Phosphine-phosphite rhodium catalysts

In 1992, Takaya et al. reported on the use of a chiral diphosphite derived from bisnaphthol in the asymmetric hydroformylation of vinyl acetate [11], but the enantioselectivity achieved was only 50%. They noted that diphosphites led to more stable hydroformylation catalysts than diphosphines. This observation prompted Takaya, Nozaki et al. to synthesize the chiral phosphine-phosphite ligands (R,S)- and (R,R)-BINAPHOS (24, figure 17), which were expected to combine the high enantioselectivity obtained with diphosphines such as BINAP in asymmetric hydrogenation, with the apparently efficient coordination of the phosphite moiety [38]. Indeed, the Rh(I) complex of Ci-symmetric (R,S)-BINAPHOS provided much higher enantioselectivities than either C2 symmetric diphosphine ligands or diphosphite ligands, viz. more than 90% ee for a wide variety of both functionalized and internal aUcenes [38,39,40,41]. 

Ever since the successful results obtained for (R,S)-BINAPHOS, the preparation of other unsymmetrical ligands to be used in the asymmetric hydroformylation has become a recurrent theme in the literature [3, 4, 36, 42]. The following section will discuss those ligands, which are stracturally related to BINAPHOS, while other phosphine-phosphite systems will be discussed separately in section 5.3.5. 

---

In this system higher CO pressures lead to lower linearities simply by shifting the complex equilibria to the species containing less phosphine. When isomerization plays a role under the reaction conditions applied, higher CO pressures will also give lower initial linearities by suppression of reaction (8i), the isomerization reaction as outlined above for the carbonyl or phosphite rhodium catalysts. These two effects can be clearly distinguished by monitoring the isomerization reaction when alkenes other than ethene and propene are used. 

Trialkylphosphine-modified cobalt catalysts and unmodified rhodium catalysts show similar olefin isomerization activity, but the most common tertiary phosphine-modified rhodium catalysts are generally ineffective in double-bond isomerization (77-79). On the other hand, phosphite-modified rhodium catalysts were found to be highly effective at alkene isomerization (80-83). 

Some rhodium catalysts for hydroformylation of epoxides have been claimed by Union Carbide and Shell (Scheme 6.106) [17]. By the use of phosphites (e.g., Alkanox 240, BIPHEPHOS) or phosphine-modified rhodium catalysts, the selectivity was enhanced and the reaction temperature could be lowered. Noteworthy, occasionally a higher partial pressure of Hg in the syngas mixture with a total pressure of 90 bar was employed. 

High enantioselectivities and regioselectivities have been obtained using both mono- and 1,2-disubstituted prochinal olefins employing chiral phosphine phosphite (33,34) modified rhodium catalysts. For example, i7j -2-butene ia the presence of rhodium and (12) (33) gave (3)-2-meth5ibutanal ia an optical yield of 82% at a turnover number of 9.84. ... 

After the discovery of the high ee provided by rhodium/diphosphite and rhodium/phosphine-phosphite complexes, with total conversion in aldehydes and high regioselectivities, rhodium systems became the catalysts of choice for asymmetric hydroformylation. Important breakthroughs in this area have been the use of rhodium systems with chiral diphosphites derived from... 

Based on the precedent of Van Leeuwen and Roobeek, livinghouse and co-workers screened a variety of electron-deficient phosphine/phosphite ligands for the rhodium-catalyzed [4-1-2] reaction, which provided an alternative catalyst system for the formation of 5,6- and 6,6-ring systems [13]. The most notable of these was the tris-(hexafluoro-2-propyl) phosphite-modified rhodium complex, which was applicable to both carbon- and oxygen-tethered substrates, and also provided the first example of a facial-directed diastereoselective intramolecular rhodium-catalyzed [4-i-2] reaction (Eq. 4). 

Under these circumstances, the third-generation catalysts, Rh(i)-chiral bis-phosphites and Rh(i)-chiral phosphine-phosphites, were developed in 1992-1993. Apart from the asymmetric matter, it was reported in the 1980s that rhodium(i) complexes of phosphites, especially those bearing bulky substituents, showed high activities in... 

Figure 7 A rhodium complex of chiral phosphine-phosphite ligand (H,S)-BINAPHOS used as a catalyst for asymmetric hydroformylation.

Also, bulky phosphite-modified rhodium catalysts are highly reactive for the hydroformylation of unsaturated fatty acid esters [23]. The catalyst was able to yield turnover numbers (TON) of 400-500 when moderate conditions with 20 bar synthesis gas pressure and 100°C were applied. These phosphites, like tris (2-ferf-butyl-methyl) phosphite, have higher activity than phosphines like triphenylphosphine. 

A system kinetically very similar to the phosphine-free rhodium carbonyl catalyst is obtained with bulky phosphites (Fig 6.3). At temperatures from 50 to 80°C, and CO and H2 partial pressures ranging from 10 to 70 bar, the rate of aldehyde formation is first order in H2 and approximately minus one order in CO. The reaction rate is independent of the concentration of 1-octene at conversions below 30%. The reaction was found to be first order in rhodium concentration and insensitive to the phosphite/rhodium ratio, provided that the absolute concentration was sufficiently high to generate a hydride complex from the pentanedionate precursor (reaction 9). 

The electronic and steric properties of the phosphine ligand(s) can have dramatic effects on the rate and selectivity of the rhodium catalysts. As mentioned above, electron-rich alkylated phosphines generally have a negative effect on the rate and regioselectivity, while more electron deficient phosphines such as PPhs and phosphites (see Electron Deficient Compound) generate more active and regioselective catalysts (Table 3). 

Soon after the initial discovery of the hydroformylation activity of RhH(CO)(PPh3)3 it was found that ligands have a profound influence on the activity and selectivity of the rhodium catalysts. The majority of the numerous published patents and publications are concerned with the effect of ligands and it is therefore impossible to review this matter here. Instead we have made a choice that includes a study of the electronic and steric effects of phosphines and phosphites and a few examples of bidentate ligands. A few trends will be briefly summarized. 

---