Spectroscopic Studies of Working Carbonylation Reactions

We are now in a position, having considered the various model studies on individual reaction steps, to look more closely at what is observed in working catalytic MeOH and MeOAc carbonylation reactions. In some ways the anhydrous carbonylation of MeOAc to AC2O is easier to describe because, in the absence of HjO, there is little or no water gas shift chemistry competing with the main carbonylation cyde. 

In a later publication, workers at Eastman demonstrated more clearly the role of reductants in maintaining catalyst activity in MeOAc carbonyiation to AC2O. They also used a HP IR transmission cell external to an autoclave to show that a portion of the Rh was present as inactive [Rh(CO)2l4] in batch reactions and that by adding H2 to the reaction it was converted to [Rh(CO)2l2], thus increasing the catalytic activity, (Eq. (47)) [5]. 

As already described, commercially useful rates in Rh catalysed MeOAc carbonyiation to AC2O can only be achieved in the presence of substantial amounts of F. The kinetic studies of MeOAc carbonyiation to AC2O coupled with the various spectroscopic observations and model reactions show that overall the observed carbonyiation rate is controlled by the kinetics of formation and reaction of Acl as indicated in the reaction sequence below, (Eq. (48)), (Eq. (21)) and (Eq. (12)). 

The Eastman group clearly described how this was due to promotion of reaction of intermediate Acl with MeOAc, (Eq. (12)). The HP NMR experiments described above show a good correlation between activity in Mel/MeOAc exchange and promotional activity in catalytic carbonylation. At the same time, the model studies of r promotion of oxidative addition suggest that the formation of Acl may also be accelerated under process conditions (Eq. (48)). Thus, F salts can contribute to accelerating both steps that can be rate controlling in the carbonylation of MeOAc to AC2O. 

Finally, we have also seen from the studies of Acl chemistry and its reaction with Rh complexes that, at all times in working AC2O processes, Acl will be present and available to reform Rh(C(0)Me)-species. This can explain the HP NMR observation that formation of EDA by hydrogenation of AC2O increases as batch reactions proceed. It is toward the completion of carbonylation of MeOAc to AC2O, or the approach to equilibrium to be more precise, that [Acl] will be greatest and thus also presumably the Rh(C(0)Me)-species, which are hydrogenated to EDA. Eor these reasons all the reaction steps are shown as reversible equilibria. 

---