Asymmetric synthesis of anti- and syn-2,3-diamino esters using sulfinimines. Water and concentration effects

Article abstract of DOI:10.1021/ol063058c

(Chemical Equation Presented) In the absence of water, an excess of the lithium enolate of N-(diphenylmethylene)glycine ethyl ester (4) adds to sulfinimines to give syn-2,3-diamino esters 6, and in the presence of water, this enolate gives the anti-2,3-diamino esters 5. These unusual results are interpreted in terms of those factors that inhibit the retro-Mannich fragmentation in the diamino esters.

Full text of DOI:10.1021/ol063058c

ORGANIC
LETTERS
2007
Vol. 9, No. 5
833-836
Asymmetric Synthesis of anti- and
syn-2,3-Diamino Esters Using
Sulfinimines. Water and Concentration
Effects
Franklin A. Davis,* Yanfeng Zhang, and Hui Qiu
Department of Chemistry, Temple UniVersity, Philadelphia, PennsylVania 19122
fdaVis@temple.edu
Received December 18, 2006
ABSTRACT
In the absence of water, an excess of the lithium enolate of N-(diphenylmethylene)glycine ethyl ester (4) adds to sulfinimines to give syn-
2,3-diamino esters 6, and in the presence of water, this enolate gives the anti-2,3-diamino esters 5. These unusual results are interpreted in
terms of those factors that inhibit the retro-Mannich fragmentation in the diamino esters.
The beneficial effects of stoichiometric and substoichiometric
amounts of water on organic and organometallic reactions
have occasionally appeared in the literature and were often
the result of serendipitous observations by the investigators.
In their excellent review on the subject, Ribe and Wipf
grouped these effects into three categories: (1) water as a
hydrolyzing agent leading to secondary products that serve
as catalysts or promoters; (2) water as an internal quenching
agent to drive chemical equilibria; and (3) water as a Lewis
acid activator or co-activator.1 We describe here an unusual
example of how water influences the diamino ester anti/syn
selectivity resulting from the addition of glycine enolates to
sulfinimines (N-sulfinyl imines).
2,3-Diamino acids are key structural units of biologically
active compounds and valuable synthetic intermediates.2 In
2004, we disclosed a new one-pot method for the asymmetric
synthesis of syn- and anti-2,3-diamino esters (+)-3 and (-)-
5a, involving addition of the prochiral lithium enolates of
ethyl (dibenzylamino)acetate (1) and N-(diphenylmethylene)-
glycine ethyl ester (4) to (S)-(+)-N-(benzylidene)-p-toluene-
sulfinamide (2a) (Scheme 1).3 The mechanistic rationale for
the high syn/anti selectivity was the formation of the (E)-
and (Z)-enolates by 1 and 4, respectively. However, in
subsequent studies aimed at exploring and expanding the
scope of the methodology for the formation of the anti-2,3-
diamino ester 5, the reaction became unreliable with the anti/
syn selectivity varying from 1:1 to better than 33:1. It soon
became evident that the water content in the THF was an
important factor determining the selectivity (Table 1).
In the absence of water, all four 2,3-diamino ester isomers
1
were detected by H NMR (Table 1, entry 1). However, as
the ratio of water to LDA increased, there was dramatic
improvement in the formation of the anti product reaching
a maximum anti/syn ratio of 33:1 and an isolated yield of
(-)-5a of 86% when the H2O/LDA ratio was approximately
1:1 (Table 1, entry 5). As the H2O/LDA ratio increased to
2:1, the ratio of 5a/6a was maintained along with the yield
(Table 1, entry 7). Even when the H2O/LDA ratio was 4:1,
the anti/syn ratio was still 33:1, but the yield had diminished
(Table 1, entry 9). This is a general phenomenon also being
observed for sulfinimines derived from aromatic aldehydes
(1) Ribe, S.; Wipf, P. Chem. Commun. 2001, 299.
(2) Viso, A.; Fernandez de la Pradilla, R.; Garcia, A.; Flores, A. Chem.
ReV. 2005, 105, 3167.
(3) Davis, F. A.; Deng, J. Org. Lett. 2004, 6, 2789.
10.1021/ol063058c CCC: $37.00
© 2007 American Chemical Society
Published on Web 01/30/2007

Scheme 1
Table 1. Influence of Water on the Formation of
anti-2,3-Diamino Ester (-)-5a
anti/syn
5/6 ratiod
entry
2, (X)
2a (H)
H2Ob/LDAc ratio
% yielde
f
f
1
2
0:1
0.5:1
1:1
1:1
3
4
5
6
7
8
0.7:1
16:1
25:1
33:1
33:1
33:1
33:1
80
86
86
86
85
78
0.98:1
1.09:1
1.35:1
2.0:1
3.1:1
9
3.9:1
33:1
60
h
10
11
12
13
14
15
16
17
18
19
LiOHg
2.0:1i
1.5:1 MeOHj
1.5:1 t-BuOHj
0:1k
no reaction
no reaction
10:1
1:1.3
1:5
1:33
25:1
25:1
1:25
h
f
f
f
0:1k,l
1.9:1
2.0:1
0:1k,l
86
80
87
85
85
2b (Cl)
2c (CF3)
2c (CF3)
2d (MeO)
1.9:1
25:1
a These results are the average of at least two experiments. The water
content was determined by Karl Fisher titration. cEnolate concentration was
0.02 M unless otherwise noted. dDetermined by 1H NMR on the crude
reaction mixture. eIsolated yield of the major diastereoisomer. fIn addition
to 5 and 6, the other two diamino ester isomers, (SS,2S,3R) and (SS,2S,3R)
were detected in the 1H NMR of the crude reaction mixture. These isomers
could not be separated. gThe equivalent of added LiOH was 1.6 compared
to (+)-2a. hStarting materials were recovered. iThe -78 °C water-THF-
LDA solution was warmed to rt for 10 min and cooled back to -78 °C
prior to addition of 4. jAlcohol was used in the place of water. k5 equiv of
the enolate compared to (+)-2a and a 2.0 M LDA solution was used.
lEnolate concentration was 0.14 M.
b
having electron-withdrawing groups 2b (X ) Cl) and 2c
(CF3) (Table 1, entries 16 and 17) and an electron-donating
group 2d (MeO) (Table 1, entry 19). Replacement of water
with methanol resulted in a 10:1 anti/syn ratio of 5a/6a, but
the anti/syn ratio for tert-butanol was 1:1.3 with the other
amino ester isomers being detected (Table 1, entries 12 and
13).
Experimentally, the reaction was conducted by first cooling
the THF-H2O solution to -78 °C followed by addition of
the appropriate amount of LDA. N-(Diphenylmethylene)-
glycine ethyl ester (4), 1 equiv in water-free THF, cooled to
-78 °C was then added via cannula to the LDA solution.
After stirring at -78 °C for 1 h, 0.63 equiv of the sulfinimine
(S)-(+)-2 in water-free THF was introduced at -78 °C to
the yellow enolate solution. The reaction mixture was
quenched after 1.5 h at -78 °C by addition of aqueous NH4-
Cl solution. Diamino esters (-)-5a-d were isolated by
chromatography.
The syn products (+)-6a and (+)-6c were prepared by
reacting 5 equiv of the water-free preformed enolate solution
of 4 with (+)-2a or (+)-2c at a concentration of 0.14 M.
The syn products were isolated in 86% and 85% yields,
respectively, following chromatography (Table 1, entries 15
and 18).4 At a lower molar concentration of the enolate, the
syn/anti ratio was much poorer (Table 1, entry 14).
The reaction of water with LDA is expected to result in
LiOH and diisopropylamine. However, when the addition
reaction was carried out with 1.6 equiv of LiOH instead of
LDA, there was no reaction (Table, entry 10). Furthermore,
if the -78 °C THF-LDA-H2O solution is warmed to room
temperature for 10 min and cooled back to -78 °C prior to
addition of 4, there is also no reaction (Table, entry 11). As
the water content increases, the enolate was quenched, albeit
slowly, resulting in reduced yields (Table 1, entries 7-9).
Although these results imply that water and LDA could
coexist at -78 °C, another possibility is that some LDA-
hydroxide-diisopropylamine aggregate is the active base
species.
In the solid state, X-ray crystal structures of lithium
enolates show dimers, tetramers, and hexamers.5 However,
as pointed out by Collum and co-workers, the assignment
of such structures in solution is difficult.6 For these reasons,
it is not possible to provide a structure for the H2O-LDA
or the enolate species at this time. However, Willard and
MacEwan in a study of the cocrystallization of lithium or
potassium tert-butoxide with a Li+ or K+ preformed enolate
of 3,3-dimethyl-2-butanone in THF identified a unique
aggregated species by X-ray diffraction analysis. This species
contained an encapsulated hydroxide at the center of the
aggregate.7
(5) Seebach, D. Angew. Chem., Int. Ed. 1988, 27, 1624.
(6) McNeil, A. J.; Toombes, G. E. S.; Gruner, S. M.; Lobkovsky, E.;
Collum, D. B.; Chandramouli, S. V.; Vanasse, B. J.; Ayers, T. A. J. Am.
Chem. Soc. 2004, 126, 16559.
(4) The syn stereochemistry of (+)-6a was determined as previously
described in ref 3.
(7) Williard, P. G.; MacEwan, G. J. J. Am. Chem. Soc. 1989, 111,
7671.
834
Org. Lett., Vol. 9, No. 5, 2007

Scheme 2
To determine whether the anti and syn products are formed
entries 5 and 6). Significantly, when anti-5a was treated
with 5 equiv of the glycine enolate, syn-6a was isolated in
84% yield (Table 2, entry 4). However, similar reaction of
this enolate with syn-6a had no effect (Table 2, entries 7
and 8).
under kinetic or thermodynamic control, they were subjected
to treatment with LDA. Reaction of anti-5a at -78 °C with
1.1 equiv of LDA in water-free THF for 1.5 h resulted in
isomerization and detection of all four amino ester isomers
(Table 2, entry 1). There was no effect on the stereochemistry
To explain these novel results, we make the reasonable
assumption that 7, the Z-enolate of 4, adds to sulfinimine (S)-
(+)-2a to give the anti-2,3-diamino ester anion (2S)-8, which
is the kinetically favored product (Scheme 2).3,8,9 The retro-
Mannich fragmentation of (2S)-8 regenerates (+)-2 and enol-
ate 7 and can explain the formation of the various 2,3-diamino
ester isomers in water-free THF and on reaction of (-)-5a
and (+)-6a with 1.1 equiv of LDA (Tables 1 and 2). Retro-
Mannich fragmentations of sulfinimine-derived sulfinamide
products are usually not observed because the N-sulfinyl
group stabilizes anions at nitrogen.10,11 However, the com-
bination of steric hindrance and the stability of enolate 7
favor this fragmentation. When the H2O-LDA species is
used to generate enolate 7, it apparently stabilizes the anion
Table 2. Reaction of anti-5a and syn-6a with Bases at -78 °C
for 1.5 ha
conditions
(equiv of LDA
to 5a or 6a)
diamino
entry ester
productsb
[% yield]c
1
anti-5a LDA (1.1:1)
5a/6a/other isomers
(10:9:6:8)d [80], 2a [9], 4 [9]
2
3
H2O-LDAe (1.1:1) 5a [77], 2a [10], 4 [10]
LDA-4f (1.1:1)
5a/6a/other isomers
(5:1:1:1)d [82]
6a [84]
5a/6a/other isomers
(10:10:5:8)d [86], 2a [6], 4 [6]
4
5
LDA-4f (5:1)
syn-6a LDA (1.1:1)
6
7
8
H2O-LDAe (1.1:1) 6a [80], 2a [10], 4 [10]
LDA-4f (1.1:1)
LDA-4f (5:1)
(8) The lithium enolate of 4 is generally considered to have the (Z)-
geometry as a consequence of intramolecular chelation of the lithium ion
with both the enolate oxygen and the pair of electrons on the sp2 nitrogen
atom. Numerous experimental results have been rationalized in terms of
this chelated enolate structure. See: (a) McIntosh, J. M.; Leavitt, R. K.;
Mishra, P.; Cassidy, K. C.; Drake, J. E.; Chadha, R. J. Org. Chem. 1988,
53, 1947 and references cited therein. (b) Alvarez-Ibara, C.; Csaky, A. G.;
Colmenero, B.; Ouiroga, M. L. J. Org. Chem. 1997, 62, 2478. (c) Ezquerra,
J.; Pedregal, C.; Merino, I.; Florez, J.; Barluenga, J.; Garcia-Granda, S.;
Llorca, M.-A. J. Org. Chem. 1999, 64, 6554. (d) Ref 2.
6a [84]
6a [90]
a Water-free THF used unless otherwise noted. bDetermined integration
of the p-tolyl methyl on the crude reaction mixtures. cIsolated yield. dIsomers
could not be separated. eRatio of water to LDA (2.5:1). fPreformed enolate
of 4 used.
(9) (a) O’Donnell, M. J. Acc. Chem. Res. 2004, 37, 506. (b) O’Donnell,
M. J. Aldrichimica Acta 2001, 34, 3.
(10) For leading references, see: (a) Zhou, P.; Chen, B. C.; Davis, F.
A. Tetrahedron 2004, 60, 8003. (b) Davis, F. A. J. Org. Chem. 2006, 71,
8993.
(11) For recent reviews on the chemistry of sulfinimines, see: (a) Morton,
D.; Stockman, R. A. Tetrahedron 2006, 62, 8869. (b) Senanayake, C. H.;
Krishnamurthy, D.; Lu, Z.-H.; Han, Z.; Gallou, I. Aldrichimica Acta 2005,
38, 93. (c) Ref 9. (d) Ellman, J. A.; Owens, T. D.; Tang. T. P. Acc. Chem.
Res. 2002, 35, 984.
of anti-5a on reaction with H2O/LDA (2.5:1), and it was
recovered in 77% yield (Table 2, entry 2). Retro-Mannich
products 2a and 4 were also isolated in ca. 10% yield. Similar
results were noted for syn-6a with these base combinations.
With LDA, all four isomers were detected, but with H2O-
LDA, the syn isomer was recovered in 80% yield in addition
to small amounts of the retro-Mannich products (Table 2,
Org. Lett., Vol. 9, No. 5, 2007
835

at nitrogen in (2S)-8 preventing the retro-Mannich reaction
and on workup gives anti-(-)-5a. We suggest that intramo-
lecular chelation in syn-10, which is in equilibrium with anti-
(2S)-8, is sterically favored over similar chelation in anti-9
inhibiting the retro-Mannich fragmentation in the former. In
the presence of enolate 7, anti-(-)-5a gives the thermody-
namically favored syn product (+)-6a on workup (Scheme
2).
In summary, manipulation of the water and enolate concen-
trations makes it possible to prepare either the anti- or syn-
2,3-diamino esters (-)-5 or (+)-6, respectively, from a com-
mon enolate precursor N-(diphenylmethylene)glycine ethyl
ester (4) and a chiral sulfinimine. The anti-2,3-diamino ester
(-)-5 is favored under the LDA-H2O conditions because
the retro-Mannich fragmentation in inhibited. An excess of
the enolate of 4 provides the thermodynamically favored syn-
2,3-diamino ester (+)-6 and is explained in terms of
intramolecular chelation in syn-10 that inhibits the retro-
Mannich fragmentation.
Acknowledgment. We thank David Collum (Cornell
University), Martin J. O’Donnell (Indiana Purdue Univer-
sity), David Dalton (Temple University), Scott Sieburth
(Temple University), and Paul Williard (Brown University)
for helpful discussions. Additionally, we thank J. Deng
(Temple University) for preliminary studies and H. Zhang
(Temple University) for aid with the Karl Fisher titrations.
NIH-NIGMS (GM57870) supported this work.
Supporting Information Available: Experimental details
for enolate preparation and addition. This material is available
free of charge via the Internet at http://pubs.acs.org.
OL063058C
836
Org. Lett., Vol. 9, No. 5, 2007