Scope and limitations of the nitro-Mannich reaction for the stereoselective synthesis of 1,2-diamines

Article abstract of DOI:10.1021/jo048304h

(Chemical Equation Presented) The acetic acid-promoted addition of lithium nitropropanate and the Lewis acid-catalyzed [Sc-(OTf)₃, Cu(OTf) ₂, or Ti(OiPr)₄] addition of trimethylsilyl nitropropanate to a range of heteroaromatic and simple aliphatic aldimines gave anti-rich (～3-19:1) β-nitroamines in >95% yields as the kinetic products. It was found that a nonpolar N-imine protecting group was essential for reactivity with the o-methoxybenzyl (OMB) group giving better selectivities and yields than p-methoxybenzyl (PMB) or p-methoxyphenyl (PMP) in the Lewis acid-catalyzed addition reactions. Reduction with SmI₂, treatment with COCl ₂, followed by OMB deprotection gave diastereomerically pure cis-imidazolidinones in 55-79% overall yield from imine. Preliminary results have shown that acetic acid can catalyze the reaction of N-OMB-benzylideneamine with nitropropane, used as solvent, to give the thermodynamically more stable syn-β-nitroamine product.

Full text of DOI:10.1021/jo048304h

Scope and Limitations of the Nitro-Mannich Reaction for the
Stereoselective Synthesis of 1,2-Diamines
James C. Anderson,* Alexander J. Blake,† Gareth P. Howell, and Claire Wilson†
School of Chemistry, University of Nottingham, Nottingham, NG7 2RD, United Kingdom
j.anderson@nottingham.ac.uk
Received September 23, 2004
The acetic acid-promoted addition of lithium nitropropanate and the Lewis acid-catalyzed [Sc-
(OTf)3, Cu(OTf)2, or Ti(OiPr)4] addition of trimethylsilyl nitropropanate to a range of heteroaromatic
and simple aliphatic aldimines gave anti-rich (∼3-19:1) â-nitroamines in >95% yields as the kinetic
products. It was found that a nonpolar N-imine protecting group was essential for reactivity with
the o-methoxybenzyl (OMB) group giving better selectivities and yields than p-methoxybenzyl (PMB)
or p-methoxyphenyl (PMP) in the Lewis acid-catalyzed addition reactions. Reduction with SmI2,
treatment with COCl2, followed by OMB deprotection gave diastereomerically pure cis-imidazoli-
dinones in 55-79% overall yield from imine. Preliminary results have shown that acetic acid can
catalyze the reaction of N-OMB-benzylideneamine with nitropropane, used as solvent, to give the
thermodynamically more stable syn-â-nitroamine product.
SCHEME 1
Introduction
The development of stereoselective reactions that cre-
ate carbon-carbon bonds and heteroatom functionality
can provide valuable building blocks for organic synthe-
sis. Due to the importance of the 1,2-diamine structural
motif in biologically active natural products, medicinal
agents, and more recently as chiral auxiliaries and chiral
ligands in asymmetric catalysis,1 we set out to develop a
general method to synthesize this class of compounds
that did not rely upon the use of naturally occurring
amino acids or amino alcohols. Guided by the notion that
a closed six-membered transition state structure could
help to maximize stereocontrol, due to the associated
conformational energies of the defined six-membered
ring, we concentrated on searching for potential R-aza
carbanions possessing an extra donor group (X in 1,
Scheme 1) that would add to imines. We eventually
discovered that under very precise conditions lithium
nitronates would add to imines in the presence of a
Brønsted acid to give â-nitroamines 2 in good yield and
diastereoselectivites in certain cases (Scheme 1).2
nitro-Mannich reaction3 that had previously been de-
scribed in a different guise over 100 years ago,4 but for
which no stereoselective examples had ever been re-
ported.5 Reduction of the nitro function is complicated
with retroaddition and/or â-elimination of 2,6 but sa-
marium di-iodide was found to perform the reduction in
As this protocol starts with a preformed imine, it
represents a stepwise method of conducting the classic
(2) Adams, H.; Anderson, J. C.; Peace, S.; Pennell, A. M. K. J. Org.
Chem. 1998, 63, 9932.
(3) Baer, H. H.; Urbas, L. In The Chemistry of the Nitro and Nitroso
Groups; Patai, S., Ed.; Interscience, New York, 1970; Part 2, p 117.
(4) Henry, L. Bull. Cl. Sci., Acad. R. Belg. 1896, 32, 33.
† Corresponding authors for X-ray crystal structures.
(1) Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem., Int. Ed.
Engl. 1998, 37, 2580.
10.1021/jo048304h CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/14/2004
J. Org. Chem. 2005, 70, 549-555
549

Anderson et al.
SCHEME 2
high yield and with no erosion in stereochemical in-
tegrity.6b,7 This stepwise protocol that combines an imine,
derived from an aromatic or aliphatic carbaldehyde, with
a nitro compound under nonbasic conditions represents
a more high yielding and stereoselective synthesis of 1,2-
diamines than the condensation of amines with formal-
dehyde and nitroalkanes under reducing conditions,
which had been reported over fifty years ago.8 An
asymmetric variant of the nitro-Mannich reaction with
N-phosphinoylaldimines was then developed by Shiba-
saki et al.9 They employed their heterobimetallic asym-
metric complexes which contain both Brønsted basic and
Lewis acidic sites,10 first with nitromethane9 and then
extended to other nitroalkanes,11 but only with imines
derived from aromatic aldehydes. Our results had led us
to postulate that activation of the imine by protonation
is a prerequisite for nitronate addition. Extending this
idea with the knowledge of metal-catalyzed additions of
silyl ketene acetals and silyl enol ethers to imines12 and
the fluoride induced addition of silyl nitronates to car-
bonyl compounds,13 we reported that Sc(OTf)3 catalyzed
the addition of 1-trimethylsilyl nitropropanate to aryl and
aliphatic imines in good yield with variable diastereose-
lectivity (Scheme 2).14 We found that PMP-protected
imines (3) gave generally higher diastereoselectivities
than the analogous PMB-protected imines and could be
removed after reduction of the nitro functionality and
oxazilidinone formation. Throughout our studies the
sensitivity of â-nitroamines 2 and 4 toward retro-addition
meant that reduction with SmI2 to stable monoprotected
1,2-diamines should be performed immediately.
FIGURE 1. Imine substrates.
tioselectivity using bisoxazoline-type chiral ligands.15
These catalysts were simultaneously developed for the
efficient asymmetric addition of nitroalkanes to N-PMP
R-imino esters in the presence of catalytic Et3N.16 The
high selectivity of these two processes can be attributed
in part to the bidentate chelation of the R-imino esters
to the Lewis acid metal centers. The â-nitro-R-amino
esters generated by Jørgensen seem more stable than
simple â-nitroamines (2) and can be readily reduced with
Raney Ni/H2 in good yields with no erosion of stereo-
chemical integrity. Other workers have shown that Yb-
(OiPr)3 can catalyze the addition of nitromethane to
N-sulfonylaldimines derived from aromatic aldehydes.17
The catalytic asymmetric synthesis of ICI-199441 and
CP-99994 by Shibasaki et al. using the nitro-Mannich
reaction has demonstrated that the products from the
nitro-Mannich reaction are useful synthetic building
blocks and should stimulate their use in other target
syntheses.18 Most recently chiral organic Brønsted acid
catalysts have been developed to provide aromatic â-ni-
troamines in good enantioselectivities.19,20
Despite the considerable advances made in this area
there remain limitations to the general applicability of
the nitro-Mannich reaction. We set out to develop a
flexible methodology that would allow the synthesis of a
variety of 1,2-diamines through the coupling of a range
of nitroalkanes with aromatic and aliphatic imines. This
is in direct contrast to Shibasaki’s system, which uses
aromatic aldimines, and Jørgensen’s, which is necessarily
restricted to R-imino esters. This paper reports our efforts
to tackle this difficult goal and the results prepare the
ground for our asymmetric studies which are currently
under investigation.
After this report Jørgensen et al. reported that the Cu-
catalyzed addition of trimethylsilyl nitronates to N-PMP
R-imino esters could give optically active â-nitro-R-amino
acids and R,â-diamino acid derivatives in high yield, with
generally good diastereoselectivity and excellent enan-
Results and Discussion
(5) (a) Hurd, C. D.; Strong, J. S. J. Am. Chem. Soc. 1950, 72, 4813-
4814. (b) Kozlov, L. M.; Fink, E. F. Tr. Kaz. Khim. Tekhnol. Instrum.
S. M. Kirova 1956, 21, 163-166
Inspired by Jørgensen’s work with R-imino esters (5,
Figure 1)15,16 and that of Yamamoto’s group concerning
the use of o-alkoxy aniline-derived aldimines in Lewis
acid-catalyzed Mannich-type reactions,21 we focused on
the bidentate coordination of imine substrates to Lewis
acid metal centers which would reduce the degrees of
freedom of the reaction system. To mimic this effect, but
(6) (a) Akhtar, M. S.; Sharma, V. L.; Seth, M.; Bhaduri, A. P. Indian
J. Chem. 1988, 27B, 448-451. (b) Sturgess, M. A.; Yarberry, D. J.
Tetrahedron Lett. 1993, 34, 4743-4746.
(7) Kende, A. S.; Mendoza, J. S. Tetrahedron Lett. 1991, 32, 1699-
1702.
(8) (a) Lambert, A.; Rose, J. D. J. Chem. Soc. 1947, 1511-1513. (b)
Senkus, M. J. Am. Chem. Soc. 1946, 68, 10-12. (c) Johnson, H. G. J.
Am. Chem. Soc. 1946, 68, 12-14. (d) Johnson, H. G. J. Am. Chem.
Soc. 1946, 68, 14-18.
(9) Yamada, K.; Harwood: S. J.; Gro¨ger, H.; Shibasaki, M. Angew.
Chem., Int. Ed. Engl. 1999, 38, 3504.
(10) Shibasaki, M.; Sasai, H.; Arai, T. Angew. Chem., Int. Ed. Engl.
1997, 36, 1236 and references therein.
(15) Knudsen, K. R.; Risgaard, T.; Nishiwaki, N.; Gothelf, K. V.;
Jørgensen, K. A. J. Am. Chem. Soc. 2001, 123, 5843.
(16) Nishiwaki, N.; Knudsen, K. R.; Gothelf, K. V.; Jørgensen, K.
A. Angew. Chem., Int. Ed. 2001, 40, 2992.
(11) Yamada, K.; Moll, G.; Shibasaki, M. Synlett 2001, 980.
(12) For pertinent references in these areas see: (a) Cordova, A.
Acc. Chem. Res. 2004, 37, 102. (b) Josephsohn, N. S.; Snapper, M. L.;
Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 3734.
(13) Colvin, E. W.; Beck, A. K.; Seebach, D. Helv. Chim. Acta 1981,
64, 2264.
(17) Qian, C.; Gao, F.; Chen, R. Tetrahedron Lett. 2001, 42, 4673.
(18) Tsuritani, N.; Yamada, K.; Yoshikawa, N.; Shibasaki, M. Chem.
Lett. 2002, 276.
(19) Okino, T.; Nakamura, S.; Furukawa, T.; Takemoto, Y. Org. Lett.
2004, 6, 625.
(20) Nugent, B. M.; Yoder, R. A.; Johnston, J. N. J. Am. Chem. Soc.
2004, 126, 3418.
(14) Anderson, J. C.; Peace, S.; Pih, S. Synlett 2000, 850.
550 J. Org. Chem., Vol. 70, No. 2, 2005

Stereoselective Synthesis of 1,2-Diamines
TABLE 1. Comparison of the Imine Protecting Group
entry imine
methoda
t
yield of 11 (%)b anti/sync
1
2
3
4
5
6
7
8
10a
10a
10a
10a
10b
A
1 h
>95
>95
>95
>95
90
90/10
90/10
80/20
80/20
90/10
65/35
90/10
85/15
B Cu(OTf)2 5 min
B Sc(OTf)3
5 min
B Ti(Oi-Pr)4 5 min
A
1 h
2 h
2 h
10b B Sc(OTf)3
10c B Sc(OTf)3
10d B Cu(OTf)2 2 h
70
90
>95
FIGURE 2. OMB protected imines.
a Method A: AcOH (2.4 equiv) was added to a solution of lithium
1-nitropropanate (1.4 equiv) and imine (1.0 mmol) in THF at -78
°C then rt. Method B: 1-trimehtylsilyl nitropropanate (2.4 equiv)
was added to imine (0.40 mmol) and Lewis acid (5 mol %) in THF
at -78 °C. b Crude yield estimated from the mass balance and 1H
NMR. c From crude 1H NMR rounded to the nearest 5%.
ketimine derived from acetone. Despite considerable
effort we were unable to synthesize significant quantities
of analogous imines derived from acetophenone or ben-
zophenone due to incomplete reactions and difficulties
with purification.
Each of the imines (Figure 2) were subjected to both
the acetic acid-promoted and Lewis acid-promoted nitro-
Mannich reaction (eq 2, Table 2). Purification of the
without limiting variation in the imine substrate, we in-
vestigated imines bearing potentially chelating protecting
groups (Figure 1). A range of possible candidates were
synthesized and screened in both the acetic acid-
promoted nitro-Mannich reaction (Scheme 1) with nitro-
propane and the Lewis acid-catalyzed nitro-Mannich
reaction (Scheme 2) with trimethylsilyl nitropropanate
previously developed by us.
â-nitroamines 19 by chromatography was not possible
due to degradation and the data were collected from the
crude reaction mixtures. The relative stereochemistry
Imines 6-9 were found to be wholly unreactive under
nitro-Mannich reaction conditions employing either AcOH
or Lewis acids Cu(OTf)2, Sc(OTf)3, and Ti(Oi-Pr)4. Con-
versely the N-(2-methoxybenzyl) (OMB)-protected imine
10a was found to be highly reactive under both sets of
reaction conditions to give anti-11a in good yields and
diastereoselectivities (Table 1, entries 1-4). The ad-
ditional R-heteroatoms in imines 6-9 could be diminish-
ing the electrophilicity of the imine carbon through
electron donation or there are alternate binding modes
of the Lewis acid, which may not include the imine
nitrogen, that lead to inactive complexes. To compare the
efficiency of the OMB group we compared similar reac-
tions with PMB (10b entries 5 and 6) and PMP (10c entry
7) protected imines of benzaldehyde (eq 1). The OMB
imine 10a gave consistently higher selectivities and/or
yields under both sets of reaction conditions. To explore
whether the increased selectivity attributed to the OMB
group was due to sterics or the presence of a coordinating
group in the ortho position, we synthesized the N-(2-
methylbenzyl)-protected imine 10d and subjected it to
the Cu(OTf)2 Lewis acid-catalyzed reaction conditions.
The results (entry 8) did not illuminate the coordination
mode of 10a in these reactions.
1
was assigned based on H NMR coupling constants and
was confirmed by single-crystal X-ray analysis of deriva-
tives (vide supra). The diastereomeric ratios were mea-
1
sured from distinct signals in the H NMR spectrum.22
The diastereoselectivities of OMB-imines 12 and 16 are
far superior in comparison to our previous studies with
the corresponding PMB or PMP imines14 and were
another justification for us choosing to optimize diaste-
reoselectivites for OMB-imines.
General patterns in the data (Table 2) show both
methods A and B to be typically high yielding with
method A (acetic acid promoted) showing higher anti-
diastereoselectivity in most cases. Within method B,
among the Lewis acid catalysts Cu(OTf)2, Sc(OTf)3, and
Ti(Oi-Pr)4, the diastereoselectivity seems independent of
the Lewis acid used, but the use of Ti(Oi-Pr)4 often
resulted in the lowest conversion. Pyrrole imine 14 was
practically inert under all reaction conditions (entries
9-12). Possibly the existing hydrogen bonding between
the pyrrole NH and the imine lone pair interferes with
coordination of an acid. This was supported by the
N-methyl analogue 15 undergoing satisfactory reaction
(entries 13-16). Imine 16 derived from a straight chain
aldehyde gave up to a 90:10 diastereomeric ratio (entries
17-20) whereas the more sterically demanding isopropyl
imine 17 gave diastereomeric ratios closer to 1:1 (entries
21-24).
The crude â-nitroamines were subjected to SmI2 reduc-
tion to give 1,2-diamines to confirm the yields and sense
of diastereoselectivity assigned (eq 3, Table 3). We took
Mindful of our goal to develop a broad and robust
method for nitro-Mannich coupling we synthesized and
screened a range of OMB imines (Figure 2). Imines 12-
15 are heterocyclic-Schiff bases, 16 and 17 are aldimines
derived from simple aliphatic aldehydes, and 18 is a
(21) Saito, S.; Hatanaka, K.; Yamamoto, H. Tetrahedron 2001, 57,
875.
this opportunity to survey other common reducing sys-
J. Org. Chem, Vol. 70, No. 2, 2005 551

Anderson et al.
anti/sync
TABLE 2. Addition of Nitronate Species to Imines
entry
imine
methoda
t
19
R
yield of 19 (%)b
1
2
12
12
12
12
13
13
13
13
14
14
14
14
15
15
15
15
16
16
16
16
17
17
17
17
18
18
18
18
A
1 h
b
b
b
b
c
c
c
c
d
d
d
d
e
e
e
e
f
2-furyl
2-furyl
2-furyl
2-furyl
91
>95
>95
89
95/5
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
5 min
5 min
5 min
2 h
85/15
80/20
65/35
95/5
3
4
5
2-thiophenyl
2-thiophenyl
2-thiophenyl
2-thiophenyl
2-pyrrole
2-pyrrole
2-pyrrole
2-pyrrole
2-(N-methyl)pyrrole
2-(N-methyl)pyrrole
2-(N-methyl)pyrrole
2-(N-methyl)pyrrole
n-hexyl
71
6
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
A
B Cu(OTf)2
B Sc(OTf)3
B Ti(Oi-Pr)4
45 min
45 min
45 min
6 h
86
75/25
80/20
75/25
7
>95
72
8
9
<5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
3 h
<5
3 h
<5
3 h
<5
3 h
89
75/25
75/25
75/25
80/20
90/10
90/10
85/15
80/20
60/40
50/50
65/35
50/50
1 h
>95
>95
60
1 h
1 h
1 h
>95
>95
>95
>95
92
5 min
5 min
5 min
1 h
f
n-hexyl
f
f
n-hexyl
n-hexyl
g
g
g
g
h
h
h
h
i-Pr
5 min
5 min
5 min
1 h
5 min
5 min
5 min
i-Pr
90
i-Pr
>95
82
i-Pr
Me2
93
Me2
>95
90
Me2
Me2
89
a Method A: AcOH (2.4 equiv) was added to a solution of lithium 1-nitropropanate (1.4 equiv) and imine (1.0 mmol) in THF at -78 °C
then rt. Method B: 1-trimehtylsilyl nitropropanate (2.4 equiv) was added to imine (0.40 mmol) and Lewis acid (5 mol %) in THF at -78
°C. b Crude yield estimated from the mass balance and 1H NMR. c From crude 1H NMR rounded to the nearest 5%.
TABLE 3. Isolated Yields of 1,2-Diamine Derivatives
system for this reduction as long as the reagent was
freshly prepared.6b We encountered difficulties in purify-
ing the 1,2-diamines with significant loss of mass upon
chromatography and found bulb-to-bulb distillation un-
satisfactory. The 1,2-diamines visibly degraded in the
atmosphere over 24 h. As a consequence we found it
convenient to treat the crude reduction mixture with
phosgene to generate the cyclic ureas, OMB-imidazoli-
dinones (eq 3, Table 3). The isolated yield quoted is for
three steps from imine to diastereomerically pure 20.
The three-step sequence of nitro-Mannich coupling of
OMB-imines with a nitroalkane, followed by reduction
and cyclic urea formation can give cis-diamine derivatives
in moderate to good overall yields. It appears that
â-nitroamines derived from ketimines suffer from ret-
roaddition during the reduction step when an OMB
protecting group is used (20h). In our original acetic acid-
promoted reaction the nitroamine from 2-nitropropane
and N-PMB benzylideneamine furnished a moderate
yield of the desired 1,2-diamine (48%).2 This suggests that
diamines possessing tertiary centers can only be derived
from secondary nitro compounds and not ketimines. The
Single-crystal X-ray analysis of 20a-c,e,g confirmed the
cis-stereochemistry which we had assumed from 1H NMR
analysis.27 Diamine derivative 20f was assigned cis by
20
R1
yield (%)a
20a
20b
20c
20e
20f
20g
20h
Ph
74
2-furyl
87b
59b
69b
83b
26c
2-thiophenyl
2-(N-methyl)pyrrole
n-hexyl
i-Pr
Me2
d
-
a Yield over three steps from imine to diastereomerically pure
cis-20. b Minor isomer removed by recrystallization. c Separated
by chromatography from a 1:1 mixture of nitro-Mannich product
after cyclic urea formation. d N-OMB-isopropylamine isolated in
76% yield after reduction.
tems, which had been used for the reduction of nitroal-
kanes, on crude â-nitroamine 11a. Standard Fe or Sn
metal reduction under acidic conditions gave a complex
mixture of products.23 Treatment with LiAlH4 furnished
OMB-benzylamine (55%) from reduction of the iminium
ion from retroaddition of nitropropane from 11a.24 The
use of nickel borohydride25 gave unreacted starting
material among a mixture of degradation products.
Hydrogenolysis26 with Pt/H2 or with Raney nickel alone
gave recovered starting material. Raney nickel reduction
in the presence of atmospheric H2 gave only ∼10%
reduced product along with recovered starting material
after 24 h.15 In keeping with our previous studies we
found SmI2 in THF/MeOH to be an effective and mild
1
analogy of its H NMR spectrum.
To complete the synthesis of 1,2-diamine derivatives
we wanted to demonstrate the removal of the OMB
protecting group. Inspection of the literature revealed no
precedent for cleavage of N-OMB bonds so we studied
analogous methods used for the cleavage of N-PMB
groups using 20a (R ) Ph). The use of one-electron donors
CAN28 and DDQ29 was ineffective, either giving no
reaction or eliminated product 22. Hydrogenolysis was
(22) See Supporting Information.
(23) Johnson, K.; Degering, E. F. J. Am. Chem. Soc. 1939, 61, 3194.
(24) Bunce, S. C.; Clemans, S. D.; Bennett, B. A. J. Org. Chem. 1975,
40, 961.
(25) Osby, J. O.; Ganem, B. Tetrahedron Lett. 1985, 26, 6413.
(26) Iffland, D. C.; Cassis, F. A. J. Am. Chem. Soc. 1952, 74, 6284.
552 J. Org. Chem., Vol. 70, No. 2, 2005

Stereoselective Synthesis of 1,2-Diamines
TABLE 4. Removal of the OMB Protecting Group
tration of 23 is attainable. This is supported by the
observation that decreasing the level of nitropropane
from solvent to 5 or 10 equiv in CH2Cl2, THF, or MeCN
results in <5% reaction in 16 h.
20
R1
yield of 21 (%)a yield of 22 (%)a
20a Ph
95
87
94
-
95
-
20b 2-furyl
The nitro-Mannich reaction of OMB-imine 10a in neat
nitropropane was significantly accelerated by the addi-
tion of substoichiometric amounts of AcOH. We found
that the optimum amount of AcOH for the most syn-
selective reaction appeared to be 20 mol %, where syn-
11a was formed in 71% yield of the reaction mixture (eq
5, anti-11a/syn-11a/24, 20:75:5). We may expect a Brøn-
sted acid to raise the levels of 23 by analogy with keto/
enol tautomerism and thus promote this reaction. Elimi-
nation product 24 was formed in exclusively the E-form32
and is derived from standard acid-catalyzed Henry
elimination of OMB-NH2.
20c
20e
20f
2-thiophenyl
2-(N-methyl)pyrrole
n-hexyl
75
20g i-Pr
a Isolated yield.
SCHEME 3
similarly ineffective and the use of harsher conditons
employing AlCl3 in refluxing anisole30 caused complete
degradation. Clean deprotection of the OMB group was
achieved by refluxing in neat TFA (eq 4) to give high
yields of 21 (Table 4).31 This method proved successful
for most of the diamines 20, except N-methylpyrrole
derivative 20e underwent facile elimination to 22 and
the i-Pr diamine 20g proved unstable to the reaction
conditions.
Conversion of syn-rich 11a to the 1,2-diamine deriva-
tive, imidazolidinone trans-20a, proceeded as before by
reduction with SmI2 followed by cyclization with phos-
gene in 41% isolated yield from starting imine. This
variant of the nitro-Mannich reaction is the subject of
further investigations to engineer a general syn-selective
process.
We have investigated the Lewis acid-catalyzed and
AcOH-promoted nitro-Mannich reaction as a diastereo-
selective route to certain 1,2-diamine derivatives. Only
carbon benzyl or phenyl N-imine protecting groups were
successful with OMB better than PMB and PMP. The
anti-rich â-nitroamines were reduced to 1,2-diamines
with SmI2 and the OMB group readily removed by
refluxing TFA from the imidazoline derivatives in high
isolated yields and diastereomeric purity. Work is con-
tinuing to understand the sense and magnitude of
diastereoselection under the different reaction conditions
reported in this paper. The Lewis acid-catalyzed ex-
amples will form the basis of enantioselective studies
with the use of chiral multidentate ligands and Cu(II),
Sc(III), or Ti(IV) complexes. Preliminary results show
that acetic acid can catalyze a syn-selective nitro-Man-
nich reaction with OMB-imine in neat nitropropane and
we are currently looking at other Brønsted acids and
bases which may facilitate an efficient and general
enantio- and syn-selective process.
The two nitro-Mannich protocols presented above give
inherently anti-â-nitroamines, identical with similar
processes reported in the literature.11,15,16,18 We have
investigated modifications of our protocols to produce syn
products. We found that imine 10a bearing an OMB
protecting group underwent a slow nitro-Mannich reac-
tion in neat nitropropane at room temperature. After 16
h, 35% of imine had been consumed giving â-nitro-amine
product 11a with an anti/syn ratio of 4:1. Conversion
tended to a maximum of ∼75% (>72 h), but with an
eroded diastereoselectivity of 2:1. We attribute this
reaction to the small concentration of nitronic acid
tautomer 23 that can protonate imine 10a to allow
addition of the nitronate anion (Scheme 3).2
The nitro/nitronic acid tautomeric equilibrium lies
heavily in favor of the nitro species and we postulate that
it is only in neat nitropropane that a significant concen-
(27) Crystallographic data (excluding structure factors) for cis-20a
and 20c,e,g have been deposited with the Cambridge Crystallographic
Data Centre as supplementary publication numbers CCDC 242114-
7, respectively. Copies of the data can be obtained free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax
+44-(0)1223-336033 or e-mail deposit@ccdc.cam.ac.uk]. Although we
have clearly established the stereochemistry by single-crystal X-ray
diffraction methods for compound 20b, there were significant problems
with crystal quality and the resulting structure was not of sufficient
quality for publication.
Experimental Section
General Procedure for the Synthesis of Imines. Rigor-
ously dried 4 Å molecular sieves (1.0 g per mmol) were added
to a solution of amine in DCM (5 mL per mmol) under N2 at
room temperature and after a period of 5 min carbonyl
compound (1 equiv) was added. The mixture was left to stir
for 14 h at room temperature before filtration through Celite
and removal of solvents in vacuo, to yield the crude imine.
Imines were judged >95% pure by 1H NMR, but were purified
if possible. For spectroscopic data see the Supporting Informa-
tion.
(28) Yoshimura, J.; Yamaura, M.; Suzuki, T.; Hashimoto, H. Chem.
Lett. 1983, 1001.
(29) Singh, S. B. Tetrahedron Lett. 1995, 36, 2009.
(30) Akiyama, T.; Takesue, Y.; Kumegawa, M.; Nishimoto, H.; Ozaki,
S. Bull. Chem. Soc. Jpn. 1991, 64, 2266.
(31) This method was developed from the known cleavage of PMB
groups from amides with neat TFA at room temperature. Brooke, G.
M.; Mohammed, S.; Whiting, M. C. Chem. Commun. 1997, 1511.
(32) Kawai, Y.; Inaba, Y.; Tokitoh, N. Tetrahedron: Asymmetry 2001,
12, 309.
J. Org. Chem, Vol. 70, No. 2, 2005 553

Anderson et al.
General Procedure for the AcOH-Promoted Nitro-
Mannich Reaction (method A, Tables 1 and 2). A solution
of n-butyllithium (2.5 M in hexanes, 0.56 mL, 1.40 mmol) was
added to a solution of 1-nitropropane (0.13 mL, 1.40 mmol) in
THF (20 mL) over 10 min at -78 °C under N2 and the mixture
was stirred for a 10 min. A solution of imine (1.00 mmol) in
THF (10 mL) was then added, the mixture was stirred for 10
min at -78 °C, and then AcOH (0.14 mL, 2.4 mmol) was added
and the mixture was stirred for a further 20 min before being
allowed to warm to 25 °C over 30 min. Saturated aq NaHCO3
(10 mL) and Et2O (20 mL) were then added, and the organic
phase was separated, washed with satd aq NaHCO3 (10 mL),
dried (MgSO4), and evaporated to yield crude â-nitroamine.
These were analyzed by 1H NMR. Purification (chromatogra-
phy on silica or alumina) was attempted with some crude
mixtures but gave very little of the desired material. MS under
a variety of techniques causes retroaddition and no collection
was made.
d, J ) 13.4 Hz), 3.34-3.42 (2H, m), 6.40-6.55 (1H, m), 6.66-
6.73 (2H, m), 6.82-6.90 (2H, m), 7.08 (2H, m); 13C NMR δ 10.3,
23.0, 47.6, 54.9, 60.1, 95.4, 110.4, 120.7, 125.5, 126.8, 127.1,
128.7, 130.3, 143.6, 154.9, 158.1.
19e: Isolated as a yellow oil. NMR data for major anti
1
(1S*,2R*) H NMR δ 0.71 (3H, t, J ) 7.2 Hz), 1.52 (1H, m),
1.97 (2H, m), 2.95 (3H, s), 3.26 (3H, s), 3.62 (2H, m), 3.80 (2H,
m), 4.12 (1H, d, J ) 7.2 Hz), 4.41 (1H, ddd, J ) 11.1, 7.4, 3.8
Hz), 6.19-6.28 (5H, m), 6.45-6.51 (2H, m), 6.80 (2H, m), 7.02-
7.15 (4H, m); 13C NMR δ 10.3, 23.5, 33.3, 47.1, 54.8, 56.5, 94.1,
107.7, 108.0, 110.4, 120.6, 122.9, 128.3, 128.6, 130.1, 154.0,
158.2; NMR data for minor syn (1R*, R*) 1H NMR δ 0.56 (3H,
t, J ) 7.3 Hz), 1.18 (1H, m), 2.90 (3H, s), 3.29 (3H, s), 3.62
(2H, m), 3.80 (2H, m), 4.16 (1H, d, J ) 10.0 Hz), 4.61 (1H, td,
J ) 10.2, 2.7 Hz), 5.98 (1H, m), 6.19-6.28 (5H, m), 6.45-6.51
(2H, m), 6.80 (2H, m), 7.02-7.15 (4H, m).
19f: Isolated as a yellow oil. NMR data for major anti
(1S*,2R*) 1H NMR δ 0.87 (3H, t, J ) 7.2 Hz), 0.97 (3H, t, J )
7.4 Hz), 1.10-1.48 (8H, m), 1.81-2.16 (2H, m), 2.89 (1H, m),
3.78 (1H, d, J ) 13.1 Hz), 3.84 (4H, m), 4.47 (1H, ddd, J )
10.6, 5.9, 3.5 Hz), 6.88 (1H, d, J ) 8.0 Hz), 6.93 (1H, td, J )
7.4, 0.9 Hz), 7.20-7.30 (2H, m); 13C NMR δ 10.8, 14.0, 22.1,
22.8, 25.4, 30.8, 31.8, 47.5, 55.2, 59.4, 93.6, 110.2, 120.6, 128.5,
128.6, 130.1, 157.7.
General Procedure for the Lewis Acid-Catalyzed Ni-
tro-Mannich Reaction (method B, Tables 1 and 2). A
solution of imine (0.40 mmol) and Lewis acid catalyst [Cu-
(OTf)2, Sc(OTf)3, or Ti(Oi-OPr)4, 5 mol %) in THF (1 mL) was
stirred for 5 min at room temperature and then cooled to -78
°C. A solution of O-trimethylsilyl nitropropanate in THF (3
mL) was then added over a period of 15 min via cannula. The
mixture was left to stir at -78 °C for 30 min before filtration
through poly(4-vinylpyridine) to remove Cu catalysts, or
neutral alumina to remove Sc or Ti catalyst. Removal of
solvent in vacuo gave the crude â-nitroamine. These were
19g: Isolated as a yellow oil. NMR data for major anti
(1S*,2R*) 1H NMR δ 0.57 (3H, d, J ) 6.6 Hz), 0.67-0.81 (9H,
m), 1.27 (1H, dqd, J ) 14.7, 7.3, 3.7 Hz), 1.53 (1H, hepd, J )
6.8, 4.3 Hz), 1.74 (1H, ddq, J ) 14.6, 10.7, 7.2 Hz), 2.76 (1H,
dd, J ) 7.8, 4.3 Hz), 3.29 (3H, s), 3.90 (2H, s), 4.27 (2H, m),
6.50, (2H, dd, J ) 8.1, 2.2 Hz), 6.87 (2H, tdd, J ) 7.4, 3.2, 1.0
Hz), 7.08 (2H, tdd, J ) 8.0, 3.3, 1.7 Hz), 7.19 (1H, dd, J ) 7.3,
1.7 Hz); 13C NMR δ 10.5, 16.4, 20.3, 24.4, 29.9, 50.4, 54.7, 65.0,
94.2, 110.5, 120.9, 128.6, 129.2, 130.3, 157.9; NMR data for
minor syn (1R*,2R*) 1H NMR δ 0.67-0.81 (9H, m), 1.64 (1H,
hepd, J ) 6.8, 3.7 Hz), 1.88 (1H, dqd, J ) 14.7, 7.4, 3.2 Hz),
1.97 (1H, ddq, J ) 14.6, 10.9, 7.2 Hz), 2.88 (1H, dd, J ) 7.9,
3.7 Hz), 3.31 (3H, s), 3.78 (2H, d, J ) 13.0 Hz), 4.27 (2H, m),
6.50, (2H, dd, J ) 8.1, 2.2 Hz), 6.87 (2H, tdd, J ) 7.4, 3.2, 1.0
Hz), 7.08 (2H, tdd, J ) 8.0, 3.3, 1.7 Hz), 7.36 (1H, dd, J ) 7.3,
1.7 Hz); 13C NMR δ 10.8, 16.1, 20.2, 24.0, 30.1, 50.7, 54.7, 65.3,
93.6, 110.4, 120.9, 128.9, 129.4, 130.3, 157.9.
1
analyzed by H NMR. Purification (chromatography on silica
or alumina) was attempted with some crude mixtures but gave
very little of the desired material. MS under a variety of
techniques causes retroaddition and no t collection was made.
11a: Isolated as a yellow oil. NMR data for anti (1S*,2R*)
1H NMR δ 0.60 (3H, t, J ) 7.4 Hz), 1.67 (1H, dqd, J ) 14.9,
7.4, 3.0 Hz), 1.97 (1H, ddq, J ) 14.7, 10.9, 7.3 Hz), 3.31 (3H,
s), 3.54 (1H, d, J ) 13.5 Hz), 3.80 (1H, d, J ) 13.5 Hz), 4.04
(1H, d, J ) 6.1 Hz), 4.39 (1H, ddd, J ) 11.1, 6.1, 3.0 Hz), 6.49
(1H, m), 6.82 (1H, m), 6.86 (2H, m), 7.00-7.20 (5H, m); 13C
NMR δ 10.4, 22.5, 47.5, 54.8, 64.4, 95.2, 110.6, 120.9, 127.8,
128.7, 128.9, 130.0, 130.6, 139.2, 158.1; NMR data for syn
1
(1R*,2R*) H NMR δ 0.50 (3H, t, J ) 7.4 Hz), 1.53 (1H, ddq,
1
19h: Isolated as a yellow oil. H NMR δ 0.78 (3H, t, J )
J ) 14.9, 10.8, 7.2 Hz), 3.28 (3H, s), 3.52 (1H, d, J ) 13.9 Hz),
3.78 (1H, d, J ) 13.9 Hz), 3.97 (1H, d, J ) 9.7 Hz), 4.46 (1H,
ddd, J ) 11.0, 9.7, 3.3 Hz), 6.49 (1H, m), 6.82 (1H, m), 6.86
(2H, m), 7.00-7.20 (5H, m);13C NMR δ 10.2, 20.1, 47.2, 54.7,
64.5, 95.8, 110.5, 120.6, 127.8, 128.6, 128.8, 130.2, 130.6, 139.1,
158.1.
7.4 Hz), 1.08 (3H, s), 1.16 (3H, s), 1.49 (1H, m), 2.01 (1H, m),
3.48 (3H, s), 3.76 (1H, d, J ) 12.7 Hz), 3.86 (1H, d, J ) 12.8
Hz), 4.40 (1H, dd, J ) 11.9, 2.4 Hz), 6.64 (1H, d, J ) 8.1 Hz),
6.99 (1H, t, J ) 7.4 Hz), 7.20 (1H, t, J ) 9.7), 7.47 (1H, d, J )
7.4 Hz); 13C NMR δ 11.0, 21.9, 23.8, 41.3, 54.7, 55.3, 97.7, 110.3,
120.9, 128.1, 128.3, 129.5, 129.8, 157.7.
11b: For characterization data see ref 2.
11c: For characterization data see ref 13.
General Procedure for Reduction of the Nitro Group6b
and Subsequent Formation of Imidazolidinone. A triple
evacuation/N2 fill was carried out on a suspension of samarium
metal (40 mesh, 5.26 g, 35.0 mmol) and 1,2-diiodoethane (9.16
g, 32.5 mmol) in a Schlenk tube. The mixture was diluted with
THF (50 mL) and stirred for 1 h under N2. A further 150 mL
of THF was added and the mixture was stirred for 2 h until
an intense, deep blue solution was obtained. Crude â-nitroam-
ine (5.00 mmol) in MeOH/THF (1:1, 30 mL) was added and
the mixture was stirred for 16 h at room temperature. A
solution of oxalic acid (8.60 g) in water (100 mL) was added
and the mixture was filtered (Celite) and the organics removed
in vacuo. The resulting aqueous phase was basified to pH >12
(2 M, NaOH) and extracted with EtOAc. The combined
organics were washed with satd aq Na2S2O3 and satd brine
before drying (MgSO4) and evaporation to yield crude, mono-
protected 1,2-diamine.
A solution of phosgene (20% in toluene, 2.65 mL, 5.00 mmol)
was added dropwise to a solution of the crude diamine from
above in CH2Cl2 (30 mL) and Et3N (1.50 mL, 10.0 mmol) at
-20 °C. After addition the mixture was allowed to warm to
25 °C over 30 min, 5 M NaOH (20 mL) was added, and the
organic phase was separated, dried (MgSO4), and evaporated
to a yellow/brown highly viscous oil. After being dissolved in
11d: Prepared using general procedure B, isolated as a
yellow oil. NMR data for major anti (1S*,2R*) 1H NMR δ 0.75
(3H, t, J ) 7.4 Hz), 1.75-2.05 (2H, m), 2.21 (3H, s), 3.38 (1H,
d, J ) 12.6 Hz), 3.52 (1H, d, J ) 12.6 Hz), 4.05 (1H, d, J ) 7.2
Hz), 4.41 (1H, m), 7.10-7.30 (9H, m); 13C NMR δ 10.4, 19.1,
25.2, 49.9, 65.0, 95.0, 126.2, 127.9, 128.4, 129.1, 129.3, 129.4,
130.8, 137.2, 138.1, 139.3.
19b: Isolated as a yellow oil. NMR data for major anti
1
(1S*,2R*) H NMR δ 0.67 (3H, t, J ) 7.4 Hz), 1.88 (2H, m),
3.28 (6H, s), 3.62 (2H, m), 3.80 (2H, m), 4.13 (1H, d, J ) 7.3
Hz), 4.51 (1H, ddd, J ) 10.1, 7.5, 3.9 Hz), 6.00 (1H, m), 6.06
(1H, m), 6.50 (2H, m), 6.84 (2H, m), 6.95-7.25 (6H, m); 13C
NMR δ 10.1, 23.7, 47.4, 54.8, 58.7, 92.8, 108.8, 110.4, 120.9,
128.3, 128.4, 128.7, 129.9, 142.5, 152.0, 157.9; NMR data for
minor syn (1R*,2R*) 1H NMR δ 0.57 (3H, t, J ) 7.3 Hz), 1.11
(1H, m), 1.62 (1H, m), 3.28 (6H, s), 3.62 (2H, m), 3.80 (2H, m),
4.09 (1H, d, J ) 9.9 Hz), 4.61 (1H, td, J ) 10.2, 3.3 Hz), 5.83
(1H, d, J ) 3.2 Hz), 5.96 (1H, m), 6.50 (2H, m), 6.84 (2H, m),
6.95-7.25 (6H, m).
19c: Isolated as a yellow oil. NMR data for major anti
1
(1S*,2R*) H NMR δ 0.59 (3H, t, J ) 7.4 Hz), 1.74 (1H, m),
1.95 (1H, m), 3.31 (3H, s), 3.62 (1H, d, J ) 13.5 Hz), 3.92 (1H,
554 J. Org. Chem., Vol. 70, No. 2, 2005

Stereoselective Synthesis of 1,2-Diamines
the minimum amount of hot EtOAc/petroleum ether, the
product imidazolidinone was crystallized and obtained as a
solid. Further recrystallization (EtOAc/petroleum ether mix-
tures) to remove the minor diastereoisomer was carried out
where indicated. For spectroscopic data see the Supporting
Information.
General Procedure for the Removal of the 2-Meth-
oxybenzyl Group. A solution of OMB-imidazolidinone (0.5
mmol) in TFA (5 mL) was stirred at reflux (∼95 °C). After
consumption of starting material (TLC) the TFA was removed
in vacuo to yield a dark red, viscous oil. After addition of EtOAc
and washing with saturated NaHCO3, the organic phase was
dried (MgSO4) and evaporated to give a viscous, yellow oil.
Flash-column chromatography (1:2 EtOAc/petroleum ether)
gave the imidazolidinone product as a white solid.
21a:. According to the general procedure above 20a (155 mg,
0.50 mmol) gave 21a as a white solid (94 mg, >99%): mp
191.0-193.0 °C (lit.2 mp 192.0-194.0 °C); all other spectro-
scopic data were in agreement with those already published
by us.2
(EI+) 196 (32%, M+), 136 (22%), 135 (100%), 88 (34%), 78 (65%);
HRMS C9H12N2OS calcd 196.0670, found 196.0679.
21f: According to the general procedure above 20f (152 mg,
0.5 mmol) gave 21f as a white solid (92 mg, >99%): mp 85.1-
1
87.5 °C; IR νmax (solid) 3217, 1698, 1463, 770 cm-1; H NMR
(CDCl3) δ 0.87-0.98 (6H, m), 1.18-1.40 (6H, m), 1.41-1.59
(4H, m), 3.61 (1H, q, J ) 7.0 Hz), 3.71 (1H, ddd, J ) 8.9, 7.8,
4.6 Hz), 4.84 (1H, br s), 4.98 (1H, br s); 13C NMR (CDCl3) δ
10.8, 14.0, 22.5, 22.7, 26.0, 29.6, 31.8, 56.2, 57.8, 164.5; m/z
(EI+) 184 (6%, M+), 155 (68%, M+ - Et), 113 (100%, M+
-
n-pen), 75 (12%), 58 (32%); HRMS C10H20N2O calcd 184.1576,
found 184.1569.
General Procedure for Thermodynamic Reactions. To
a solution of imine 10a (0.40 mmol) in 1-nitropropane (2.50
mL) was added acetic acid (20 mol %) and the mixture was
stirred for 18 h at room temperature. Filtration through poly-
(4-vinylpyridine) and removal of solvent at ∼40 °C in vacuo
gave the crude â-nitroamine product, which was analyzed by
NMR spectroscopy: 95% conversion, anti-11a/syn-11a/24, 20:
75:5.
21b: According to the general procedure above 20b (150
mg, 0.5 mmol) gave 21b as a white solid (78 mg, 87%): mp
182.0-184.0 °C; IR νmax (solid) 3206, 1698, 1455, 1260, 1148,
Acknowledgment. This work is part of the Ph.D.
Thesis of G. P. H., University of Nottingham, 2004. We
thank the EPSRC and GlaxoSmithKline for financial
support, Drs. A. Ross and R. Lawrence for helpful
discussions, Mr. T. Hollingworth and Mr. D. Hooper for
providing mass spectra, and Mr. T. J. Spencer for
microanalytical data.
1
1003, 750, 719 cm-1; H NMR (CDCl3) δ 0.84 (3H, t, J ) 7.5
Hz), 1.22 (2H, m), 3.90 (1H, dt, J ) 8.3, 5.6 Hz), 4.88 (1H, d,
J ) 8.2 Hz), 5.25 (1H, br s), 5.60 (1H, br s), 6.33 (2H, m), 7.36
(1H, s); 13CNMR (CDCl3) δ 10.7, 24.4, 54.4, 58.7, 108.1, 110.5,
142.5, 152.2, 163.5; m/z (EI+) 180 (100%, M+), 109 (53%), 94
(26%), 67 (30%), 59 (26%); HRMS C9H12N2O2 calcd 180.0899,
found 180.0893.
21c: According to the general procedure above 20c (158 mg,
0.5 mmol) gave 21c as a white solid (92 mg, 94%): mp 161.0-
Supporting Information Available: General experimen-
tal, spectroscopic data for compounds 10a-d, 12-18, cis- and
163.0 °C; IR νmax (solid) 3199, 1693, 1456, 1236, 773, 687 cm-1
;
1
trans-20a, 20b,c,e,f, and cis- and trans-20g and copies of H
1H NMR (CDCl3) δ 0.84 (3H, t, J ) 7.4 Hz), 1.23 (2H, m), 3.87
(1H, q, J ) 6.1 Hz), 5.12 (1H, d, J ) 8.0 Hz), 5.23 (1H, br s),
5.49 (1H, br s), 6.98 (2H, m), 7.26 (1H, s); 13C NMR (CDCl3) δ
10.8, 22.2, 56.3, 59.2, 125.1, 125.6, 128.0, 142.0, 163.8; m/z
and 13C NMR data. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO048304H
J. Org. Chem, Vol. 70, No. 2, 2005 555