3,3′-disubstituted BINAP ligands: Synthesis, resolution, and applications in asymmetric hydrogenation

Article abstract of DOI:10.1021/ol051413d

(Chemical Equation Presented) A novel family of BINAP ligands were prepared with alkoxy- and acetoxy-derived substituents in the 3,3′-positions. They were prepared through a convergent synthesis starting from readily available 4-bromo-2-naphthol. These ligands afforded excellent enantioselectivities in the asymmetric hydrogenation of substituted olefins. The presence of the 3,3′-substituents was shown to be beneficial by a direct comparison with the parent unsubstituted BINAP.

Full text of DOI:10.1021/ol051413d

ORGANIC
LETTERS
2005
Vol. 7, No. 17
3765-3768
3,3′-Disubstituted BINAP Ligands:
Synthesis, Resolution, and Applications
in Asymmetric Hydrogenation
J. Matthew Hopkins, Sean A. Dalrymple, Masood Parvez, and Brian A. Keay*
Department of Chemistry, UniVersity of Calgary, 2500 UniVersity DriVe NW,
Calgary, Alberta, T2N 1N4, Canada
keay@ucalgary.ca
Received June 17, 2005
ABSTRACT
A novel family of BINAP ligands were prepared with alkoxy- and acetoxy-derived substituents in the 3,3
′
-positions. They were prepared through
a convergent synthesis starting from readily available 4-bromo-2-naphthol. These ligands afforded excellent enantioselectivities in the asymmetric
hydrogenation of substituted olefins. The presence of the 3,3′-substituents was shown to be beneficial by a direct comparison with the parent
unsubstituted BINAP.
Catalytic asymmetric synthesis is one of the most powerful
methods for the construction of an array of enantiomerically
enriched compounds. The field of ligand design, specifically
the design of new chiral bisphosphine ligands, is crucial for
the further development of highly efficient transition metal-
catalyzed processes. Despite the fact that a diverse array of
phosphine ligands are known, continued entries into this class
are of great importance. We report here a convergent
approach to a novel class of substituted BINAP ligands (2-
5), containing substitution at the 3,3′-positions. Greatly
enhanced levels of enantioselectivity in the asymmetric
hydrogenation of substituted olefins over that of unsubstituted
BINAP (1) illustrate the potential benefits of this extra
substitution.
The biaryl family of bisphosphine ligands, pioneered by
BINAP (1, Figure 1),1 have proven to be excellent ligands
Figure 1. 3,3′-Disubstituted biaryl bisphosphine ligands.
(1) (a) Noyori, R.; Takaya, H. Acc. Chem. Res. 1990, 23, 345. (b)
Ohkuma, T.; Kitamura, M.; Noyori, R. Catalytic Asymmetric Synthesis;
Ojima, I., Ed.; VCH: New York, 2000. (c) Noyori, R. Asymmetric Catalysis
In Organic Synthesis; Wiley: New York, 1994.
for many transition metal-catalyzed asymmetric reactions.2
Given the fact that BINAP is not a superior ligand for all of
10.1021/ol051413d CCC: $30.25
© 2005 American Chemical Society
Published on Web 07/21/2005

these processes, it is obvious that varied electronic and
structural properties of new ligands can lead to greater
selectivity. We envisioned using the successful biaryl
framework and incorporating structural variations at a
position close to the site of metal coordination, the 3,3′-
positions, as an entry into a new class of ligands. In a recent
review of modified BINAP ligands, Lemaire states that the
3 and 3′ are the positions for which introduction of further
substitution could most influence the electronic density of
the phosphorus donor and the steric environment around the
catalytic site.3 Looking at the positive results that have been
obtained in many cases with 3,3′-disubstituted BINOL
ligands,4,5 and given the fact that this substitution pattern is
absent from current modified BINAP ligands,6,7 it was felt
that an approach to this class would be beneficial. Previous
disclosures of 3,3′-disubstituted biaryl bisphosphine ligands
have been made for the analogous biphenyl class of ligands.
Zhang’s o-Ph-hexaMeO-BIPHEP8 (6) and o-Ph-MeO-
BIPHEP9 (7) have proven to be successful in the asymmetric
hydrogenation of cyclic enamides and dehydroamino acids.
Results from our lab using BIPHEP derivatives 8-15 have
proven to be successful catalysts in asymmetric Heck and
hydrogenation reactions.10
Scheme 1. Synthesis of 3,3′-Disubstituted BINAP Derivatives
Our synthesis began from readily available 4-bromo-2-
naphthol (16),11 which was treated with diphenylphosphinic
chloride in the presence of triethylamine and DMAP to afford
phosphinate 17 (Scheme 1). Regioselective migration of the
diphenylphosphinyl unit upon treatment of 17 with LDA
afforded phosphine oxide 18.12 The existence of a free OH
group proved to be important, as it provided an attachment
point for a chiral auxiliary, which upon Ullmann coupling,
would facilitate the formation of diastereomers for separation
of the generated axial isomers. Reaction of 18 with (+)-
lactic acid-derived (S)-2-acetoxypropanoyl chloride13 af-
forded 19, which underwent subsequent Ullmann coupling
to generate a mixture of diastereomers 20 and 21. The
coupling proceeded with a moderate diastereoselectivity
(≈2:1 by 31P NMR), and the major diastereomer could be
isolated readily by column chromatography14 and trituration
with tert-butylmethyl ether. Facile saponification of the chiral
auxiliary afforded optically pure (S)-3,3′-(OH)2-BINAP(O)
22, which could readily be converted into a variety of
substrates. The phosphine oxides 23-26 were formed by
standard methods, which after reduction with trichlorosilane
yielded phosphines 2-5. The absolute stereochemistry of
the axis of chirality for the major diastereomer from the
Ullmann coupling was determined to be Sax from the X-ray
crystal structure of the phosphine selenide derived from (S)-2
upon crystallization from EtOAc/hexanes (Figure 2).15
To better understand the effects of the 3,3′-substituents
on metal coordination, the newly synthesized ligands were
reacted with a stoichiometric amount of (PhCN)2PdCl2 to
generate the corresponding palladium dichloride adducts (see
Table 1). X-ray crystal structure analysis16 of (S)-27 and (S)-
28 (Supporting Information) allowed for a direct comparison
with the known BINAP structure.17 It was observed that the
average Pd-P and Pd-Cl bonds remained fairly constant
for all of the ligands and that the chiral pocket found in the
(2) Tang, W.; Zhang, X. Chem. ReV. 2003, 103, 3029.
(3) (a) Berthod, M.; Mignani, G.; Woodward, G.; Lemaire, M. Chem.
ReV. 2005, 105, 1801. (b) For a recent review on biaryl bisphosphine ligands,
see: Shimizu, H.; Nagasaki, I.; Saito, T. Tetrahedron 2005, 105, 857.
(4) (a) Chen, Y.; Yekta, S.; Yudin, A. Chem. ReV. 2003, 103, 3155. (b)
Kocovsky, P.; Vyskocyl, S.; Smrcina, M. Chem. ReV. 2003, 103, 3213. (c)
Pu, L. Chem. ReV. 1998, 98, 2405.
(5) Brunel, J. M. Chem. ReV. 2005, 105, 857.
(6) For a 4,4′-disubstituted BINAP, see: Ngo, H. L.; Lin, W. J. Org.
Chem. 2005, 70, 1177. For a 5,5′-derivative, see: Deng, G.; Fan, Q.; Chen,
X.; Liu, D.; Chan, A. S. Chem. Commun. 2002, 1570. For a 6,6′-derivative,
see: Ngo, H. L.; Hu, A.; Lin, W. Chem. Commun. 2003, 1912. For a 7,7′-
derivative, see: Che, D.; Andersen, N. G.; Lau, S. Y. W.; Parvez, M.; Keay,
B. A. Tetrahedron: Asymmetry 2000, 11, 1919.
(7) To the best of our knowledge, no report of the synthesis, characteriza-
tion or applications of any 3,3′-disubstituted BINAP ligands have been
reported to date. While the Lemaire review of ref 3 cites a patent that outlines
a synthetic proposal for such ligands, no experimental detalis were provided.
Zhang, X. (The Penn State Research Foundation). PCT. Int. Appl. WO 02/
40491, 2002.
(8) Tang, W.; Chi, Y.; Zhang, X. Org. Lett. 2002, 4, 1695.
(9) Wu, S.; He, M.; Zhang, X. Tetrahedron: Asymmetry 2004, 15, 2177.
(10) (a) Gorobets, E.; Sun, G.; Wheatley, B. M. M.; Parvez, M.; Keay,
B. A. Tetrahedron Lett. 2004, 45, 3597. (b) Gorobets, E.; Wheatley, B. M.
M.; Hopkins, J. M.; McDonald, R.; Keay, B. A. Tetrahedron Lett. 2005,
46, 3843.
(11) Newman, M. S.; Sankaran, V.; Olson, D. R. J. Am. Chem. Soc.
1976, 98, 3237.
(12) This regioselective migration has been observed previously for the
nonbrominated npahthalene; see: Dhawan, B.; Redmore, D. J. J. Org. Chem.
1991, 56, 833.
(14) Rf of 16 ) 0.29, Rf of 17 ) 0.18 (5% MeOH/CHCl3).
(15) Crystal data for the phosphine selenide of (Sax)-2: orthorhombic
P212121; a ) 11.4539(12) Å, b ) 15.6457(17) Å, c ) 23.672(7) Å, R )
90°, â ) 90°, γ ) 90°, V ) 4242.1(14) Å3; Z ) 4; R ) 0.032; Rw ) 0.066.
The absolute structure was determined by the Flack method (Flack, H. D.
Acta Crystallogr. 1983, A39, 876). The Flack parameter for the inverted
structure was 1.003(6). Therefore, the inverted structure was rejected as
the one present in the crystal.
(13) Babudri, F.; Fiandanese, V.; Marchese, G.; Punzi, A. Tetrahedron
1999, 55, 2431.
3766
Org. Lett., Vol. 7, No. 17, 2005

been placed upon the metal, which will most certainly alter
substrate binding for further transformation.
Ligands (S)-2-5 were used in the hydrogenation of
substituted olefins, and the results were compared directly
to those with unsubstituted BINAP (1) under identical
reaction conditions. Table 2 contains results for the hydro-
Table 2. Asymmetric Hydrogenation of 2-Acetamidoacrylic
Acid Derivatives with Ligands (S)-1-5
entrya
ligand
olefin
ee (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
(S)-1
29c
30
31
29c
30
31
29c
30
31
29c
30
31
29c
30
31
33.8 (R)
21.2 (R)
14.8 (R)
94.9 (S)
98.0 (S)
23.2 (S)
98.6 (S)
>99 (S)
32.8 (S)
>99 (S)
>99 (S)
74.7 (S)
92.3 (S)
96.3 (S)
44.8 (S)
Figure 2. X-ray crystal structure of the phosphine selenide of (S)-
2.
(S)-2
(S)-3
(S)-4
(S)-5
Table 1. Structural Parameters for PdCl2 Complexes of
(S)-1-3
a All reactions proceeded with complete conversion. Reactions of olefins
29 and 30 used Rh(COD)2OTf as the Rh source. Reactions of olefin 31
used Rh(NBD)2BF4 as the Rh source. b For 33, determined by chiral GC
on a Cyclodex B column; for 34, determined by chiral HPLC on a Chiralcel
OD column. c Ee determined upon conversion to the methyl ester.
average average
ligand
torsion
Pd-P
Pd-Cl bite angle angle
entry 3,3′-substituent
(Å)
(Å)
(deg)
(deg)
1
2
3
H
OiPr
OMe
2.245
2.290
2.270
2.350
2.338
2.355
92.77
89.91
88.09
68.42
62.64
60.25
genation of acrylic and cinnamic acid derivatives. The
catalyst was prepared in situ by stirring a solution of the Rh
precursor and the phosphine ligand prior to introduction of
the olefin and being pressurized under H2. These new ligands
proved to be very selective for the hydrogenation of the
acrylate derivatives when compared to BINAP. Very low
levels of enantioselectivity were observed for BINAP (entries
1 and 2), whereas introduction of groups in the 3,3′-positions
led to dramatic increases in selectivity. Of the four deriva-
tives, the OBn analogue (S)-5 gave the lowest selectivity of
the three (entries 13 and 14), while the OMe and OCOtBu
derivatives produced nearly enantiopure products (entries 7,
8, 10, and 11). It is interesting to note that the 3,3′-
disubstituted ligands provided the product of the opposite
configuration to that of BINAP, despite the fact that all of
the ligands were of the Sax configuration. When a more
hindered olefin was subjected to similar reduction conditions
(Table 2, R1 ) Ph), the enantioselectivity levels were still
greater than that of BINAP (entry 3 vs entries 6, 9, 12, and
15) but were much lower compared to the results obtained
with the disubstituted olefin. The OCOtBu ligand (S)-4
provided the best results with 74.7% ee, and as was observed
BINAP structure, with one of the phenyl substituents on each
of the phosphorus atoms undergoing a π-stacking interaction
with the binaphthalene backbone, was maintained upon
introduction of the 3,3′-substituents. The ligand bite angles
( P-Pd-P) were compared, and for both the complexes, a
tightening of this angle was observed compared to BINAP.
Directly related to this is the torsion angle ( C2-C1-C1′-
C2′) formed by the twisting of the binaphthalene backbone.
A decrease in this angle was again observed for the 3,3′-
complexes. These structural data suggest that a rigid chiral
pocket still remains upon incorporation of groups in the 3,3′-
positions of BINAP but that significant steric constraints have
(16) Crystal data for (S)-27: orthorhombic P212121; a ) 14.653(2) Å, b
) 17.076(2) Å, c ) 18.442(2) Å, R ) 90°, â ) 90°, γ ) 90°, V )
4614.5(10) Å3, Z ) 4, R ) 0.040; Rw ) 0.085. Crystal data for (S)-28:
triclinic P1; a ) 12.359 (5) Å, b ) 12.929(5) Å, c ) 14.434 (5) Å, R )
76.435(5)°, â ) 69.693(5)°, γ ) 74.968(5)°, V ) 2062.3(14) Å3, Z ) 2, R
) 0.0399; Rw ) 0.1120.
(17) Ozawa, F.; Kubo, A.; Matsumoto, Y.; Hayashi, T.; Nishioka, E.;
Yanagi, K.; Moriguchi, K. Organometallics 1993, 12, 4188.
Org. Lett., Vol. 7, No. 17, 2005
3767

for the less hindered olefin, the ligands with 3,3′-substituents
afforded the product with the opposite stereochemistry to
that of BINAP of the same axial configuration.
The asymmetric hydrogenation of enamides is another
reaction often used to probe the efficacy of new bisphosphine
ligands. Table 3 contains results for the hydrogenation of a
This additional substitution leads to a different coordination
environment around the metal center upon ligation, leading
to enhanced asymmetric induction. The well studied mech-
anism of hydrogenation for these types of olefins involves
the formation of two diastereomeric Rh-olefin complexes.18
It was shown that the minor diastereomeric complex of this
mixture actually leads to the observed major product. Given
the reversal of stereochemistry observed with our ligands, it
appears that the introduction of 3,3′-substituents alters the
equilibrium mixture of the two diastereomeric olefin com-
plexes, allowing the major one to react with H2 and lead to
the product with the opposite stereochemistry. Further studies
to probe this mechanism are currently under way.
Table 3. Asymmetric Hydrogenation of
N-Acetyl-phenylethenamide with Ligands (S)-1-5
In summary, we have devised a short, convergent synthesis
to this novel class of phosphine ligands and have demon-
strated their potential for improvement in processes in which
BINAP itself gives poor results. We are continuing to pursue
the synthesis of further derivatives of this class with new
3,3′-substituents and expanding the scope of their applications
as well as investigating the cause for the reversal in absolute
configuration of the hydrogenation products. These results
will be reported in due time.
entrya
ligand
ee (%)b
1
2
3
4
5
(S)-1
(S)-2
(S)-3
(S)-4
(S)-5
10.7 (R)
89.9 (S)
26.9 (S)
85.6 (S)
59.6 (S)
a All reactions proceeded with complete conversion. b Determined by
chiral GC using a Cyclodex B column.
Acknowledgment. Financial support for this work was
porvided by Merck Frosst Canada and by the Natural
Sciences and Engineering Research Council of Canada
through a CRD grant to B.A.K. and an NSERC postgraduate
scholarship to J.M.H. The Alberta Ingenuity Fund is also
thanked for a studentship to J.M.H. Thanks are extended to
S.A.D. and M.P. for crystal structure determination.
simple enamide. The catalyst was again prepared in situ, and
while the results obtained with this substrate did not provide
levels of induction as high as those in Table 2, significant
improvements over BINAP were observed for ligands (S)-2
and (S)-4. It was again observed that the opposite product
enantiomer was formed for the 3,3′-disubstituted ligands
versus that with BINAP, despite the fact that all of the ligands
were of the Sax configuration. These results further suggest
that the incorporation of groups ortho to the phosphine
substituents can have a beneficial effect on enantioselectivity
for processes in which the lack of substitution provides low
levels of enantioselectivity.
Supporting Information Available: Full experimental
procedure for the synthesis of ligands 2-5, along with
spectral data, and a general procedure for the hydrogenation
of substituted olefins and chiral GC data. This material is
available free of charge via the Internet at http://pubs.acs.org.
As these early results indicate, substituents at the 3,3′-
positions of BINAP can indeed have a beneficial effect on
enantioselectivity when compared to the unsubstituted ligand.
OL051413D
(18) Landis, C. R.; Halpern, J. J. Am. Chem. Soc. 1987, 109, 1746.
3768
Org. Lett., Vol. 7, No. 17, 2005