On the efficacy of propeller-shaped, C3-symmetric triarylphosphines in asymmetric catalysis

Article abstract of DOI:10.1039/a806811i

Ligand sets 1 and 2 were prepared and examined for evidence of C₃-symmetric propeller-shaped conformations in solution, and for their ability to induce enantioselectivity in an allylation reaction.

Full text of DOI:10.1039/a806811i

View Article Online / Journal Homepage / Table of Contents for this issue
On the efficacy of propeller-shaped, C3-symmetric triarylphosphines in
asymmetric catalysis
Mark T. Powell, Alexander M. Porte and Kevin Burgess*
Department of Chemistry, Texas A & M University, PO Box 300012, College Station, TX 77842-3012, USA.
E-mail: burgess@chemvx.tamu.edu
O
OH
R1
Ligand sets 1 and 2 were prepared and examined for
evidence of C3-symmetric propeller-shaped conformations
Br
Br
Me
Me
in solution, and for their ability to induce enantioselectivity
in an allylation reaction.
i
R1
3
There are conflicting arguments with regard to the potential of
optically pure, C3-symmetric triarylphosphines in asymmetric
syntheses. Some researchers may correctly point to the value of
C2-ligands1 and claim that application of C3-ligands is a logical
extrapolation of the field. C3-Symmetric arrangements of three
aromatic groups around a central atom can adopt stable
enantiomeric propeller-shaped conformations that might pro-
vide chiral pockets to facilitate enantiodiscrimination.2 How-
ever, the contrary argument is also convincing. Asymmetric
induction cannot increase indefinitely with the symmetry of the
chiral directing group because a perfectly spherical object
would be useless for inducing a chiral environment.
ii
OR2
OR2
Me
P
Br
Me
iii, iv
R1
3
R1
4
1 R1 = H; 2 R1 = OMe
a R2 = Me; b R2 = Ph;
c R2 = 2,6-Me2C6H3
Experimentally, the value of optically active, propeller-
shaped, C3-symmetric phosphines is hard to assess. This is
because of synthetic difficulties associated with obtaining the
requisite ligands, and due to a lack of techniques to recognize
rigid propeller conformations in solution. Sharpless and co-
workers, for instance, prepared triarylphosphine cage structures
(an example is shown below) by relatively difficult synthetic
routes.3,4 They then found that in an optically active complex,
this ligand stereomutates between enantiomeric propeller-
shaped conformations at room temperature. This ligand design
therefore did not facilitate a test of the efficacy of propeller-
shaped C3-symmetric ligands, hence their value remained
questionable. The work described in this manuscript deals with
attempts to address this issue using phosphines 1 and 2. The
tenet of this project is that a chiral substituent on the aromatic
rings could be easily installed, and may lead to stable
C3-symmetric propeller-shaped aryl arrays in the ligand.
Scheme 1 Reagents and conditions: i, BH3·SMe2, 5 mol% 4,5,6,7-tetra-
hydro-1-methyl-3,3-diphenyl-1H,3H-pyrrolo[1,2]-[1,3,2]oxazaborolebor-
ane (CBS),8 CH2Cl2, 225 °C, 12 h (94% and 96.2% ee for 1; 85%, 99% ee
for 2); ii, alkylation (yields range from 70 to 99%, e.g. MeI, NaH, DMF for
1a); iii, BuLi, Et2O, 230 °C, 1 h; iv, PCl3, Et2O, 230 to 25 °C (yields
typically 30–50% for steps iii and iv)
t
Scheme 1 outlines the route by which the ligand set 1a–1c
(and later 2a–2c) was obtained. The route diverges from the
common chiral alcohol intermediate 3 hence this strategy is
more efficient than ones that rely on different starting materials
for each phosphine prepared.
Evidence for preferred stereoisomeric propeller-shaped con-
formations in solution is hard to obtain. Crystallographic studies
of derivatives such as complex 5a (Fig. 1) indicated the desired
conformations exist in the solid state, but these observations can
give no indication of their dynamic behavior in solution.
Consequently, a set of circular dichroism (CD) spectra was
recorded to elucidate solution state conformations. Chiral
ordering of the aromatic groups should be accompanied by
Fig. 1 Comparison of normalized ellipticities (i.e. ellipticities per mole of
aromatic ring) for compounds 4a, 5a and 6a
Chem. Commun., 1998
2161

View Article Online
OCOBn
+
O
5 mol % {Pd(C3H5)Cl}2
N
O
O
10 mol % ligand
CH2Cl2, 0 °C. ca. 36 h
N
H
O
The data presented here suggest that phosphines 2 can exist in
conformations in which the aromatic groups are ordered in
propeller-shaped arrays, and that these same phosphines give
significant induction in an allylation reaction. However, it is
unlikely that perfectly C3-symmetric conformations predom-
inate for ligand 2 in complexes because the meta-substituent can
adopt orientations that are exo and endo with respect to the
metal. Work now in progress concerns a ligand system for
which this is not a possibility.
We thank Marcel Jaspars for some informative preliminary
experiments. Support for this work was provided by The
Petroleum Research Fund administered by The American
Chemical Society, and by The Robert A. Welch Foundation.
K. B. thanks the NIH Research Career Development Award,
and The Alfred P. Sloan Foundation for a fellowship.
Fig. 2 Phosphines 1 and 2 in an allylation reaction (unoptimized yields,
measured by GC using an internal standard: 1a, 72%; 1b, 33%; 1c, 54%; 2a,
50%; 2b, 51%; 2c, 47%)
increased molar ellipticities at wavelengths corresponding to
aromatic absorptions for palladium complex 5a relative to the
aryl bromide starting material 4a.5 Fig. 1 shows that an
increased ellipticity was not observed for complex 5a derived
from ligand 1a. This led us to suppose that conformational
rigidity could be increased by incorporation of a para-
substituent to disfavor free rotation about the bond to the meta-
chiral center. Consequently, ligand set 2 was prepared and
selected derivatives were examined by CD. Fig. 1 indicates that
complex 6a derived from ligand 2a does indeed show an
enhanced ellipticity relative to intermediate 4a.
Notes and References
1 J. K. Whitesell, Chem. Rev., 1989, 89, 1581.
2 C. Moberg, Angew. Chem., Int. Ed. Engl., 1998, 37, 248.
3 C. Bolm and K. B. Sharpless, Tetrahedron Lett., 1988, 29, 5101.
4 C. Bolm, W. D. Davis, R. L. Haltermann and K. B. Sharpless, Angew.
Chem., Int. Ed. Engl., 1988, 27, 835.
5 J. W. Canary, C. S. Allen, J. M. Castagnetto and Y. Wang, J. Am. Chem.
Soc., 1995, 117, 8484.
6 K. Burgess, D. Moye-Sherman and A. M. Porte, Molecular Diversity and
Combinatorial Chemistry, American Chemical Society, Washington DC,
1996, pp. 128–136.
7 A. M. Porte, J. Reibenspies and K. Burgess, J. Am. Chem. Soc., in
press.
8 D. J. Mathre, A. S. Thompson, A. W. Douglas, K. Hoogsteen, J. D.
Carroll, E. G. Corley and E. J. J. Grabowski, J. Org. Chem., 1993, 58,
2880.
High throughput parallel screens6 were used to test ligands 1
and 2 in the palladium mediated allylation reaction illustrated
below. Thus reactions were run simultaneously in wells
contained in a cooled aluminium block, then analyzed using an
autosampler/chiral HPLC apparatus. Details of this approach
applied in other studies from our group have been documented.7
Fig. 2 shows the data obtained. The enantioselectivities reached
an optimum value of 82%. We think that this level of induction
by the distal chiral meta-substituents would not be possible
unless ordering of the aromatic rings were operative. Enhanced
ellipticities when a para-substituent is present (i.e. 2a vs. 1a)
correlates with dramatically increased enantioselectivities. On
average, higher enantioselectivities tend to be observed for the
ligands 2 than for series 1.
Received in Bloomington, IN, USA, 24th April 1998; 8/06811I
2162
Chem. Commun., 1998