Asymmetric aza-Henry reactions from N-p-tolylsulfinylimines

Article abstract of DOI:10.1021/ol051580d

(Chemical Equation Presented) N-Sulfinylimines derived from aromatic or aliphatic aldehydes and ketones react with nitromethane and NaOH in a highly diastereoselective manner under mild conditions. In the presence of TBAF, the reaction rates are strongly increased and the stereoselectivity is inverted. This method provides enantiomerically pure β-nitroamines derived from enolizable aldimines and ketimines, which so far are hardly accessible by aza-Henry reactions.

Full text of DOI:10.1021/ol051580d

ORGANIC
LETTERS
2005
Vol. 7, No. 20
4407-4410
Asymmetric Aza-Henry Reactions from
N-p-Tolylsulfinylimines
Jose´ Luis Garc´ıa Ruano,* Markus Topp, Jesu´s Lo´pez-Cantarero, Jose´ Alema´n,
Modesto J. Remuin˜a´n, and M. Bele´n Cid*
Departamento de Qu´ımica Orga´nica, UniVersidad Auto´noma de Madrid, Cantoblanco,
28049 Madrid, Spain
joseluis.garcia.ruano@uam.es; belen.cid@uam.es
Received July 6, 2005
ABSTRACT
N-Sulfinylimines derived from aromatic or aliphatic aldehydes and ketones react with nitromethane and NaOH in a highly diastereoselective
manner under mild conditions. In the presence of TBAF, the reaction rates are strongly increased and the stereoselectivity is inverted. This
method provides enantiomerically pure
aza-Henry reactions.
â-nitroamines derived from enolizable aldimines and ketimines, which so far are hardly accessible by
Optically pure â-nitroamines, obtained by aza-Henry (or
nitro-Mannich) reactions, are attractive targets in asymmetric
synthesis, mainly due to their easy conversion into vicinal
diamines. They have remarkable structural motifs, not only
because of their occurrence in biologically active compounds,
including natural products, but also as valuable building
blocks, chiral auxiliaries, and metal ligands.1
Diastereoselective aza-Henry reactions have been de-
scribed by Anderson2 and Bernardi3 using different condi-
tions for aromatic aldimines. In the first of these papers, the
authors comment on the difficulties associated with the use
of aliphatic imines as starting materials. However, these
reactions have been studied on racemic substrates, and to
our knowledge, there are no general studies involving
diastereoselective aza-Henry reactions yielding optically pure
compounds.4
heterobimetallic catalyst.5 Jørgensen described the addition
of different nitronates to R-iminoesters, catalyzed by chiral
copper complexes to provide precursors of R,â-diamino
carboxylic acids in high diastereo- and enantioselectivity.6
In addition to these strategies, based on metal Lewis acid
activation of the imine and Brønsted base activation of the
nucleophile, several organocatalytic approaches have recently
been published. Johnston7 has reported the proton-catalyzed
reaction of N-Boc aromatic aldimines with nitroethane in
the presence of a chiral bisamidine ligand providing â-ni-
troamines in a highly diastereo- and enantioselective manner.
The bifunctional thiourea-based chiral organocatalysts, de-
(4) (a) A diastereoselective aza-Henry type reaction has been described
using R-amidoalkylphenyl sulfones as precursors to generate in situ the
corresponding N-acylimines: Foresti, E.; Palmieri, G.; Petrini, M.; Profeta,
R. Org. Biomol. Chem. 2003, 1, 4275. (b) We have found two inconspicuous
examples of aliphatic chiral non activated imines using either a chiral amine
or an aldehyde component in: Baricordi, N.; Benetti, S.; Biondini, G.; de
Risi, C.; Pollini, G. P. Tetrahedron Lett. 2004, 45, 1373.
(5) (a) Yamada, K.; Harwood: S. J.; Gro¨ger, H.; Shibasaki, M. Angew.
Chem. 1999, 111, 3713; Angew. Chem., Int. Ed. 1999, 38, 3504. (b) Yamada,
K.; Moll, G.; Shibasaki, M. Synlett 2001, 980.
The catalytic enantioselective aza-Henry reaction was first
reported by Shibasaki using N-phosphinoylimines and a
(1) Lucet, D.; le Gall, T.; Mioskowski, C. Angew. Chem. 1998, 110,
2724; Angew. Chem., Int. Ed. 1998, 37, 2581.
(2) (a) Adams H.; Anderson, J. C.; Peace, S.; Pennell, A. M. K. J. Org.
Chem. 1998, 63, 9932. (b) Anderson, J. C.; Peace, S.; Pih, S. Synlett 2000,
6, 850. (c) Anderson, J. C.; Blake, A. J.; Howell, G. P.; Wilson, C. J. Org.
Chem. 2005, 70, 549.
(3) Bernardi, L.; Bonini, B. F.; Capito, E.; Dessole, G.; Comes-Franchini,
M.; Fochi, M.; Ricci, A. J. Org. Chem. 2004, 69, 8168.
(6) (a) Nishiwaki, N.; Knudsen, K. R.; Gothelf, K. V.; Jørgensen, K. A.
Angew. Chem. 2001, 113, 3080; Angew. Chem., Int. Ed. 2001, 40, 2992.
(b) Knudsen, K. R.; Risgaard, T.; Nishiwaki, N.; Gothelf, K. V.; Jørgensen,
K. A. J. Am. Chem. Soc. 2001, 123, 5843.
(7) Nugent, B. J.; Yoder, R. A.; Johnston, J. N. J. Am. Chem. Soc. 2004,
126, 3418.
10.1021/ol051580d CCC: $30.25
© 2005 American Chemical Society
Published on Web 09/03/2005

signed by Takemoto8 and Jacobsen,9 have also been shown
to be efficient for catalyzed reactions of N-phosphinoylimines
and N-Boc-imines. Recently Jørgensen has reported high ee’s
for the reaction of tertiary nitro compounds with R-imino-
esters using a combination of chiral Lewis acids and chiral
organocatalysts.10
solvents (THF, CH2Cl2, toluene, nitromethane), bases,15 and
Lewis acids [CuF2, Yb(OiPr)3, Yb(Cl)3, Yb(OTf)3].
Table 1. Optimization of the Reaction Conditions
The main drawback for all of these conceptually brilliant
enantioselective aza-Henry reactions is the restricted scope
of the starting substrates, using only aromatic aldimines,
which seriously limits their synthetic usefulness.11 As this
limitation could be due to the difficulties associated in the
synthesis of the enolizable activated imines so far used in
these reactions, we reasoned that it would be overcome by
using other activating groups of the CdN bonds lacking this
synthetic restriction.
t
T
conva dea (%) yield b
entry
1
conditions
(h) (°C) (%)
2a/3a
(%)
NaOH (5 equiv)
4 Å MS
NaOH (5 equiv)
4 Å MS
NaOH (5 equiv)
4 Å MSd
120
24
rt
40
40
95
98
91
94:6
85 (62)c
2
3
94:6
90 (65)c
N-Sulfinylimines have shown to be one of the most
efficient imine derivatives in asymmetric processes involving
reactions with nucleophiles. As a consequence, they have
been successfully applied, mainly by Davis12 and Ellman,13
in the synthesis of a broad variety of structurally diverse
nitrogen-containing molecules. However, to our knowledge,
they have never been used as activated imines in the aza-
Henry processes. Despite the presumably lower activation
induced by the sulfinyl group at the CdN bond, with respect
to that of the usually employed in these reactions, the use of
the N-sulfinylimines as substrate of the aza-Henry reactions
would have the advantage of the few structural restrictions
found in their synthesis in optically pure form, even when
they have enolizable protons. These features prompted us
to evaluate their behavior as substrates in aza-Henry reac-
tions. We report herein the initial study of conditions
concerning the use of N-sulfinylimines as starting products
in the asymmetric aza-Henry reactions with nitromethane.
It has allowed us the synthesis of a wide variety of
â-nitroamines derived from N-sulfinylaldimines and ketimines,
with or without an enolizable proton, in a highly stereose-
lective manner. This strategy provides some advantages in
terms of handling and substrate variety.
96
83:17 59
4
5
6
Yb(OiPr)3 (1 equiv)d 200
Yb(OiPr)3 (1 equiv) 144
rt
rt
rt
20
16
95
77:23 14
83:17
93:7
e
75
Yb(OiPr)3 (1 equiv)
NaOH (5 equiv)
TBAF (1 equiv)
TBAF (1 equiv)
TBAF (0.2 equiv)
TBAF (1 equiv)
NaOH (5 equiv)
24
7
8
9
0.1 rt
0.3
0.3 rt
rt
100
100
100
100
36:64 99
36:64 99
37:63 95
38:62 95
0
10
1
a Determined by 1H NMR. b After silica gel chromatography. c Isolated
yield of 2a after crystallization in ether. d Solvent: THF, MeNO2 (20 equiv).
e Not determined.
As shown in Table 1, the best results were obtained with
NaOH (5 equiv) in the presence of powdered 4 Å molecular
sieves using nitromethane as solvent. After 120 h at room
temperature (entry 1), a 94:6 mixture of nitroamines 2a and
3a was obtained in 85% yield (95% conversion). The mixture
can be purified by chromatography, which indicate the
isomers are stable to the retro-aza-Henry reaction, and the
major isomer could be isolated in 62% yield after crystal-
lization in ether. By increasing the temperature to 40 °C,
the reaction times are shortened (24 h) with slightly improv-
ing yields without affecting the selectivity (entry 2). Lower-
ing the temperature did not improve the stereoselectivity but
dramatically increased the reaction times. Slower reactions
and very low stereoselectivity are the result of using THF
as solvent (entry 3).
The use of lanthanide Lewis acids such as Yb(OiPr)3,
which have proven to be excellent catalysts for the activation
of N-sulfonylimines in their aza-Henry reactions in THF,16
had no positive influence in THF (entry 4) or nitromethane
(entry 5) as the solvent. Nevertheless, when it was used in
combination with NaOH a small decrease in reaction time
was observed at room temperature (entry 6). The highest
increase of the reaction rate was promoted by the addition
of TBAF which reduced the reaction time to 20 min at 0 °C
N-Sulfinylimines 1a-i were obtained by condensation of
the corresponding aldehydes and ketones with (S)-N-p-
tolylsulfinylamide following the Ti(OEt)4 Davis protocol12
with slight modifications in the workup.14 Optimization of
their reactions with nitromethane were examined on the
aromatic p-tolylsulfinylaldimine 1a by exploring several
(8) Okino T.; Nakamura, S.; Furukawa T.; Takemoto, Y. Org. Lett. 2004,
6, 625.
(9) Yoon T. P.; Jacobsen, E. N. Angew. Chem., Int. Ed. 2005, 44, 466.
(10) Knudsen, K. R.; Jørgensen, K. A. Org. Biomol. Chem. 2005, 3,
1362.
(11) During the preparation of the present manuscript, a case of
enantioselective nitro-Manich addition of nitroethane into two alkyl-derived
aldimines was published: Anderson, J. C.; Howell, G. P.; Lawrence, R.
M.; Wilson C. J. Org. Chem. 2005, 70, 5565. Nevertheless, the authors
state the moderate stability of these two prepared aldimines.
(12) (a) Davis, F. A.; Zhou, P.; Chen, B.-C. Chem. Soc. ReV. 1998, 27,
13 and references cited therein. (b) Zhou, P.; Chen, B.-C.; Davis, F. A.
Tetrahedron 2004, 60, 8003 and references cited therein. (c) Davis, F. A.;
Yang, B. J. Am. Chem. Soc. 2005, 127, 8398.
(13) (a) Ellman, J. A.; Owens, T. D.; Tang, T. P. Acc. Chem. Res. 2002,
35, 984. (b) Weix, D. J.; Shi, Y.; Ellman, J. A. J. Am. Chem. Soc. 2005,
127, 1092.
(14) Garc´ıa Ruano, J. L.; Alema´n, J.; Cid, M. B.; Parra, A. Org. Lett.
2005, 7, 179.
(15) Bases of lithium (LiOtBu, LiH, LiNH2, LiHMDS, Li2CO3, LiOH),
sodium (NaOtBu, NaHMDS, NaOH), and potassium (KOH, KOtBu) have
been explored.
(16) Qian, Ch.; Gao, F.; Chen, R. Tetrahedron Lett. 2001, 42, 4673.
4408
Org. Lett., Vol. 7, No. 20, 2005

(entry 7). With this catalyst the selectivity is inverted, with
3a obtained as the major isomer, but the de observed was
low and changing the temperature had little effect (entry 8).
The use of substoichiometric amounts of TBAF does not
substantially modify the results (compare entries 7 and 9),
which reveals the catalytic role of this additive, whereas the
addition of NaOH to the reactions catalyzed by TBAF does
not improve the results (entry 10). Reaction of N-tert-
butylsulfinylbenzylidenimine under the conditions of entry
7 was slower than that of 1a (16 h) but evolved with similar
stereoselectivity.
Table 3. Aza-Henry Reaction of Nitromethane and
N-Sulfinylimines with TBAF
t
conva
(%)
yield
(%)
entry
R1
R2 s.m./prod (h)
2/3a
1
2
3
4
5
6
7
8
9
C6H5
p-CNC6H4
p-CNC6H4
p-MeOC6H4
p-MeOC6H4
Me
t-Bu
Ph
i-Pr
H
H
H
H
H
H
H
Me
Me
a
b
b
c
c
e
g
h
i
0.3
23
0.1
0.25
0.1
0.3
19
2.5
4
98 36:64 99
60 30:70 56
90 30:70
99 38:62 90
99 38:62
99 34:66 97
83 26:74 62
100 36:64 79
100 38:62 76
With the optimized conditions in hand, we examined the
scope and limitations of the reaction of a wide variety of
N-sulfinylimines with nitromethane by using NaOH (Table
2) and TBAF (Table 3) as reagents. As expected, reactions
b
c
b
c
Table 2. Aza-Henry Reaction of Nitromethane and
(S)-N-Sulfinylimines 1b-i with NaOH
a Determined by 1H NMR. b 1 equiv of TBAF was used. c Not deter-
mined.
for aldimines. It is noteworthy that 2h, bearing a quaternary
stereogenic center (entry 9), is formed in notable yield and
stereoselectivity. The instability of 1i under NaOH during
the long times required for reaction with nitromethane
(entries 10 and 11) must be responsible for the low yield of
2i. It would also explain the formation of 30-38% of the
sulfinamide resulting in the hydrolysis of ketimine 1i.
Reactions catalyzed by substoichiometric amounts of
TBAF (Table 3) show a dramatic increase in the reactivity
as well as an inversion of the stereoselectivity. The few
minutes required for aromatic N-sulfinylaldimines 1a and
1c (entries 1 and 4, Table 3) and the aliphatic N-sulfinylal-
dimine 1e (entry 6) contrast with the much longer times of
their reactions with NaOH (Table 2), with isolated yields
higher than 90% in all cases. Despite the expected lower
reactivity of bulky 1g under TBAF (after 19 h only 83%
conversion was observed, entry 7), it is much higher than
under NaOH (compare entry 7, Table 3 ,with entry 7, Table
2). Remarkable are the results obtained from 1h and 1i, which
complete their evolution in less than 4 h affording 2h and
2i in very good yields (entries 8 and 9, Table 3), revealing
that both starting imines are stable enough under these
reaction conditions. The most intriguing result is that obtained
for 1b which reacts more slowly under TBAF (0.2 equiv)
than under NaOH (compare entry 2 in Tables 2 and 3).
Nevertheless, this inversion of reactivity is not observed
when 1 equiv of TBAF was used. Thus, reactions of 1b and
1c are both instantaneous (entries 3 and 5, Table 3). This
observation could suggest that the effect of TBAF was
partially inhibited by the presence of the CN substituent
(entries 2 and 3, Table 3). All reactions provide mixtures of
diastereoisomers with 3 being the major ones. The stereo-
selectivity is only moderate (30-48% de), and it is not
significantly improved by decreasing the temperature (reac-
tions performed at 0 °C afford mixtures of composition
similar to that indicated in Table 3). The electronic density
t
conva
(%)
yield
(%)
entry
R1
C6H5
p-CNC6H4
p-MeOC6H4
PhCHdCH
Me
i-Pr
t-Bu
t-Bu
Ph
R2 s.m./prod (h)
2/3a
1
2
3
4
5
6
7
8
9
H
H
H
H
H
H
a
b
c
d
e
f
24
12b
48
12b
12b
21
98 94:6
95 94:6
80 92:8
90
75
76c
77 89:11h 69d
95 94:6
84 97:3h 69e
68 97:3
46 95:5
79 (66)d
H
g
g
h
i
160
160
46f (36)d
j
Hg
Meg
Me
Meg
72 100 85:15 60
96 100i 94:6
48 100i 93:7
10 i-Pr
11 i-Pr
25
32
i
a Determined by 1H NMR. b rt. c 19% of 1c was recovered. d Isolated
yield of 2a after crystallization in ether. e 11% of 1f was recovered. f 19%
of 1g was recovered. g In the presence of 1 equiv of Yb(OiPr)3 at rt.
h Determined by HPLC after chromatography. i 30-38% of p-tolylsulfina-
mide was detected in the crude by HPLC. j Not determined.
of 1b and 1c catalyzed by NaOH are, respectively, faster
and slower than that of 1a (compare entries 2 and 3 with 1,
Table 2) but the stereoselectivity is very similar. The R,â-
unsaturated imine 1d also provided the â-nitroamine 2d with
no traces of 1,4-conjugate addition product (entry 4, Table
2). More interestingly, the enolizable aliphatic N-sulfinyl-
aldimines 1e and 1f also evolve into the aza-Henry products
in similar yields and stereoselectivities than those observed
for the aromatic aldimines (entries 5 and 6, Table 2). The
bulky N-sulfinylimine 1g requires longer reaction times
(entry 7, Table 2), and the addition of Yb(OiPr)3 did not
improve the reactivity (entry 8). N-Sulfinylketimines 1h-i
exhibit lower reactivity (entries 9-11), thus requiring Yb-
(OiPr)3 catalysis and longer reaction times to get their
complete evolution under conditions similar to those used
Org. Lett., Vol. 7, No. 20, 2005
4409

at the aromatic ring does not have significant influence on
the stereoselectivity. Thus, imines 1a-c afford quite similar
de (entries 1-5).
rate is also evident in the presence of 5 equiv of NaOH (entry
10, Table 1), which immediately would capture the proton
from HF, suggests other possibilities. The basic strength of
the F- (pKa ) 3.17) is enough to activate the nucleophile
by conversion of a significant part of the nitromethane into
its conjugated base. Although this activation is a necessary
condition (tetrabutylammonium hydroxide shows the same
accelerating influence as TBAF, whereas the reaction does
not work in the presence of tetrabutylammonium bromide)
the activation of the N-sulfinylimine by the ammonium cation
seems much more important. This activation has been
reported, among others, for aldehydes17 in their condensation
with nitroalkanes (Henry reaction) and for enolates18 in their
alkylation reactions, which is attributed to the formation of
a contact ion pair O-‚‚‚N+ between the reagent and the
catalyst. The same explanation could be used for explaining
the strong increase of the reactivity observed for N-
sulfinylimines, with the sulfinyl oxygen involved in the
interaction. The formation of this contact ion pair precludes
the formation of the transition states postulated in Figure 1,
thus justifying the observed changes in the stereoselectivity.
In summary, we have proved that stable and easily
available N-sulfinylimines react with nitromethane anion in
a highly diastereselective manner under mild conditions
providing enantiomerically pure versatile â-nitroamines
derived from aromatic or aliphatic aldehydes and ketones
which had not been accessible by other previously described
methods. These reactions are strongly catalyzed by am-
monium ions, which can afford the opposite diastereomer
as the major in moderate selectivity.
The configuration of compound 2a was established by
X-ray crystallography of a racemic sample (see the Sup-
porting Information). As the conditions used in these
reactions do not affect the configuration of the sulfinyl sulfur,
we can unequivocally assign the (SS,S) configuration to 2a
as well as to the major diastereoisomers obtained in all
reactions performed with NaOH, because it is expected that
similar stereochemical evolution takes place for all the
imines. The â-nitroamines 3 are therefore assigned as
(Ss, R).
The high selectivity of the addition of nitromethane in the
presence of NaOH could be attributed to the formation of a
rigid transition state involving both reacting partners coor-
dinated to the metal (Figure 1). The important role of the
Figure 1. Transition states involved in the aza-Henry reactions
carried out in the presence of NaOH.
metal in the stereoselectivity has been proven by performing
the reaction of 1a with nitromethane in the presence of 5
equiv of 15-crown-5. Under these conditions, a complex
reaction mixture was obtained in which 2a and 3a were
detected in a 1:1 ratio. The S configuration of the imine
favors the transition state I leading to the (SS,S) isomer since
the transition state II affording the (SS,R) one would be
destabilized by the steric interactions of the bulky p-tolyl
group (Figure 1).
Concerning the role of the TBAF, a simultaneous and
double activation of the nucleophile (nitromethane) and the
electrophile (N-sulfinylimine) would explain its strong
increase in reactivity. Initially, we thought that the F- was
responsible for the double activation. It would form the
conjugate base of the nitromethane and the resulting HF
would provoke the protonation and, therefore, activation of
the imine. However, the fact that the increase of the reaction
Acknowledgment. We thank the Spanish Government
for financial support (Grant No. BQU2003-04012). J.A. and
M.B.C. thank the Ministerio de Ciencia y Tecnolog´ıa for a
predoctoral fellowship and Ramo´n y Cajal contract, respec-
tively. M.T. thanks EC for an Erasmus fellowship.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for compounds 2a-i and
3a-i and X-ray ORTEP diagram of 2a. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL051580D
(17) (a) Hanessian, S.; Devasthale, P. V. Tetrahedron Lett, 1996, 37,
987. (b) Corey E. J.; Zhang F.-Y. Angew. Chem., Int. Ed. 1999, 38, 1931
and references cited therein.
(18) (a) Corey, E. J.; Xu, F.; Noe, M. C. J. Am. Chem. Soc. 1997, 119,
12414. (b) Corey, E. J.; Bo, Y.; Busch-Petersen, J. J. Am. Chem. Soc. 1998,
120, 13000 and references cited therein.
4410
Org. Lett., Vol. 7, No. 20, 2005