cyclobutane aminals (Scheme 1(2)). Owing to the absence of
hydrogen on the spirocyclic aminal carbon of these inter-
mediates, elimination of HBr should not be possible and, in-
stead, a rearrangement reaction involving ring strain induced
cyclobutane ring expansion by C-to-N migration might
occur to yield bicyclic amidines.6ꢀ8 A further driving force
for the 1,2-carbon migration would be provided by the neigh-
boring nitrogen, which stabilizes positive charge develop-
ment at carbon center from which 1,2-migration takes place
in a manner that is similar to the driving force for the Schmidt
reaction.
ethylenediamine (2a). Reaction of these substances in di-
chloromethane for 1 d followed by solvent removal leads
to the isolation of spirocyclic aminal 3 in nearly quanti-
tative yield. Treatment of the crude product mixture
containing 3 with NBS promotes formation of bicyclic
amidine 4a in a 93% yield (Scheme 2). When NCS is
employed in place of NBS as the halogen source, the
reaction proceeds to give a comparable yield of 4a, while
NIS promotes a lower yielding reaction. The rearrange-
ment reaction of 3 occurs spontaneously even at room
temperature and in the absence of additives, such as silver
salts.3
Scheme 1. Reactions of Aminals
Scheme 2. Test Reaction
Following this study, which demonstrates the feasibility
of the new bicyclic amidine forming rearrangement reac-
tion of N-halo intermediates generated from spirocyclic
aminals, we carried out experiments designed to explore the
substrate scope of the process. A one-pot procedure was
employed for these processes, involving initial condensation
of the cyclobutanone with the selected diamine followed by
oxidative rearrangement promoted by treatment with NBS.
As the results displayed in Table 1 demonstrate, reaction
takes place to produce a bicyclic amidine in excellent yield
when 1,2-diphenyl ethylenediamine (2a) and 2,2-dimethyl-
1,3-diaminopropane (2b) are used for spirocyclic aminal
formation. The presence of substituents at C-3 of the cyclo-
butane substrate was found to have no effect on the effi-
ciency of this process (Table 1, entries 2, 4, 6, and 7). In addi-
tion, reaction of the aminal derived from diamine 2a and
3-phenylcyclobutanone (1b) generates amidine 4b as a 3:1
mixture of diasteromers. Sterically bulky 3,3-disubstituted
cyclobutane derivatives also undergo this reaction to give the
corresponding bicyclic amidines in high yields (Table 1,
entries 3 and 5). Furthermore, substrates possessing BocN,
BnO, and TBDPSO groups also react under the NBS pro-
moted reaction conditions to produce products in equally
high yields (Table 1, entries 5ꢀ7). However, the aminal de-
rived from condensation of cyclopentanone with diamine 2a
does not undergo the rearrangement reaction under the con-
ditions. This finding suggests that ring strain present in the
cyclobutane ring serves as an important driving force for the
rearrangement process.
Bicyclic amidines are an important class of organic bases,
as is exemplified by the utilzation of 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU) and 1,5-diazabicyclo[4.3.0]non-
5-ene (DBN) in promoting a variety of synthetically useful
organic transformations.9 Existing methods for the prepara-
tion of bicyclic amidines, involving cyclization reactions of
lactam derivatives, most often require several steps and harsh
reaction conditions.10 Consequently, a new method for the
efficient preparation of these substances is an important goal.
Initial studies to explore this proposal were conducted
using the reaction of cyclobutanone (1a) with 1,2-diphenyl
(6) For a review on the reactions of cyclobutanes, see: Seiser, T.;
Saget, T.; Tran, D. N.; Cramer, N. Angew. Chem., Int. Ed. 2011, 50,
7740.
(7) For our studies on the reaction of cyclobutanols with hypervalent
iodine reagents, see: (a) Fujioka, H.; Komatsu, H.; Miyoshi, A.; Murai,
K.; Kita, Y. Tetrahedron Lett. 2011, 52, 973. (b) Fujioka, H.; Komatsu,
H.; Nakamura, T.; Miyoshi, A.; Hata, K.; Ganesh, J.; Murai, K.; Kita,
Y. Chem. Commun. 2010, 46, 4133.
(8) Although cyclobutane ring is strained, the rearrangement reac-
tion of cyclobutyl amine was reported to need treatment of silver salt.
See ref 3.
(9) For examples, see: (a) Ghosh, N. Synlett 2004, 574. (b) Taylor,
J. E.; Jones, M. D.; Williams, J. M. J.; Bull, S. D. Org. Lett. 2010, 12,
5740. (c) Miura, M.; Toriyama, M.; Kawakubo, T.; Yasukawa, K.;
Takido, T.; Motohashi, S. Org. Lett. 2010, 12, 3882. (d) Price, K. E.;
ꢀ
Larrivee-Aboussafy, C.; Lillie, B. M.; McLaughlin, R. W.; Mustakis, J.;
Hettenbach, K. W.; Hawkins, J. M.; Vaidyanathan, R. Org. Lett. 2009,
11, 2003. (e) Baidya, M.; Mayr, H. Chem. Commun. 2008, 1792. (f)
Birman, V. B.; Li, X.; Han, Z. Org. Lett. 2006, 9, 37. (g) Shieh, W.-C.;
ꢁ
Dell, S.; Repic, O. J. Org. Chem. 2002, 67, 2188. (h) K. Aggarwal, V.;
Mereu, A. Chem. Commun. 1999, 2311.
(10) For synthesis of DBU and DBN type bicyclic amidines, see: (a)
Kumagai, N.; Matsunaga, S.; Shibasaki, M. Angew. Chem., Int. Ed.
2004, 43, 478. (b) Ostendorf, M.; Dijkink, J.; Rutjes, F. P. J. T.;
Hiemstra, H. Eur. J. Org. Chem. 2000, 115. (c) Kotsuki, H.; Sugino,
A.; Sakai, H.; Yasuoka, H. Heterocycles 2000, 53, 2561. (d) Dijkink, J.;
Eriksen, K.; Goubitz, K.; van Zanden, M. N. A.; Hiemstra, H. Tetra-
hedron: Asymmetry 1996, 7, 515. (e) Convery, M. A.; Davis, A. P.; Dunne,
C. J.; MacKinnon, J. W. Tetrahedron Lett. 1995, 36, 4279 and ref 9f.
The reaction of aminal 5a, derived by condensation of
o-aminobenzylamine (2c) was next investigated. The reaction
of the aminal 5a is of interest because two products can
possibly be produced by a difference in the reaction course.
Specifically, 1,2,3,9-tetrahydropyrrolo[2,1-b]quinazoline (6a)
Org. Lett., Vol. 14, No. 3, 2012
773