbon monoxide
no mechanistic study has been reported. as a function of time. A given amount of
We examined the kinetics of methanol IrC14. Hz0 (0.06425 mmol) and a mixture
carbonylation by iridium catalyst, to com- of methyl iodide (0.75-20 mmol), methanol
pare with that by rhodium catalyst under (50-250 mmol), and acetophenone (20 ml)
similar conditions, and found that the were placed in the autoclave, and the mix-
kinetics are significantly different from each ture was heated up to a reaction tempera-
other in nature, suggesting different rate- ture (125-175°C). Carbonylation was ini-
determining steps. Thus the isolation of an tiated by introducing carbon monoxide to
intermediate complex was attempted to a desired pressure (7-50 kg/cm2), which

96
0021-9517/78/0511-0096$02.00/O
Copyright 0 1978 by Academic Press. Inc.
All rights of reproduction in any form reserved.
 MECHANISTIC STUDY OF METHANOL CARBONYLATION 97

TABLE 1
Analytical Results for Carbonylation Productsa
Reaction conditions Carbonylation products Calculated results (%)
(mmol)
Temp. IrClr.HzO CHtOH C0llV.b Ma%3 co
PC) (mmol) (mmol) AcOH AcOCHa CHaOH CHzI balance’ balanced

125 0.06 250 2.9 28 190 8.1 12 99 107

140 0.06 250 4.0 42 148 7.7 18 94 106
156 0.06 250 6.5 60 104 7.4 26 92 115
173 0.06 250 7.0 83 50 7.3 36 90 112
173 0.25 250 126 39 2.5 7.5 66 87 93
173 0.25 125 55 28 Trace 8.2 66 90 94
173 0.25 85 37 22 1.0 8.5 69 96 95
173 0.25 50 19 12 1.5 7.5 G3 96 105

a Conditions: CHsI. 10 mmol; ChHnCOCHa. 20 ml ; pC0, 30 kg/cm* (at reaction temp.) ; time, 2 hr.
6 ([AcOH] + [AcOCHsll/[CHsOHli,it. X 100 (AC = CHaCO).
c ([AcOH] + 2[AcOCHal+ [CHsOHl)/[CHaOHli.it. X 100.
d ([AcOHl + CAcOCHd l/CCOl.ba. X 100.

was kept constant during the run by duced up to 30 kg/cm2, and both the tem-
supplying carbon monoxide from a high- perature and pressure were kept constant
pressure gas reservoir (100 ml). The product for 20 min, during which period about 20
solutions were analyzed by gas chroma- mmol of carbon monoxide was consumed.
tography for acetic acid, methyl acetate, The autoclave was rapidly (in a few min-
methanol, and methyl iodide. utes) cooled to 0°C and carbon monoxide
Isolation of acetyl iridium(III) complex. was purged out. The orange-yellow product
IrC14. Hz0 (0.5 mmol) and a mixture of solution was transferred into a flask (100
methyl iodide (5 mmol), methanol (125 ml) containing a mixture of triphenyl-
mmol), and acetophenone (10 ml) were phosphine (5 mmol), methanol (20 ml),
placed in the autoclave, and it was heated and acetophenone (5 ml) under a nitrogen
up to 140°C. Carbon monoxide was intro- atmosphere. The solution was stirred at
25% for 1 hr, and then cooled to O”C, a

r yellow crystal being precipitated. The pre-

cipitate was washed with methanol, and
recrystallized from dichloromethane and
methanol.

RESULTS AND DISCUSSION

O/
i//____ Kinetics
Dependence of the carbonylation rate on

.30
($5-6 the amount of methanol was examined in
acetophenone at 173°C under 30 kg/cm2
pressure. Carbonylation proceeded without
induction time, suggesting a rapid forma-
0 tion of active species. Analytical results for
0 30 60
products are given in Table 1. As is the
case with rhodium catalysts, the initial
FIG. 1. Time courses of CO absorption at various
product in the presence of large amounts
amounts of CHaOH. CHaOH (mmol) : l ,50; @,85;
0, 125; 0, 250. CHJ, 10 mmol; IrCl~.H20, 0.25 of methanol is methyl acetate, while pro-
mmol; PhCOCH*, 20 ml; $0, 30 kg/cmz; 173°C. duction of acetic acid increases with the
98 MATSUMOTO, MIZOROKI, AND OZAKI

extent of conversion as the product water order rate constant, 2.8 hr-‘, is obtained
accumulates in the reaction mixture. The from the slope.
selectivity of carbonylation to form acetic In this way, the first-order rate con-
acid or methyl acetate is close to 100% so stants were obtained under different reac-
that the rate of carbonylation can be ob- tion conditions, and are summarized in
tained from consumption of carbon monox- Table 2. It is obvious that the rate con-
ide, as shown in Fig. 1, as a function of stant is proportional to the amount of
reaction time. iridium catalyst, as expected, while inde-
It is obvious that the initial rate is pro- pendent of carbon monoxide pressure above
portional to the initial amount of methanol. 15 kg/cm2 and also independent of the
In order to confirm the proportionality to amount of methyl iodide above 5 mmol
methanol during runs, the amount of (CH31/Ir > 20), in marked contrast to the
methanol remaining at time t, [MeOH],, result with rhodium catalysts (6) that the
is estimated by the relation, [MeOHJ = rate of carbonylation linearly increases
[MeOH& - 2[CO] $, where [MeOHIo and with a CH31/Ir ratio of 20 to 160. It is
[CO], are the amounts of methanol at noteworthy that first-order dependence on
time aero and the carbon monoxide con- methanol is invariably found for runs with
sumed up to time t, respectively. This rela- lower methyl iodide concentration or carbon
tion is valid for the initial period where monoxide pressure, again in marked con-
methyl acetate is the predominant product., trast with the result with rhodium cata-
in conformity with the equation lysts (6) that the rate of carbonylation is
2MeOH + CO ---) MeCOOMe + HZO. independent of methanol concentration.
The activation energy of the carbonylation
The first-order plot with respect to the is obtained from the rate constants given
estimated amount of methanol gives a in Table 2 to be 8.9 kcal/mol, which is con-
straight line, at least for the initial 0.5~hr siderably lower than that obtained previ-
period, as shown in Fig. 2. Thus a first- ously with a rhodium catalyst in the same
solvent, 19 kcal/mol (6). Thus the rate-
0.6 determining step of carbonylation with the
iridium catalyst is likely different from that
i0 with the rhodium catalysts. The first-order
kinetics in methanol as well as the zero-
order kinetics in both carbon monoxide and
0.4 - P methyl iodide strongly suggested that
methanol is involved as a reactant in the
rate-determining step of carbonylation with
the iridium catalyst.

Relevance to the Mechanism and Character-

ization of the Iridium Complex Isolated
In the accepted scheme of methanol
carbonylation with rhodium catalysts (3,
5), methyl iodide is oxidatively added to
0 15 30 rhodium(I) complex to give methyl rho-
Time(min) dium(II1) complex, which is believed to be
FIQ. 2. Fir&order plot of the rate. CH,OH rapidly transformed into acetyl rhodium-
(mmol): .,50; a, 85; 0, 125; 0,250. (III) complex. Two competitive reaction
 MECHANISTIC STUDY OF METHANOL CARBONYLATION 99

TABLE 2

Apparent First-Order Rate Constants, d[CO]/dt = d[AcOCHt]/dt = kl[CHaOH],

in Acetophenone (20 ml)

Temp. CHlOH CHJ I&la. Hz0 pC0 at

F.3 (mmol) (mmol) (mmol) reaction (h:$
temperature”
(k/cm*)

173 125 10 0.08 30 0.62

173 125 10 0.13 30 1.3
173 125 10 0.25 30 2.8
173 125 20 0.25 30 2.8
173 125 10 0.25 30 2.8
173 125 5 0.25 30 2.8
173 125 2.5 0.25 30 1.9
173 125 1.3 0.25 30 1.4
173 125 0.75 0.25 30 0.75
173 125 10 0.25 50 2.8
173 125 10 0.25 30 2.8
173 125 10 0.25 15 2.6
173 125 10 0.25 11 1.5
173 125 10 0.25 7 0.54
173 250 10 0.06 30 0.42
156 250 10 0.06 30 0.27
140 250 10 0.06 30 0.19
125 250 10 0.06 30 0.12

a Pressures of carbon monoxide at the reaction temperature.

paths have been suggested for the fate of the iridium complex from the reaction
acetyl rhodium(II1) complex, i.e., (1) re- mixture.
ductive elimination of acetyl iodide fol- The yellow iridium complex (Calcd for
lowed by methanolysis and (2) direct C39H3302P2121r:C, 45.0; H, 3.19; I, 24.4.
methanolysis of acetyl rhodium (III) com- Found : C, 45.4 ; H, 3.33 ; I, 24. I%‘,)! which
plex. If we apply the above reaction scheme was isolated from the carbonylation product
to carbonylation with the iridium catalyst, by stabilizing with triphenylphosphine as
the most probable rate-determining step described in Experimental, was identified
would be the methanolysis of acetyl iridium to be a mixture of stereoisomers of acetyl
complex to give methyl acetate and a iridium(II1) complex, CH3COIr12(CO)
hydrid complex. In this case, the large (PPhs)z. The IR spectrum of the complexes
part of iridium, during catalysis in the with KBr showed two strong and broad
presence of an excess amount of methyl bands, 2020 and 1620 cm-‘, in the carbonyl
iodide, should be in the oxidation state of stretching region, both of which consisted
Ir (III), with the oxidat,ive addition of of several neighboring bands. The bands
methyl iodide being very fast in conformity at 2020 and 1620 cm-’ can be reasonably
with the zero-order kinetics in methyl assigned t.o terminal carbonyls and acetyl
iodide. The positive order observed with frequencies, respectively, because an analo-
lower CHJ/Ir ratios may be understood gous acetyl iridium(II1) complex prepared
as a thermodynamic control of the amount from the oxidative addition of acetyl irid-
of Ir(II1) complexes. In view of the above ium(II1) complex prepared from the oxida-
consideration, it was attempted to isolate tive addition of acetyl bromide to iridium(I)
100 MATSUMOTO, MIZOROKI, AND OZAKI

complex, CHaCOIrBrz (CO) (PEtzPh) 2, was served kinetics. But if an iridium(I) com-
reported to have two strong bands at 2036 plex is the predominant species, with the
and 1633 cm-’ (7). ‘H NMR of the iridium oxidative addition of methyl iodide being
complexes in CD&% solution showed sev- rate-determining, the kinetics should be
eral peaks (8.8-9.47) due to methyl pro- first order in methyl iodide, as is the case
tons, indicating a mixture of stereoisomers. with the rhodium catalyst (6). Thus it may
The mass spectrum of the iridium complex be concluded that the oxidation state of
taken by the field desorption method’ the iridium catalyst is Ir(II1) during car-
showed three strong fragment peaks at bonylation, with the acethyl iridium(II1)
869-872( f l), 779-782 (f l), and 262, whose complex being a plausible form (8).2
mass numbers corresponded to CH&OIr12-
(CO) (PPh3) (870, 872), IrI (CO) (PPh3) 2 REFERENCES
(778, 780) (lslIr, 38.5%; lg31r, 61.5%), and
1. Pa&k, F. E., and Roth, J. F., Chem. Commun.,
PPh3(262), respectively, although no parent 1578 (1968).
peak (1040, 1042) was observed. All these 2. Schultz, R. G., and Montgomery, P. D., J. Catal.
spectroscopic data are reasonably explained 13, 105 (1969).
as a mixture of stereoisomers of acetyl irid- 3. Roth, J. F., Craddock, J. H., Her&man, A., and
Paulik, F. E., Chemtech., 600 (1971).
ium complexes, CH$OIrL(CO) (PPh3)z.
The isolation of acetyl iridium(II1) com- 4. Robinson, K. K., Her&man, A., Craddock, J. H.,
and Roth, J. F., J. Catal. 27, 389 (1972).
plex from the reaction mixture is reason- 6. For&r, D., J. Amer. Chem. Sot. 98, 846 (1976).
able if the methanolysis of this complex is 6. Matsumoto, T., Mori, K., Mizoroki, T., and
rate-determining in the catalytic cycle, a.s Ozaki, A., Bull. Chem. Sot. Japan 50 (9), 2337
(1977).
discussed above. However, there is a pos-
7. Chatt, J., Johnson, N. P., and Shaw, B. L., J.
sibility that the intermediate complex has Chem. Sot. A, 604 (1967).
undergone a transformation under the isola- 8. Brodzki, B., Denis, B., and Pannetier, G., J.
tion conditions, rendering the isolated com- Mol. Cat&. 2, 149 (1977).
plex to be one of the possible forms of the 2 After the completion of this report, D. Brodzki
intermediate. For example, CHJrL (CO) 3 et al. reported the isolation of anionic Ir(II1)
also is a possible form, since it can be readily complexes such as [PPha]+[Ir(C0)214]- and
transformed into the acetyl complex during [CH~PPh~]+[Ir(CO)&]- from the carbonylation
products of methanol catalyzed by iridium(I) com-
isolation and is consistent with the ob-
plexes using acetic acid or propionic acid as the
1 Taken at the Japan Electron Optics Laboratory solvent (CH,I/Ir(I) ~200, pC0 -60 kg/cm2,
with JMS-OISG-2. 19OOC) (8).