Catalytic asymmetric oxidation of tert-butyl disulfide. Synthesis of tert-butanesulfinamides, tert-butyl sulfoxides, and tert-butanesulfinimines

Article abstract of DOI:10.1021/ja9809206

The first example of the catalytic asymmetric oxidation of tert-butyl disulfide (1) is described. The product, tert-butyl tert-butanethiosulfinate (2) is obtained with 91% enantiomeric excess in yields of ≤92% on scales as large as 1 mol. The application of H₂O₂ as stoichiometric oxidant in the presence of 0.25 mol% of VO(acac)₂ and 0.26 mol% of a chiral Schiff base ligand, 6a, is both convenient and cost-effective. Thiosulfinate ester 2 is chemically and optically stable and serves as an excellent precursor to chiral tert-butanesulfinyl compounds by the stereospecific nucleophilic displacement of tert-butyl thiolate. Addition of LiNH₂ in liquid ammonia and THF provides tert-butanesulfinamide (3; 91% yield). A single recrystallization provides enantiomerically pure 3 in 71-75% overall yield from disulfide 1. Enantiomerically pure thiosulfinate ester 2 also reacts readily and stereospecifically with Grignard reagents, organolithiums, lithium amides, and lithium imine salts to provide enantiomerically pure chiral sulfoxides, sulfinamides, and sulfinimines in good yield.

Full text of DOI:10.1021/ja9809206

VOLUME 120, NUMBER 32
AUGUST 19, 1998
©
Copyright 1998 by the
American Chemical Society
Catalytic Asymmetric Oxidation of tert-Butyl Disulfide. Synthesis of
tert-Butanesulfinamides, tert-Butyl Sulfoxides, and
tert-Butanesulfinimines
†
‡
Derek A. Cogan, Guangcheng Liu, Kyungjin Kim, Bradley J. Backes, and
Jonathan A. Ellman*
Contribution from the Department of Chemistry, UniVersity of California, Berkeley, California 94720
ReceiVed March 18, 1998
Abstract: The first example of the catalytic asymmetric oxidation of tert-butyl disulfide (1) is described. The
product, tert-butyl tert-butanethiosulfinate (2) is obtained with 91% enantiomeric excess in yields of g92%
on scales as large as 1 mol. The application of H2O2 as stoichiometric oxidant in the presence of 0.25 mol %
of VO(acac)2 and 0.26 mol % of a chiral Schiff base ligand, 6a, is both convenient and cost-effective.
Thiosulfinate ester 2 is chemically and optically stable and serves as an excellent precursor to chiral tert-
butanesulfinyl compounds by the stereospecific nucleophilic displacement of tert-butyl thiolate. Addition of
LiNH2 in liquid ammonia and THF provides tert-butanesulfinamide (3; 91% yield). A single recrystallization
provides enantiomerically pure 3 in 71-75% overall yield from disulfide 1. Enantiomerically pure thiosulfinate
ester 2 also reacts readily and stereospecifically with Grignard reagents, organolithiums, lithium amides, and
lithium imine salts to provide enantiomerically pure chiral sulfoxides, sulfinamides, and sulfinimines in good
yield.
Introduction
being utilized as versatile chiral nitrogen intermediates for the
preparation of a range of chiral amines, including R-branched
There have been numerous reports in which the sulfinyl group
has acted as a removable source of diastereoselection in
asymmetric synthesis. Sulfoxides have long been employed for
a wide-range of stereoselective carbon-carbon bond-forming
3
4
5
amines, R- and â-amino acids, aziridines, and amino phos-
6
phonic acids. In most applications, arenesulfinyl auxiliaries,
in particular p-toluenesulfinyl auxiliaries, have been used
because practical methods for the preparation of arenesulfinyl
1
2
reactions, while sulfinamides and sulfinimines are increasingly
7
compounds have been developed. Yet, in the few comparative
†
studies of arenesulfinyl compounds with their tert-butanesulfinyl
counterparts, such as in additions of lithiated sulfoxides to R,â-
Present address: NTH 2122, GlaxoWellcome Inc., P.O. Box 13398,
Research Triangle Park, NC 27709-3398.
‡
8
9
Present address: Roche Research Center, 340 Kingsland St., Nutley,
unsaturated esters or methyl iodide, and reactions of sulfin-
NJ 07110.
(
1) (a) Solladi e´ , G. Synthesis 1981, 185-196. (b) Carre n˜ o, M. C. Chem.
(3) (a) Liu, G.; Cogan, D.; Ellman, J. A. J. Am. Chem. Soc. 1997, 119,
9913-9914. (b) Moreau, P.; Essiz, M.; M e´ rour, J.-Y.; Bouzard, D.
Tetrahedron: Asymmetry 1997, 8, 591-598. (c) Annunziata, R.; Cinquini,
M.; Cozzi, F. J. Chem. Soc., Perkin Trans. 1 1982, 341-343. (d) Yang,
T.-K.; Chen, R.-Y.; Lee, D.-S.; Peng, W.-S.; Jiang, Y.-Z.; Mi, A.-Q.; Jong,
T.-T. J. Org. Chem. 1994, 59, 914-921.
ReV. 1995, 95, 1717-1760. (c) Patai, S. The Chemistry of Sulphinic Acids,
Esters and their DeriVatiVes; John Wiley & Sons: Chichester, England,
1
990. (d) Barbachyn, M. R.; Johnson, C. R. In Asymmetric Synthesis;
Morrison, J. D., Scott, J. W., Eds.; Academic Press: New York, 1983; Vol.
, p 227.
2) Davis, F. A.; Portonovo, P. S.; Reddy, R. E.; Reddy, G. V.; Zhou, P.
Phosphorus, Sulfur, Silicon Relat. Elem. 1997, 120&121, 291-303.
4
(
(4) Review: Davis, F. A.; Zhou, P.; Chen, B.-C. Chem. Soc. ReV. 1998,
27, 13-18.
S0002-7863(98)00920-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/31/1998

8012 J. Am. Chem. Soc., Vol. 120, No. 32, 1998
Cogan et al.
Scheme 1
selective synthesis of R-branched amines from 3 (eq 2). In this
(2)
imines with sulfur ylides,5b the tert-butanesulfinyl auxiliary
provided higher levels of diastereoselectivity. To best elaborate
upon the role of the tert-butanesulfinyl moiety as an auxiliary,
methods for the preparation of enantiomerically pure tert-
butanesulfinyl compounds, including sulfinamides, sulfinimines,
and sulfoxides, must be developed.
procedure, the sulfinamide 3 is readily condensed with aldehydes
to provide sulfinimines in high yield. Additions of Grignard
reagents to the sulfinimines proceed with high levels of
diastereoselection and with no competitive addition at sulfur.
In contrast, arenesulfinamides must be N-silylated in order to
Our interest in the synthesis of tert-butanesulfinyl compounds
arose during explorations of diastereoselective alkylations of
N-acylsulfinamide enolates (Scheme 1). The goal has been the
development of a support-bound sulfinamide linker that serves
both as a chiral auxiliary and an activatable linkage functionality
for nucleophile-mediated release of products from solid sup-
ports.10 In initial solution studies, the N-acyl derivatives of the
more sterically hindered tert-butanesulfinamide provided levels
of diastereoselection superior to that of the N-acyl derivatives
of arenesulfinamides.11 Unfortunately, the chiral precursor, tert-
butanesulfinamide (3) had never been isolated.12 As a result,
there was no information regarding the chemical or configura-
tional stability of enantiomerically pure 3.
13
condense with most aldehydes, and unstabilized Grignard
reagents have been reported to attack at sulfur for the related
3b
p-toluenesulfinimines.
Thus, in addition to differences in
diastereoselection, there are distinct differences in chemical
reactivities between tert-butanesulfinyl compounds and their
arenesulfinyl counterparts.
The application of thiosulfinate ester 2 as a chiral tert-
butanesulfinyl intermediate was instrumental in developing the
first practical route to multigram quantities of enantiopure
sulfinamide 3. In this article, we describe the series of
experiments that led us to recognize tert-butanethiosulfinate,
2
, as a precursor for a range of enantiomerically pure tert-
butanesulfinyl compounds. We further define optimized reac-
tion parameters for the highly catalytic asymmetric oxidation
of tert-butyl disulfide (1) to provide access to large quantities
of 2. Finally, we describe the scope and limitations of
nucleophilic additions to 2, efficiently affording enantiomerically
pure tert-butanesulfinyl compounds, including sulfinamides,
sulfinimines, and sulfoxides in high yield.
In a preliminary account,3 we described a method for
preparing enantiomerically enriched tert-butyl tert-butanethio-
sulfinate (2) in the first example of an asymmetric catalytic
oxidation upon a disulfide (eq 1). Reaction of LiNH2 with 2
a
Results
(1)
Initial Efforts. At the time we began our efforts, Alcudia
and co-workers had described the most practical method for
the preparation of enatiomerically pure tert-butyl sulfoxides and
tert-butanesulfinimines by the addition of nucleophiles to
diastereomerically pure diacetone-D-glucose (DAG) tert-bu-
provided the sulfinamide 3 in high yield with absolute stereo-
specificity. In addition, another application of chiral tert-
butanesulfinyl compounds was described in the general stereo-
1
4
tanesulfinate esters. Therefore, this technology was applied
to the synthesis of the sulfinamide 3 by addition of lithium bis-
(trimethylsilyl)amide to (D)-DAG-(RS)-tert-butanesulfinate esters
(
5) (a) Davis, F. A.; Zhou, P.; Reddy, G. V. J. Org. Chem. 1994, 59,
(eq 3). After treatment with KF-Al2O3, the enantiomerically
3
243-3245. (b) Ruano, J. L. G.; Fern a´ ndez, I.; Catalina, M. d.; Cruz, A.
A. Tetrahedron: Asymmetry 1996, 7, 3407-3414. (c) Davis, F. A.; Zhou,
P. Tetrahedron Lett. 1994, 35, 7525-7528. (d) Davis, F. A.; Reddy, G.
V.; Liang, C.-H. Tetrahedron Lett. 1997, 38, 5139-5142.
(6) (a) Mikolajczyk, M.; Lyzwa, P.; Drabowicz, J. Tetrahedron: Asym-
metry 1997, 8, 3991-3994. (b) Lefebvre, I. M.; Evans, S. A., Jr. J. Org.
Chem. 1997, 62, 7532-7533. (c) Mikolajczyk, M.; Lyzwa, P.; Drabowicz,
J.; Wieczorek, M. W.; Blaszczyk, J. J. Chem. Soc., Chem. Commun. 1996,
1
503-1504.
(
7) (a) Andersen, K. K. In The Chemistry of Sulfones and Sulfoxides;
(3)
Patai, S., Rappoport, Z., Stirling, C. J. M., Eds.; John Wiley & Sons: New
York, 1988; Chapter 3, pp 56-94. (b) Kagan, H. B.; Rebiere, F. SYNLETT
1
3
3
990, 643-649. (c) Solladi e´ , G. Synthesis 1981, 185-196.
(
8) Casey, M.; Manage, A. C.; Gairns, R. S. Tetrahedron Lett. 1989,
0, 6919-6922.
9) Durst, T.; Viau, R.; McClory, M. R. J. Am. Chem. Soc. 1971, 93,
077-3078.
10) For examples of enolate alkylation and derivatizable cleavage from
solid support, see: (a) Backes, B. J.; Ellman, J. A. J. Am. Chem. Soc. 1994,
16, 11171-11172. (b) Backes, B. J.; Virgilio, A. A.; Ellman, J. A. J. Am.
Chem. Soc. 1996, 118, 3055-3056.
11) Backes, B. J.; Ellman, J. A. Abstracts of Papers, 213th National
(
(
pure tert-butanesulfinamide was isolated in modest yield as a
stable, white crystalline solid. Significantly, no racemization
1
(13) Davis, F. A.; Reddy, R. E.; Szewczyk, J. M.; Reddy, G. V.;
Portonovo, P. S.; Zhang, H.; Fanelli, D.; Reddy, R. T.; Zhou, P.; Carroll,
P. J. J. Org. Chem. 1997, 62, 2555-2563.
(14) (a) Fern a´ ndez, I.; Khiar, N.; Llera, J. M.; Alcudia, F. J. Org. Chem.
1992, 57, 6789-6796. (b) Khiar, N.; Fern a´ ndez, I.; Alcudia, F. Tetrahedron
Lett. 1994, 35, 5719-5722.
(
Meeting of the American Chemical Society, San Francisco, CA; American
Chemical Society: Washington, DC, 1997; ORGN 066.
(12) In one case, 3 has been prepared in situ and directly consumed:
Bleeker, I. P.; Engberts, J. B. F. N. J. Org. Chem. 1981, 46, 1012-1014.

Asymmetric Oxidation of tert-Butyl Disulfide
J. Am. Chem. Soc., Vol. 120, No. 32, 1998 8013
Chart 1. Methods for the Resolution of tert-Butanesulfinamide 3
or decomposition of the crystalline sulfinamide has been
observed over at least 18 months.15 Although 3 could be
prepared by this procedure, it was readily apparent that the DAG
approach was not practical for the efficient production of large
quantities of sulfinamide. In particular, this method requires a
chromatographic separation of a diastereomerically enriched
mixture of high-molecular-weight DAG-sulfinates. Other di-
astereomeric sulfinyl intermediates have been described, but
these routes also require chromatographic separations and
Scheme 2
16
employ expensive and high-molecular-weight chiral auxiliaries.
Methods for the resolution of racemic sulfinamide 3, which
can be accessed in high yield on large scale from extremely
inexpensive reagents, were next investigated (Chart 1). Treat-
ment of disulfide 1 with H2O2 followed by Cl2, according to
the procedure of Prinzbach, provides tert-butanesulfinyl chloride,
Scheme 3
1
7
(
rac)-4. Treatment of (rac)-4 with concentrated ammonium
hydroxide provided racemic tert-butanesulfinamide ((rac)-3) in
3% overall yield (eq 4). Unfortunately, attempts to resolve
9
(4)
(
rac)-3 with inexpensive chiral acids, including camphorsulfonic,
base ligands is straightforward,20 enabling numerous derivatives
to be evaluated.
tartaric, mandelic and malic acids, were unsuccessful. Kinetic
resolution procedures were explored, such as oxidation of
racemic sulfinamide as precedented by related resolutions of
sulfoxides,18 but were quickly dismissed, since at least half of
the material would necessarily be discarded. In the same way,
enzymatic resolution by acylation of the sulfinamide was also
briefly explored but abandoned.
Catalytic Methods. Catalytic asymmetric oxidation at sulfur
would be more attractive than the aforementioned routes but
only if high selectivities and catalytic efficiencies could be
achieved with inexpensive and nontoxic stoichiometric oxidants.
Such a catalyst system had recently been reported for the
oxidation of thioethers by Bolm and Bienewald.19 Employing
Schiff base-vanadium complexes, oxidations proceeded in up
to 85% enantiomeric excess (ee), with catalyst loadings as low
as 0.01% and with hydrogen peroxide as a clean and inexpensive
stoichiometric oxidant (Scheme 2). Preparation of the Schiff
A number of sulfur-containing oxidation substrates that would
provide sulfinamide 3 could be envisioned. An obvious choice
is tert-butanesulfenamide (5), prepared in one step from the very
inexpensive tert-butyl disulfide (1) (Scheme 3). An equally
practical route is the direct oxidation of 1 to tert-butyl tert-
butanethiosulfinate (2) (Scheme 3), followed by nucleophilic
21
amide addition, known to occur with inversion. Notably, tert-
22
butanethiosulfinate esters are optically and chemically stable
in contrast to other thiosulfinate derivatives.
2
3
For the pilot oxidations of disulfide 1 and sulfenamide 5,
ligand 6a (eq 5, Table 1), formed by the condensation of 3,5-
di-tert-butylsalicylaldehyde with (S)-tert-leucinol, was used. The
(20) (a) Hayashi, M.; Inoue, T.; Miyamoto, Y.; Oguni, N. Tetrahedron
1994, 50, 4385-4398. (b) Hayashi, M.; Tanaka, K.; Oguni, N. Tetrahe-
dron: Asymmetry 1995, 6, 1833-1836. (c) Hayashi, M.; Kaneda, H.; Oguni,
N. Tetrahedron: Asymmetry 1995, 6, 2511-2516. (d) Aratani, T. Pure Appl.
Chem. 1985, 57, 1839-1844.
(21) (a) Sagramora, L.; Koch, P.; Garbesi, A.; Fava, A. J. Chem. Soc.,
Chem. Commun. 1967, 985-986. (b) Mikolajczyk, M.; Drabowicz, J. J.
Chem. Soc., Chem. Commun. 1976, 220-221. (c) Kishi, M.; Ishihara, S.;
Komeno, T. Tetrahedron 1974, 30, 2135-2142.
(22) (a) Block, E.; O’Connor, J. J. Am. Chem. Soc. 1974, 96, 6, 3921-
3929. (b) Block, E.; O’Connor, J. J. Am. Chem. Soc. 1974, 96, 3929-
3944.
(
15) Other sulfinamides are less optically stable: Booms, R. E.; Cram,
D. J. J. Am. Chem. Soc. 1972, 94, 5438-5446
16) (a) Evans, D. A.; Faul, M. M.; Colombo, L.; Bisaha, J. J.; Clardy,
J.; Cherry, D. J. Am. Chem. Soc. 1992, 114, 5977-5985. (b) Rebi e´ re, F.;
Samuel, O.; Ricard, L.; Kagan, H. B. J. Org. Chem. 1991, 56, 5991-5999.
(
(
17) Netscher, T.; Prinzbach, H. Synthesis 1987, 683-688.
(18) (a) Komatsu, N.; Hahizume, M.; Sugita, T.; Uemura, S. J. Org.
Chem. 1993, 58, 7624-7626. (b) Lattanzi, A.; Bonadies, F.; Senatore, A.;
Soriente, A.; Scettri, A. Tetrahedron: Asymmetry 1997, 8, 2473-2478.
(23) Takata, T.; Endo, T. In The Chemistry of Sulfinic Acids, Esters and
Their DeriVatiVes; Patai, S., Ed.; John Wiley & Sons: Chichester, England,
1990; pp 527-575.
(19) Bolm, C.; Bienewald, F. Angew. Chem., Int. Ed. Engl. 1995, 34,
2
640-2642.

8014 J. Am. Chem. Soc., Vol. 120, No. 32, 1998
Cogan et al.
Table 1. Effect of Varying the Salicylaldehyde Portion of
Ligands 6 on the Oxidation of Disulfide 1 (eq 5)
Table 2. Effect of Varying the Amino Alcohol Portion of Ligands
6 for the Oxidation of Disulfide 1a
a
ligand 6
ligand 6
conv,
%
ee, %c
(config)
R1
R2
conv, %b
ee, %c
3
4
R5
b
R (config)
R (config)
6
6
6
6
6
6
6
6
6
6
6
a
b
c
d
e
f
g
h
i
t-Bu
t-Bu
t-Bu
t-Bu
H
t-Bu
94
18
85
88
50
17
15
10
12
23
60
82
45
79
83
46
6a
6l
t-Bu (S)
i-Pr (S)
Ph (R)
Bn (S)
Ph (S)
H
Me (R)
(1S,2R)-1-amino-2-indanol
i-Pr (S)
t-Bu (S)
Me (S)
H
H
H
H
Ph (R)
Me (S)
Ph (S)
H
H
H
H
H
H
H
H
94
65
47
69
60
40
85
81
16
10
16
82 (R)
60 (R)
25 (S)
53 (R)
44 (S)
19 (R)
41 (R)
58 (S)
53 (R)
19 (R)
41 (R)
NO
2
OMe
H
H
6m
6n
6o
6p
6q
6r
6s
d
OMe
H
<10d
NO
Br
Cl
2
Br
Br
Cl
<10
25
d
<10
50
42
H
H
H
Me
Me
Me
j
k
Br
Me
6t
6u
2-hydroxy-1-naphthaldimine
a
a
All reactions were performed at 23 °C at 0.8 M in CH
2
Cl
2
with
All reactions were performed at 23 °C at 0.8 M in CH Cl with
2
1
2
b
b
2
% VO(acac)
2
and 3% ligand. Conversions determined by GC analysis
2% VO(acac) and 3% ligand. Conversions determined by H NMR.
Enantiomeric excess determined by chiral HPLC analysis.
2
c
c
with tridecane as an internal standard. Enantiomeric excess determined
by chiral HPLC analysis. Impurities limited accuracy of ee determi-
d
nation.
At 0.8-1.5 M in disulfide, investigations toward ligand
optimization were initiated. The Schiff base ligands, 6a-r, were
easily prepared by condensation of salicylaldehydes with amino
alcohols. Previous work with these ligands has indicated a
necessity for sterically bulky substituents ortho to the phenol,
but electronic effects about the salicylaldehyde portion vary
widely among different transformations and substrates.1
Therefore, it would be prudent to screen a number of different
ligands.
first oxidations (2-mmol scale) were performed with 4 mol %
of VO(acac)2 and 6 mol % of 6a with CH2Cl2 as solvent (0.5
M in substrate). Oxidation of 5 proceeded with only 4% ee.
Preliminary oxidation studies employing alternative catalyst
systems were equally disappointing. In contrast, the vanadium
catalyst system afforded thiosulfinate ester (R)-2 with 82% ee.
Clearly, oxidation of tert-butyl disulfide, followed by amide
displacement, was the more promising approach.
Even without optimization, this oxidation of 1 marks the
highest optical purity for the asymmetric oxidation of a disulfide.
Previous reports have described stoichiometric oxidations with
chiral peracids,
conditions.
9,20
2
4
The ligands 6a-j were chosen to probe the effects of
2
substituents at R (eq 5, Table 1). A linear relationship of
2
electron donation from the R substituent with either enantio-
21a,25
26
selection or conversion is not evident, although the electron-
a chiral oxaziridine, or modified Sharpless
Thiosulfinate esters other than 2 have been
2
7
withdrawing nitro substituent (6b) is clearly undesirable. Not
2
surprisingly, there appears to be no steric effect at R (6a,d).
obtained with a high degree of stereospecificity by acid treatment
The remainder of Table 1 describes data for ligands derived
from various commercially available salicylaldehydes. Substi-
of enantiomerically enriched sulfinamides in the presence of
thiols.28
1
tution at R is critical, with the large tert-butyl substituent being
Optimization of Disulfide Oxidation. We chose to first
improve the catalytic efficiency of the oxidation of disulfide 1.
When the oxidation of 1 was performed under the same reaction
conditions, but with a reduction in the amount of VO(acac)2
from 4 to 1 mol %, thiosulfinate ester 2 was produced in only
essential for high conversions and enantioselections.
Various amino alcohols were next condensed with 3,5-di-
tert-butyl salicylaldehyde and its ketone derivative to examine
3
4
steric effects at R and R (eq 6, Table 2). None of these ligands
performed better than 6a. Interestingly, substantial steric
3
8% conversion with a 71% ee. However, by increasing the
3
4
demand at R or R had a favorable influence upon both
enantioselection and conversion. Ketimine ligands have pro-
vided improved selectivities in the Ti-catalyzed reactions of
concentration of disulfide from 0.5 to 1.5 M, the oxidation
proceeded with 89% conversion and 81% ee. The disulfide
concentration can be further increased to 2 M without a
deleterious effect on yield or enantioselection, but higher
concentrations reduced the solubility of the catalyst.
2
9
diketene with aldehydes.
However, for the oxidation of
disulfide 1, poorer selectivities were observed with ketimine
ligands 6s-u. On the basis of the results from the ligand-
screening efforts, ligand 6a was chosen for further optimization
of oxidation conditions due to superior enantioselection, high
conversions, and the availability of its amino alcohol and
aldehyde precursors.
Solvent effects were also found to be important. Data
summarized in Table 3 clearly indicate that oxidations performed
in chlorinated solvents proceeded with the highest enantiose-
(24) The absolute sense of stereochemistry was assigned by addition of
MeMgBr, known to proceed with inversion (vide infra), to deliver (S)-tert-
butyl methyl sulfoxide.
(
25) (a) Savage, W. E.; Fava, A. J. Chem. Soc., Chem. Commun. 1965,
4
3
17-418. (b) Kice, J. L.; Large, G. B. Tetrahedron Lett. 1965, 3537-
541.
(
26) Davis, F. A.; Jenkins, R. H. J.; Awad, S. B.; Stringer, O. D.; Watson,
W. H. J. Am. Chem. Soc. 1982, 104, 5412-5418.
27) Nemecek, C.; Du n˜ ach, E.; Kagan, H. B. New J. Chem. 1986, 10,
(
7
61-764.
(
28) Drabowicz, J.; Mikolajczyk, M. Tetrahedron Lett. 1985, 26, 5703-
(29) Hayashi, M.; Inoue, T.; Oguni, N. J. Chem. Soc., Chem. Commun.
1994, 341-342.
5
706.

Asymmetric Oxidation of tert-Butyl Disulfide
J. Am. Chem. Soc., Vol. 120, No. 32, 1998 8015
Table 3. Effect of Solvent on ee for the Oxidation of Disulfide 1a
In agreement with this conclusion, oxidations performed with
ligands of varying enantiopurity exhibited a linear dependence
of product ee with ligand ee. Therefore, aggregation of
diastereomeric 6a-VO complexes either does not occur or does
not have an effect on enantioselectivity. A nonlinear depen-
dence of product ee upon catalyst ee in unrelated systems has
been attributed to formation diastereomeric aggregates with
ee, %b
ee, %b
solvent
CH Cl
CHCl
solvent
(CH Cl)
CCl
CH
t-BuOH
CF
i-PrOAc
2
2
82
88
44
62
74
29
2
2
81
nr
c
3
4
MeCN
MeNO
3
C
6
H
5
70
11
nr
2
c
C H
6 3
Cl
3
3
C
6
H
5
EtOAc
13
32
different reactivities. The same linear dependence on catalyst
a
All reactions were performed at room temperature with 2%
ee was observed in the 6-VO-catalyzed oxidations of thio-
b
19
VO(acac)
2
and 3% ligand 6a. Enantiomeric excess determined by
ethers.
c
chiral HPLC analysis. The catalyst system was not soluble in these
solvents.
The addition of hydroperoxides to V(IV) or V(V) species
commonly results in the formation of vanadium peroxide
3
3
complexes [L-VO(O2)]. Such complexes have been widely
Table 4. Effect of Temperature on ee and Yield for the Oxidation
a
34
of Disulfide 1
studied both as oxidants for synthesis and as bioinorganic
3
3,35
mechanistic models.
There are a number of L-VO(O2)
VO(acac)
2
, %
temp, °C
yield, %
ee, %
77b
81
91
91
91
91
91
complexes with ligands similiar to 6a (Schiff base, O-N-O
4
4
1
0
0
0
0
.0
.0
.0
.25
.25
.25
.10
32
23
20
20
15
10
15
46
89
98
93
93
85
57
3
5a,36
b
donors).
In one case, Butler and co-workers reported that
peroxide oxidation of a VO-Schiff base complex to the
peroxy-vanadium complex could be monitored by a blue shift
in the UV-visible spectrum. Treatment of a 6a-VO solution
in CHCl3 with H2O2 resulted in a similar blue shift in the UV-
visible spectrum, from a local maximum at 302 nm to λmax at
264 nm. Infrared data taken in CHCl3 showed no shift in the
a
Except where noted, experiments were performed at 1.5 M in
b
CHCl
3
2
with 0.5 mol of 1, with 1:1.1 VO(acac) /ligand 6a. Experiments
-1
VdO stretch of VO(acac)2 (929 cm ) upon addition of 6a.
were performed at 1.0 M in CH
2
2
Cl with 2 mmol of 1.
Addition of disulfide 1 and H2O2 to the 6a-VO complex
-
1
resulted in a very minor shift in the VdO stretch (931 cm ).
Although these data are consistent with the production of a
vanadium peroxide complex, other active oxidants are still
conceivable. One possible mechanism is via the oxaziridine
of Schiff base 6a. Page and co-workers have formed oxaziri-
dines by the oxidation of imines with H2O2, that have in turn
lectivities. The most favorable results are observed with
chloroform as solvent, resulting in thiosulfinate ester 2 with a
substantially better ee than for oxidations performed in CH2-
Cl2.
In addition to solvent effects, temperature, stir rate, and even
the size and shape of the reaction flask had significant impacts
upon the yield and enantioselection of the oxidation of 1. The
effects of temperature and catalyst loadings are summarized in
Table 4. The oxidation proceeds smoothly in a very narrow,
but convenient temperature range without substantial loss in
yield when as little as 0.25% VO(acac)2 and 0.26% ligand 6a
is employed.
37
oxidized thioethers. However, 6a and H2O2 do not oxidize 1
or thioethers, and the imine stretching frequency in the IR
spectrum of 6a is not affected by VO(acac)2 and H2O2, making
this route seem unlikely.
Nucleophilic Additions to tert-Butyl tert-Butanethiosulfi-
nate (2). Our interest in thiosulfinate 2 was primarily as a chiral
tert-butanesulfinyl intermediate to which LiNH2 could be added
to afford sulfinamide 3. In the same way, carbanions, N-
substituted metal amides, and metal imine salts could be used
to displace the thiolate of 2 to deliver sulfoxides, sulfinamides,
and sulfinimines directly. Such nucleophilic displacements
Oxidations of thioethers by VO(acac)2 and ligands 6a,b are
most effective when rapidly stirred.19 However, the oxidation
of disulfide 1 does not proceed under these conditions. Instead,
stirring must be performed sufficiently slowly that the aqueous-
organic interface is not broken. The surface area at the interface
21
typically proceed with inversion at the sulfinyl sulfur.
3
0
is also important. With common and inexpensive glassware,
the oxidation of disulfide 1 can be reproducibly performed on
a 1-mol scale with only 0.25% catalyst and 460 mL of CHCl3,
providing thiosulfinate ester 2 with 91% ee in 92% conversion
and isolated in 88% yield.
Although it has not been unequivocally demonstrated, nu-
cleophilic displacements are believed to proceed via an addi-
tion-elimination mechanism similar to additions to carboxylic
38
acid derivatives (Scheme 4). Both incoming and departing
ligands of the proposed sulfurane intermediates are believed to
reside in the apical positions so that an overall inversion is
Mechanistic Considerations. There are a number of simi-
larities between the oxidation of thioethers described by Bolm
and Bienewald and the oxidation of disulfide 1.19 As was the
case for thioether oxidations, VO(acac)2 alone does not catalyze
the oxidation of 1. Use of 2 equiv of VO(acac)2 with respect
to ligand 6a had no effect on either the rate of oxidation or the
enantioselection. The apparent ligand acceleration eliminates
39
observed. In this scheme, Berry psuedorotation of the
purported sulfurane intermediate would account for deterioration
(
32) (a) Oguni, N.; Matsuda, Y.; Kaneko, T. J. Am. Chem. Soc. 1988,
1
10, 7877-7878. (b) Kitamura, M.; Okada, S.; Noyori, R. J. Am. Chem.
Soc. 1989, 111, 4028-4036.
(33) Butler, A.; Clague, M. J.; Meister, G. E. Chem. ReV. 1994, 94, 625-
6
38.
the impact of vanadium species not complexed with the chiral
ligand and precludes the need for excess ligand.31
(34) Hirao, T. Chem. ReV. 1997, 97, 2707-2724.
(35) (a) Clague, M. J.; Keder, N. L.; Butler, A. Inorg. Chem. 1993, 32,
The use of 2 equiv of ligand with respect to VO(acac)2 also
had no impact on either the oxidation rate or the enantioselec-
tion. This suggests that, at a 1:1 ligand/V ratio, the formation
of a 6a-VO complex is favored, and that excess ligand does
not result in the formation of higher-order 6a-VO complexes.
4754-4761. (b) Rehder, D. Angew. Chem., Int. Ed. Engl. 1991, 30, 148-
167. (c) Hamstra, B. J.; Colpas, G. J.; Pecoraro, V. L. Inorg. Chem. 1998,
3
7, 949-955 and references therein.
(
36) (a) Mimoun, H.; Mignard, M.; Brechot, P.; Saussine, L. J. Am. Chem.
Soc. 1986, 108, 3711-3718. (b) Nakajima, K.; Kojima, M.; Toriumi, K.;
Saito, K.; Fujita, J. Bull. Chem. Soc. Jpn. 1989, 62, 760-767. (c) Schmidt,
H.; Baskirpoor, M.; Rehder, S. J. Chem. Soc., Dalton Trans. 1996, 3865-
3870.
(30) See the Experimental Section for details.
(
31) Berrisford, D. J.; Bolm, C.; Sharpless, K. B. Angew. Chem., Int.
(37) Page, P. C. B.; Heer, J. P.; Bethell, D.; Collington, E. W.; Andrews,
D. M. SYNLETT 1995, 773-775.
Ed. Engl. 1995, 34, 1059-1070.

8
016 J. Am. Chem. Soc., Vol. 120, No. 32, 1998
Cogan et al.
Table 5. Addition of Nucleophiles to Thiosulfinate Ester 2
Scheme 4
nucleophile
LiNHBn
product (config)a
yield, %
racemization, %
b,f
7a (R)
7b (R)
7c (R)
8 (R)
89
98
91
92
0
91
82
55
81
88
98
0
<0.1
d,f
Li(c-NC
H
5 10
)
<0.1
b
LiNHPh
1.0
<0.1
b,f
LiNdC(Me)Ph
LiHMDS
MeMgBr
EtMgBr
i-PrMgBr
vinyl-MgBr
MeLi
c,f
9a (S)
9b (S)
9c (S)
9d (S)
9a (S)
9e (R)
<0.1
c,f
<0.1
e
b,f
<0.1
c,f
<0.1
c,f
PhLi
LiCH CO Et
2 2
<0.1
of stereospecificity observed under some conditions. Alterna-
tively, an SN2-like mechanism is also consistent with the
stereochemical outcome of these displacements.
a
Stereochemical assignments are consistent with known compounds
(
9a, 9b, 9e) or are correlated with known compounds (vide infra).
b
c
When 2 was treated with commercially available metal amides
Enantiomeric excess determined by chiral HPLC analysis. Enantio-
meric excess determined by chiral GC analysis. d Enantiomeric excess
(LiNH2, NaNH2) in a number of solvents (tetrahydrofuran
determined by HPLC analysis of 9e obtained after 7b was treated with
(THF), dioxane, dimethoxyethane (DME), diethyl ether) at -78
PhLi. e All methods for determining optical purity, including chiral
°
C, there was no appreciable addition after several hours.
HPLC and GC, and NMR with various chiral shift reagents failed.
There was no racemization evident at the level of detection of the
analytical methods employed.
f
Warming the reaction mixture to ambient temperature over 12
h provided only small quantities of nearly racemic sulfinamide
3
. Addition of (Et)2NMgBr to a thiosulfinate ester had also
2
1b
enantiopure (minor isomer not detected by chiral HPLC) 3 in
good yield (77-78%).
been reported to proceed with some loss of optical activity.
Racemization may have occurred through the proposed sulfurane
intermediate by psuedorotation or by an alternative mechanism.
Interestingly, LiNH2 provided more highly enantiomerically
enriched product (8-12% ee) than the less soluble NaNH2 (0%
ee). The ee of the recovered thiosulfinate was also nearly
racemic. It appeared likely that, at low amide salt concentra-
tions, the tert-butanethiolate byproduct could competitively
racemize the thiosulfinate ester (eq 7).
Recrystallization readily affords 3 of >99% ee. However,
for additions upon thiosulfinate ester 2 where the tert-butane-
sulfinyl product is not crystalline, it is advantageous to employ
enantiomerically pure 2. Careful crystallization of thiosulfinate
ester 2 twice from hexanes provides (R)-2 with >99% ee and
40
5
2% recovery.
Addition of lithium amides in THF at -78 °C provided the
sulfinamides, 7a,b (Table 5), with high levels of stereospecificity
and presumably the product of inversion at sulfur. In contrast
to the addition of LiNH2 into thiosulfinate ester 2, THF is an
effective solvent for the synthesis of 7a,b due to the higher
solubility of the N-substituted amide salts. Although the
addition of lithium anilide to 2 proceeded with poor conversion
at -78 °C and with substantial racemization at higher temper-
atures, addition in ammonia and THF proceeded smoothly to
afford sulfinamide 7c in high yield. The lithium imine salt
prepared by addition of methyllithium to benzonitrile also
provided the sulfinimine, 8, stereospecifically in high yield as
only one isomer. In contrast, lithium bis(trimethylsilyl)amide
did not react with thiosulfinate ester 2. None of these
sulfinamide derivatives prepared by nucleophilic additions to 2
have been prepared by other means. Moreover, these additions
are the first examples of reaction of a thiosulfinate ester with
the lithium salts of ammonia, primary amines, and imines.
Carbon nucleophiles also react with thiosulfinate esters to
provide sulfoxides. The first account of the attempted addition
(
7)
Indeed, treatment of 2 with 25 mol % of lithium thiolate
resulted in nearly complete racemization at 0 °C over the course
of 90 min. In contrast, the racemization process is very slow
at -78 °C. If nucleophilic displacement of thiolate proceeds
quickly at low temperatures, highly stereospecific additions may
be achieved. To boost the solubility of the metal amides at
lower temperatures, the use of additives such as crown ethers,
and more polar solvents such as TMEDA, was investigated
without success. However, when performed in ammonia with
THF as cosolvent, the reaction of LiNH2 with thiosulfinate ester
2
occurred almost instantaneously (eq 8). The amide addition
2
1b
of MeMgI into 2 did not succeed.
A later attempt with
(38) (a) Mikolajczyk, M. Phosphorus Sulfur 1986, 27, 31-42. (b)
Mikolajczyk, M.; Drabowicz, J. In Topics in Stereochemistry; Allinger, N.
L., Eliel, E. L., Wilen, S. H., Eds.; John Wiley & Sons: New York, 1982;
Vol. 13, pp 333-468. (c) Okuyama, T. In The Chemistry of Sulfinic Acids,
Esters and Their DeriVatiVes; Patai, S., Ed.; John Wiley & Sons: Chichester,
England, 1990; pp 623-638.
(8)
(
39) Holmes, R. Acc. Chem. Res. 1979, 12, 257-265.
(40) A cation effect may also be responsible. Addition of LiHMDS in
upon 2 (91% ee) has been performed numerous times on 0.5
mol or larger scale in high yield and without racemization (91%
yield, 91% ee). In each case, a single recrystallization provided
DAG sulfinate esters is stereospecific, but NaHMDS proceeds with
considerable racemization. Kim, K.; Backes, B. J.; Ellman, J. A., unpub-
lished results.

Asymmetric Oxidation of tert-Butyl Disulfide
J. Am. Chem. Soc., Vol. 120, No. 32, 1998 8017
Scheme 5
thiosulfinate ester 2 in high yield. Addition of LiNH2 in liquid
ammonia and THF provides 3 (91% yield). A single recrys-
tallization provides enantiomerically pure product in 71-75%
overall yield from disulfide 1. Thiosulfinate ester 2 also reacts
readily with Grignard reagents, organolithiums, other lithium
amides, and metal imine salts to give chiral sulfoxides,
sulfinamides, and sulfinimines in good yield and high levels of
stereospecificity. This expedient route to enantiomerically pure
tert-butanesulfinyl compounds should have significant utility
in asymmetric synthesis.
Experimental Section
General Methods. Unless otherwise noted, all reagents were
obtained from commercial suppliers and were used without further
purification. All solvents were distilled under nitrogen from the
following drying agents immediately before use: THF, diethyl ether,
and DME were distilled from sodium/benzophenone ketyl; dichlo-
romethane and chloroform were distilled from calcium hydride; toluene
was distilled from sodium. Chromatography was carried out using
Merck 60 230-400-mesh silica gel. IR spectra of liquids were recorded
as thin films on NaCl plates and IR spectra of solids were recorded as
KBr pellets. Chemical shifts in NMR spectra are expressed in ppm.
Scheme 6
All NMR spectra were obtained at room temperature in CDCl
TMS as an internal standard.
3
with
The general procedures for the preparation of racemic 3, and ligands
a-r have been previously described.3a The preparation of 3-tert-
6
butyl-5-nitrosalicylaldehyde and spectral data for all the Schiff base
ligands are included in the Supporting Information. The ligand
p-tolyl-MgBr did proceed in high yield.27 In our hands, the
addition of Grignard reagents provided the corresponding known
sulfoxides 9a,b and 9d without racemization (Table 5). The
hindered Grignard reagent i-PrMgCl reacts with 2 in lower yield,
but the product sulfoxide’s ee could not be determined. Alkyl
and aryllithiums also react with 2 with absolute stereospecificity
42
precursors 3-tert-butyl-5-methoxysalicylaldehyde, 3-tert-butylsalicyl-
aldehyde,43 and (2-hydroxy-3,5-di-tert-butyl)phenylmethyl ketone
29
were prepared as previously described.
(R)-(+)-tert-Butyl tert-Butanethiosulfinate (2). 1-Mol Scale. Into
a 100 × 190 mm crystallizing dish containing a 2.5-in. Teflon-coated
magnetic stir bar44 was added 663 mg (2.50 mmol) of VO(acac)
and
. The
resulting blue-green solution was stirred for ∼10 min before 178 g
1.00 mol) of tert-butyl disulfide was added. The dish was immersed
2
8
68 mg (2.60 mmol) of ligand 6a followed by 475 mL of CHCl
3
(MeLi, 9a; PhLi, 9e). Unfortunately, the less nucleophilic
lithium enolate of ethyl acetate did not react with thiosulfinate
ester 2.
(
in a water bath, and the stir rate set for the fastest rate that would not
break the phase interface. At this time, 130 mL (1.15 mol) of cold
Stereochemical Correlations. The addition of MeMgBr to
4
1
2
affords the known (S)-tert-butyl methyl sulfoxide (9a)
3
0% aqueous hydrogen peroxide was added in a slow steady stream.
indicating that 2 is the R isomer. The addition of lithium
piperidide to (R)-2 is expected to deliver 7d with an R
configuration at sulfur (Scheme 5). Indeed, the reaction of
sulfinamide 7b with PhLi delivered the expected (S)-9e, the
product of two inversions at sulfur. The product of the addition
of PhLi to 2 provided (R)-9e as expected for only one inversion.
The configuration of sulfinamide 3 was assigned R since 3
obtained from thiosulfinate 2 and from the known (D)-DAG-
The cooling bath temperature was maintained at 15-20 °C. The color
of the organic phase became dark brown, and the aqueous phase became
yellow. After 40 h, the color of the organic phase faded to yellow and
the aqueous phase became orange. Brine (75 mL) was added, and the
layers were shaken and then separated. The aqueous layer was extracted
2 2 2
once with CH Cl , and the combined organic layers were dried (Na -
SO ) and concentrated. NMR analysis of the crude material indicated
4
a 92% conversion. Kugelrohr distillation at 30 °C (5 µTorr) removed
unreacted disulfide. Kugelrohr distillation at 37 °C (5 µTorr) subse-
quently delivered 189 g (88%) of 2 as a clear colorless oil that
crystallized upon standing at room temperature. The ee was determined
to be 91% by chiral HPLC analysis (Diacel Chiralpak AS column, 97:3
(RS)-tert-butanesulfinate ester was identical. Reduction of
sulfinamide 7a also provides 3 as the R isomer (Scheme 6).
Similarly, hydrazinolysis of sulfinimine 8 provided 3 with the
expected R configuration at sulfur. Therefore, the additions of
amide salts also react with thiosulfinate ester 2 with inversion.
hexanes/2-propanol; 1 mL/min, 258 nm; (S)-2, r
) 7.9 min). Crystallization of 2 (91.0 g of 91% ee) twice from hexanes
95 mL, 70 mL) at -20 °C delivered 47.6 g (52%) of (R)-2 as white
t
) 6.5 min; (R)-2, r
t
(
Conclusion
crystals (99.8% ee). The absolute configuration of the major enantiomer
was determined to be R by addition of MeMgBr to enantiomerically
pure 2, providing the (S)-tert-butyl methyl sulfoxide with inversion of
The development of the novel yet practical asymmetric
oxidation of 1 has been instrumental in our synthesis of 3, which
serves as a versatile chiral nitrogen synthon. On 1 mol of 1,
oxidation with H2O2 as stoichiometric oxidant and only 0.25
mol % VO(acac)2 and 0.26 mol % of a chiral Schiff base ligand
provides 2 in 91% ee and g92% yield. Thiosulfinate ester 2 is
also chemically and optically stable, showing no signs of
racemization or decomposition over an extended period of time
41
23
the stereochemistry: [R]
D
2 2
+150° (c 0.55, CH Cl ); mp 30-32 °C;
-1
1
IR 1046, 3190 cm ; H NMR (400 MHz) δ 1.38 (s, 9H), 1.56 (s,
(42) Larrow, J. F.; Jacobsen, E. N. J. Org. Chem. 1994, 59, 1939-1942.
(
43) (a) Tramposch, K. M.; Kung, H. F.; Blau, M. J. Med. Chem. 1983,
2
6
6, 121-125. (b) Zwanenburg, D. J.; Reynen, W. A. P. Synthesis 1976,
24-625.
(44) To attain the optimal interface surface area, the crystallizing dish
(
>8 months). Strong nucleophiles react stereospecifically with
described here was effective. For 0.5-mol-scale oxidations of disulfide 1, a
simple 500-mL Erlenmeyer flask provided the best interface area. Stirring
was always performed with a magnetic stir bar; overhead mechanical stirrers
break the organic-aqueous interface.
(41) Drabowicz, J.; Dudzinski, B.; Mikolajczyk, M.; Wieczorek, M. W.;
Majzner, W. R. Tetrahedron: Asymmetry 1998, 9, 1171-1178.

8018 J. Am. Chem. Soc., Vol. 120, No. 32, 1998
Cogan et al.
9
H); 13C NMR (101 MHz) δ 24.1, 32.2, 48.5, 59.3. Anal. Calcd for
and the mixture was slowly warmed to room temperature with
evaporation of ammonia. The remaining material was diluted with 200
C
8
H18OS
2
: C, 49.44; H, 9.33. Found: C, 49.26; H, 9.29.
R)-(+)-tert-Butanesulfinamide (3). Method 1: From Bis(tri-
methylsilyl)-tert-Butanesulfinamide. To a solution of (D)-DAG-(R )-
tert-butanesulfinate ester (510 mg, 1.4 mmol) in 10 mL of THF at -78
C was added 2.8 mL of 1 M LHMDS in THF. The mixture was
allowed to warm to room temperature while stirring overnight. To
(
mL of CH
solution was dried (Na
2
Cl
2
and then filtered through a Celite pad. The resulting
SO ) and concentrated to provide unpurified
S
2
4
(R)-3 of 91% ee. Recrystallization from hexanes provide 10.7 g (82%)
of (R)-3 of >99% ee (minor enantiomer not detected).
°
(R)-(-)-1-tert-Butanesulfinylpiperidine (7b). To a solution of
0.508 mL (5.13 mmol) of piperidine in 12 mL of THF at -78 °C was
added 2.05 mL of BuLi (2.40 M, 5.13 mmol). Stirring was continued
for 45 min. To the resulting slurry was slowly added 398 mg of 2
(>99% ee; minor enantiomer not detected) in 2 mL of THF. The
reaction vessel was maintained at -78 °C for 1 h, whereupon excess
4
5
the solution at 0 °C was added 2 g of wet KF-Al
After being shaken vigorously for 1 h, the mixture was filtered and
concentrated in vacuo. Chromatography (3% Et N in EtOAc) furnished
7 mg (51%) of 1 as a white crystalline solid.
Method 2: From (R)-2. A 5-L three-necked round-bottomed flask
2
O
3
in one portion.
3
8
saturated aqueous NH
temperature. The mixture was concentrated under a gentle stream of
, and the resulting residue was dissolved in 15 mL of CH Cl and 3
mL of brine. The layers were shaken and separated, and the aqueous
layer was extracted twice with 3 mL of CH Cl . The combined organic
layers were dried (Na SO ) and concentrated. Chromatography (gradi-
ent elution; 5-10% acetone in CH Cl ) afforded 380 mg (98%) of 7b
as a clear oil, whose ee was determined to be >99% (minor isomer
4
Cl was added and the mixture warmed to room
equipped with a mechanical stirrer, an ammonia condenser, and a
nitrogen inlet was charged with 500 mL of liquid ammonia. A few
crystals of Fe(NO ) were added, and lithium wire (8.1 g, 1.16 mol)
3 3
were slowly added in ∼500-mg portions, each portion being added as
the blue color disappeared. A -78 °C bath was periodically raised to
the bottom of the flask to abate refluxing. When all the lithium wire
was added, resulting in a gray suspension, the flask was submerged
into the -78 °C bath. After 30 min, a solution of 2 (90.2 g, 0.465
mol) with 91% ee in 175 mL of THF was slowly added over the course
of 1 h. Once the addition was complete, the mixture was stirred an
N
2
2
2
2
2
2
4
2
2
23
not detected) after treatment with PhLi: [R]
D
-6.7° (c 1.0, acetone);
IR 1080, 1359, 1453 cm ; H NMR (400 MHz) δ 1.10 (s, 9H), 1.48-
.55 (m, 6H), 2.90-2.95 (m, 2H), 3.05-3.10 (m, 2H); 13C (101 MHz)
δ 23.0, 24.3, 26.1, 47.8, 58.1. Anal. Calcd for C 19NOS: C, 57.10;
-
1 1
1
4
additional 15 min before 74.5 g (1.40 mol) of NH Cl was added slowly
9
H
and carefully. The cold bath was removed, and stirring continued until
the mixture reached ambient temperature. The remaining volatile
material was removed under aspirator pressure. To the remaining
residue was added 50 mL of water with swirling. To the resulting
mixture was added 500 mL of EtOAc. After vigorous stirring, the
organic layer was decanted away. The EtOAc was washed with 25
mL of brine. This process was repeated three more times, and the
H, 10.12; N, 7.40. Found: C, 57.02; H, 10.06; N, 7.50.
Addition of PhLi to 7b. A solution of 110 mg (0.580 mmol) of
7b in 1 mL of THF was chilled to -78 °C and added dropwise to
phenyllithium (1.8 M in cyclohexanes/ether; 0.645 mL, 1.16 mmol) in
3 mL at -50 °C. The mixture was stirred for 1 h at -50 °C before 2
mL of saturated aqueous NH
removed under a gentle stream of N
4
Cl was added and then the THF was
. Chiral HPLC analysis indicated
washed organic layers were combined and dried (Na
2 4
SO ). The solvent
2
was removed in vacuo, and the residue recrystallized once from hexanes
to provide 43.4 g (77%; 75% overall from tert-butyl disulfide) of
enantiomerically pure (R)-3 (HPLC, Diacel Chiralpak AS column, 90:
that sulfoxide isolated from this procedure was the S enantiomer with
>99% ee (minor enantiomer not detected).
(R)-(-)-N-Phenyl tert-Butanesulfinamide (7c). To 10 mL of
1
0 hexanes/ethanol; 1.2 mL/min, 222 nm; (R)-3, r
t
) 6.6 min; (S)-3, r
t
3 3
ammonia at -78 °C was added a crystal of Fe(NO ) followed by 86
2
3
)
9.4 min): [R]
D
1
+4.9° (c 1.0, CHCl
3
); mp 101-102 °C; IR 1032,
mg (3.0 mmol) of lithium metal. The resulting blue solution was stirred
-
1
1
364, 1474 cm ; H NMR (400 MHz) δ 1.18 (s, 9H), 3.82 (br s, 2H);
with slow warming to ∼-40 °C until the blue color was replaced by
13
C NMR (101 MHz) δ 22.1, 55.3. Anal. Calcd for C
9.64; H, 9.15; N, 11.56. Found: C, 39.74; H, 9.14; N, 11.24.
R)-(-)-N-Benzyl tert-Butanesulfinamide (7a). To a 50-mL round-
4
H11NOS: C,
2
a gray precipitate. The resulting LiNH mixture was chilled to -78
3
°C, and aniline (1.17 mL, 12.8 mmol) was added slowly. After stirring
for 1 h at -78 °C, a solution of 800 mg (4.12 mmol) of 2 (90.2% ee)
in 2 mL of THF was added very slowly to the pink lithium anilide
solution. This mixture was stirred for 2 h at -78 °C, at which time
(
bottomed flask was added 0.48 g (4.48 mmol) of benzylamine. Next
was added 7.5 mL of THF, and the flask was placed into a -78 °C
bath. While the solution was vigorously stirred, 1.50 mL (3.73 mmol)
of 2.5 M butyllithium in hexanes was added slowly. The mixture was
stirred at -78 °C for 30 min before 150 mg (0.77 mmol) of 2 in 1.5
mL of THF was added. The mixture was then gradually warmed to
room temperature and stirred overnight. Saturated NaCl was added,
and the aqueous layer was extracted with EtOAc (3 × 30 mL). The
excess NH
removed by slowly warming the solution while under a steady but gentle
flow of N . The resulting residue was filtered with 75 mL of CH Cl
and the filtrate was concentrated. Chromatography of the red solution
gradient elution; 3-10% acetone in CH Cl ) afforded 730 mg of 7c
as white crystals with an 89.2% ee (HPLC, Diacel Chiralcel OD column,
0:10 hexanes/2-propanol, 1 mL/min, 254 nm; (R)-7c, r ) 5.4 min;
S)-7c, r ) 7.7 min). Recrystallization from EtOAc/hexanes furnished
the enantiomerically pure product (minor enantiomer not detected):
4
Cl crystals were carefully added. The volatile material was
2
2
2
,
(
2
2
9
t
2 4
combined organic portions were dried over Na SO and concentrated.
(
t
Chromatography (50:50 EtOAc/hexanes) provided 144 mg of 7a as
fluffy white crystals. Analysis by chiral HPLC indicated that there
was no racemization (HPLC, Diacel Chiralcel OD column, 90:10
2
3
-1
1
[
R]
D
-181° (c 1.0, CHCl
3
); IR 1044, 1061, 1498, 1600 cm ; H
NMR (400 MHz) δ 1.32 (s, 9H), 5.65 (s, 1H), 6.96-7.03 (m, 3H),
hexanes/2-propanol, 1 mL/min, 260 nm; (S)-7a, r
t
) 6.0 min; (R)-7a,
7
.12-5.25 (m, 2H); 13C (101 MHz) δ 22.4, 56.4, 118.2, 122.7, 129.3,
r
t
) 9.6 min). Recrystallization from hexanes furnished enantiomeri-
2
3
142.0. Anal. Calcd for C H NOS: C, 60.88; H, 7.66; N, 7.10.
cally pure product (minor enantiomer not detected): [R]
.0, CHCl
400 MHz) δ 1.25, (s, 9), 3.47 (br m, 1), 4.26 (dd, J ) 8.0, 13.7 Hz,
D
1
-35.9° (c
10 15
); mp 71-72 °C; IR 1076, 1365, 1459, 2963 cm- ; H NMR
1
Found: C, 61.01; H, 7.60; N, 7.00.
1
(
3
(R)-(-)-N-(1-Phenylethylidene)-tert-butanesulfinamide (8). To
566 µL (5.55 mmol) of benzonitrile in 12 mL of THF at 0 °C was
added 3.80 mL (5.33 mmol) of 1.4 M ethereal MeLi solution. The
mixture was stirred for 45 min and was then cooled to -78 °C. After
addition of 431 mg (2.22 mmol) of (R)-2 (>99% ee; minor isomer not
detected) in 2 mL of THF, the mixture was stirred at -78 °C for 1 h.
Brine was added to quench the reaction followed by EtOAc (15 mL).
The aqueous layer was extracted with EtOAc (2 × 5 mL). The
1
3
1
H), 4.37, (dd, J ) 4.7, 13.7, 1H), 7.27-7.35 (m, 5); C NMR (101
MHz) δ 22.6, 49.3, 55.8, 127.5, 128.0, 128.5, 138.4. Anal. Calcd for
17NOS: C, 62.52; H, 8.11; N, 6.63. Found: C, 62.32; H, 8.11;
N, 6.60.
Reduction of 7a. A 2-L three-necked round-bottomed flask in a
40 °C bath was equipped with a mechanical stirrer, an ammonia
condenser, and a nitrogen inlet. The flask was charged with 400 mL
of ammonia and then 13.2 g of NH Cl (246 mmol) and 25.0 g (119
11
C H
-
2 4
combined organic layers were dried (Na SO ) and concentrated.
4
Chromatography (20:80 EtOAc/hexanes) provided 456 mg (92%) of
enantiopure (minor isomer not detected) 8 (HPLC, Diacel Chiralpak
mmol) of 7a (91% ee) in 200 mL of THF. Lithium pieces (2.06 g
total, 296 mmol) were slowly added to the mixture in 500-mg portions.
After 30 min, another 13 g of NH Cl was then very carefully added,
4
OD column, 90:10 hexanes/2-propanol, 1.0 mL/min, 250 nm; (R)-8, r
t
1
)
6.1 min; (S)-8, r
t
) 7.6 min): [R]23
D
3
-0.9° (c 1.0, CHCl ); H NMR
1
3
(
45) Bergbreiter, D. E.; Lalonde, J. J. J. Org. Chem. 1987, 52, 1601-
(500 MHz) δ 1.32 (s, 9H), 2.77 (s, 3H), 7.26-7.89 (m, 5H); C NMR
(125 MHz) δ 22.5, 24.2, 57.4, 127.2, 128.4, 131.6, 138.7, 176.4. Anal.
1
603.

Asymmetric Oxidation of tert-Butyl Disulfide
J. Am. Chem. Soc., Vol. 120, No. 32, 1998 8019
Calcd for C12H17NSO: C, 64.54; H, 7.67; N, 6.27. Found: C, 64.66;
CHCl
3
); IR 1043, 1368, 1464, 2964 cm-1; 1H NMR (400 MHz) δ 1.11
H, 7.47; N, 6.32
(d, J ) 6.9, 3H), 1.13 (s, 9H), 1.19 (d, J ) 7.1, 3H), 2.77 (dq, J ) 6.9,
Hydrazinolysis of 8. A solution of 237 mg (1.06 mmol) of
enantiopure 8 and 0.26 mL (5.3 mmol) of hydrazine in 1 mL of THF
was warmed to 50 °C for 20 min. The mixture was concentrated, and
the resulting biphasic mixture was dissolved in 8 mL of CH
then washed once with water. The organic layer was dried (Na
7.1, 1 H); 13C NMR (101 MHz) δ 15.0, 20.2, 23.1, 44.2, 53.80. Calcd
7
for C H16OS: C, 56.71; H, 10.88. Found: C, 56.56; H, 10.97.
(S)-(+)-tert-Butyl vinyl sulfoxide (9d) was prepared by the general
procedure from 400 mg (2.06 mmol) of (R)-2 (90.2% ee) and 5.2 mL
of vinyl-MgBr (1.0 M in THF) in 5 mL of THF. Chromatography
2
Cl
2
2
and
SO
4
)
and concentrated. Chromatography (1:1 hexanes/EtOAc to elute
hydrazones, then gradient elution; 1-5% MeOH in EtOAc) delivered
2 2
(gradient elution; 5-10% acetone in CH Cl ) provided 219 mg of 9d
(81%) as a clear colorless oil, and with an 90.2% ee (HPLC, Diacel
Chiralpak AS column, 90:10 hexanes/2-propanol, 1.2 mL/min, 254 nm;
8
5 mg (66%) of (R)-3 (>99% ee; minor enantiomer not detected).
General Procedure for the Addition of Grignard Reagents to 2.
2
3
retention times: (R)-9d, 7.5 min; (S)-9d, 10.1 min): [R]
D
-321° (c
+283° (c 1, acetone); IR 1056, 1367,
1458 cm ; H NMR (400 MHz) δ 1.15 (s, 9H), 5.94 (d, J ) 10.0,
16b
23
In a round-bottomed flask under a nitrogen atmosphere, 2 (1 equiv)
was dissolved in THF, resulting in a 0.2 M solution. The flask was
placed into a -78 °C cold bath, and a solution of Grignard reagent
1.0, acetone) (lit. (R)-9d: [R]
D
-
1
1
1
3
1H), 6.00 (d, J ) 16.6, 1H), 6.49 (dd, J ) 10.0, 16.6, 1H); C (101
MHz) δ 22.7, 54.5, 123.6, 136.4. Anal. Calcd for C H19NOS: C,
9
(
2.5 equiv) was added slowly with stirring. The resulting mixture was
stirred for 4 h. Once complete as judged by TLC, a minimum of
saturated NH Cl in water was carefully added. The cold bath was
54.50; H, 9.15. Found: C, 54.26; H, 9.29.
4
(R)-(+)-tert-Butyl Phenyl Sulfoxide (9e). Thiosulfinate 2 (98.5%
ee; 345 mg, 1.77 mmol) was dissolved in 9 mL of THF, and the solution
placed into a -78 °C cold bath. To the mixture was added 2.67 mL
removed, and the mixture was transferred to a separatory funnel and
EtOAc (about twice the volume of THF) was added. The layers were
shaken and separated, and the aqueous layer was washed twice with
of phenyllithium (2.0 M in cyclohexane/Et
became dark brown. The reaction was judged complete by TLC within
30 min, and the cold bath removed. Saturated NH Cl (1 mL) was added
2
O; 5.34 mmol). The mixture
EtOAc. The combined organic layers were dried over Na
concentrated. Column chromatography delivered the desired sulfoxides.
S)-(+)-tert-Butyl methyl sulfoxide (9a) was prepared by the general
procedure from 100 mg (0.52 mmol) of (R)-2 (>99% ee) and 0.43 mL
of MeMgBr (3.0 M in Et O; 1.29 mmol) in 300 µL of THF.
Chromatography (gradient elution; 1-2% MeOH in EtOAc) provided
6 mg (91%) of 9a (>99% ee; capillary GC: Astec Chiraldex G-BP
2 4
SO and
4
(
while the mixture was still cold. The mixture was stirred until it reached
room temperature. After addition of 1 mL of water and 18 mL of
EtOAc, the layers were shaken and separated. The aqueous layer was
extracted twice with EtOAc (20 mL). The combined organic layers
were dried (Na SO ), filtered, and concentrated. Column chromatog-
2 4
raphy (50:50 hexanes/EtOAc) provided 314 mg of 9e (98%) as a white
2
5
column, 20 psi He, 90-180 °C at 2 °C/min; (R)-9a, r
S)-9a, r
corresponded to those found in the literature: [R]
t
) 11.7 min;
(
t
) 12.3 min) as a colorless oil, whose spectroscopic data
crystalline solid with 98.5% ee (Chiral HPLC: Chiralpak AS column,
1
6a
23
D
+7.6° (c 1.0,
90:10 hexanes/ethanol, 1.0 mL/min, 256 nm; (S)-9e, r
9e, r ) 9.5 min). Recrystallization from hexanes furnished the
t
) 7.5 min, (R)-
1
6a
23
-1
1
CHCl
400 MHz) δ 1.21 (s, 9H), 2.34 (s, 3H); 3C NMR (101 MHz) δ 22.4,
1.5, 52.5.
S)-(-)-tert-Butyl ethyl sulfoxide (9b) was prepared by the general
procedure from 100 mg (0.52 mmol) of (R)-2 (>99% ee) and 0.43 mL
of EtMgBr (3.0 M in Et O) in 300 µL of THF. Chromatography
gradient elution; 1-2% MeOH in EtOAc) provided 57 mg (82%) of
3
) (lit. [R]
D
+7.8° (c 1, CHCl
3
)); IR 1044 cm ; H NMR
t
1
23
(
enantiomerically pure product (minor enantiomer not detected): [R]
D
) (lit.1 (S)-9e: [R]
6b
23
3
+181° (c 1.0, CHCl
3
D
3
-175° (c 1, CHCl )); IR
cm 1034, 1365, 1401, 1443; H NMR (400 MHz) δ 1.14 (s, 9H),
7.44-7.49 (m, 3H), 7.53-7.58 (m, 2H); C NMR (101 MHz) 22.7,
-
1
1
(
13
2
55.7, 126.2, 128.3, 131.0, 139.9.
(
9
b (>99% ee; capillary GC: Astec Chiraldex G-BP column, 20 psi
) 87.2 min; (R)-9b, r ) 89.4
min) whose spectroscopic data corresponded to those found in the
Acknowledgment. We thank Prof. Carsten Bolm for helpful
discussions. The support of the NSF, Abbott Laboratories,
Pharmacopia & Upjohn, and Berlex Biosciences is gratefully
acknowledged. D.A.C. thanks Abbott Laboratories for a graduate
fellowship.
He, 50-180 °C at 0.4 °C/min; (S)-9b, r
t
t
46
23
-1 1
literature: [R]
D
3
-85.9° (c 1.0, CHCl ); IR 1046 cm ; H NMR (400
13
MHz) δ 1.20 (s, 9H), 1.35 (t, J ) 7.44, 3 H), 2.36-2.52 (m, 2H);
NMR (101 MHz) δ 8.4, 22.8, 39.0, 52.7.
C
(
S)-(-)-tert-Butyl 2-propyl sulfoxide (9c) was prepared by the
general procedure from 97 mg (0.50 mmol) of (R)-2 (>99% ee) and
.75 mL of i-PrMgCl (2.0 M in THF) in 2 mL of THF. Chromatog-
raphy provided 40 mg of 9c (55%), whose spectroscopic data
Supporting Information Available: Experimental details
for the synthesis of 3-tert-butyl-5-nitrosalicylaldehyde, as well
as 400-MHz H and 101-MHz C NMR spectral data for ligands
6a-u (4 pages print). See any current masthead page for
ordering information and Web access instructions.
0
1
13
4
5
23
corresponded to that found in the literature: [R]
D
-62.8° (c 1.0,
(46) Drabowicz, J.; Lyzwa, P.; Popielarczyk, M.; Mikolajczyk, M.
Synthesis 1990, 937-938.
JA9809206