It cannot be overstated how important the hydroformylation reaction is and it is best illustrated by the vast volume (14 million metric tons per year), producing annually compounds with a value of 15 billion euros. The reaction involves the homologation of an olefin by addition of carbon monoxide and hydrogen producing aldehydes as primary products. This provides a pool of aldehyde compounds that can be easily further converted to the corresponding alcohols, acids, imines, amines and aldol condensation products (e.g. to produce 2-ethylhexy alcohols, an important industrial building block).
The hydroformylation reaction can be applied to both petrochemical feedstocks (of which the propene hydroformylation is the largest in size) and bio-based feedstocks (such as fatty acids and terpenes). InCatT has been actively involved in developing catalyst systems for these transformations.[i] [ii] [iii]
Nowadays, the majority of known hydroformylation reactions involve rhodium phosphine or phosphite complexes. These catalysts typically operate under conditions that can be easily applied to standard high pressure equipment (10-80 bar, between 40-140 °C). Before the discovery of the rhodium catalysts in the 1960s, cobalt catalyst was the working horse which operates under more sever conditions: HCo(CO)4 (140−200 °C and 100−300 bar) and the phosphine-modified cobalt system (180−200 °C and 50−150 bar). Despite the price of the catalytic metal (cobalt price is 0.01% of rhodium), the CAPEX and OPEX has limited its use to dedicated large-scale industrial processes for the hydroformylation of medium and long-chain olefins. Recently, Stanley et al has reported the first improvements for the cobalt hydroformylation in 50 years and we have published a blog on these latest developments.[iv] This finding has revived and intensified the research into cobalt hydrofomylation reaction and now the research teams of prof. Elena Gusevskaya and prof. Matthias Beller has published a joined publication (doi.org/10.1021/acssuschemeng.1c00205) on the use of stable and inexpensive phosphine oxides to promote the formation of active cobalt catalysts for the hydroformylation of olefins.
The authors show that the addition of phosphine oxides dramatically increases the reactivity of the system: up to 91% of the substrate was consumed depending on the nature of the phosphine oxide. Particularly, the diadamantylalkyl phosphine oxides were effective in the cobalt hydroformylation under mild conditions (40 bar (CO/H2 = 1:3), 40 °C) yielding 89% conversion with low overreduction to alkanes and alcohols (2% and 6%, respectively). Bidentate P8 provided similar high conversion but displayed a much higher isomerization activity (16%). Interestingly, simple triphenylphosphine oxide (4% relative to cobalt) resulted in 90% conversion and 92% selectivity to aldehydes under the benchmark conditions. The regioselectivity (l/b ratio of 76:24) is similar to the observed Rh/PPh3 catalyst system.
The hydroformylation reaction has been studied in detail and shows that the syngas composition is important to reach high conversion. Reducing the CO:H2 ratio from 1:3 to 1:1 reduced the conversion from 90% to just 35%. Interestingly, the hydroformylation reaction could be performed with syngas pressures as low as 5 bars with only a limited effect on the activity (69% conversion). The reaction temperature (between 40 and 60 °C) did not result in a different catalytic outcome. The authors noted that the regioselectivity could be improved by changing the syngas composition. Detailed studies show that the presence of triphenylphosphine oxide reduces the induction period to just a few minutes after which the hydroformylation reactions starts. In the presence of triphenylphosphine the formation of active catalyst species was completely inhibited. The authors suggest that strong acceleration is a result of a fast disproportionation reaction of the precatalyst Co2(CO)8 to give the active HCo(CO)4 species.
The tested substrate scope is quite extended in which terminal, linear internal, cyclic olefins are smoothly converted. This process can also be applied to substrates with electron-withdrawing substituents. Particularly, the perfluoroalkyl-substituted olefin was transformed into the linear aldehyde product (l/b = 99:1) in high yield and regioselectivity in contrast to the rhodium system. The hydroformylation of 3,3-dimethylbut-1-ene led exclusively to the linear product (l/b = 99:1), which was explained by the tBu steric hindrance. Finally, tri- and tetrasubstituted olefins could be smoothly converted by an isomerization-hydroformylation sequence.
This report shows that the old working horse that typically that was believed (for more than 80 years!) to exclusively perform under very drastic conditions can now be applied under very mild conditions (as low as 5 bar and 40 °C). In the presence of simple, commercially available triphenylphosphine oxide a range of olefins can be hydroformylated with Co2(CO)8. Employing slightly higher temperatures, even cyclic, internal, or sterically encumbered olefins can be hydroformylated in good yields. Interestingly, the authors show that the addition of phosphine oxides dramatically reduces the preactivation time of the classic cobalt precatalyst.
About InCatT (www.incatt.nl): InCatT B.V. is a company specialized in catalyst screening and catalyst development from initial catalyst-lead finding to process optimization. Over the years we have worked with different industries ranging from Flavor & Fragrance, Bio-based industry, Pharmaceutical and Bulk chemical industry to solve their most challenging projects.
Article: “Cobalt-Catalyzed Hydroformylation under Mild Conditions in the Presence of Phosphine Oxides”
By: Fábio G. Delolo, Ji Yang, Helfried Neumann, Eduardo N. dos Santos, Elena V. Gusevskaya*, and Matthias Beller*
CS Sustainable Chem. Eng. 2021http
[i] Pim R. Linnebank Stephan Falcão Ferreira Alexander M. Kluwer, Joost N. H. Reek, Chem. Eur.J., 2020, 26,8214 –8219
[ii] Alexander M. Kluwer, Michael J. Krafft, Ingo Hartenbach, Bas de Bruin, Wolfgang Kaim, Top Catal, 2016, 59, 1787–1792
[iii] Alexander M. Kluwer, Chretien Simons, Quinten Knijnenburg, Jarl Ivar van der Vlugt, Bas de Bruin, Joost N. H. Reek, Dalton Trans., 2013, 42, 3609–3616
[iv] Drew M. Hood, Ryan A. Johnson, , Alex E. Carpenter, Jarod M. Younker, David J. Vinyard, George G. Stanley, Science, 2000, 367, 542–548