Mild and selective reduction of imines: Formation of an unsymmetrical macrocycle

Article abstract of DOI:10.1021/jo049197u

During investigations of 5, a [3 + 3] Schiff-base macrocycle with six imines, a partially reduced Schiff-base macrocycle, 6, possessing one CH ₂NH and five imine groups was obtained. Control experiments and deuterium labeling indicate that the macrocycle is reduced by a benzimidazoline generated during the reaction. Benzimidazolines may be convenient reagents for the mild and selective reduction of imines.

Full text of DOI:10.1021/jo049197u

Mild and Selective Reduction of Imines: Formation of an
Unsymmetrical Macrocycle
Amanda J. Gallant, Brian O. Patrick, and Mark J. MacLachlan*
Department of Chemistry, University of British Columbia, 2036 Main Mall,
Vancouver, BC, Canada V6T 1Z1
mmaclach@chem.ubc.ca
Received May 13, 2004
During investigations of 5, a [3 + 3] Schiff-base macrocycle with six imines, a partially reduced
Schiff-base macrocycle, 6, possessing one CH2NH and five imine groups was obtained. Control
experiments and deuterium labeling indicate that the macrocycle is reduced by a benzimidazoline
generated during the reaction. Benzimidazolines may be convenient reagents for the mild and
selective reduction of imines.
Conjugated macrocycles are a rapidly growing area of
interest because they may form the basis of catalytic,1
magnetic,2 liquid crystalline,3 pharmaceutical,4 and su-
pramolecular materials.5 Schiff-base condensation is a
convenient route for the construction of conjugated mac-
rocycles such as expanded porphyrins (e.g., Texaphyrin)
and Robson-type macrocycles (e.g., 1).6,7 Indeed, many
examples of macrocycles that form by [2 + 2] and [3 + 3]
Schiff-base condensation reactions are now known.8 We
have recently reported on new giant metallomacrocycles
that form by [3 + 3] condensation reactions.9
Macrocycle 1 is an attractive ligand to form new
magnetic and catalytic materials as it may bind to two
metals.10 These metal complexes can be prepared directly
by condensation of a 2,6-diformylphenol with a diamine
in the presence of a transition metal template or by
reaction of the preformed macrocycle with a metal salt.2,11
Unfortunately, the synthesis of these fully conjugated
organic macrocycles is usually complicated by in situ
reduction.12-14 For example, the reaction to form 1 yields
(1) (a) Vigato, P. A.; Tamburini, S.; Fenton, D. E. Coord. Chem. Rev.
1990, 106, 25-170. (b) Schmidt, J. A. R.; Mahadevan, V.; Getzler, Y.
D. Y. L.; Coates, G. W. Org. Lett. 2004, 6, 373-376. (c) Nestler, O.;
Severin, K. Org. Lett. 2001, 3, 3907-3909. (d) Nelson, A. P.; DiMagno,
S. G. J. Am. Chem. Soc. 2000, 122, 8569-8570.
(2) Dziembowska, T.; Guskos, N.; Typek, J.; Szymczak, R.; Likodi-
mos, V.; Glenis, S.; Lin, C. L.; Wabia, M.; Jagodzinska, E.; Fabrycy, E.
Mater. Res. Bull. 1999, 34, 943-954.
(3) Sessler, J. L.; Callaway, W.; Dudek, S. P.; Date, R. W.; Lynch,
V.; Bruce, D. W. Chem. Commun. 2003, 2422-2423.
(4) (a) Pandey, S. K.; Gryshuk, A. L.; Graham, A.; Ohkubo, K.;
Fukuzumi, S.; Dobhal, M. P.; Zheng, G.; Ou, Z.; Zhan, R.; Kadish, K.
M.; Oseroff, A.; Ramaprasad, S.; Pandey, R. K. Tetrahedron 2003, 59,
10059-10073. (b) Dairou, J.; Vever-Bizet, C.; Brault, D. Antiviral Res.
2004, 61, 37-47. (c) Lottner, C.; Knuechel, R.; Bernhardt, G.; Brunner,
H. Cancer Lett. 2004, 203, 171-180.
(5) (a) Campbell, K.; Kuehl, C. J.; Ferguson, M. J.; Stang, P. J.;
Tykwinski, R. R. J. Am. Chem. Soc. 2002, 124, 7266-7267. (b) Zhao,
D.; Moore, J. S. Chem. Commun. 2003, 807-818. (c) Ho¨ger, S.; Bonrad,
K.; Mourran, A.; Beginn, U.; Mo¨ller, M. J. Am. Chem. Soc. 2001, 123,
5651-5659.
(9) Ma, C.; Lo, A.; Abdolmaleki, A.; MacLachlan, M. J. Org. Lett.
2004, 6, 3841-3844.
(10) (a) Okawa, H.; Furutachi, H.; Fenton, D. E. Coord. Chem. Rev.
1998, 174, 51-75. (b) Bosnich, B. Inorg. Chem. 1999, 38, 2554-2562.
(c) Brooker, S. Coord. Chem. Rev. 2001, 222, 33-56. (d) Gavrilova, A.
L.; Bosnich, B. Chem. Rev. 2004, 104, 349-383.
(11) For examples of metal complexes of Robson-type macrocycles,
see: (a) Huang, W.; Gou, S.; Hu, D.; Chantrapromma, S.; Fun, H.-K.;
Meng, Q. Inorg. Chem. 2002, 41, 864-868. (b) Adams, H.; Bailey, N.
A.; Bertrand, P.; Rodriguez de Barbarin, C. O.; Fenton, D. E.; Gou, S.
J. Chem. Soc., Dalton Trans. 1995, 275-279. (c) Lacroix, P.; Kahn,
O.; Theobald, F.; Leroy, J.; Wakselman, C. Inorg. Chim. Acta 1988,
142, 129-134. (d) Diril, H.; Chang, H. R.; Zhang, X.; Larsen, S. K.;
Potenza, J. A.; Pierpont, C. G.; Schugar, H. J.; Isied, S. S.; Hendrickson,
D. N. J. Am. Chem. Soc. 1987, 109, 6207-6208. (e) Atkins, A. J.; Black,
D.; Finn, R. L.; Marin-Becerra, A.; Blake, A. J.; Ruiz-Ramirez, L.; Li,
W.-S.; Schroeder, M. Dalton Trans. 2003, 1730-1737. (f) Thompson,
L. K.; Mandal, S. K.; Tandon, S. S.; Bridson, J. N.; Park, M. K. Inorg.
Chem. 1996, 35, 3117-3125. (g) Handa, M.; Farida, A.; Thompson, L.
K.; Hayashibara, C.; Sugimori, T.; Hiromitsu, I.; Kasuga, K. Mol. Cryst.
Liq. Cryst. Sci. Technol., Sect. A 2000, 342, 75-80. (h) Brooker, S.;
Croucher, P. D.; Davidson, T. C.; Smith, P. D. Polyhedron 2000, 19,
1887-1894.
(6) Sessler, J. L.; Seidel, D. Angew. Chem., Int. Ed. 2003, 42, 5134-
5175.
(7) Pilkington, N. H.; Robson, R. Aust. J. Chem. 1970, 23, 2225-
2236.
(8) For examples of macrocycles formed by Schiff-base condensation,
see: (a) Gawron´ski, J.; Kołbon, H.; Kwit, M.; Katrusiak, A. J. Org.
Chem. 2000, 65, 5768-5773. (b) Kuhnert, N.; Rossignolo, G. M.; Lopez-
Periago, A. Org. Biomol. Chem. 2003, 1, 1157-1170. (c) Jazwinski, J.;
Lehn, J.-M.; Me´ric, R.; Vigneron, J.-P.; Cesario, M.; Guilhem, J.;
Pascard, C. Tetrahedron Lett. 1987, 28, 3489-3492. (d) Shimakoshi,
H.; Kai, T.; Aritome, I.; Hisaeda, Y. Tetrahedron Lett. 2002, 43, 8261-
8264. (e) Menif, R.; Martell, A. E. J. Chem. Soc., Chem. Commun. 1989,
1521-1523. (f) Kwit, M.; Gawron´ski, J. Tetrahedron: Asymmetry 2003,
14, 1303-1308. (g) Gao, J.; Martell, A. E. Org. Biomol. Chem. 2003, 1,
2795-2800. (h) Zhao, D.; Moore, J. S. J. Org. Chem. 2002, 67, 3548-
3554. (i) Depree, C. V.; Beckmann, U.; Heslop, K.; Brooker, S. Dalton
Trans. 2003, 3071-3081. (j) Chadim, M.; Budeˇsˇinsky´, M.; Hodacˇova´,
J.; Za´vada, J.; Junk, P. C. Tetrahedron: Asymmetry 2001, 12, 127-
133. (k) Sessler, J. L.; An, D.; Cho, W.-S.; Lynch, V. J. Am. Chem. Soc.
2003, 125, 13646-13647.
(12) (a) Ustynyuk, Y. A.; Borisova, N. E.; Nosova, V. M.; Reshetova,
M. D.; Talismanov, S. S.; Nefedov, S. E.; Aleksandrov, G. A.; Eremenko,
I. L.; Moiseev, I. I. Russ. Chem. Bull., Int. Ed. 2002, 51, 488-498. (b)
Borisova, N. E.; Reshetova, M. D.; Ustynyuk, Y. A. Russ. Chem. Bull.,
Int. Ed. 2004, 53, 181-188.
(13) (a) Aguiari, A.; Bullita, E.; Casellato, U.; Guerriero, P.; Tam-
burini, S.; Vigato, P. A. Inorg. Chim. Acta 1992, 202, 157-171. (b)
Brychcy, K.; Dra¨ger, K.; Jens, K.-J.; Tilset, M.; Behrens, U. Chem. Ber.
1994, 127, 1817-1826. (c) Kumar, D. S.; Alexander, V. Inorg. Chim.
Acta 1995, 238, 63-71. (d) Bowden, F. L.; Ferguson, D. J. Chem. Soc.,
Dalton Trans. 1974, 460-462.
(14) (a) Atwood, D. A.; Jegier, J. A.; Lindholm, N. F.; Martin, K. J.;
Rutherford, D. J. Coord. Chem. 1996, 38, 305-318. (b) Singer, A. L.;
Atwood, D. A. Inorg. Chim. Acta 1998, 277, 157-162. (c) Chiu, S.-M.;
Wong, T.-W.; Man, W.-L.; Wong, W.-T.; Peng, S.-M.; Lau, T.-C. J. Am.
Chem. Soc. 2001, 123, 12720-12721.
10.1021/jo049197u CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/13/2004
J. Org. Chem. 2004, 69, 8739-8744
8739

Gallant et al.
SCHEME 1. Synthesis of Robson-Type Macrocycle 1
SCHEME 2. Synthesis of Macrocycles 5 and 6
mostly the partially reduced macrocycle 2, Scheme 1. A
few research groups have studied these reduced products,
but the reducing agent in these reactions is still
disputed.12-15
protons (6.2-7.2 ppm). This is in contrast to macrocycle
5a, which shows only one imine resonance (8.54 ppm),
one phenol OH (13.23 ppm), and two aromatic protons
as a result of its average D3h symmetry. The 1H and 13
C
We are interested in the supramolecular chemistry of
macrocycles, such as 5a, which aggregate into tubular
assemblies in solution when combined with small cat-
ions.16,17 These red, crystalline compounds are conve-
niently prepared in high yield by reacting 3,6-diformyl-
catechol 3 with substituted o-phenylenediamine 4 in
CHCl3/MeCN at 80 °C, Scheme 2.17
NMR spectra show new peaks at 4.4 and 48.5 ppm,
respectively, consistent with a CH2NH group. The IR
spectrum of this byproduct was very similar to that of
5a, showing an intense CdN stretch (1609 cm-1) and no
carbonyl group. In addition, a new peak at 3605 cm-1
was observed and is characteristic of the N-H stretch
for a secondary amine. ESI-MS revealed the mass of this
byproduct to be greater than that of macrocycle 5a by
exactly 2 amu, Figure S1. Single-crystal X-ray diffraction
confirmed that the byproduct was indeed a macrocycle
and not an oligomeric compound, Figure 1. Together, the
structural and spectroscopic data indicate the byproduct
to be monoreduced macrocycle 6a.
During our investigations of macrocycle 5a, we isolated
a deep red crystalline byproduct (ca. 10-20% by mass)
by fractional crystallization of the crude reaction mixture.
1
The H NMR spectrum of this compound showed five
imine (8-8.5 ppm) and six hydrogen-bonded phenol (11-
15 ppm) resonances, as well as numerous aromatic
The crystal structure of 6a confirms that this byprod-
uct is cyclic and shows that the compound is nonplanar
with a single molecule of water hydrogen-bonded to the
center of the macrocycle. A plane of symmetry runs
through the molecule, reflecting one-half of the macro-
cycle onto the other.18 The six hydroxyl groups of the
(15) Tian, Y.; Tong, J.; Frenzen, G.; Sun, J.-Y. J. Org. Chem. 1999,
64, 1442-1446.
(16) (a) Huck, W. T. S.; van Veggel, F. C. J. M.; Reinhoudt, D. N.
Recl. Trav. Chim. Pays-Bas 1995, 114, 273-276. (b) Akine, S.;
Taniguchi, T.; Nabeshima, T. Tetrahedron Lett. 2001, 42, 8861-8864.
(17) Gallant, A. J.; MacLachlan, M. J. Angew. Chem., Int. Ed. 2003,
42, 5307-5310.
8740 J. Org. Chem., Vol. 69, No. 25, 2004

Selective Reduction of Macrocycles
FIGURE 1. Molecular structure (from single-crystal X-ray diffraction) of monoreduced macrocycle 6a with H2O coordinated in
the center (thermal ellipsoids shown at 33% probability).
macrocycle are tilted out of planarity with two of the diol
moieties oriented upward and the third downward rela-
tive to the center of the macrocycle. Within the unit cell,
the macrocycles are organized into planes, but the pores
of the macrocycles are not aligned into channels.
It is not surprising that this macrocycle is nonplanar.
Even the fully conjugated macrocycle (with no or different
alkoxy groups) is nonplanar as deduced by single-crystal
X-ray diffraction.16b,19 These macrocycles likely twist to
relieve interatomic interactions within the macrocycle,
but there may also be an electronic or solvent effect (i.e.,
due to the water molecule hydrogen-bonded in the
center).
postulated that either starting material 3 or 4 could serve
as reducing agent, forming a quinone or diimine, respec-
tively. Others have postulated that the amine may be
responsible for in situ reduction in the synthesis of
related macrocycle 1, forming direduced macrocycle 2.15
The combination of equimolar concentrations of com-
pound 3 and macrocycle 5a in acid-free CHCl3 (dried with
K2CO3) produced no reaction. However, using commercial
CHCl3, which contains residual acid, results varied but
1
usually a mixture of products was obtained by H NMR
spectroscopy, including monoreduced macrocycle 6a and
the 1:1 condensation product 7a (Scheme 3). The reaction
of diamine 4a with macrocycle 5a showed only a 1:2
condensation product 8a (and unreacted 5a) with no
evidence (1H NMR) of monoreduced macrocycle 6a in both
commercial and acid-free CHCl3 (even after reflux for 24
h), as shown in Scheme 3. Compounds 7 and 8, which
are possible intermediates in the macrocycle formation
and may also form via hydrolytic fission of the macro-
cycle, could be independently prepared by the controlled
reaction of 3 and 4 in the appropriate stoichiometry.21
These results suggest that neither 3 nor 4 is indepen-
dently responsible for the reduction of 5.
Mechanistic Studies
As imines are usually reduced using considerable
pressures of hydrogen or in the presence of a metal
catalyst,20 it was surprising to isolate the unsymmetrical
monoreduced macrocycle 6a without the addition of any
formal reducing agent or catalyst. We therefore conducted
experiments to identify the in situ reducing agent and
to investigate the reduction process itself. Initially, we
(18) It was impossible to identify exactly which imine was reduced
because of disorder. The macrocycle appears to pack equally well with
the imine or amine in any position, as there is a crystallographic mirror
plane through the middle of the macrocycle (i.e., it appears more
symmetrical as a result of disorder).
(21) The 1:1 condensation product 7 containing both an amine and
an aldehyde is quite reactive, readily undergoing condensation. At-
tempts to purify this compound with short alkoxy chains failed; 7 with
short alkoxy chains (4, 5, and 6 carbon atoms) could be observed by
1H NMR spectroscopy but could only be isolated with long alkoxy
chains (∼12 or 14 carbon atoms). As well, it always contained an
impurity of ∼10% of compound 8 (confirmed by 1H NMR spectroscopy
and MS).
(19) Gallant, A. J.; Hui, J.-K.; Zahariev, F. E.; Wang, A.;
MacLachlan, M. J. Manuscript in preparation.
(20) Adams, J. P. J. Chem. Soc., Perkin Trans. 1 2000, 125-139.
J. Org. Chem, Vol. 69, No. 25, 2004 8741

Gallant et al.
SCHEME 3. Reactions of 3 and 4a with Macrocycle 5a
Reaction of equimolar quantities of diol 3 and diamine
4a in commercial CHCl3 at 50 °C afforded mostly
monoreduced macrocycle 6a. In the presence of added
macrocycle 5a, the major product was still monoreduced
species 6a. As these reactions may involve formation of
the 1:1 adduct, we combined compound 722 with 5a at
50 °C, and after 7 h the 1H NMR spectrum showed that
5a was consumed (>80%) with monoreduced macrocycle
as the major product. When the same three experiments
were performed using acid-free CHCl3, no monoreduced
product was observed. These results indicate that re-
sidual acid must be present in the reaction solvent for
macrocycle reduction. Moreover, an intermediate con-
densation product (e.g., 7) needs to be present to permit
the reduction.
The formation of 6 proceeds slowly (days) in com-
mercial CHCl3 from diol 3 and diamine 4 but noticeably
faster (hours) when acid is added. As macrocycle 5 forms
quickly in solution, the acid may catalyze the hydrolysis
of 5 to yield significant amounts of 7 that would not
otherwise be present to allow for reduction. Unsym-
metrical macrocycle 6a could be prepared in 53% yield
by the addition of less than 5% p-toluenesulfonic acid to
3 and 4a in CHCl3, keeping the reaction mixture at
approximately 50 °C. Monoreduced macrocycles with
different alkoxy substituents (e.g., 6b,c) could also be
prepared by the same procedure, demonstrating the
generality of this reaction. Alternatively, monoreduced
macrocycle 6a could be obtained by the addition of less
than 5% p-toluenesulfonic acid to macrocycle 5a in
CHCl3. In this case, however, the reaction leads to a
mixture of products and a lower yield of 6a.
We propose that a benzimidazole is also responsible
for the reduction of macrocycle 5 but that the preformed
macrocycle is being reduced directly. That is, macrocycle
6 is not assembled from reduced components. Benzimi-
dazoline 9 formed from the 1:1 condensation product 7
is responsible for the in situ reduction of macrocycle 5 to
afford compound 6.
Benzimidazolines, generated in situ by the reaction of
a diamine with an aldehyde, are known to be selective,
mild reagents for reducing CdC bonds.23 In our reaction,
benzimidazoline 9 is likely generated from 7 and reacts
with macrocycle 5 to afford 6, yielding benzimidazole 10
as a byproduct, Scheme 4.
A deuterium labeling experiment was conducted to
provide insight into the reaction mechanism. Diol 3-d2
with deuterium-labeled formyl groups and macrocycle
5a-d6 with deuterium-labeled imines were synthesized.
When diol 3-d2 and phenylenediamine 4a were combined
with a catalytic amount of acid and macrocycle 5a-d6 in
Ustynyuk et al. have postulated that a benzimidazoline
may be involved in the formation of 2.12 They did not
believe that macrocycle 1 was being reduced but rather
that 2 formed through recombination of reduced inter-
mediates. It was reported that in the first step, the
diformylphenol and o-phenylenediamine condense to
form a 1:1 condensation product. This species then
undergoes disproportionation in the second step to yield
a benzimidazole and a reduced 1:1 condensation product.
In the final step, two reduced 1:1 condensation products
combine to form macrocycle 2. Although this does ratio-
nalize the formation of 2, it is not clear why a monore-
duced macrocycle is not also obtained as a product.
(22) Because of the difficulty of purifying 7 with short alkoxy chains,
7 with tetradecyloxy chains was used for these experiments.
8742 J. Org. Chem., Vol. 69, No. 25, 2004

Selective Reduction of Macrocycles
SCHEME 4. Proposed Mechanism for Formation of 6 by in Situ Reduction of Macrocycle 5
CDCl3, compound 6a-d7 was observed to be the major
Supporting the proposed mechanism, benzimidazole 10
product within a few hours. The absence of a peak at 4.4
has been isolated from the p-toluenesulfonic acid cata-
lyzed reaction of 3 with 4a. Its identity was verified by
ESI-MS and NMR experiments. Presumably, this fluo-
rescent product is formed as a byproduct in the reduction
of macrocycle 5a to afford 6a. Efforts to isolate benzimi-
dazoline 9 or benzimidazole 11 from the reaction mix-
tures have been unsuccessful. It has also not been
possible to prepare 10 or 11 separately by oxidation
(including the use of catalytic FeCl3).25
1
ppm in the H NMR spectrum, where the characteristic
CH2NH resonance is observed in 6a-c, indicated that
only deuterium atoms were on the methylene carbon of
the reduced imine in 6a-d7. This suggests that the
reduction is selective and does not involve HD formation.
We also conducted an experiment in which the OH and
NH2 groups in 3 and 4a, respectively, were deuterated
(by exchange with D2O). When treated with catalytic acid
to generate 6a, the product showed no significant deu-
terium incorporation into the methylene adjacent to the
amine as determined by 1H NMR spectroscopy. This
further confirms that regioselective HD transfer occurs
during the reaction.
Although the isolation of benzimidazole 10 lends
support to 9 being the reducing agent, it is possible that
other condensation products are also responsible for the
reduction. For example, a bis(benzimidazoline) formed
from 8 may be involved. There may also be oligomeric
species that are involved but could not be isolated. Their
presence in reactions to form 6 is difficult to discern
because of the complexity of the 1H NMR spectrum of 6,
but ESI-MS analysis of a reaction mixture showed major
peaks assigned to both benzimidazoles 10 and 11, provid-
ing evidence for their involvement in the reaction. The
same species (but with butyl substituents in the place of
hexyl groups) were observed in the ESI-MS spectrum of
the crude reaction mixture to form 6c. The peak assigned
to 11 in the MS is also observed as a fragment of both
macrocycles 5a and 6a, but compound 10 is not.26 The
ESI-MS experiments indicated that 10 was present in
Analogous to the reduction of CdC bonds,23b reduction
may take place by either a stepwise mechanism with
hydride (H-) transfer from C-H of 9 followed by proton
(H+) transfer from NH of 9 or by the concerted addition
of H- and H+ from benzimidazoline 9. The acid catalyst
may protonate the imine of 5, thereby promoting H-
abstraction by the carbon adjacent to the iminium cation
of 5.24
(23) (a) Chikashita, H.; Nishida, S.; Miyazaki, M.; Itoh, K. Synth.
Commun. 1983, 13, 1033-1039. (b) Chikashita, H.; Nishida, S.;
Miyazaki, M.; Morita, Y.; Itoh, K. Bull. Chem. Soc. Jpn. 1987, 60, 737-
746. (c) Risitano, F.; Grassi, G.; Foti, F.; Bilardo, C. Heterocycles 2001,
55, 1311-1314. (d) Chikashita, H.; Morita, Y.; Itoh, K. Synth. Commun.
1985, 15, 527-533. (e) Itoh, K.; Ishida, H.; Chikashita, H. Chem. Lett.
1982, 1117-1118.
(25) (a) Coville, N. J.; Neuse, E. W. J. Org. Chem. 1977, 42, 3485-
3491. (b) Mackman, R. L.; Hui, H. C.; Breitenbucher, J. G.; Katz, B.
A.; Luong, C.; Martelli, A.; McGee, D.; Radika, K.; Sendzik, M.; Spencer,
J. R.; Sprengeler, P. A.; Tario, J.; Verner, E.; Wang, J. Bioorg. Med.
Chem. Lett. 2002, 12, 2019-2022.
(24) Lewis acids such as [CpFe(CO)2(THF)]+ coordinate to imines
to generate electrophilic iminium ions that react with nucleophiles.
For example, see: Mayer, M. F.; Hossain, M. M. J. Org. Chem. 1998,
63, 6839-6844.
J. Org. Chem, Vol. 69, No. 25, 2004 8743

Gallant et al.
the crude reaction mixture and did not form during the
isolation procedure.
direduced macrocycle 2, with no evidence of mono-, tri-,
or fully reduced macrocycles (or other direduced isomers).
This selectivity in the reduction of macrocyclic imines
suggests that benzimidazolines may be mild and selective
reducing agents for other imines.
The driving force for the reduction likely arises from
the formation of a stable, aromatic benzimidazole and
has been suggested in the formation of 2. Macrocycle 5
with 48 π electrons is not aromatic, so the reduction does
not break aromaticity. The reduction may also afford a
relief of strain from interatomic repulsion or torsions
within the macrocycle to lead to a stable monoreduced
product. It is very likely that hydrogen bonding to solvent
or water trapped within the macrocycle plays a role in
stabilizing the monoreduced macrocycle relative to the
fully conjugated precursor.
Benzimidazole 10 is presumably formed from benz-
imidazoline 9, which is in equilibrium with 7, during the
in situ reduction of macrocycle 5. Recent calculations
indicate that the reaction to form the benzimidazoline
from the 1:1 condensation compound (i.e., 7) should be
endothermic.12b High-temperature NMR experiments in
C2D2Cl4 and C2D2Cl4/EtOH showed no evidence for a
benzimidazoline. Similar experiments with the 1:2 con-
densation compound 8 failed to show any evidence for
equilibrium with a bis(benzimidazoline). Either the equi-
librium concentration of benzimidazoline is too low to
detect by 1H NMR spectroscopy or the equilibrium is acid-
catalyzed (acid was not added to the reaction as this was
shown previously to afford macrocycle 6a).
Remarkably, the reaction is selective and stops after
monoreduction of the macrocycle. No polyreduced mac-
rocycles were observed (1H NMR) even after several days
or under a variety of conditions. ESI-MS of the reaction
mixture after nearly complete monoreduction (by 1H
NMR spectroscopy) showed only macrocycles 5a and 6a
(plus benzimidazoles 10 and 11), with no further reduced
products present (Figure S2). We anticipated that the
imines would be reduced nonselectively, especially those
that are separated across the macrocycle, but this is not
the case. After monoreduction, the macrocycle may be
stabilized by additional hydrogen bonding from the
amine, or there may be stabilization from hydrogen-
bonding solvent molecules (e.g., H2O) in the interior of
the macrocycle. Interestingly, our attempts (and those
of others) to synthesize macrocycle 1 usually afford the
Conclusions
We have synthesized and characterized the unsym-
metrical macrocycle 6. On the basis of isolation of a
benzimidazole byproduct and deuterium labeling experi-
ments, we have shown that 6 is obtained by the reduction
of 5 with a benzimidazoline intermediate generated in
situ from diol 3 and diamine 4. This discovery may shed
light on the reduction of other Schiff-base macrocycles
(such as 1) and may lead to improved synthetic routes
to these compounds. More generally, in situ generation
of benzimidazolines may offer a convenient, purely
organic reducing agent for mild and selective reduction
of imines.
Acknowledgment. We thank Prof. Marco Ciufolini
and Dr. Yun Ling for helpful discussions. We are
grateful to the University of British Columbia and the
Natural Sciences and Engineering Research Council of
Canada (NSERC) for funding and for graduate scholar-
ships to A.J.G.
Supporting Information Available: Crystallographic
data for 6a (in CIF format), experimental procedures, 13C NMR
data for 6a, ESI-MS data (S2) for a reaction to form 6a, and
1H NMR spectra for new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
(26) We have used tandem mass spectrometry to verify the frag-
mentation pattern of macrocycles 5a and 6a by MS. By analyzing the
mass spectra of macrocycles 5 with various chain lengths, the peaks
are easily assigned. One fragment of [5a + H]+ and [6a + H]+ is a
species with the same chemical composition as bis(benzimidazole) 11.
Compound 10 is never observed as a fragment of [5a + X]+ or [6a +
X]+ (X ) H, Na, K, Rb, Cs).
JO049197U
8744 J. Org. Chem., Vol. 69, No. 25, 2004