Enantioselective oxidation of secondary alcohols using a chiral nitroxyl (N-oxoammonium salt) catalyst

Full text of DOI:10.1021/jo9522320

1
194
J . Org. Chem. 1996, 61, 1194-1195
En a n tioselective Oxid a tion of Secon d a r y
Alcoh ols Usin g a Ch ir a l Nitr oxyl
Sch em e 1
(
N-Oxoa m m on iu m sa lt) Ca ta lyst
Scott D. Rychnovsky,* Terri L. McLernon, and
Hemaka Rajapakse
Department of Chemistry, University of California,
Irvine California 92717, and the Department of Chemistry,
University of Minnesota, Minneapolis, Minnesota 55455
Received December 19, 1995
Optically pure secondary alcohols are some of the most
important chiral intermediates in organic synthesis.
They have been prepared by many methods, including
enantioselective reduction of prochiral ketones1 and
enzymatic resolution of their racemates, usually by
acetylation or deacylation.2 Resolution of secondary
alcohols has also been achieved by enantioselective
oxidation using the redox enzyme horse liver alcohol
dehydrogenase (HLADH).3 Several other enantioselec-
tive chemical oxidations have been reported, but each one
has been limited by modest selectivity and/or low turn-
over.4,5 During the course of this investigation, Bobbitt
reported the preparation of chiral nitroxides derived from
and concluded that the transition state for oxidation step
is a cyclic fragmentation of the alkoxide-N-oxoammo-
6
,10
nium salt complex similar to a Cope elimination.
With
(
+)-dihydrocarvone and their use as enantioselective
5
their fast and efficient turnover using inexpensive oxi-
dants, and their highly-ordered transition states for
oxidation, N-oxoammonium salts are promising leads for
the development of enantioselective catalysts.
oxidants. We report the first efficient, enantioselective
oxidation of secondary alcohols using a nonenzymatic
catalyst.
In designing an enantioselective oxidation, we were
attracted to the very efficient catalytic oxidations medi-
ated by nitroxyl radicals. In the presence of a bulk
oxidant, nitroxyl radicals are oxidized to N-oxoammo-
nium salts that in turn rapidly oxidize alcohols to
aldehydes or ketones. The resulting hydroxylamines are
The preparation of the optically pure nitroxide catalyst
(-)-(S)-3,5-dihydro-3,3,5,5-tetramethyl-4H-dinaphth[2,1-
c:1′,2′-e]azepine-N-oxyl (1) is shown in Scheme 1. Ni-
troxide 1 was identified as a promising lead because of
its similarity to the many very selective transition
metal-BINAP catalysts.1 Azepine 2 (>92% ee) was
11
reoxidized by the bulk oxidant to N-oxoammonium salts
1
2
to complete the catalytic cycle.6 A number of bulk
prepared by the procedure of Hawkins and Fu.
The
7
permethylation was carried out in two phases: alkylation
of an N-nitroso derivative followed by Grignard additions
to nitrones 4 and 5. Nitrosation of 2 proceeded in almost
quantitative yield. Bisalkylation was carried out using
KH and excess MeI in refluxing THF to give the dimethyl
oxidants have been used including m-CPBA and an
8
electrochemical oxidation, but the most convenient
system uses buffered, commercial bleach.9 TEMPO or
substituted TEMPO catalysts are normally used at 1 mol
%
0
, and the reactions are complete in less than 10 min at
1
3
°C.9 Semmelhack has studied the reaction mechanism
azepine 3 in 74% overall yield after hydrolysis.
All
attempts to deprotonate the dimethyl N-nitroso inter-
mediate failed, presumably due to poor proton alignment
(
1) Noyori, R. Asymmetric Catalysis in Organic Synthesis; J ohn
Wiley & Sons, Inc.: New York, 1994.
2) Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic
Chemistry; Pergamon: Oxford, U.K., 1994.
3) For a discussion of resolutions using HLADH, see: Danieli, B.;
1
4
with the acidifying nitroso and aryl groups. Oxidation
(
15
2 2
of optically pure 3 with sodium tungstate and H O gave
the nitrone 4, and methyl Grignard addition followed by
reoxidation gave the nitrone 5 in 51% overall yield. A
final methyl Grignard addition and oxidation gave the
desired nitroxide 1 in 51% yield. The optical purity of
nitroxide (-)-(S)-1 was assumed to be >97% ee on the
(
Lesma, G.; Passarella, D.; Riva, S. In Advances in the Use of Synthons
in Organic Chemistry; Donodoni, S., Ed.; J AI Press: Greenwich, CT,
993; Vol. 1, pp 143-219.
(
1
4) For example, see: (a) Ohkubo, K.; Hirata, K.; Yoshinaga, K.
Chem. Lett. 1976, 577-578. (b) Beckett, M. A.; Homer, R. B. Inorg.
Chim. Acta 1986, 122, L5-L7. (c) Ishii, Y.; Suzuki, K.; Ikariya, T.;
Saburi, M.; Yoshikawa, S. J . Org. Chem. 1986, 51, 2822-2824. (d)
Perkins, M. J .; Berti, C.; Brooks, P. J .; Grierson, L.; Grimes, J . A.-M.;
J enkins, T. C.; Smith, S. L. Pure Appl. Chem. 1990, 62, 195-200.
15,16
basis of the optical purity of the dimethylazepine 3.
2
4
); [lit. [R]20
+574.8° (c ) 0.7, CHCl )]. Maigrot, N.; Mazaleyrat,
(11) 2: [R]
D
+574.4° (c ) 0.79, CHCl
3
D
+620° (c ) 0.78,
1
2a
20
(
5) Bobbitt reported enantioselective oxidations of 1-phenylethanol
turnover ) 0.3, S ) 3.3) and cis-1,2-cyclohexanedimethanol (turnover
3.6, S ) 2.2). Ma, Z.; Huang, Q.; Bobbitt, J . M. J . Org. Chem. 1993,
8, 4837-4843.
6) Semmelhack, M. F.; Schmid, C. R.; Cortes, D. S. Tetrahedron
Lett. 1986, 27, 1119-1122.
7) (a) Cella, J . A.; Kelley, J . A.; Kenehan, E. F. J . Org. Chem. 1975,
0, 1860-1862. (b) Ganem, B. J . Org. Chem. 1975, 40, 1998-2000.
8) (a) Semmelhack, M. F.; Chou, C. S.; Cortes, D. S. J . Am. Chem.
CHCl
3
)
and [R]
D
3
(
)
5
J . P.; Welvart, Z. J . Org. Chem. 1985, 50, 3916-3918.
(12) (a) Hawkins, J . M.; Fu, G. C. J . Org. Chem. 1986, 51, 2820-
2822. (b) Hawkins, J . M.; Lewis, T. A. J . Org. Chem. 1992, 57, 2114-
2121. (c) Hawkins, J . M.; Lewis, T. A. J . Org. Chem. 1994, 59, 649-
652.
(
(
(13) Meyers reported an independent synthesis of 3 using his
enantioselective aryl coupling and alkylation: Meyers, A. I.; Nguyen,
T. H. Tetrahedron Lett. 1995, 36, 5873-5876.
4
(
Soc. 1983, 105, 4492-4494. (b) Inokuchi, T.; Matsumoto, S.; Torii, S.
(14) Fraser, R. R.; Boussard, G.; Postescu, I. D.; Whiting, J . J .;
Wigfield, Y. Y. Can. J . Chem. 1973, 51, 1109-1115.
(15) The hydrochloride salt of 3 was recrystallized twice to ensure
J . Org. Chem. 1991, 56, 2416-2421.
(
9) (a) Anelli, P. L.; Biffi, C.; Montanari, F.; Quici, S. J . Org. Chem.
2
4
13
23
1
987, 52, 2559-2562. (b) Anelli, P. L.; Montanari, F.; Quici, S. Org.
React. 1990, 69, 212-219.
10) For an alternate mechanistic proposal see: Ma, Z; Bobbitt, J .
M. J . Org. Chem. 1991, 56, 6110-6114.
optical purity. 3: [R]
D
+391.7° (c ) 1.00, CHCl
3 D
); [lit. [R] +365°
(c ) 1.0, CH
2
Cl
2
)]. Mosher’s analysis showed none of the minor
(
enantiomer.
2
3
(16) 1: [R]
D 3
-610° (c 0.18, CHCl ).
0
022-3263/96/1961-1194$12.00/0 © 1996 American Chemical Society

Communications
J . Org. Chem., Vol. 61, No. 4, 1996 1195
hols to ketones at a convenient rate. The oxidations in
Table 1 were carried out with 0.5-1.0 mol % of 1, 0.6-
0
.7 equiv of NaOCl (0.35 M, pH 8.6), and 0.1 equiv of
KBr in a rapidly stirred CH
at 0 °C. The efficiency of the resolution is characterized
by S () k /k
), the selectivity factor.18 The S values range
2 2 2
Cl /H O mixture for 30 min
S
R
from 3.9 to 7.1 for simple benzylic alcohols with a
reproducibility of about (0.5. There is a slow uncata-
lyzed oxidation19 that reduces the apparent S value, but
this can be overcome by increasing the amount of catalyst
(entries 1 and 2) from 0.5 to 1.0% which increases S from
5
.0 to 7.1. Further increasing the catalyst concentration
to 2.0% does not improve S. The optical purity of the
recovered alcohol is dependent on the conversion; an S
value of 6.0, for example, leads to recovered starting
material with an ee of 89% at 70% conversion (entry 6).
Alkynyl alcohols are also resolved under these conditions,
but with modest S values (e.g., entry 9). A single
example using a primary alcohol with an adjacent chiral
F igu r e 1. ORTEP representation of the nitroxide 1 (50%
probability ellipsoids). The axial (Me ) and equatorial (Me )
A E
methyl groups are labeled. The nitrogen atom is pyramidal,
and the oxygen atom is disordered in the structure.
5
center shows that resolution is possible, but the S factor
Ta ble 1. En a n tioselective Oxid a tion s a n d Resolu tion s of
20
is too low to be useful (entry 10). Among the aromatic
Ra cem ic Alcoh ols w ith (-)-(S)-1
substrates, entries 1-8, ortho-substituted benzyl alcohols
are a bit more selective than unsubstituted rings, and
the 1,2,3,4-tetrahydro-1-naphthol (entry 8) is unselective.
The selectivity in this series correlates well with the
ground state conformation of the benzyl alcohol: sub-
strates where the benzyl hydrogen eclipses the aromatic
ring are oxidized selectively, whereas substrates with the
benzyl hydrogen perpendicular to the aromatic ring are
less selective or unselective. The S enantiomer is oxi-
dized preferentially with each aromatic alcohol. The
identity of the fast reacting enantiomer is not obvious
from inspection of molecular models. However, the
transition state for the fast S reacting isomer is predicted
to be 0.57 kcal/mol more stable than for the slow reacting
R isomer using an AM1 approximation, in good agree-
ment with the experimental results.21
We have described the first efficient, enantioselective
oxidation of secondary alcohols using a nonenzymatic
catalyst. The system is both rapid and chemically
efficient, using ca. 1 mol % nitroxide catalyst (-)-(S)-1
with bleach as the bulk oxidant. The selectivity factors
are good for a first-generation system. We are investi-
gating other catalysts to improve the selectivity and to
extend the oxidation to other classes of substrates.
Ack n ow led gm en t. Support was generously pro-
vided by the Camille and Henry Dreyfus Foundation,
the Alfred P. Sloan Foundation, Pfizer, Bristol-Myers
Squibb, and Zeneca.
a
Eq (S)-1 is the molar equivalent of the catalyst in each
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures for a typical oxidation and for the preparation of
nitroxide catalyst (-)-(S)-1 (7 pages).
reaction. The configurations were determined by Mosher’s ester
analysis. Enantiomeric excess determined by GC analysis with
b
c
a Chiraldex G-TA (30 m × 0.32 mm) column. Enantiomeric excess
d
determined by Mosher’s ester analysis. Configuration determined
J O9522320
by comparison to the Sharpless asymmetric epoxidation product.
e
Optical purity by rotation: observed [R]D 44.6° (lit.22 [R]D 78°).
(
17) Anelli, P. L.; Banfi, S.; Montanari, F.; Quici, S. J . Org. Chem.
f
For greater accuracy, the S value was determined at lower
1
989, 54, 2970-2972.
g
conversion: 66% ee at 54% conversion. Configuration assigned
by analogy to the other examples.
(18) S ) ln[(1 - C)(1 - ee)]/ln[(1 - C)(1 + ee)] where ee is the
fractional enantiomeric excess and C is the conversion. For an excellent
discussion of kinetic resolutions, see: Kagan, H. B.; Fiaud, J . C. Top.
Stereochem. 1988, 18, 249-330.
The nitroxide oxidation catalyst (-)-(S)-1 was character-
ized by X-ray crystallography, and its structure is shown
in Figure 1.
Enantioselective oxidations of alcohols using nitroxide
catalyst (-)-(S)-1 are shown in Table 1. The TEMPO
catalyst oxidizes primary alcohols faster than secondary
alcohols under Anelli’s conditions, but the oxidation still
(19) Lee, G. A.; Freedman, H. H. Tetrahedron Lett. 1976, 1641.
(20) We thank Mr. Timothy Richardson for performing this experi-
ment.
(21) The transition state for the oxidation of methoxide with
dimethyl-N-oxoammonium cation was located at 6-31G*. The transition
state atoms were frozen, and the complete transition state for the
reaction of (R)- and (S)-1-phenylethanol with the N-oxoammonium salt
of 1 was assembled around this frozen core and minimized at the AM1
level. The calculated TS energy for (S)-1-phenylethanol was 161.70
kcal/mol, while the TS energy for (R)-1-phenylethanol was 162.27 kcal/
mol. Details of these calculations will be presented elsewhere.
9
,17
works with secondary alcohols.
Nitroxide 1 is both
more hindered and more selective than TEMPO. While
TEMPO rapidly oxidizes unsubstituted secondary alco-
hols, nitroxide 1 oxidizes only activated secondary alco-
(22) Balfe, M. P.; Downer, E. A. W.; Evans, A. A.; Kenyon, J .; Poplett,
R.; Searle, C. E.; Tarnoky, A. L. J . Chem. Soc. 1946, 787-803.