The Rhodium Catalyst System

This preferential formation of 1 1 adduct to form 1,4-hexadiene in a mixture of ethylene and butadiene was further studied by Cramer (4). He concluded that the results appeared to be the consequence of thermodynamic control reactions through a relatively stable 7r-crotyl Rh complex. 

The most general and comprehensive reaction mechanism of the 1 1 codimerization has been reported by Cramer (4). The results were based on reaction properties measured in an alcoholic medium under relatively mild conditions. 

Cramer has isolated this 77-crotyl complex in its stable dimeric form (1) by interacting butadiene with an ethyl-Rh111 chloride (4)  

The rate equation for the dimerization of ethylene (5) can be used to describe the codimerization in the presence of large excesses of butadiene. The rate of the addition reaction as measured by the disappearance of ethylene is represented in Eq. (5). It is first order in ethylene, proton, chloride, and rhodium. 

Effect of Butadiene Concentration on the Rate of Addition of Ethylene to Butadiene at 50°C  

---

In 1968 Wilkinson discovered that phosphine-modified rhodium complexes display a significantly higher activity and chemoselectivity compared to the first generation cobalt catalyst [29]. Since this time ligand modification of the rhodium catalyst system has been the method of choice in order to influence catalyst activity and selectivity [10]. 

The acetic anhydride process employs a homogeneous rhodium catalyst system for reaction of carbon monoxide with methyl acetate (36). The plant has capacity to coproduce approximately 545,000 t/yr of acetic anhydride, and 150,000 t/yr of acetic acid. One of the many challenges faced in operation of this plant is recovery of the expensive rhodium metal catalyst. Without a high recovery of the catalyst metal, the process would be uneconomical to operate. 

The major advantage of the use of two-phase catalysis is the easy separation of the catalyst and product phases. FFowever, the co-miscibility of the product and catalyst phases can be problematic. An example is given by the biphasic aqueous hydro-formylation of ethene to propanal. Firstly, the propanal formed contains water, which has to be removed by distillation. This is difficult, due to formation of azeotropic mixtures. Secondly, a significant proportion of the rhodium catalyst is extracted from the reactor with the products, which prevents its efficient recovery. Nevertheless, the reaction of ethene itself in the water-based Rh-TPPTS system is fast. It is the high solubility of water in the propanal that prevents the application of the aqueous biphasic process [5]. 

One approach that has been successfully used to separate the catalyst from the product aldehyde is to use a biphasic system in which the rhodium catalyst is soluble in water and the product is soluble in an organic phase. This approach is used by Hoechstdlhone-Poulenc to produce more than 600,000 t/year of butyraldehyde (a lower aldehyde) (2). Unfortunately, this process caimot be used to produce higher aldehydes because the water solubihty of the higher olefins that are the feedstock is very low, which dramatically reduces the reaction rate. 

Despite the very attractive properties of the rhodium-based system, no commercial plants used it because the low stability of the catalyst meant that the catalyst separation problem prevented commercialisation. Very recently, this situation has changed with the introduction of rhodium-based plant by Sasol in South Africa which uses technology developed by Kvaemer Process Technology (now Davy Process Technology). This batch continuous plant produces medium-long chain aldehydes and the separation is carriedoutbylow pressure distillation [16-18]... 

Aldehyde dimer may undergo dehydration to give an a, -unsaturated carbonyl. From butanal, the conjugated carbonyl is ethylpropylacrolein (Equation 2.10). The conjugated system of this material competes for coordination sites on the rhodium catalyst so that hydroformylation inhibition is observed.[8] The formation of 2-ethylhex-2-enal can be limited by minimizing the concentration of dimers. Dimers are removed along with the product in a liquid recycle separation system. 

The compound of the distinct three oxo processes, all rhodium-based, enables a highly efficient recovery system to be achieved. Figure 5.17 describes the TPPTS manufacture and its use for the preparation of the rhodium catalyst, using either freshly introduced Rh acetate or recycled Rh 2-ethylhexanoate. The recycle technique of the RCH/RP process and its performance is depicted earlier. Spent Rh-TPPTS solutions are worked-up (see Figures 5.18 and 5.19), the resulting TPPTS returns to the RCH/RP process. The rhodium portion passes also a work-up stage and is reformulated as Rh 2-ethylhexanoate. This Rh salt may serve all various oxo processes of the oxo loop and will compensate for possible Rh losses as mentioned earlier. 

The reaction was not influenced by the type of ionic liquid, since no significant differences were observed when the reaction was carried out in [BMIM][PF6] or [BMIMJtn-CgHnOSOa]. In Table 7.5 the most relevant results for the hydroformylation of propene with the different rhodium catalyst systems in the ionic liquid [BMIM][PF6] are compiled. 

With reference to the homogeneous catalyst systems thus far reported for the synthesis of hydrocarbons/chemicals from carbon monoxide and hydrogen, only the anionic rhodium systems of Union Carbide show any appreciable shift activity. With neutral species of the type M3(CO)12 (M = Ru or Os), only small quantities of carbon dioxide are produced under the synthesis conditions (57). 

Quite different types of rhodium compound can give very similar reaction rates in a system which shows a kinetic dependence on the rhodium catalyst concentration. In particular, rhodium(III) halides and rho-dium(I) phosphine complexes give almost identical reaction rates after an initial induction period. Thus, in the case of these two systems, it appears that a common species is being formed. 

Rhodium (I) complexes of chiral phosphines have been the archetypical catalysts for the hydrocarbonylation of 1-alkenes, with platinum complexes such as (61) making an impact also in the early 1990s[1461. More recently, rhodium(I)-chiral bisphosphites and phosphine phosphinites have been investigated. Quite remarkable results have been obtained with Rh(I)-BINAPHOS (62), with excellent ee s being obtained for aldehydes derived for a wide variety of substrates1 471. For example, hydroformylation of styrene gave a high yield of (R)-2-phenylpropanal (94% ee). The same catalyst system promoted the conversion of Z-but-2-ene into (5)-2-methylbutanal (82% ee). 

It is now nearly 40 years since the introduction by Monsanto of a rhodium-catalysed process for the production of acetic acid by carbonylation of methanol [1]. The so-called Monsanto process became the dominant method for manufacture of acetic acid and is one of the most successful examples of the commercial application of homogeneous catalysis. The rhodium-catalysed process was preceded by a cobalt-based system developed by BASF [2,3], which suffered from significantly lower selectivity and the necessity for much harsher conditions of temperature and pressure. Although the rhodium-catalysed system has much better activity and selectivity, the search has continued in recent years for new catalysts which improve efficiency even further. The strategies employed have involved either modifications to the rhodium-based system or the replacement of rhodium by another metal, in particular iridium. This chapter will describe some of the important recent advances in both rhodium- and iridium-catalysed methanol carbonylation. Particular emphasis will be placed on the fundamental organometallic chemistry and mechanistic understanding of these processes. 

An alternative strategy for catalyst immobilisation uses ion-pair interactions between ionic catalyst complexes and polymeric ion exchange resins. Since all the rhodium complexes in the catalytic methanol carbonylation cycle are anionic, this is an attractive candidate for ionic attachment. In 1981, Drago et al. described the effective immobilisation of the rhodium catalyst on polymeric supports based on methylated polyvinylpyridines [48]. The activity was reported to be equal to the homogeneous system at 120 °C with minimal leaching of the supported catalyst. The ionically bound complex [Rh(CO)2l2] was identified by infrared spectroscopic analysis of the impregnated resin. 

To elucidate the use of TMS systems for the isomerizing hydroformylation, PC was chosen as the solvent for the rhodium catalyst, because the best selectivity to n-nonanal of 95% with a conversion on trans-4-octene of also 95% was achieved in this solvent under single-phase conditions. Dodecane was used as a non-polar solvent for the extraction of the product and p-xylene served as the mediator between the catalyst and the product phase [24]. Appropriate operation points for the reaction within this solvent system were determined by cloud titrations. 

However, the TMS-system PC/dodecane/p-xylene has still some severe limitations. Via ICP-investigations a strong rhodium leaching of 47% of the rhodium catalyst was detected. Furthermore, we observed a correlation between the amount of the mediator p-xylene and the amoimt of leaching. The more p-xylene used, the more rhodium is transferred into the non-polar do-decane phase. Therefore, catalyst recycling in these systems is impossible at the moment. 

Oligomerization and polymerization of terminal alkynes may provide materials with interesting conductivity and (nonlinear) optical properties. Phenylacetylene and 4-ethynyltoluene were polymerized in water/methanol homogeneous solutions and in water/chloroform biphasic systems using [RhCl(CO)(TPPTS)2] and [IrCl(CO)(TPPTS)2] as catalysts [37], The complexes themselves were rather inefficient, however, the catalytic activity could be substantially increased by addition of MesNO in order to remove the carbonyl ligand from the coordination sphere of the metals. The polymers obtained had an average molecular mass of = 3150-16300. The rhodium catalyst worked at room temperature providing polymers with cis-transoid structure, while [IrCl(CO)(TPPTS)2] required 80 °C and led to the formation of frani -polymers. 

Similar to the above case, hydroformylation of 1-hexene using a water-soluble rhodium catalyst [RhH(CO)(TPPMS)3] gave lower yields when a-cyclodextrin was added to the biphasic reaction system [14]. Again, the reason was suspected in the interaction between the cyclodextrin and the rhodium catalyst. 

In 2000, we demonstrated that the planar-chiral phosphaferrocene PF-PPhj is a useful ligand for rhodium-catalyzed asymmetric isomerizations of several allylic alcohols, providing the first catalyst system that furnishes the target aldehyde in >60% ee (Eq. 6) [7]. It appears that, in order to obtain high enantiomeric excess (>0% ee), the olefin should bear a relatively bulky substituent (for example, Pr Eq. 6). 

As previously mentioned, Chung and co-workers have also developed a highly efficient coordinatively unsaturated catalyst system, which has general applicability to multiple heteroatom tethers (Scheme 12.8) [7]. This was the first example of a single rhodium catalyst system capable of facilitating [4-1-2] annulations for all three of the most common tethers (O, N, and C), including some intermolecular examples. 

---