K. Tanaka et al. / Tetrahedron Letters 53 (2012) 5899–5902
5901
of the efficient reaction, producing triazabicyclo[3.3.1]nonanes in
modest to excellent yields under mild conditions and through quite
simple operations.
tion, that is, by the aerobic oxidation of benzyl amine, could poten-
1
2
tially be the photo-induced effectors and/or radical mediators.
The detailed mechanism of the formal [4+4] reaction, the acceler-
ation effects of the amines, and the theoretical treatments of the
1
0
The mechanism of the reaction was suggested by the following
two facts: 1) the unsaturated benzylimine itself did not participate
in the reaction without the extra amine, and 2) the mono-1,4-ad-
ducts of an imino nitrogen to another imine, such as 6 (Scheme 3)
were observed as the intermediates, based on the extensive NMR
analysis of the reaction between the aldehyde 1 and the benzyl-
amine (a) (see time-course NMR data in Supplementary data).
Therefore, the extra amine should mediate the conjugate addition
of an imino nitrogen, namely, by enhancing the nucleophilicity of
the nitrogen atom by forming the transient aminoacetal, as pro-
posed in Scheme 3. Thus, the benzylimine could be transformed to
the mono-1,4-adduct 6 by the help of the excess benzylamine. After
producing the iminium intermediate 7 (path A), it then would par-
ticipate in the second intramolecular 1,4-addition. Bis-acetal forma-
tion of the iminium 8 finally led to the most stable isomer 1a with
two equatorial ester substituents on both 6-membered rings in
1
3
reaction will be reported in due course.
It is rather surprising that the facile preparation of the triazabi-
cyclo[3.3.1]nonane derivatives via formal [4+4] reaction described
here, for example, 1a, has been overlooked until now, but we sus-
pect that the synthesis of imines is usually carried out by treat-
ment of the aldehyde with exact equimolar quantities of the
amine, and this point has not previously been investigated to
greater depth. Even if the reactant ratios were adjusted by chance,
1
as occurred accidentally in our research, the simplified H NMR of
the symmetrical 1a with the marked absence of vinyl protons (see
the NMR spectra in the Supplementary data) might be assumed to
be a polymerized byproduct via D–A, ene-reactions, or some other
mechanism, due to the instability of the alkylamine-derived unsat-
urated imines. Further expansion of the scopes, detailed mechanis-
tic investigation, and application to the natural and the bioactive
compounds are currently in progress.
1
a. Intramolecular aminoacetal formation and the subsequent
eight-membered ring closure in 6 (path B), which after all gives
the same iminium intermediate 8, could be the alternative pathway
to 1a. Compound 1a with the rigid triazabicyclo[3.3.1]nonane skel-
eton, once produced, was thermodynamically so stable that any for-
mation of 1g and/or the amine-scrambled products could not be
observed by the treatment with excess allylamine (3 equiv) under
the reaction conditions in Table 1. Although the proton NMR signals
corresponding to the aminoacetals could not be clearly observed in
the reaction mixture (Supplementary data), the acetal formation
was therefore considered to be the rate-limiting step of whole
Acknowledgments
We thank Dr. Kouzou Matsumoto, Department of Chemistry,
Osaka University, for X-crystallographic analysis of the triazabicy-
clo[3.3.1]nonane derivative. This work was supported in part by
Grants-in-Aid for Scientific Research No. 19681024 and 19651095
from the Japan Society for the Promotion of Science, New Energy
and Industrial Technology Development Organization (NEDO, pro-
ject ID: 07A01014a), and a MEXT Grant-in-Aid Project for Scientific
Research on Innovative Areas, Chemical Biology of Natural Prod-
ucts: Target ID and Regulation of Bioactivity (No. 24102518).
[
4+4] process, and hence the triazabicyclo[3.3.1]nonane formation.
It is a very interesting observation that (1) our [4+4] reaction
smoothly proceeds when utilizing the unsaturated imines which
contain the electron rich benzyl or allyl groups on the nitrogen,
and also (2) the reaction is notably slowed by blocking the room
light (room light contains UV light to some extent) or in the pres-
ence of a radical scavenger. Our trials using the chromatographi-
Supplementary data
Supplementary data (experimental details, characterization and
crystallographic data, and copies of MS and NMR spectra) associ-
1
1
cally purified and/or distilled reagents and solvents,
still
afforded the triazabicyclo[3.3.1]nonane 1a in the similar efficiency
to those observed in Table 1. Nevertheless, molecular oxygen as
well as very small amounts of metals or the other byproducts with
References and notes
the extended
p
-systems being generated in the course of the reac-
1
.
.
(a) Tanaka, K.; Fukase, K.; Katsumura, S. Chem. Rec. 2010, 10, 119–139; (b)
Tanaka, K.; Fukase, K.; Katsumura, S. Synlett 2011, 2115–2139.
2
(a) Tanaka, K.; Katsumura, S. Org. Lett. 2000, 2, 373–375; (b) Tanaka, K.; Mori,
H.; Yamamoto, M.; Katsumura, S. J. Org. Chem. 2001, 66, 3099–3110; (c) Tanaka,
K.; Katsumura, S. J. Am. Chem. Soc. 2002, 124, 9660–9661; (d) Tanaka, K.;
Kobayashi, T.; Mori, H.; Katsumura, S. J. Org. Chem. 2004, 69, 5906–5925; (e)
Kobayashi, T.; Nakashima, M.; Hakogi, T.; Tanaka, K.; Katsumura, S. Org. Lett.
2006, 8, 3809–3812; (f) Kobayashi, T.; Hasegawa, F.; Tanaka, K.; Katsumura, S.
Org. Lett. 2006, 8, 3813–3816; (g) Sakaguchi, T.; Kobayashi, T.; Hatano, S.;
Tsuchikawa, H.; Fukase, K.; Tanaka, K.; Katsumura, S. Chem. Asian J. 2009, 4,
Nucleophilic activation?
Bn
path A
CO2Me
Bn
CO2Me
Bn
N
BnHN
N
path B
NH2
BnNH2 (a)
1
6
N
Bn
N
MeO2C
MeO C
2
Bn
1573–1577.
path B
3. (a) Tanaka, K.; Masuyama, T.; Hasegawa, K.; Tahara, T.; Mizuma, H.; Wada, Y.;
Watanabe, Y.; Fukase, K. Angew. Chem., Int. Ed. 2008, 47, 102–105; (b) Tanaka,
K.; Fukase, K. Org. Biomol. Chem. 2008, 6, 815–828; (c) Tanaka, K.; Fukase, K.
Mini-Rev. Org. Chem. 2008, 5, 153–162; (d) Tanaka, K.; Minami, K.; Tahara, T.;
Fujii, Y.; Siwu, E. R. O.; Nozaki, S.; Onoe, H.; Yokoi, S.; Koyama, K.; Watanabe, Y.;
Fukase, K. ChemMedChem 2010, 5, 841–845; (e) Tanaka, K.; Minami, K.; Siwu, E.
R. O.; Nozaki, S.; Watanabe, Y.; Fukase, K. J. Carbohydr. Chem. 2010, 29, 118–
path A
Bn
N
CO2Me
Bn
CO2Me
N
7
Bn
Bn
N
Bn
H2N
N
H
N
Bn
MeO2C
132; (f) Tanaka, K.; Siwu, E. R. O.; Minami, K.; Hasegawa, K.; Nozaki, S.;
MeO2C
Bn
Nucleophilic activation?
Kanayama, Y.; Koyama, K.; Chen, C. W.; Paulson, J. C.; Watanabe, Y.; Fukase, K.
Angew. Chem., Int. Ed. 2010, 49, 8195–8200; (g) Tanaka, K.; Kitadani, M.; Fukase,
K. Org. Biomol. Chem. 2011, 9, 5346–5349.
CO2Me
4.
We initially assigned the products as the 1,2-diazetidines and reported in 90th
Annual Meeting of The Chemical Society of Japan, March, 2010, Tokyo, 40th
Congress of heterocyclic Chemistry, Japan, October, 2010, Sendai, International
Chemical Congress of Pacific Basin Societies (PACIFICHEM 2010), December,
2010, Hawaii, and 99th Symposium on Organic Synthesis, Japan, June, 2011,
Tokyo. The structures for these specific compounds should be corrected as
shown in this Communication.
Bn
N
N
1
a
Bn
Bn
Bn
N
N
MeO2C
8
N
H
CO2Me
N
Bn
MeO2C
Most stable structure
Scheme 3. Proposed pathway to triazabicyclo[3.3.1]nonane 1a through amine-
mediated formal [4+4] reaction of benzylimine.
5. (a) Katritzky, A. R.; Baker, V. J.; Ferguson, I. J.; Patel, R. C. J. Chem. Soc. Perkin
Trans. 2 1979, 143–150; (b) Shim, S. C.; Kim, D.-W.; Moon, S.-S.; Chae, Y. B. J.