High regioselectivity in propylene hydroformylation using rhodium-bisphosphite catalysts is due to properties of the SRS diastereomer

Article abstract of DOI:10.1039/b105391b

The large cone angle and bite angle of the SRS ligand diastereomer in biphenol-based Rh-bisphosphite catalysts lead to high linear regioselectivity in the hydroformylation of propylene.

Full text of DOI:10.1039/b105391b

View Article Online / Journal Homepage / Table of Contents for this issue
High regioselectivity in propylene hydroformylation using
rhodium-bisphosphite catalysts is due to properties of the SRS
diastereomer
John R. Briggs and Gregory T. Whiteker*
Corporate R&D, The Dow Chemical Company, 3200 Kanawha Turnpike, South Charleston, WV 25303,
USA. E-mail: whitekgt2@dow.com
Received (in Cambridge, UK) 26th June 2001, Accepted 14th September 2001
First published as an Advance Article on the web 18th October 2001
The large cone angle and bite angle of the SRS ligand
diastereomer in biphenol-based Rh-bisphosphite catalysts
lead to high linear regioselectivity in the hydroformylation of
propylene.
a negative optical rotation; the second peak had a positive
optical rotation. These enantiomers were physically separated
by preparative chiral HPLC and then re-analyzed. Slow
1
racemization of these enantiomers occurred (t = 53 min) at
2
34 °C by rotation about the central bridging biaryl bond.
Rhodium-catalyzed hydroformylation of a-olefins is widely
used for the industrial synthesis of alcohols and carboxylic acid
derivatives.1 Biphenol-based bisphosphite ligands, such as 1,
Interconversion of diastereomers of 1, therefore, most often
occurs via dibenzo[d,f][1,3,2]dioxaphosphepin epimerization
with a rate faster than the rate of hydroformylation.
Given the possibility that diastereomeric forms of 1 could
contribute to the observed catalytic behavior, we sought to study
the effect of isolated configurationally-stable bisphosphite
diastereomers in propylene hydroformylation. Bisphosphite 2
was selected for study due to the presence of 6,6A-methyl
substituents which prevent biaryl rotation. In addition, bis-
phosphite 2 contains 3,3A-tert-butyl substituents, which are
necessary for high hydroformylation regioselectivity using
ligands such as 1. The synthesis of two of the three possible
diastereomeric bisphosphites, 2, is depicted in Scheme 1. Acid-
catalyzed alkylation of enantiomerically-pure9 (R)-3 led to (R)-
4. Synthesis of (RRR)-2 and (SRS)-2 as pure enantiomers was
performed by reaction of (R)-4 with the phosphorochloridite
prepared by reaction of PCl3 with the requisite enantiomer of
3.
offer increased reaction rates and dramatic improvements in
linear/branched regioselectivity (l/b) and chemoselectivity over
conventional phosphine ligands in rhodium-catalyzed hydro-
formylation.2 Unfortunately, a comprehensive mechanistic
explanation of the factors that control hydroformylation
regioselectivity has remained elusive. Several ligand structural
features have been identified to be important in determining
regiochemistry in hydroformylation, including steric bulk,3
electrophilicity,4 and, in bidentate ligands, bite angle.5 In this
communication we show that in this class of bisphosphite
ligands, diastereomeric configuration can also have a profound
effect on catalytic hydroformylation regiochemistry and activ-
ity.
Rhodium-catalyzed propylene hydroformylation† was per-
formed using (RRR)-2 and (SRS)-2 to investigate the effects of
relative ligand stereochemistry. Hydroformylation using Rh–
(SRS)-2 resulted in a high regioselectivity (l/b = 46) with an
Rhodium-catalyzed hydroformylation of a-olefins using 1
produces linear aldehydes with high regioselectivities while
tolerating a wide variety of functional groups.6 Structurally
related optically-active bisphosphites have been developed
which result in high regio- and enantioselectivities for asym-
metric hydroformylation.7 Due to the presence of three axial
elements of chirality, bisphosphite 1 can potentially adopt three
diastereomeric forms. The notation for these diastereomers
(RRR, SRR, SRS) uses the middle descriptor for the central
bridging biphenoxide moiety. Interconversion of these
diastereomers can occur by rotations about the biaryl axes.
Previous NMR studies have demonstrated facile epimerization
of the cyclic dibenzo[d,f][1,3,2]dioxaphosphepin moiety in
related bisphosphites.8 This rotational process can interconvert
all three possible diastereomers (RRR, SRR, SRS).
Rotation about the central tert-butyl substituted biphenyl axis
could also interconvert RRR and RSR diastereomers, however,
chiral HPLC studies reveal this process is slow. Analytical
chiral stationary phase HPLC (Chiralcel OD, 99+1 hexane–
propan-2-ol) separated 1 into two fractions (1+1 ratio) which
exhibited identical UV spectra. When the column eluent was
monitored using a polarimetric detector, the first peak exhibited
Scheme 1 Reagents and conditions: (i) isobutylene, nitrobenzene, triflic
acid (5 mol%); (ii) for (RRR)-2: R-6,6A-dimethylbiphenylphosphochloridite
(2 equiv.), NEt3, THF; (iii) for (SRS)-2: S-6,6A- dimethylbiphenylphospho-
chloridite (2 equiv.), nBuLi, THF.
2174
Chem. Commun., 2001, 2174–2175
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b105391b

View Article Online
ences between linear and branched alkyl or acyl intermediates
and could account for the high regioselectivity for linear
aldehyde observed with ligands such as 1.
Previous observations of increased hydroformylation re-
gioselectivity with increased natural bite angle have been
complicated by the relationship between cone angle and bite
angle for chelating ligands.5,12 Since large bite angle diphos-
phines and bisphosphites typically have larger cone angles than
their smaller bite angle analogs, the inherent effect of the natural
bite angle on hydroformylation regioselectivity is unclear.
Recent calculations by Bo and van Leeuwen suggest that steric
effects play the dominant role in determining hydroformylation
regioselectivity.13 Although the natural bite angles of (RRR)-2
and (SRS)-2 correlate with their hydroformylation regiose-
lectivity, the very large difference in cone angles for these two
diastereomers suggests that the high hydroformylation re-
gioselectivity of (SRS)-2 is predominantly due to an increase in
steric influence of the bisphosphite ligand.
Fig. 1 Molecular mechanics structures of [(RRR)-2]Rh(CO)2H (left) and
[(SRS)-2]Rh(CO)2H (right). Hydrogen atoms are omitted for clarity.
average rate of 513 turnovers h21 over the course of 6 h. Under
identical conditions, Rh–(RRR)-2 gave a much lower rate (110
turnovers h21) and regioselectivity (l/b = 4) than the SRS
diastereomer. Bisphosphite 1 exhibited a rate (434 turnovers
h21) and regioselectivity (l/b = 53)2 which was very similar to
that observed with (SRS)-2. These results demonstrate that
rhodium complexes of the RRR and SRS diastereomers of ligand
2 lead to inherently different regioselectivities and rates for
propylene hydroformylation. The similarity between the cata-
lytic performance of (SRS)-2 and 1 suggests that during
catalysis, 1 exists predominantly in the SRS configuration.
Similar differences between bisphosphite diastereomers have
been reported by van Leeuwen for asymmetric styrene hydro-
formylation, where the branched isomer predominates.10 Reetz
has recently reported that asymmetric alkene hydrogenation
with biphenyl-based bisphosphite ligands displays an analogous
difference between ligand diastereomers.11 Our results obtained
using diastereomeric structures of 2 indicate that achiral
transformations mediated by bisphosphites are likewise ef-
fected by ligand structure.
Molecular mechanics calculations‡ of [(SRS)-2]Rh(CO)2H
and [(RRR)-2]Rh(CO)2H were performed to identify the factors
which could lead to their dramatically different hydroformyla-
tion behavior. The strain energies calculated for these two
diastereomeric complexes were essentially identical (within 2
kcal mol21). The natural bite angle12 for the SRS diastereomer
was calculated to be 117°, whereas the RRR diastereomer was
calculated to have a natural bite angle of 111°. Consistent with
previous correlations5 between hydroformylation regioselectiv-
ity and chelate natural bite angle, the more regioselective SRS
diastereomer of 2 exhibits a larger natural bite angle than the
less regioselective RRR diastereomer. In addition to differences
in natural bite angle, these calculations also revealed a
significant difference in cone angle between these two diaster-
eomeric catalysts (Fig. 1). The cone angles at each phosphorus
atom in (RRR)-2 and (SRS)-2 were calculated to be 104° and
157°, respectively. The RRR (low l/b) catalyst adopts an open
geometry around the equatorial CO ligand, which should result
in minimal energy differences between linear and branched
alkyl or acyl intermediates. The SRS diastereomer, which
dominates hydroformylation catalysis, is calculated to form a
more crowded complex in which the two dibenzo[d,f]-
[1,3,2]dioxaphosphepin moieties flank the equatorial CO li-
gand. This arrangement should result in large energy differ-
Notes and references
† Propylene hydroformylation experiments were performed in tetraglyme
solution in a stirred autoclave at constant pressure (148 ppm Rh (1.4 mM),
L+Rh(CO)2(acac) = 2.0, 82 °C, 85 psi propylene, 45 psi 1+1 H2–CO).
‡ Molecular mechanics calculations were performed using the augmented
MM3 force field implemented in the CAChe software package. Natural bite
angles were calculated by using a P–Rh–P bending force constant of 0 kcal
mol21 deg22. The authors are indebted to the reviewers for insight
regarding these calculations.
1 Rhodium Catalyzed Hydroformylation (Catalysis by Metal Complexes,
Vol. 22), ed. P. W. N. M. van Leeuwen and C. Claver, Kluwer Academic
Press, Dordrecht, 2000.
2 E. Billig, A. G. Abatjoglou and D. R. Bryant, US Pat., 4,769,498,
1988.
3 P. W. N. M. van Leeuwen and C. F. Roobeek, J. Organomet. Chem.,
1983, 258, 343.
4 W. R. Moser, C. J. Papite, D. A. Brannon, R. A. Duwell and S. J.
Weininger, J. Mol. Catal. A: Chem, 1987, 41, 271; J. D. Unruh and J. R.
Christenson, J. Mol. Catal. A: Chem, 1982, 14, 19.
5 C. P. Casey, G. T. Whiteker, M. G. Melville, L. M. Petrovich, J. A.
Gavney and D. R. Powell, J. Am. Chem. Soc., 1992, 114, 5535; M.
Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer and P. W. N. M. van
Leeuwen, Organometallics, 1995, 14, 3081.
6 G. D. Cuny and S. L. Buchwald, J. Am. Chem. Soc., 1993, 115, 2066.
7 J. E. Babin and G. T. Whiteker, US Pat. 5,360,938, 1994; G. J. H.
Buisman, P. C. J. Kamer and P. W. M. N. van Leeuwen, Tetrahedron:
Asymmetry, 1993, 4, 1625.
8 G. T. Whiteker, A. M. Harrison and A. G. Abatjoglou, J. Chem. Soc.,
Chem. Commun., 1995, 1805.
9 S. Kanoh, N. Tamura, M. Matoi and H. Suda, Bull. Chem. Soc. Jpn.,
1987, 60, 2307.
10 G. J. H. Buisman, L. A. van der Veen, A. Klootwijk, W. G. J. de Lange,
P. C. J. Kamer, P. W. N. M. van Leeuwen and D. Vogt, Organome-
tallics, 1997, 16, 2929.
11 M. Reetz and T. Neugebauer, Angew. Chem., Int. Ed. Engl., 1999, 38,
179.
12 C. P. Casey and G. T. Whiteker, Isr. J. Chem., 1990, 30, 299.
13 J. J. Carbo, F. Maseras, C. Bo and P. W. N. M. van Leeuwen, J. Am.
Chem. Soc., 2001, 123, 7630.
Chem. Commun., 2001, 2174–2175
2175