A convenient synthesis of 4-aminoaryl substituted cyclic imides

Article abstract of DOI:10.1016/S0040-4039(03)00167-9

An efficient one-step synthesis of 4-nitro-N-aryl substituted glutarimides, succinimides and maleimides in polyphosphoric acid is described together with the subsequent reduction to the corresponding anilines. The scope and limitation of this cyclocondensation are presented.

Full text of DOI:10.1016/S0040-4039(03)00167-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 2133–2136
A convenient synthesis of 4-aminoaryl substituted cyclic imides
Werner W. K. R. Mederski,* Manfred Baumgarth, Martina Germann, Dieter Kux and
Thomas Weitzel
Merck KGaA, Preclinical Pharmaceutical Research, 64271 Darmstadt, Germany
Received 13 November 2002; revised 17 December 2002; accepted 16 January 2003
Abstract—An efficient one-step synthesis of 4-nitro-N-aryl substituted glutarimides, succinimides and maleimides in polyphospho-
ric acid is described together with the subsequent reduction to the corresponding anilines. The scope and limitation of this
cyclocondensation are presented. © 2003 Elsevier Science Ltd. All rights reserved.
As part of our research program to obtain new
inhibitors of coagulation factor Xa (fXa), we have
synthesized and evaluated 4-aminophenylpiperidones
(surrogates for the S4 binding pocket) such as benz-
amidine 1a.1 In order to extend the scope of this class
of compounds, cyclic imides of type 1b were envisaged
(Fig. 1).
mal and chemical dehydration to the corresponding
cyclic imides. One of the precedents utilized aromatic
diamines which afforded the respective 2-aminoaryl
derivatives in one step.8 However, cyclic anhydrides
were applied in all cases.
In an attempt to prepare 4-chloroimidazo[4,5-c]-
pyridines with a butyric acid side chain in the 2 posi-
tion, we have found that the conversion of 2-chloro-3,4-
diaminopyridine 2 with glutaric acid 3a in hydrogen
chloride unfortunately led to the 4-oxo compound 4 in
high yield (Scheme 1). However, utilizing polyphospho-
ric acid (PPA), only cyclic imide 5 could be obtained in
moderate yield. This is obviously due to the nature of
PPA, which usually has a phosphorus pentoxide con-
tent of 82–85% and is therefore a versatile acidic dehy-
drating reagent.9 This result led us to investigate the
generality of this reaction.
N-Substituted cyclic imides are key moieties in natu-
rally occurring alkaloids and in synthetic compounds
that have antispasmodic effects2 and show antinocicep-
tive properties,3 respectively.
The formation of N-substituted cyclic imides has been
studied comprehensively and some procedures have
been developed and implemented. Most of the transfor-
mation to such ring structures involves a three-step
synthesis: (1) reaction of a cyclic anhydride with an
aniline, (2) conversion of the intermediate monoamide
to an activated ester with N,N%-disuccinimidyl oxalate,4
acetic anhydride3,5,6 or thionyl chloride,7 and (3) ther-
Herein, we report on the use of PPA as a more suitable
and convenient reagent to prepare N-aryl substituted
Figure 1. Representative potent fXa inhibitor 1a and his putative congener 1b.
Keywords: condensations; cyclisation; N-imides; nitro compounds.
* Corresponding author. E-mail: mederski@merck.de
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00167-9

2134
W. W. K. R. Mederski et al. / Tetrahedron Letters 44 (2003) 2133–2136
Scheme 1. Selective reaction of compound 2 with glutaric acid 3a to 4 or 5, respectively.
cyclic imides from diacids. The scope and limitation of
this reaction are presented together with the subsequent
reduction of the intermediate nitro compounds.
In order to test the influence of the substitution pat-
tern of the dicarboxylic acids some selected, commer-
cially available diacids were reacted with 4-nitroaniline
6a under standard conditions. 3,3-Dimethylglutaric
acid 3c could be converted to the glutarimide 7c in
high yield whereas the 3-phenyl derivative failed in
this reaction (data not shown). However, with digly-
colic acid 3d and iminodiacetic acid 3e a new route to
C2-symmetric analogues of Tro¨ger’s base was discov-
ered.11 Both acids underwent exclusively condensation
with 4-nitroaniline 6a to the corresponding
methanodibenzo[b,f ]diazocine 8 (Scheme 2) in low to
moderate yield.12 This condensation afforded the hith-
erto inaccessible nitro analogue of Tro¨ger’s base. No
trace of the corresponding cyclic imides could be
detected.
We initially investigated the effect of PPA on the reaction
of dicarboxylic acids 3a–c with different 4-nitroanilines
6a–j, and the results are summarized in Table 1.
As shown in Table 1, amidation and following ring
closure afforded cyclic imides 7a–p in excellent yields and
purities.10 Heating the respective 4-nitroanilines at 80°C
for 12 h in polyphosphoric acid the desired glutar- and
succinimides were prepared in one step, respectively. The
condensation proceeded smoothly for anilines bearing
either an electron withdrawing or an electron donating
group at the 2 or 3 position.
Table 1. Reaction of 6a–j with dicarboxylic acids 3a–c to 7a–p and subsequent reduction to 8a–o
Entry
6
3
R
A
7
8
Yielda 7(8)%
1
2
3
4
5
6
7
8
6a
6a
6a
6b
6c
6d
6e
6f
3a
3b
3c
3b
3b
3b
3b
3b
3b
3b
3b
3a
3a
3a
3a
3a
H
H
H
2-Cl
2-Me
2-OMe
2-CF3
3-CF3
2-NO2
2,3-diCl
2-Ph
CH2
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
8a
8b
8c
8d
8e
8f
8g
8h
8ib
8j
8k
–
8l
8m
8n
8o
97(85)
91(75)
98(71)
89(83)
93(57)
89(95)
97(70)
93(58)
92(48b)
91(86)
89(59)
95(–)
bond
C(Me)2
Bond
Bond
Bond
Bond
Bond
Bond
Bond
Bond
CH2
9
6g
6h
6i
10
11
12
13
14
15
16
6j
2-Br
6c
6d
6e
6f
2-Me
2-OMe
2-CF3
3-CF3
CH2
CH2
CH2
CH2
98(93)
97(84)
98(90)
95(85)
a Isolated yield; structures were confirmed by 1H NMR as well as HRMS or elemental analyses.
b R=2-NH2.

W. W. K. R. Mederski et al. / Tetrahedron Letters 44 (2003) 2133–2136
2135
Scheme 2. Reaction of compound 6a with diglycolic acid 3d and iminodiacetic acid 3e to analogue 8 of Tro¨ger’s base.
Scheme 3. Reaction of amines 6k–m with glutaric acid 3a to the open chain acids 10a–c and amines 6a and 6e with maleic acid
to pyrrolediones 11a and 11b.
In summary, we have developed a novel, facile method-
ology for the one step synthesis of 4-nitroaryl substi-
tuted cyclic imides in high purity and yield. However,
application to other aryl and heteroaryl compounds has
to be confirmed. The subsequent reduction to the corre-
sponding anilines offers an easy entry into the design of
factor Xa inhibitors bearing a cyclic imide in the S4
pocket. The results of this investigation will be reported
in due course.
We next expanded this cyclocondensation to maleic
acid (Scheme 3). Starting from anilines 6a and 6e the
preparation of maleimides 11a and 11b proceeded in
good to moderate yields, respectively.
A different course of reaction was observed for aniline
substrates 6k–m with a hydroxy-, carboxy- or benzoyl
substituent at the 2 position (Scheme 3). In these cases
the intermediate monoamides 10a–c were formed as the
sole products in moderate yields. The outcome of this
reaction can be better ascribed to electronic effects than
to steric factors.
References
1. Dorsch, D.; Mederski, W.; Tsaklakidis, C.; Cezanne, B.;
Gleitz, J.; Barnes, C. PCT Int. Appl., WO 02 57236
(Chem. Abstr. 2002, 137, 125096).
2. Calixto, J. B.; Yunes, R. A.; Neto, A. S. O.; Valle, R. M.
R.; Rae, G. A.; Braz, J. Med. Biol. Res. 1984, 17, 313.
3. Stiz, D. S.; Souza, M. M.; Golin, V.; Neto, R. A. S.;
Correˆa, R.; Nunes, R. J.; Yunes, R. A.; Cechinel-Filho,
V. Pharmazie 2000, 55, 12.
Having obtained this overall encouraging result, we
investigated the conversion of glutaric acid 3a with
other aniline substrates. We screened a variety of nitro,
chloro, cyano and ethylanilines under standard condi-
tions and examined the crude product by HPLC MS.
2-Nitro-, 3-nitro-, 4-ethyl-, 3-chloro- and 4-chloroani-
lines were transformed into the corresponding cyclic
imides in 38–78% yield. The products of 2-ethyl, 3-
ethyl, 2-chloroanilines and the parent aniline contained
a mixture of the intermediate monoamides and the
cyclic imides and the reaction of the aminobenzonitriles
afforded a mixture of undefined products.
4. Kometani, T.; Fitz, T.; Watt, D. S. Tetrahedron Lett.
1986, 27, 919.
5. Wanner, M. J.; Koomen, G.-J. Tetrahedron 1991, 47,
8431.
6. Akula, M. R.; Kabalka, G. W. Synth. Commun. 1998, 28,
2063.
The nitro group of a variety of succinimides or glu-
tarimides 7a–k,m–p in Table 1 was reduced to the
corresponding anilines 8a–o, respectively, under hydro-
genation conditions in good to excellent yield.13 In the
case of dinitro derivative 7i, both groups were con-
verted to the amine to give compound 8i in moderate
yield. In general, this reduction led to the starting
materials (R=H, Fig. 1) for the synthesis of factor Xa
inhibitors of type 1b.
7. Caulfield, W. L.; Gibson, S.; Rae, D. R. J. Chem. Soc.,
Perkin Trans. 1 1996, 545.
8. Al-Khathlan, H.; Zimmer, H. J. Heterocyclic Chem.
1988, 25, 1047.
9. For a review concerning PPA, see: Rowlands, D. A.
Synth. Reagents 1985, 6, 156.
10. Representative experimental procedure: A mixture of 4-
nitro-phenylamine 6a (1.0 g, 7.2 mmol), glutaric acid 3a
(0.951 g, 7.2 mmol) and 86% polyphosphoric acid (10 g)

2136
W. W. K. R. Mederski et al. / Tetrahedron Letters 44 (2003) 2133–2136
was heated at 80°C for 12 h. After cooling, water (50 ml)
MS (EI) m/z 312 (100, M+). Anal. calcd for C15H12N4O4:
C, 57.69; H, 3.87; N, 17.94. Found: C, 57.50; H, 4.00; N,
17.90.
was added, the precipitate was filtered off and dried
which afforded the 1-(4-nitro-phenyl)-piperidine-2,6-
dione 7a (1.63 g, 97%) as a colorless solid: mp 205–
13. Representative experimental procedure: A mixture of 1-
(4-nitro-phenyl)-piperidine-2,6-dione 7a (1.0 g, 4.3
mmol), Raney Ni (0.1 g), and tetrahydrofuran (10 ml)
was stirred at ambient temperature under hydrogen (1
atm) for 2 h. The reaction mixture was filtered through
Celite, and the solvent was removed under vacuum to
furnish 1-(4-amino-phenyl)-piperidine-2,6-dione 8a (0.74
g, 85%) as a colorless solid: mp 214–216°C; 1H NMR
(DMSO-d6) l 6.67 (d, J=8.7 Hz, 2H), 6.53 (d, J=8.7 Hz,
2H), 5.11 (s, 2H), 2.67 (t, J=6.5 Hz, 4H), 2.0–1.86 (m,
2H); MS (EI) m/z 204 (100, M+). Anal. calcd for
C11H12N2O2: C, 64.69; H, 5.92; N, 13.72. Found: C,
64.50; H, 6.00; N, 13.85.
1
207°C; H NMR (DMSO-d6) l 8.30 (d, J=9.0 Hz, 2H),
7.47 (d, J=9.0 Hz, 2H), 2.74 (t, J=6.5 Hz, 4H), 2.09–
1.95 (m, 2H); MS (EI) m/z 234 (63, M+), 206 (100). Anal.
calcd for C11H10N2O4: C, 56.41; H, 4.30; N, 11.96.
Found: C, 56.46; H, 4.36; N, 12.16.
11. Tro¨ger, J. J. Prakt. Chem. 1887, 36, 225.
12. For the synthesis of other analogues of Tro¨ger’s base, see:
Jensen, J.; Wa¨rnmark, K. Synthesis 2001, 1873. Cf. spec-
1
tral data of Tro¨ger’s base analogue 8: mp 258–260°C; H
NMR (DMSO-d6) l 8.01 (dd, J=6.2 and J=2.6 Hz, 2H),
7.96 (d, J=2.6 Hz, 2H), 7.37 (d, J=8.8 Hz, 2H), 4.82 (d,
J=17.2 Hz, 2H), 4.45 (d, J=17.2 Hz, 2H), 4.36 (s, 2H);