Stereoselective generation of E- and Z-disubstituted amide enolates. Reductive enolate formation from bicyclic thioglycolate lactams [17]

Full text of DOI:10.1021/ja0058280

J. Am. Chem. Soc. 2001, 123, 2091-2092
Scheme 1. Synthesis of 5,7- and 6,7-Bicyclic Lactams
2091
Stereoselective Generation of E- and Z-Disubstituted
Amide Enolates. Reductive Enolate Formation from
Bicyclic Thioglycolate Lactams
Jeffrey M. Manthorpe and James L. Gleason*
Department of Chemistry, McGill UniVersity
a MsCl, NEt3, CH2Cl2. b MeO2CCH2SH, NaH, DMF. c LiOH, THF/
H2O. d TFA (or HCl/Et2O). e EDC-HOBT (or 2-chloro-1-methylpyridnium
iodide), NEt3, CH2Cl2. 53% yield (5 steps) for 2. 46% yield (5 steps) for
3.
801 Sherbrooke St. West, Montreal, QC, Canada H3A 2K6
ReceiVed NoVember 27, 2000
The formation of enolates is a process that is fundamental to
a multitude of chemical transformations. In many cases, the
stereochemistry of an enolate (E or Z) is an integral part of
stereoselective reactions (e.g., syn/anti control in aldol reactions).
For monosubstituted ester and ketone enolates, stereochemistry
can often be influenced by judicious choice of solvent, base, and
temperature.1 For monosubstituted tertiary amide enolates, mini-
mization of A-1,3 interactions usually favors Z-enolate formation.2
Stereocontrol in disubstituted enolates is a more difficult task and
must often be evaluated on a case-by-case basis. Highest levels
of stereocontrol are usually associated with cyclic frameworks,3
including metal chelates,4 while control based on differential steric
environments is less reliable.5,6 We have initiated a project to
develop stereoselective quaternary carbon forming reactions based
on enolate transformations. The goal is to develop a general
method that does not rely on specific enolate features such as
chelating functionality or a large steric difference between enolate
substituents. In this communication, we report a method for
controlling enolate geometry in disubstituted amide enolates where
the E/Z selectivity is dependent only on the geometry and
stereochemistry of the enolate precursor.
Our design utilizes a two-electron reduction of R,R-dialkylated
bicyclic thioglycolate lactams to provide disubstituted amide
enolates (Figure 1).7 Assuming that (a) two alkyl groups (R1 and
R2) are installed stereoselectively at the R-position, (b) the O-C-
C-S dihedral angle is held as close to 90° as possible by the
bicyclic system, and (c) significant bond rotation does not occur
about the carbonyl-carbon/R-carbon bond during the two-electron
reduction process, the E/Z stereochemistry of the enolate should
be controlled by the relative positions of R1 and R2 in the starting
lactam. Importantly, this should afford kinetic E/Z stereocontrol
that is independent of the relative stabilities of the two enolates
and does not depend on a large difference in size of the two alkyl
groups. Significantly, switching the position of R1 and R2 by
inverting the order of their installation should lead to a reversal
of enolate geometry. In many regards, our model resembles the
preferred transition state for deprotonation adjacent to a carbonyl
group, with sulfur transposed for hydrogen. The significant
difference is that deprotonation is a concerted (two-electron)
process whereas the reductive process undoubtedly involves two
separate one-electron-transfer steps and thus bond rotation is a
potential competing process in the intermediate radical anion
resulting from C-S bond scission.
Molecular modeling calculations (MM2) using a Monte Carlo
conformational search (Macromodel) were used to identify
suitable candidates for this stereoselective reduction process.
Several classes of bicyclic thioglycolate lactams were analyzed
for desirable O-C-C-S dihedral angles both at the ground state
and as a weighed average of all stable conformations within 2
kcal/mol of the ground state. From these calculations, the 5,6-,
5,7- and 6,7-bicyclic lactams 1-3 were identified as candidates
for study (see Table 1). Of these, the 5,7- and 6,7-bicyclic lactams
2 and 3 appear to be reasonable candidates (O-C-C-S dihedral
angles of 120-150°), while the 5,6-bicyclic lactam 1 has an
Table 1. Alkylation of Bicyclic Lactams
lactam R1-X yield
dea
R2-X product yield
deb
1
1
1
1
n-PrI
MeI
allyl-Br 88%
MeI
84%
88%
88% MeI
81% n-PrI
85% MeI
4a
4b
4c
4d
94%
88%
91%
86%
72%
72%
79%
78%
allyl-Br
2
2
2
2
2
2
2
2
n-PrI
MeI
76%
86%
88% MeI
93% n-PrI
85% MeI
5a
5b
5c
5d
5e
5f
90%
72%
85%
96%
90%
86%
95%
87%
allyl-Br 88%
MeI
BnBr
MeI
n-PrI
EtI
Figure 1. Model for stereoselective enolate generation.
allyl-Br
96% >95% MeI
88% >99%
(1) (a) Ireland, R. E.; Mueller, R. H.; Willard, A. K. J. Am. Chem. Soc.
1976, 98, 2868. (b) Fataftah, Z. A.; Kopka, I. E.; Rathke, M. W. J. Am. Chem.
Soc. 1980, 102, 3959. (c) Corey, E. J.; Gross, A. W. Tetrahedron Lett. 1984,
24, 495. (d) Ireland, R. E.; Wipf, P.; Armstrong, J. D. J. Org. Chem. 1991,
56, 650.
(2) Evans, D. A.; Takacs, J. M. Tetrahedron Lett. 1980, 21, 4233.
(3) Romo, D.; Meyers, A. I. Tetrahedron 1991, 47, 9503.
(4) (a) Frater, G. HelV. Chim. Acta 1979, 62, 2825, 2829. (b) Schultz, A.
G.; Macielag, M.; Sundararaman, P.; Taveras, A. G.; Welch, M. J. Am. Chem.
Soc. 1988, 110, 7828.
(5) Vedejs, E.; Kruger, A. W.; Lee, N.; Sakata, S. T.; Stec, M.; Suna, E J.
Am. Chem. Soc. 2000, 122, 4602.
(6) For a stereoselective enolization of tiglate oxazolidinone imides, see:
Hosoawa, S.; Sekiguchi, K.; Enemoto, M.; Kobayashi, S. Tetrahedron Lett.
2000, 41, 6429.
(7) Kamata, S.; Uyeo, S.; Haga, N.; Nagata, W. Synth. Commun. 1973, 3,
265.
BnBr
EtI
87% n-PrI
91% >99%
5g
5h
75%
76%
88%
94%
50%a,c
62%a,c
66%a,e
10%
80%
3
3
3
3
n-PrI
MeI
BnBr
MeI
95%
95%
68% MeI
87% n-PrI
6b
6a
6f
81%
81%
89%
69%d
80% >95% MeI
BnBr
6f
a Determined by integration of 1H and/or 13C NMR resonances.
b Determined by capillary gas chromatography on a Chirasil Dex
column. Unless otherwise noted, diastereomers were not separable.
c Alkylation selectivity is reversed for these substrates. These substrates
were readily separable by flash chromatography. d Lactam 6e was
isolated in 31% yield and lactam 6f was isolated in 38% yield.
10.1021/ja0058280 CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/10/2001

2092 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Communications to the Editor
Table 2. Reductive Enolization of Bicyclic Thioglycolate Lactams
O-C-C-S
O-C-C-S
series
n
m
lactam
R1
R2
dihedrala
Z/E ratiob
lactam
R1
R2
dihedrala
Z/E ratiob
1
2
1
1
1
1
4a
4c
n-Pr
Allyl
Me
Me
174
175
47:53
44:56
4b
4d
Me
Me
n-Pr
Allyl
182
178
64:36
68:32
3
4
5
6
1
1
1
1
2
2
2
2
5a
5c
5e
5g
n-Pr
Allyl
Bn
Me
Me
Me
Et
141
140
145
139
87:13
87:13
92:8
5b
5d
5f
Me
Me
Me
Et
n-Pr
Allyl
Bn
128
138
149
137
20:80
26:74
12:88
12:88
n-Pr
80:20
5h
n-Pr
7
8
2
2
2
2
6a
6e
n-Pr
Bn
Me
Me
140
143
83:17
92:8
6b
6f
Me
Me
n-Pr
Bn
133
139
37:63
53:47
a Weighted average (calculated at -78 °C) of all conformations within 2 kcal/mol of the ground state as determined by Monte Carlo calculations.
b Determined by integration of 13C resonances.
unfavorable dihedral angle (175-180°) and thus was expected
to serve as a control.
amide enolate with 87:13 selectivity. Importantly, reduction of
the stereoisomeric lactam (5b) produced the E-enolate, also with
good leVels of stereocontrol (80:20 E/Z). A similar trend was
observed for all 5,7-bicyclic lactams studied (Series 3-6, Table
2). Reduction of the 6,7-bicyclic lactams 6a and 6e produced
Z-enolates cleanly. However, E-enolate selectivity was reduced
for reduction of diatstereomers 6b and 6f. Overall, the results
are in accord with our proposed model and indicate that bond
rotation in the intermediate radical anions is at most a minor
competing factor in the reductive enolization process. The reasons
for reduced selectivity with lactams 6b and 6f are unclear, but
may reflect a higher degree of flexibility in the 6,7-bicyclic ring
system.
In conclusion, we have developed a method for stereoselective
generation of disubstituted amide enolates wherein the E/Z
selectivity is determined by the stereochemistry and geometry of
the starting bicyclic lactams. The method affords both E- and
Z-amide enolates without relying on a steric difference between
the two substituents. The high levels of alkylation selectivity of
the 5,7-bicyclic system and the excellent levels of stereocontrol
for enolate generation appear to earmark it for future development.
In this regard, it is important to note that the reduction generates
an enolate structure that is reminiscent of prolinol amide enolates2
and may serve as a template for subsequent stereoselective
transformations.12 We expect that this new enolate preparation
will find use in numerous reactions including alkylations and aldol
reactions.
Lactam 1 was prepared in two steps from prolinol following
the procedure of Ishibashi.8 Lactams 2 and 3 were prepared from
N-Boc-2-(2-pyrrolidine)ethanol and N-Boc-2-(2-piperidine)etha-
nol, respectively, by displacement of the corresponding mesylates
with methyl thioglycolate followed by a straightforward lactam-
ization protocol (Scheme 1). Sequential alkylation of lactams 1-3
was performed in THF with LDA (1.1 equiv) in the presence of
LiCl (5 equiv).9 Excellent yields and good to excellent diastereo-
selectivities were obtained for the alkylations of 5,6- and 5,7-
bicyclic lactams (Table 1). The expected stereochemical outcome
of these alkylations, wherein the electrophile approaches from
the exo-face of the bicyclic system,3 was confirmed through NOE
studies. Alkylation of the 6,7-bicyclic lactam 3 was not only less
selective but the facial selectivity in the second alkylation reaction
was reversed relative to that of the 5,7-bicyclic system.
Reduction of the thioglycolate lactams was best accomplished
by titrating with lithium di-tert-butylbiphenylide (LiDBB) in THF
at -78 °C. The resulting enolate dianions were trapped as the
silyl ketene aminals by addition of trimethylsilyl chloride followed
by warming the reaction slowly to 23 °C over 0.5 h.10 The silyl
ketene aminals were isolated by removal of solvent in vacuo and
trituration into C6D6 and the E/Z ratio of the products was
determined by 13C NMR.11 The stereochemistry for the silyl ketene
aminals derived from 5e-h (Table 2, Series 5-6) was definitively
assigned by observation of NOE enhancements. Assignments for
the remaining products were made by analogy and by comparison
of the 13C NMR shifts for the ketene aminal carbons among related
series.
Acknowledgment. Support for this research from NSERC, FCAR,
Research Corporation and Merck Frosst Inc. is gratefully acknowledged.
J.M.M. acknowledges support from FCAR in the form of a postgraduate
scholarship.
Reduction and trapping of amides 4a-d proceeded with low
levels of E/Z selectivity, consistent with the large O-C-C-S
dihedral angle in these substrates (Series 1-2, Table 2). In
contrast, reduction of lactam 5a afforded the Z-2-methylpentan-
Supporting Information Available: Experimental procedures and
characterization data for all new compounds (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(8) Ishibashi, H.; Uegaki, M.; Sakai, M. Synlett 1997, 915.
(9) For the use of LiCl in alkylations of amide enolates, see: Myers, A.
G.; Yang, B. H.; Chen, H.; McKinstry, L.; Kopecky, D. J.; Gleason, J. L. J.
Am. Chem. Soc. 1997, 119, 6496.
JA0058280
(10) Woodbury, R. P.; Rathke, M. W. J. Org. Chem. 1978, 43, 881.
(11) Generally the reduction of the amides was complete and e5% of
enolate quenching was observed. Reduction of 5d produced a byproduct (ca.
30%) that could not be fully characterized.
(12) For example, reduction of 5e with LiDBB followed by addition of
allyl bromide (3 equiv) and stirring at -78 °C for 3 h affords the corresponding
C,S-dialkylated product in 96% yield and 90:10 diastereoselectivity. Complete
studies on this alkylation reaction will be reported in due course.