Asymmetric catalytic synthesis of α-aryloxy alcohols: Kinetic resolution of terminal epoxides via highly enantioselective ring-opening with phenols

Full text of DOI:10.1021/ja9910917

6086
J. Am. Chem. Soc. 1999, 121, 6086-6087
Scheme 1
Asymmetric Catalytic Synthesis of r-Aryloxy Alcohols:
Kinetic Resolution of Terminal Epoxides via Highly
Enantioselective Ring-Opening with Phenols
Joseph M. Ready and Eric N. Jacobsen*
Department of Chemistry and Chemical Biology
HarVard UniVersity, Cambridge, Massachusetts 02138
ReceiVed April 6, 1999
Enantiopure R-aryloxy alcohols (1) are valuable targets for
asymmetric synthesis as a result of their role as key synthetic
intermediates in a variety of pharmaceutically important com-
pounds.1 In principle, access to these building blocks may be
provided by several routes, including asymmetric reduction of
aryloxy ketones2 or the ring opening of enantiopure terminal
epoxides with phenols. Of these, the latter is probably the most
versatile and direct, but available methods for the addition of
phenols to epoxides are extremely limited. No catalytic methods
have been devised for phenolic opening of terminal epoxides,3
and forcing conditions are required for the uncatalyzed reaction,
such as heating epoxide in the presence of a phenoxide salt to
high temperatures in a polar solvent. These thermal methods are
generally low-yielding and are particularly unsuitable for sensitive
substrates. Thus, despite the recent discovery of general methods
for accessing terminal epoxides in high optical purity,4 the
development of routes to enantiopure R-aryloxy alcohols via
epoxide ring-opening with phenols remains an unsolved problem.
The ready accessibility of terminal epoxides in racemic form
renders kinetic resolution of terminal epoxides with phenols a
potentially attractive route to 1 (Scheme 1, Nu ) OAr). The high
selectivities obtained in the recently reported hydrolytic kinetic
resolution of terminal epoxides with catalyst 3b4 (Scheme 1, Nu
) OH) suggested that (salen)Co(III) complexes might also serve
as effective catalysts for the enantioselective addition of phenols
to epoxides. This strategy has proven successful, and we report
here the first examples of kinetic resolution of epoxides with
phenols, with the isolation of 1-aryloxy 2-alcohols (1) in high
ee’s and yields.
Reaction of 2.2 equiv of (()-1,2-epoxyhexane (2a) with phenol
(4a) in the presence of (salen)Co(OAc) complex 3b (0.044 equiv)
in tert-butyl methyl ether (TBME) led to 61% conversion of
phenol after 55 h at room temperature, with 1-phenoxy-2-hexanol
(1a) generated in 94% ee. Encouraged by the observation of high
enantioselectivity in this reaction, we evaluated a variety of
reaction parameters with the goal of identifying a more reactive
system. The identity of the counterion for the (salen)cobalt
complex proved to be important in this context, with the perfluoro
tert-butoxide complex displaying superior reactivity. Thus, the
use of complex 3c5 under conditions otherwise identical to those
outlined above resulted in 80% conversion of phenol in 18 h and
formation of 1-phenoxy-2-hexanol as the major product in 96%
Table 1. Kinetic Resolution of Terminal Epoxides with Phenol
Catalyzed by 3ca
entry
R
equiv 1c temp (°C) yield (%)b ee (%)c
1
2
3
4
5
6
7
(CH2)3CH3 (2a)
CH2Cl (2b)
CH2O(allyl) (2c)
c-C6H11 (2d)
C(O)CH2CH3 (2e)
CO2CH3 (2f)
C6H5 (2g)
0.044
0.044
0.044
0.088
0.088
0.044
0.044
25
-15
4
-15
-20
-20
-25
97
97
93
99
96
98
e
98
99
97
97
96
96
n.d.
a Reactions run 5 M in TBME for 4 to 18 h, unless otherwise noted.
See Supporting Information for details. b Isolated yield based on phenol.
c Determined by chiral HPLC analysis or chiral GC analysis. d Reaction
run in CH3CN. e GC/MS of crude reaction mixture indicated formation
of a 2:1 ratio of regioisomeric products.
ee. Small amounts of 1,2-diol were also generated, presumably
as a result of epoxide hydrolysis with adventitious water,4 but
this pathway could be suppressed easily by the inclusion of 3 Å
molecular sieves in the reaction mixture. The optimized procedure
afforded the product in 97% isolated yield based on phenol and
98% ee (Table 1, entry 1).6,7
A series of terminal epoxides were screened in the kinetic
resolution with phenol, and results are summarized in Table 1.
Both electron-rich (entries 1 and 4) and electron-poor (entries
2,3,5 and 6) epoxides as well as epoxides with a range of steric
properties reacted with complete regioselectivity to provide the
corresponding R-aryloxy alcohols in excellent yields and ee’s.
In contrast, reaction with styrene oxide resulted in a mixture of
regioisomeric ring-opened products (entry 7). In general, the
stereoselectivities in the kinetic resolution displayed a strong
temperature dependence, such that reactions providing moderate
ee’s at room temperature could be rendered significantly more
selective simply by lowering the reaction temperature. For
example, in the reaction of phenol with methyl glycidate, the
following data were obtained: 25 °C, 85% ee; 4 °C, 90% ee;
-20 °C, 96% ee. There was correspondingly little effect of temp-
erature on reaction rate, with all of the above reactions reaching
completion within 16-24 h.
The phenolic kinetic resolution was found to have a broad
substrate scope with respect to the phenol (Table 2). Alkyl
(1) (a) Wright, J. L.; Gregory, T. F.; Heffner, T. G.; MacKenzie, R. G.;
Pugsley, T. A.; Meulen, S. V.; Wise, L. D. Bioorg. Med. Chem. Lett. 1997,
7, 1377. (b) Baker, N. R.; Byrne, N. G.; Economides, A. P.; Javeld, T. Chem.
Pharm. Bull. 1995, 1045. (c) Kirkup, M. P.; Rizvi, R.; Shankar, B. B.; Dugar,
S.; Clader, J.; McCombie, S. W.; Lin, S.; Yumibe, N.; Huie, K.; Heek, M.;
Compton, D. S.; Davis, H, R.; McPhail, A. T. Bioorg. Med. Chem. Lett. 1996,
6, 2069.
(5) Commercially available (salen)Co complex 3a was effectively oxidized
to (salen)Co(III) complex 3c simply by stirring 3a and (CF3)3COH in CH2Cl2
open to the atmosphere for 45 min and then removing the solvent by rotary
evaporation. See Supporting Information.
(2) For examples of asymmetric reductions of R-arlyoxy ketones, see: (a)
Takahashi, H.; Sakuraba, S.; Takea, H.; Achiwa, K. J. Am. Chem. Soc. 1990,
112, 5877. (b) Gooding, O.; Colin, B.; Cooper, G.; Jackson, D. J. Org. Chem.
1993, 58, 3681. (c) Yuan, R.; Watanabe, S.; Kuwabata, S.; Yoneyama, H. J.
Org. Chem. 1997, 62, 2494. (d) Kang, S. B.; Ahn, E. J.; Kim, Y.; Kim, Y. H.
Tetrahedron Lett. 1996, 37, 9317. (e) Guanti, G.; Banfi, L.; Narisano, E.
Tetrahedron Lett. 1986, 27, 3547.
(3) Shibasaki has reported the asymmetric catalytic ring opening of meso
epoxides with 4-methoxyphenol using a Ga(BINOL) catalyst system: Lida,
T.; Yamamoto, N.; Matsunaga, S.; Shigeki, M.; Woo, H.; Shibasaki, M. Angew.
Chem., Int. Ed. 1998, 37, 2223.
(4) Tokunaga, M.; Larrow, J. F.; Kakiuchi, F.; Jacobsen. E. N. Science
1997, 277, 936.
(6) General procedure for the kinetic resolutions in Table 1 and Table 2:
A 10 mL flask was charged with 86 mg (0.100 mmol) of 3c and 100 mg MS
3A. Epoxide (5.00 mmol) and phenol (2.25 mmol) were added at the indicated
reaction temperature, and then TBME (0.15 mL) was added. The reaction
was stirred at the indicated temperature until GC analysis indicated complete
conversion of phenol, at which time 75 mg (0.30 mmol) pyridinium
p-toluenesulfonate was added. The reaction mixture was filtered through a
pad of silica and washed with 50% EtOAc/hexanes. The filtrate was
concentrated and purified by chromatography on silica gel with EtOAc/hexanes
or Kugelrohr distillation under reduced pressure. The enatiomeric purity was
determined by GC or HPLC.
(7) Full experimental procedures, spectral data for new compounds, and
ee determinations are presented in the Supporting Information.
10.1021/ja9910917 CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/10/1999

Communications to the Editor
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 6087
Table 2. Kinetic Resolution of 1,2-Epoxyhexane with Phenols
Catalyzed by 3ca
Scheme 2
3c
(equiv)
temp time yield
ee
entry
R
(°C)
(h)
(%)b
(%)c
1
2
H (4a)
0.044
0.044
0.044
0.044
0.044
0.088
0.088
0.088
0.044
0.044
25
25
12
12
16
120
12
48
18
18
12
96
97
95
99
<5d
92
98
75
93
86
e
98
97
99
n.d.
99
92
99
91
99
p-CH3 (4b)
3
m-CH3 (4c)
o-CH3 (4d)
25
4
25
5
p-Br (4e)
-15
-30
4
6
o-Br (4f)
7
p-OCH3 (4g)
p-NO2 (4h)
p-(CH2)2NHBoc (4i)
1-Naphthol (4j)
8
-20
25
9
10
25
a See Supporting Information for full experimental details. b Isolated
yield based on phenol. c Determined by chiral HPLC analysis. d De-
termined by GC analysis of crude reaction mixture. e No reaction as
determinedby GC analysis.
ring-opening thus provides a highly efficient route to enantiopure
aryl glycidyl ether derivatives.
A general mechanistic pattern has begun to emerge for
asymmetric epoxide ring-opening reactions, wherein the catalyst
can serve a dual role of Lewis acid activator of the epoxide and
counterion for nucleophile delivery.3,11 In that context, we sought
to identify the active catalyst in the phenolic kinetic resolution
reaction and, in particular, to evaluate the possible intermediacy
of a (salen)Co(phenoxide) complex. Addition of 3,5-difluorophe-
nol to a CH2Cl2 solution of 3a under air led to an immediate
darkening of the solution and isolation of a black solid (3d) upon
solvent evaporation. This material exhibited mass spectral and
solution 1H, 19F, and 13C NMR (DMSO-d6) properties consistent
with a complex of formula (salen)Co(OAr)(HOAr) (Ar ) 3,5 F2-
C6H3).12 X-ray crystallographic analysis of a single-crystal grown
from heptane confirmed this formulation and revealed a rare
example of a five-coordinate square pyramidal cobalt-aryloxide
complex with a molecule of ArOH hydrogen-bonded to the
aryloxide oxygen (Figure 1).13 The characterization of 3d is
consistent with the intermediacy of a (salen)Co(aryloxide) com-
plex in the reaction with epoxides, and this is further supported
by the observation of stoichiometric aryloxide transfer from 3d
to (()-1,2-epoxyhexane (2.2 equiv) to provide the corresponding
R-aryloxy alcohol as the sole product in >99% ee.
The (salen)Co(III)-catalyzed kinetic resolution of terminal
epoxides with phenols provides a highly practical route to
1-aryloxy-2-alcohols using an operationally simple procedure and
a readily accessible catalyst. The reaction exhibits extraordinary
generality with respect to the steric and electronic properties and
degree of functionalzation of the epoxides and the phenol. This
constitutes the first case where an entire class of nucleophiless
rather than a single nucleophile such as N3- or OH-scan be used
for the kinetic resolution of terminal epoxides. This raises the
interesting possibility of applying this methodology to the
enantioselective catalytic synthesis of parallel libraries of ring-
opened products. Our efforts in this area are underway.
Figure 1. ORTEP diagram of (salen)Co(OAr)(HOAr) complex 3d (Ar
) 3,5 F2-C6H3).
substitution at the para- or meta-positions on the phenol had little
influence on yield or ee (entries 2 and 3). While ortho-substituted
alkyl phenols displayed poor reactivity, the successful use of
o-bromophenol in the kinetic resolution (entry 6) allows access
to a wide range of ortho-substituted aryl ethers through subsequent
cross-coupling reactions. Overall, phenols with a wide range of
electronic properties participated in the ring-opening reaction with
good yields and ee’s.8
Epibromohydrin is prone to undergo racemization in the
presence of bromide ion,9 and this substrate therefore emerged
as a promising candidate for dynamic kinetic resolution under
the conditions of the phenolic ring-opening reaction. Reaction of
(()-epibromohydrin (1.05 equiv) with phenol (1.00 equiv) in the
presence of (R,R)-3c and LiBr (4 mol %) in CH3CN led to
formation of bromohydrin 1i in >99% ee and 74% yield (eq 1).
Acknowledgment. This work was sponsored by NIH (GM-43214).
The crystal structure analysis was performed by Dr. Matthew S. Sigman
and Dr. Richard Staples. We thank D. Allen Annis for helpful discussions.
The exceptionally high enantioselectivity obtained for the ring-
opening product can be ascribed in part to a secondary resolution
pathway involving the ring opening of transiently formed aryl
glycidyl ether 5 (Scheme 2).10 Consistent with this mechanism,
the double addition byproduct 6 was generated in 1-2% yield.
The dynamic kinetic resolution of epibromohydrin via phenolic
Supporting Information Available: Complete experimental proce-
dures and chiral chromatographic analyses of racemic and enantiomeri-
cally enriched ring-opening products (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
(8) Lower limits for the values of krel for the reactions described in Tables
1 and 2 were determined using the equation krel ) ln[1 - c(1 + ee)]/ln[1 -
c(1 - ee)] where ee is the enantiomeric excess of the product and the
conversion c is set to be the isolated yield of the ring-opened product. The
values for krel thus determined ranged from 42 (Table 2, entry 8) to greater
than 300 (Table 1, entry 2, Table 2, entries 3, 4, 7, 9). Calculations for each
entry are presented in the Supporting Information.
(9) (a) Furrow, M. E.; Schaus, S.; Jacobsen, E. N. J. Org. Chem. 1998, 63,
6776. (b) Schaus, S. E.; Jacobsen, E. N. Tetrahedron Lett. 1996, 37, 7937.
(10) Catalytic asymmetric ring closure of 1,3-dichloro-2-propanol with op-
tically active (salen)Co complexes in the presence of K2CO3 has been reported,
but we have no evidence for such a pathway in the present system. Takfichi,
R.; Arihara, M.; Ishimori, M.; Tsuruta, T. Tetrahedron 1980, 36, 3391.
JA9910917
(11) (a) Hansen, K.; Leighton, J.; Jacobsen, E. J. Am. Chem. Soc. 1996,
118, 10924. (b) McCleland, B.; Nugent, W.; Finn, M. J. Org. Chem. 1998,
63, 6656.
(12) The 1H NMR also reveals the presence of a small amount of a C2
symmetric compound we formulate as the (salen)Co(III)(DMSO)2+ complex.
(13) For related structures, see: (a) Cesari, M.; Neri, C.; Percgo, G.; Perrotti,
E.; Zazzetta, A. J. Chem. Soc., Chem. Commun. 1970, 276. (b) Bailey, N. A.;
McKenzie, E. D.; Worthington, J. M. J. Chem. Soc., Dalton Trans. 1977,
763. (c) Huilan, C.; Deyan, H.; Tian, L.; Hong, Y.; Wenxia, T.; Chen, J.
Zheng, P.; Chen, C. Inorg. Chem. 1996, 35, 1502.