Is restricted M-P rotation a common feature of enantioselective monophos catalysts? An example of restricted Rh-P rotation in a secondary phosphine complex

Article abstract of DOI:10.1016/j.tetasy.2010.06.005

The enantioselective hydrogenation catalyst [Rh(α-CgPH) ₂(cod)]BF₄, (CgPH = 6-phospha-2,4,8-trioxa-adamantane) exists in solution as a mixture of two slowly interconverting rotamers, one with C₂-and the other with C₁-symmetry.

Full text of DOI:10.1016/j.tetasy.2010.06.005

Tetrahedron: Asymmetry 21 (2010) 1206–1209
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Is restricted M–P rotation a common feature of enantioselective monophos
catalysts? An example of restricted Rh–P rotation in a secondary
phosphine complex
a
a
b
a
a,*
Piotr Jankowski , Claire L. McMullin , Ilya D. Gridnev , A. Guy Orpen , Paul G. Pringle
a
School of Chemistry, University of Bristol, Cantocks Close, Bristol BS8 1TS, UK
Department of Chemistry, Tokyo Institute of Technology, Tokyo, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The enantioselective hydrogenation catalyst [Rh(
adamantane) exists in solution as a mixture of two slowly interconverting rotamers, one with C
a
-CgPH) (cod)]BF , (CgPH = 6-phospha-2,4,8-trioxa-
2
4
Received 25 May 2010
Accepted 3 June 2010
Available online 26 June 2010
2
- and
the other with C -symmetry.
1
Ó 2010 Elsevier Ltd. All rights reserved.
Dedicated to Professor Henri Kagan on the
occasion of his 80th birthday
1
. Introduction
throughput methods,10 were also shown to give highly enantiose-
lective hydrogenation catalysts. We suggested that the success of
7
In the early 1970s, Knowles et al.1 showed that C
monophos ligands such as L
cis-Rh(PR configuration in their complexes and the restricted
1
-symmetric,
a
may be rooted in their adoption of a
monodentate phosphines such as CAMP can support highly enan-
3 2
)
tioselective asymmetric hydrogenation catalysts. At the same time,
rotation around the Rh–P bonds in the complexes that should arise
from the bulk of the P-substituents.
2
Kagan et al. discovered that the C
2
-symmetric, bidentate diphos-
phine DIOP also gave highly enantioselective asymmetric hydroge-
nation catalysts; there followed more than 25 years of the
dominance of chelating diphosphines for asymmetric hydrogena-
La Z = alkyl or aryl
3
,4
O
P
tion. The assumed greater efficiency of the C
2
-symmetric diphos
O
Lb Z = NR2
Lc Z = OR
Z
chelate catalysts over their C -symmetric bis(monophos) ana-
1
logues was rationalised as follows: (i) the cis coordination, present
perforce in the chelates, was a critical feature in the mechanism
proposed for asymmetric hydrogenation and (ii) the restricted
Recently, we reported11 the resolution of the C
1
-symmetric cage
M–P rotation in the C
2
-chelates created a highly defined chiral
12
secondary phosphine b-CgPH and showed that, despite its appar-
ently ‘weak’ chirality (Fig. 1), it gave Rh complexes that performed
well in the asymmetric hydrogenation of dehydroamino acid deriv-
atives. We herein report further asymmetric hydrogenations with
resolved CgPH and show that the Rh–P bond rotation is slow
enough to observe rotamers by NMR spectroscopy even at room
temperature. The prospect is raised that restricted Rh–P bond rota-
tion may be common in other highly enantioselective asymmetric
hydrogenation catalysts featuring monophos ligands.
space around the metal which led to greater enantiofacial discrim-
ination. Indeed, the latter point was reinforced by the emergence of
3
5
backbone rigidity (in diphosphines such as BINAP and DuPhos ) as
an important factor in the design of asymmetric hydrogenation
catalysts.
In a prescient article in 2000, Kagan6 suggested that optically
active monophosphines were a neglected class of ligands for asym-
metric catalysis and were worthy of further study. In the last
1
0 years, there has been an eruption of interest in monophos ligands
for asymmetric hydrogenation, initiated by the discovery that Rh
catalysts derived from some C -symmetric monophosphonites L
gave superior enantiomeric excesses to their chelating C -symmet-
1
a
O
O
O
O
2
7
8
ric diphosphonite analogues. Subsequently, phosphoramidites L
b
9
and phosphites
c
L , whose ready syntheses are amenable to high
O
P
H
H
P
O
β−CgPH
α−CgPH
*
Corresponding author.
E-mail address: paul.pringle@bristol.ac.uk (P.G. Pringle).
Figure 1.
0
957-4166/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2010.06.005

P. Jankowski et al. / Tetrahedron: Asymmetry 21 (2010) 1206–1209
1207
2
. Results and discussion
The separation of b-CgPH from racemic CgPH via the diastereo-
1
1
isomeric quinine salts of CgPOOH has been previously reported.
The
study. The absolute stereochemistry in
been confirmed by X-ray crystallography.
The complex [Rh( -CgPH) (cod)]BF 1 (cod = 1,5-cyclooctadi-
ene) was obtained by treatment of [Rh(cod) ]BF with -CgPH
and purified by crystallisation from CH Cl /Et O. The asymmetric
a
-CgPH has now been separated and was used throughout this
a
-CgPOOH and -CgPH has
a
13
a
2
4
2
4
a
2
2
2
hydrogenation of the substrates A–E using 1 as the catalyst was
carried out and the results are given in Table 1. It is clear that, with
these substrates, the enantioselectivity of catalyst 1 is high.
CO Me
CO Me
CO Me
2
2
2
Figure 2. Thermal ellipsoid plot of the cation of 1a omitting all carbon-bound
hydrogen atoms. Selected geometrical data: bond lengths [Å] Rh1–P1 2.2887(4),
P1–C2 1.8735(14), P1–C9 1.8716(15), P1–H1 1.323(18); angles [°] P1–Rh1–P1A
NHCOMe
Ph
Ph
NHCOMe
CH CO Me
2
2
84.181(18), C2–P1–C9 94.58(6); torsion angle [°] H1–P1–Rh1–P1A 19.8(8), H1–P1–
A
B
C
D
E
P1A–H1A 39(1).
CO
2
H
CO
CH
2
H
2
CO
2
H
Table 1
Asymmetric hydrogenation results
NHCOMe
a
Entry
Substrate
ee %
c
c
1
2
3
4
5
A
B
C
D
E
88 (R)
87 (R)
88 (R)
95 (S)
88 (S)
b
Crystals of 1 suitable for X-ray crystallography were grown
14
2 2 2
from CH Cl /Et O and its structure is shown in Figure 2. The rho-
a
dium atom lies on a twofold axis and has a square planar coordina-
tion geometry with a P–Rh–P angle of 84.2° and a H–P–P–H torsion
angle of 39°.
Reaction conditions: S/C = 100:1, 5 bar, rt, 1 h, sol-
vent: MeOH. All conversions were 100% and enanti-
oselectivities determined by chiral GC analysis.
b
Reaction in CH
2
Cl
2
.
If the solid-state structure of 1 was maintained in solution, then
its 31P{ H} NMR spectrum would be expected to reflect the equiv-
alence of the P atoms. As shown in Figure 3, a doublet is indeed ob-
c
These values have been redetermined and are
1
11
higher than we previously reported.
1
served at d 15.6 ppm ( JRhP = 141 Hz) but in addition, two doublets
1
A variable temperature 31P NMR study of a solution 1a/1b
showed that, at low temperatures (in CH Cl ), the ratio of 1a/1b in-
of doublets are also evident (d
A
14.4 ppm,
JRhP = 144 Hz; d
X
1
31
À17.1 ppm,
is consistent with the presence of two species: 1a, with a struc-
ture similar to that in the crystal structure (Fig. 2) and a second
JRhP = 139 Hz, JPP = 35 Hz). The P NMR spectrum of
2
2
1
creased from ca. 1:1 at +23 °C to 3:1 at À80 °C and at elevated tem-
peratures (in CHCl CHCl ), the signals for 1a and 1b broadened, but
2
2
1
31
species 1b containing a cis-Rh(P
A
)(P
X
) moiety. The H-coupled
P
did not coalesce up to 110 °C. These data are consistent with 1a
and 1b being rotamers in equilibrium with each other and under-
going exchange via Rh–P bond rotation slowly on the NMR time-
scale at ambient temperatures. We propose that 1a has a similar
geometry to the C -symmetric structure in the X-ray crystal struc-
ture and 1b is a rotamer of C -symmetry rendering the two P nuclei
NMR spectrum revealed large JPH coupling (336–350 Hz) for all
three of the signals, confirming that the P–H bonds remain intact
in the coordinated P-ligands. The possibility that 1b was a dimer
of 1a was ruled out by the observations that (i) the ratio of 1a/1b
was invariant with concentration and (ii) a DOSY spectrum showed
that 1a and 1b have a similar hydrodynamic radius of 6.7 Å.
2
1
1
5
inequivalent (see below).
Figure 3. 31P{ H} NMR spectrum of a CD
1
2
2 2 4
Cl solution of [Rh(a-CgPH) (cod)]BF 1.

1
208
P. Jankowski et al. / Tetrahedron: Asymmetry 21 (2010) 1206–1209
The presence of two species in solutions of 1 was also evident
1
from the H NMR spectrum. There are three PH resonances and
six signals for the olefinic protons in accordance with the C - and
-symmetry of the proposed rotamers. The exchange rate be-
1
C
2
Rh
Rh
tween the two species can be gauged from the linewidths of the
H
CgP
PCg
PCg
PCg
PH signals leading to an estimate of the barrier to exchange being
H H
À1
H
1
5–18 kcal mol (Fig. 4).
1a
1b
Scheme 1. CgPH is optically pure (a-enantiomer).
*
Figure 4. Optimised geometry of 1b from DFT calculation (B3LYP/6-31G &LACV3P)
omitting all non carbon-bound hydrogen atoms; torsion angle [°] H1–P1–P2–H2
1
58.6, H1–P2–Rh–P2 169.5, H2–P2–Rh–P1 À0.4.
Figure 6. Thermal ellipsoid plot of the cation of 3a omitting all carbon-bound
hydrogen atoms. Selected geometrical data: bond lengths [Å] Rh1–P1 2.2722(5),
Rh1–P2 2.2982(5), P1–C2 1.868(2), P2–C12 1.868(2), P1–C9 1.871(2), P2–C19
To probe the hypothesis of restricted Rh–P bond rotation, DFT
calculations were used to isolate and optimise geometries in both
a and 1b for a variety of functional and basis set combinations.
1.857(2), P1–H1 1.35(2), P2–H2 1.29(2); angles [°] P1–Rh1–P1A 87.100(19), C2–P1–
C9 94.58(6); torsion angle [°] H1–P1–Rh1–P1A 5(1), H2–P2–Rh1–P1 16(1), H1–P1–
P1A–H1A 19(2).
1
Two rotamer energy minima were found which were less than
À1
0
.3 kcal mol apart, with the C
1
-symmetric 1b slightly lower in
-symmetric 1a. The Rh–P bond was then sequen-
tially rotated by constraining the H1–P1–Rh–P2 torsion, and leav-
ing the other H2–P2–Rh–P1 torsion ( ) free of constraints. The
energy than C
2
A diastereomeric mixture of aa/bb-[Rh(CgPH)
and b-[Rh(CgPH) (nbd)]BF meso-3 (nbd = norbornadiene) was
made by treatment of [Rh(nbd) ]BF with racemic CgPH. Crystals
of C -symmetric isomer 3a suitable for X-ray crystallography were
grown from a saturated CH Cl /Et O solution and its structure is
2 4
(nbd)]BF rac-2
a
2
4
s
2
4
rotation profile (Fig. 5, blue curve) shows minima around 0° and
s
À1
1
60° with a barrier of ca. 13 kcal mol at 90°.
2
2
2
1
7
shown in Figure 6. This meso isomer has a square planar rho-
dium(I) atom, with a P–Rh–P angle of 87° and a H–P–P–H torsion
angle of 19°.
The 31P NMR spectra for the mixture of 2 and 3 at various tem-
peratures are shown in Figure 7. At +23 °C, a single broad reso-
nance is observed which, upon heating to 80 °C (in CHCl
2
CHCl
2
),
1
18
became a doublet at d 6.6 ppm ( JRhP = 153 Hz) confirming that
1
9
the fluxionality is intramolecular. At low temperatures the spec-
trum resolves into a pattern of resonances assigned to rotamers,
2
0
rac-2a/2b and meso-3a/3b (Scheme 2). Thus the spectra for 2
and 3 at temperature T, resemble those of 1 at temperature
Rh
Rh
Figure 5. Graph showing rotation barrier for 1 (blue line) and 2 (red line). The
abscissa is the H1–P1–Rh–P2 torsion angle ( ).
s
H
CgP
PCg
PCg
PCg
H H
H
Thus the calculations support the suggestion that the explana-
31
2a
2b
3b
tion for the fluxionality observed in the P NMR spectra of 1 is
due to the interconversion of the rotamers, as shown in Scheme 1
which is slow on the NMR timescale because of restricted Rh–P
3
a
Scheme 2. The CgPH used was racemic and thus 2a/2b are the aa/bb-isomers and
3a/3b are the b-isomers.
1
6
bond rotation.
a

P. Jankowski et al. / Tetrahedron: Asymmetry 21 (2010) 1206–1209
1209
Figure 7. 31P NMR spectra (121 MHz) of the mixture of rac-2 and meso-3 at the following temperatures (from left to right): 183, 223, 298, 333, 383 K.
T + 120 °C and therefore the barrier to Rh–P bond rotation is lower
for the nbd complex and is estimated to be 10–12 kcal mol . The
calculated rotation profile for 2 is shown in Figure 5 (red curve)
2. (a) Kagan, H. B.; Dang, T. P. Chem. Commun. 1971, 481; (b) Kagan, H. B.; Dang, T.
À1
P. J. Am. Chem. Soc. 1972, 94, 6429.
3
4
.
.
Noyori, R.; Takaya, H. Acc. Chem. Res. 1990, 23, 345.
Efficient monophos ligands for asymmetric hydrogenation were occasionally
reported during this period. See for example: Guillen, F.; Fiaud, J.-C. Tetrahedron
Lett. 1999, 40, 2939.
and indeed shows a significantly lower barrier to rotation (ca.
À1
9
kcal mol ) than for 2, consistent with nbd being less bulky than
5.
6.
7.
Burk, M. J. Acc. Chem. Res. 2000, 33, 363.
cod.
Lagasse, F.; Kagan, H. B. Chem. Pharm. Bull. 2000, 48, 315.
Claver, C.; Fernandez, E.; Gillon, A.; Heslop, K.; Hyett, D. J.; Martorell, A.; Orpen,
A. G.; Pringle, P. G. Chem. Commun. 2000, 961.
3
. Conclusion
8.
van den Berg, M.; Minnaard, A. J.; Schudd, E. P.; van Esch, J.; de Vries, A. H. M.;
de Vries, J. G.; Feringa, B. J. Am. Chem. Soc. 2000, 122, 11539.
The variable temperature NMR spectra of complexes
Rh(CgPH) (diene)]BF 1–3 are consistent with restricted Rh–P
9. Reetz, M. T.; Mehler, G. Angew. Chem., Int. Ed. 2000, 39, 3889.
1
0. (a) de Vries, J. G.; Lefort, L. Chem. Eur. J. 2006, 12, 4722. and refs therein; (b)
Reetz, M. T.; Mehler, G.; Meiswinkel, A. Tetrahedron: Asymmetry 2004, 15, 2165.
and refs therein; (c) Jäkel, C.; Paciello, R. Chem. Rev. 2006, 106, 2912.
11. Hopewell, J.; Jankowski, P.; McMullin, C. L.; Orpen, A. G.; Pringle, P. G. Chem.
Commun. 2010, 46, 100.
[
2
4
rotation. It is possible that this rigidity is key to explaining the
enantioselectivities observed in asymmetric hydrogenations with
catalysts derived from the CgPH monophos ligand. Further work
is in progress to explore the generality of restricted M–P rotation
in monophos complexes and whether any conclusions can be
drawn for the design of effective monophos ligands for asymmetric
catalysis.
1
2. The designation of
shorthand for the systematic names. The
(1S,3S,5R,7R)-6-phospha-2,4,8-trioxa-1,3,5,7-tetramethyladamantane.
3. The crystal structures of -CgPOOH and -CgPH will be reported elsewhere.
4. Crystal data for [Rh(cod)( -CgPH) ]BF : orthorhombic, space group C222
a = 9.7590(3) Å, b = 16.1220(5) Å, c = 20.5296(6) Å, at 100 K, Z = 4,
a
and b to the enantiomers of CgPH has been made as
a-enantiomer used in this article is,
1
1
a
a
a
2
4
1
,
3
À1
V = 3230.02(17) Å ,
independent (Rint = 0.0241), R1 = 0.0191, Flack parameter = À0.022(15).
5. Restricted rotation in cis-M(PR complexes has been previously reported.
l = 0.690 mm , 21706 reflections collected, 4928
Acknowledgements
1
3 2
)
See: Baber, R. A.; Haddow, M. F.; Middleton, A. J.; Orpen, A. G.; Pringle, P. G.;
Haynes, A.; Williams, G. L.; Papp, R. Organometallics 2007, 26, 713–725.
6. The 30 ppm difference in the 31P NMR shifts of the signals for 2b is large and is
worthy of further investigation.
We thank Natalie Fey, Martin Murray and Craig Butts for help-
ful discussions of this work, Prof. S. Kuwata for help with the
crystal structure of L , EPSRC and CCDC for a studentship (to
a
1
1
7. Crystal data for [Rh(nbd)(
4 1
a-CgPH)(b-CgPH)]BF : monoclinic, space group P2 /c,
C.L.M.), COST action CM0802 ‘PhoSciNet’ for support, The Royal
Society for an Anglo-Japanese Exchange Grant and Johnson-
Matthey for a loan of precious metal compounds. We would also
like to thank the Centre for Computational Chemistry at the
University of Bristol for access to the clusters used for all calcula-
tions reported here.
a = 21.0504(3) Å, b = 9.23740(10) Å, c = 15.7936(2) Å, = 98.7150(10), at 100 K,
3
À1
Z = 4, V = 3035.62(7) Å ,
l = 0.732 mm , 55512 reflections collected, 7928
independent (Rint = 0.0524), R1 = 0.0309.
1
1
8. Signals for rac and meso diastereoisomers not resolved.
9. Addition of 1 equiv of CgPH to the solutions of led to broadened 31P NMR
signals for 2/3. For example, a sample at 300 K where w1/2 was ca. 800 Hz
increased to ca. 1000 Hz and a broad signal for free CgPH; at 383 K, a single
resonance was observed with no Rh–P coupling evident indicating that
intermolecular ligand exchange can be significant in the presence of added
ligand.
References
2
1
0. In principle there are two ab-diastereoisomers of the C -symmetric 3b but only
one set of signals was observed presumably because one isomer dominates or
the signals were overlapping.
1
.
(a) Knowles, W. S.; Sabacky, M. J.; Vineyard, B. D. Chem. Commun. 1972, 10; (b)
Knowles, W. S. Acc. Chem. Res. 1983, 16, 106.