Carboxylate-directed highly stereoselective homogeneous hydrogenation of cyclic olefins with Wilkinson's catalyst

Article abstract of DOI:10.1021/ol034464o

(Matrix presented) A highly efficient diastereoselective, carboxylate-directed homogeneous hydrogenation of cyclic olefins with use of Wilkinson's catalyst is described. Under the optimized reaction conditions, better than 99% de was achieved. The experimental protocol is very simple and readily amenable to scale-up.

Full text of DOI:10.1021/ol034464o

ORGANIC
LETTERS
2003
Vol. 5, No. 9
1587-1589
Carboxylate-Directed Highly
Stereoselective Homogeneous
Hydrogenation of Cyclic Olefins with
Wilkinson’s Catalyst
Mingbao Zhang,* Lei Zhu, Xin Ma, Miao Dai, and Derek Lowe
Department of Chemistry Research, Bayer Research Center,
400 Morgan Lane, West HaVen, Connecticut 06516
mingbao.zhang.b@bayer.com
Received March 17, 2003
ABSTRACT
A highly efficient diastereoselective, carboxylate-directed homogeneous hydrogenation of cyclic olefins with use of Wilkinson’s catalyst is
described. Under the optimized reaction conditions, better than 99% de was achieved. The experimental protocol is very simple and readily
amenable to scale-up.
3
Heteroatom-directed, homogeneous hydrogenation of olefins
in the presence of a transition metal catalyst provides a
powerful approach to regio- and stereochemical control in
organic synthesis.1 The directive effects arise when the
directing heteroatom, hydrogen, and the alkene simulta-
neously bind to the transition metal, permitting exclusive syn
delivery of hydrogen to the unsaturation site. The Wilkin-
son’s catalyst RhCl(PPh3)3 worked successfully with olefins
bearing an alkoxide directing group, providing the first
example of stereoselective directed hydrogenation.2 It un-
fortunately has failed to work with more commonly used
directing groups such as hydroxyl, ethers, esters, and amides.
For this reason, subsequent research in this area has been
focused on cationic rhodium and iridium catalysts such as
[Rh(nbd)(diphos-4)]BF4 and [Ir(cod)py(PCy3)]PF6 (Crab-
tree’s catalyst)4 which work effectively with a variety of
directing groups.5 Recently, we had a need for a simple and
scaleable diastereoselective hydrogenation method. We re-
visited Wilkinson’s catalyst because it is air and water stable
and readily commercially available.6 We reasoned that,
although hydroxyl and others such as esters, ethers, and
amides cannot displace the chloride ligand of the Wilkinson’s
catalyst, a key requirement for effecting a directed hydro-
genation,7 the carboxylate, a better nucleophile,8 may be able
(3) Brown, J. M.; Chaloner, P. A.; Kent, A. G.; Murrer, B. A.; Nicholson,
P. N.; Parker, D.; Sidebottom, P. J. J. Organomet. Chem. 1981, 216, 263.
(4) (a) Crabtree, R. H.; Felkin, H.; Morris, G. E. J. Organomet. Chem.
1977, 141, 205. (b) Crabtree, R. H. Acc. Chem. Res. 1979, 12, 331. (c)
Crabtree, R. H.; Demou, P. C.; Eden, D.; Mihelcic, J. M.; Parnell, C. A.;
Quirk, J. M.; Morris, G. E. J. Am. Chem. Soc. 1982, 104, 6994.
(5) (a) Stock, G.; Kahne, D. E. J. Am. Chem. Soc. 1983, 105, 1072. (b)
Schultz, A.; McCloskey, P. J. J. Org. Chem. 1985, 50, 5905. (c) Brown, J.
M.; Hall, S. A. Tetrahedron Lett. 1984, 25, 1393. (d) Brown, J. M.; Naik,
R. G. J. Chem. Soc., Chem. Commun. 1982, 348. (e) Takagi, M.; Yamamoto,
K. Tetrahedron 1991, 47, 8869.
(6) Price from 2002-2003 Aldrich catalog: Wilkinson’s catalyst,
$70 000/mol; Crabtree’s catalyst, $700 000/mol.
(7) With Wilkinson’s catalyst, displacement of the chloride ligand by
the directing group is required for efficient stereoselective delivery of
hydrogen. For more in-depth discussion on this point, see ref 1.
(1) Hoveyda, A. H.; Evans, D. A.; Fu, G. C. Chem. ReV. 1993, 93, 1307
and references therein.
(2) (a) Thompson, H. W.; Naipawer, R. E. J. Am. Chem. Soc. 1973, 95,
6379. (b) Thompson, H. W.; McPherson, E. J. Am. Chem. Soc. 1974, 96,
6232. For applications of the Thompson method in the synthesis of complex
organic molecules, see: (c) McCombie, S. W.; Cox, B.; Lin, Sue-Ing;
Ganguly, A. K.; McPhail, A. T. Tetrahedron Lett. 1990, 2083. (d) Callam,
C. S.; Lowary, T. L. Org. Lett. 2000, 2, 167. (e) Preston, S. A.; Cupertino,
D. C.; Palma-Ramirez, P.; Cole-Hamilton, D. J. J. Chem. Soc., Chem.
Commun. 1986, 977.
10.1021/ol034464o CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/29/2003

to do so, thus permitting stereochemical control.9 In this
letter, we wish to describe our findings of using the
carboxylate group in directed stereoselective hydrogenation
of cyclic olefins with Wilkinson’s catalyst.
directed hydrogenation. The two faces of the olefin in these
substrates are diastereotopic due to the preferred orientation
of the allylic proton, as shown for the indene substrate 1
(Figure 2) that minimizes interactions with the peri-ArH
Olefinic substrates 1-3 (Figure 1) were readily obtained
via Reformatsky reaction10 from the appropriate ketones and
Figure 2. Proposed origin of stereoselectivity of the carboxylate-
directed olefin hydrogenation.
Figure 1. Structures of olefinic substrates and their corresponding
hydrogenation products.
proton. The carboxylate anion enters the coordination sphere
of the rhodium complex by displacing the chloride ligand.
The net effect of such coordination is that delivery of
hydrogen atoms from the catalyst to the double bond takes
place preferentially from the same side as the carboxylate
group, leading to predominate formation of 1a.
Further optimization of the reaction conditions to increase
throughput led to the identification of the amine salts as better
counterions than the sodium salts. The amine salts generally
are more soluble in organic solvents, which permits hydro-
genation to proceed at higher concentrations. The results are
summarized in Table 1. We began by hydrogenating a
bromoesters followed by saponification. The carboxylates
were generated in situ via addition of an appropriate base.
We initially examined the hydrogenation of the sodium
carboxylate of the indene substrate 1 generated in situ by
sodium bicarbonate in an ethanol/water (2:1) mixture. With
2 equiv of sodium bicarbonate, the hydrogenation proceeded
cleanly at 60 psi with 5 mol % catalyst loading to produce
both diastereomers 1a and 1b with a ratio of 98:2 in favor
of 1a (Scheme 1). The diastereomeric ratio was determined
Scheme 1
Table 1. Diastereoselective Hydrogenation of 1 with Use of
Wilkinson’s Catalyst
cat.,
time
(h)
yield
(%)
runs mol %
solvents
base/equiv
de
96
>99
99
90
88
>99
>99
>99
>99
1
2
3
4
5
6
7c
5
5
5
1
5
5
5
EtOH/H2O NaHCO3/2.0
THF/EtOH quinine/1.0
THF/EtOH TEA/1.5
THF/EtOH TEA/1.5
THF/EtOH none
36
5
90a
30b
98a
90b
95b
20b
40b
50b
95a
12
24
18
12
36
72
24
1
by both H NMR (δ values for the peri-ArH in 1a and 1b
are 7.03 and 7.10, respectively, in DMSO-d6) and chiral
HPLC, which allowed baseline separation of all four dia-
stereomers.11
THF
TEA/1.5
THF/EtOH TEA/1.5
8d
5
THF/EtOH TEA/1.5
The stereochemistry of 1a was determined by single-
crystal X-ray crystallography. The racemate of 1a was readily
resolved by (R)-R-methyl benzylamine. The X-ray of the less
soluble salt showed an (SS)-configuration for the indane acid.
The observed stereochemical outcome of the hydrogena-
tion can be readily rationalized by using the concept of
a Isolated yield. b Conversion of substrate based on 1H NMR. c 1 atm of
H2. d 30 psi of H2.
diastereomerically pure quinine salt from the resolution of
1.12 The reaction proceeded very cleanly at 60 psi in EtOH/
THF (9:1) with 5 mol % of Wilkinson’s catalyst to yield
the S,S-enantiomer in >99% de. When a racemate of 1 was
hydrogenated in the presence of 1.0 equiv of quinine under
the same conditions, >99% de was observed (entry 2, Table
1). While quinine is not the most preferred base for our
purposes (high molecular weight and double bond), these
results clearly demonstrated the superior performance of an
amine as the base due to better solubility. We examined
(8) Nucleophilic constants for CH3OH 0.0, CH3COO- 4.3, Cl- 4.37,
CH3O- 6.29. Data from: Pearson, R. G.; Sobel, H.; Songstad, J. J. Am.
Chem. Soc. 1968, 90, 319.
(9) Base-promoted coordination of the carboxyl to Rh via the carboxylate
anion has been indicated to be responsible for enhancing enatioselectivity
and catalyst turnovers in enatioselective hydrogenations catalyzed by
rhodium complexes of chiral ligands. See: (a) Ojima, I.; Kogure, T.; Yoda,
N. J. Org. Chem. 1980, 45, 4728. (b) Valentine, D., Jr.; Sun, R. C.; Toth,
K. J. Org. Chem. 1980, 45, 3703.
(10) (a) Shriner, R. L. Org. React. 1942, 1, 1. (b) Rathke, M. W. Org.
React. 1975, 22, 423.
1588
Org. Lett., Vol. 5, No. 9, 2003

triethylamine (TEA) as a lightweight and more economical
alternative base. With 1.5 equiv of TEA, >99% de was
obtained with 5 mol % of Wilkinson’s catalyst at 60 psi
hydrogen pressure in ethanol/THF (9:1) (entry 3). Better than
90% de could be achieved even with 1 mol % of the catalyst
under these conditions (entry 4). With the free acid, only
88% de was observed in the ethanol/THF system (entry 5).
This result is not surprising, since the COOH group must
undergo dissociation to generate the COO- before it can
bind to the rhodium catalyst. This hypothesis is further
supported by the following observation: when the corre-
sponding methyl ester of 1 was subjected to hydrogenation
under these conditions (with no base present), no de was
observed. Obviously, the ester group is not capable of any
significant coordination to the transition metal in this case,
and therefore the hydrogenation was nonselective. The
ethanol/THF ratio has a very noticeable effect on the reaction
rates, and the optimized ethanol/THF ratio is 9:1 (v/v). When
pure THF was used, the hydrogenation proceeded very
slowly although with similarly high diastereoselectivity (entry
6). Addition of ethanol dramatically increased the reaction
rate. This observation can be easily explained by the high
solubility of hydrogen in ethanol.13 Yet, THF is still needed
in our system because it appears to offer better solubility
for the catalyst and substrates, thus permitting the reaction
to proceed at high concentrations. It is significant that these
optimized conditions also permit the hydrogenation to
proceed smoothly at a much lower hydrogen pressure (entries
7 and 8). The simplicity of the procedure has enabled us to
conduct the hydrogenation on greater than 100-g scale.
For comparison, we also examined the heterogeneous
hydrogenation of 1. It proceeded smoothly in ethanol
(Pearlman’s catalyst, ammonium formate as the hydrogen
source) to afford a mixture of diastereomers 1a and 1b in a
ratio of 1:2. The hydrogenation of the methyl ester of 1 also
showed similar stereoselectivity (1:3).
We also examined the homogeneous hydrogenation of the
substrates 2 and 3 using our optimized conditions. Under
these conditions, both substrates afforded the desired hy-
drogenation products (2a and 3a, respectively) with better
than 99% de.14
In summary, we have found that carboxylate group is a
very effective directing group for directed homogeneous
hydrogenation with Wilkinson’s catalyst. Under our reaction
conditions, indene and 3,4-dihydronaphthalene acids are
hydrogenated smoothly via their amine carboxylates in a
highly stereoselective manner. The ease of handling and
relatively low cost of the catalyst make this methodology a
very practical one. Furthermore, since carboxyl group is
readily interconvertable with a variety of functional groups
such as esters, amides, hydroxyl, etc., this methodology thus
constitutes a very practical alternative to the cationic catalyst-
based directed stereoselective hydrogenation.5
(11) Column, Chiracel AD, 4.6 (i.d.) × 250 mm; mobile phase A, 0.1%
TFA (trifluoroacetic acid) in hexanes, mobile phase B, 0.1% TFA in IPA
(isopropyl alcohol); method, isocratic 95% A (5%B), 20 min; flow rate,
1.5 mL/min; detector (UV), 284 nm. Retention times for the four
diastereomers are 5.163 (SR), 6.255 (RS), 10.262 (RR), and 14.399 min
(SS); the first letter denotes the absolute configuration of the carbon adjacent
to the carboxyl group. The stereochemistry assignment for each peak is
described as the following: a nonequal racemic diastereomeric mixture was
analyzed by chiral HPLC to obtain 4 baseline-resolved peaks. Peaks 3 and
4, 1 and 2 are enantiomer pairs based on UV integration. The absolute
configuration of peak 4 is determined to be SS by X-ray structural analysis.
Peak 3 can then be assigned a RR configuration with certainty. The absolute
configurations of peaks 1 and 2 were determined by another experiment.
Optically active (S)-indene acid (96% ee) from the chemical resolution was
subjected to diastereoselective hydrogenation. The indane acid obtained was
then analyzed by chiral HPLC. Due to high diastereoselectivity of the
hydrogenation (>99% de), only the (SR)-diastereomer peak should be
detectable by HPLC (retention time 5.363 min, ca. 0.97% area) in addition
to the desired SS diastereomer (major) and its enantiomer RR. The 2nd
peak from the 1st experiment (6.255 min) can then be assigned an RS
configuration with certainty.
Acknowledgment. The authors thank Jordi Benet-Buch-
holz for X-ray analysis, Laszlo Musza and Jeff Chin for
assistance in NMR spectrum analysis, and Timothy He and
Anthony Paiva for HRMS. We are indebted to Dr. Donald
Wolanin for proof reading this manuscript.
Supporting Information Available: Complete experi-
mental details and X-ray crystal structure (ORTEP repre-
sentation) of the (R)-R-methyl benzylanine salt of (SS)-1a.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) The racemate of the indene substrate (1) can be resolved chemically
with quinine prior to the stereoselective hydrogenation. The resolution
yielded the (S)-enantiomer as the less soluble diastereomeric salt in
acetonitrile (34%yield, 96-97% ee). The free acid was liberated by
dissolving the salt in 2 N aqueous HCl followed by extraction into methylene
chloride. The (SS)-enantiomer of 1a was then readily obtained in high ee
after hydrogenation.
OL034464O
(14) The stereochemistry assignments of 2a and 3a are based on analogy
to that of 1a and are consistent with nOe and coupling constants data. The
de measurements are based on 1H NMR integration. In both cases, the minor
diastereomer was not detected.
(13) CRC Handbook of Chemistry and Physics; Weast, R., Ed.; The
Chemical Rubber Co.: Cleveland, OH, 1969; p B-114.
Org. Lett., Vol. 5, No. 9, 2003
1589