N-salicylideneanilines: Tautomers for formation of hydrogen-bonded capsules, clefts, and chains

Article abstract of DOI:10.1021/jo052277t

The synthesis, characterization, and solid-state structures of new salicylaldimines are reported. Bis(N-salicylideneaniline)s (BSANs) and tris(N-salicylideneaniline)s (TSANs) are sterically encumbered compounds featuring a central six-membered ring in the

Full text of DOI:10.1021/jo052277t

N-Salicylideneanilines: Tautomers for Formation of
Hydrogen-Bonded Capsules, Clefts, and Chains
Marc Sauer, Charles Yeung, Jonathan H. Chong, Brian O. Patrick, and
Mark J. MacLachlan*
Department of Chemistry, UniVersity of British Columbia, 2036 Main Mall,
VancouVer, BC, V6T 1Z1 Canada
mmaclach@chem.ubc.ca
ReceiVed NoVember 2, 2005
The synthesis, characterization, and solid-state structures of new salicylaldimines are reported. Bis(N-
salicylideneaniline)s (BSANs) and tris(N-salicylideneaniline)s (TSANs) are sterically encumbered
compounds featuring a central six-membered ring in the keto-enamine tautomer. When extended with
additional functional groups, these molecules may form hydrogen-bonded capsules, clefts, and extended
structures. A TSAN with N-tBOC-o-phenylenediamine groups has been structurally investigated. The
complementary hydrogen-bonding motif in this molecule leads it to form dimers in solution and in the
solid state. A BSAN with N-tBOC-o-phenylenediamine substituents forms a hydrogen-bonded cleft in
solution but forms an extended hydrogen-bonded ladder assembly of cofacial dimers in the solid state.
When N-tBOC-1,8-naphthalenediamine was utilized to extend the cleft, an unusual perimidine structure
was obtained with the central core in the enol tautomer. In addition, ab initio calculations have been
used to support the assignment of the keto-enamine or enol-imine tautomers of the BSANs and TSANs
and to predict tautomerization in related BSANs and TSANs.
Introduction
of chemistry and biochemistry; it is implicated in pharmaceutical
effects, enzyme activity, stabilization of DNA, and self-
assembly.3-5
Tautomers are readily interconverted constitutional isomers,
usually distinguished by a rearrangement of a labile hydrogen
atom and a double bond (e.g., keto-enol tautomerization).1 The
equilibrium between tautomers is often rapid under normal
conditions, being catalyzed by traces of acid or base present in
most samples and solvents.2 Often one isomer is strongly favored
(acetone, for example, is 99.999% keto tautomer). Proton
tautomerization is one of the most thoroughly studied equilib-
rium processes because of its crucial importance to several fields
Over the past decade a great effort has been devoted to
explore the central role hydrogen-bonding plays in the self-
(3) (a) Claramunt, R. M.; Lopez, C.; Akorta, I.; Elguero, J.; Yang, R.;
Schulma, S. Magn. Reson. Chem. 2004, 42, 712. (b) Rodr´ıguez-Santiago,
L.; Sodupe, M.; Oliva, A.; Bertran, J. J. Am. Chem. Soc. 1999, 121, 8882.
(c) Glaser, R.; Lewis, M. Org. Lett. 1999, 1, 273. (d) Phillips, R. S.;
Sundararaju, B.; Faleev, N. G. J. Am. Chem. Soc. 2000, 122, 1008. (e)
Heyes, D. J.; Hunter, C. N.; van Stokkum, I. H. M.; van Grondelle, R.;
Groot, M. L. Nat. Struct. Biol. 2003, 10, 491. (f) Thompson, M. J.; Bashford,
D.; Noodleman, L.; Getzoff, E. D. J. Am. Chem. Soc. 2003, 125, 8186. (g)
Naumov, P.; Sekine, A.; Uekusa, H.; Ohashi, Y. J. Am. Chem. Soc. 2002,
124, 8540. (h) Chou, P.-T.; Chen, Y.-C.; Wei, C.-Y.; Chen, W. S. J. Am.
Chem. Soc. 2000, 122, 9322.
(1) Toullec, J. AdV. Phys. Org. Chem. 1982, 18, 1.
(2) (a) Pollack, S. K.; Hehre, W. J. J. Am. Chem. Soc. 1977, 99, 4845.
(b) Zimmermann, H. E. Acc. Chem. Res. 1987, 20, 263. (c) Toullec, J.;
Dubois, J. E. J. Am. Chem. Soc. 1974, 96, 3524.
10.1021/jo052277t CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/20/2005
J. Org. Chem. 2006, 71, 775-788
775

Sauer et al.
SCHEME 1. Condensation of a Salicylaldehyde with a
assembly of molecules into supramolecular aggregates. Several
elegant routes toward hydrogen-bonded solution aggregates,
engineered crystals, liquid crystals, and two- and three-
dimensional extended solids have been reported.6-9 The major
key in this respect is the design of modules that are capable of
forming multiple hydrogen bonds and allow control over the
assembly of the desired secondary structure.
Primary Amine
Molecular containers are of particular interest because of their
potential application in various areas ranging from drug delivery
and catalysis to molecular recognition and host-guest interac-
tions.10 Functionalized calixarenes represent a well-established
motif for the formation of hydrogen-bonded molecular contain-
ers, and their properties and applications have been described
in a large number of publications.11
N-Salicylideneanilines represent another interesting class of
compounds possessing intramolecular hydrogen-bonding. These
molecules are prepared by the condensation of salicylaldehyde
with primary amines, as shown in Scheme 1. The interest in
these substances arises from the fact that they can undergo
reversible proton transfer (tautomerism) leading to interesting
properties such as thermo- and photochromism, making these
molecules potential candidates for optical switches and storage
devices.12-14 They show equilibrium between the enol-imine
(OH) form and the keto-enamine (NH) form, with the enol-
imine form generally being the most stable form at room
temperature (Scheme 1).15 As a result, the OH form of
N-salicylideneanilines has been well-characterized, but until
recently, no purely NH form had been structurally character-
ized.16 In fact, the confirmation of a N-salicylideneaniline
exclusively in the NH form in the solid state has been
controversial.17-19
(4) (a) Abou-Zied, O. K.; Jimenez, R.; Romesberg, F. E. J. Am. Chem.
Soc. 2001, 123, 4613. (b) Duarte, H. A.; Carvalho, S.; Paniago, E. B.; Simas,
A. M. J. Pharm. Sci. 1999, 88, 111. (c) Susan-Resiga, D.; Nowak, T. J.
Biol. Chem. 2003, 278, 12660. (d) Forlani, L.; Mezzina, E.; Boga C.;
Forconi, M. Eur. J. Org. Chem. 2001, 2779.
The assembly of sterically crowded aromatics is an attractive
route to discotic columnar materials, foldamers, and donor-σ-
(5) (a) Lehn, J.-M. Supramolecular Chemistry: Concepts and Perspec-
tiVes; VCH: Weinheim, 1995. (b) Steed, J. W.; Atwood, J. L. Supra-
molecular Chemistry; Wiley: New York, 2000. (c) Jeffrey, G. A. An
Introduction to Hydrogen Bonding; Oxford University Press: New York,
1997. (d) Lawrence, D. S.; Jiang, T.; Levett, M. Chem. ReV. 1995, 95, 2229.
(e) Sijbesma, R. P.; Meijer, E. W. Curr. Opin. Colloid Interface Sci. 1999,
4, 24. (f) Whitesides, G. M.; Grzybowski, B. Science 2002, 295, 2418. (g)
Raymo, F. M.; Stoddart, J. F. Curr. Opin. Colloid Interface Sci. 1996, 1,
116.
(6) (a) Lehn, J.-M. Angew. Chem., Int. Ed. Engl. 1990, 29, 1304. (b)
Whitesides, G. M.; Simanek, E. E.; Mathias, J. P.; Seto, C. T.; Chin, D.
N.; Mammen, M.; Gordon, D. M. Acc. Chem. Res. 1995, 28, 37. (c)
Lawrence, D. S.; Jiang, T.; Levett, M. Chem. ReV. 1995, 95, 2229. (d) Conn,
M. M.; Rebek, J., Jr. Chem. ReV. 1997, 97, 1647. (e) Zimmerman, S. C.;
Corbin, P. S. Struct. Bonding 2000, 96, 63. (f) Archer, E. A.; Gong, H.;
Krische, M. J. Tetrahedron 2001, 57, 1139. (g) Sherrington, D. C.; Buggy,
M. Chem. Soc. ReV. 2001, 30, 83. (h) Sijbesma, R. P.; Meijer, E. W. Chem.
Commun. 2003, 5.
(7) (a) Prins, L. J.; Reinhoudt, D. N.; Timmermann, P. Angew. Chem.,
Int. Ed. 2001, 40, 2382. (b) Steiner, T. Angew. Chem., Int. Ed. 2002, 41,
48. (c) Malek, N.; Maris, T.; Simard, M.; Wuest, J. D. J. Am. Chem. Soc.
2005, 127, 5910. (d) Laliberte´, D.; Maris, T.; Wuest, J. D. J. Org. Chem.
2004, 69, 1776. (e) Wang, X.; Simard, M.; Wuest, J. D. J. Am. Chem. Soc.
1994, 116, 12119. (f) Uemura, K.; Kitagawa, S.; Fukui, K.; Saito, K. J.
Am. Chem. Soc. 2004, 126, 3817. (g) Sada, K.; Inoue, K.; Tanaka, T.;
Tanaka, A.; Epergyes, A.; Nagahama, S.; Matsumoto, A.; Miyata, M. J.
Am. Chem. Soc. 2004, 126, 1764.
(8) (a) Kato, T.; Mizoshita, N.; Kanie, K. Macromol. Rapid Commun.
2001, 28, 1127. (b) Kato, T. Handbook of Liquid Crystals; Demus, D.,
Goodby, J., Gray, G. W., Spiess, H. W., Vill, V., Eds.; Wiley-VCH:
Weinheim, 1998. (c) Felekis, T.; Tziveleka, L.; Tsiourvas, D.; Paleos, C.
M. Macromolecules 2005, 38, 1705. (d) Beginn, U. Prog. Polym. Sci. 2003,
28, 1049. (e) Kato, T. Science 2002, 295, 2414. (f) Davis, A. P.; Draper, S.
M.; Dunne, G.; Ashton, P. Chem. Commun. 1999, 2265. (g) McCubbin, J.
A.; Tong, X.; Zhao, Y.; Snieckus, V.; Lemieux, R. P. Chem. Mater. 2005,
17, 2574. (h) Jin, S.; Ma, Y.; Zimmerman, S. C.; Cheng, S. Z. D. Chem.
Mater. 2004, 16, 2975.
(10) (a) Conn, M. M.; Rebek, J., Jr. Chem. ReV. 1997, 97, 1647. (b)
Rebek, J., Jr. Chem. Soc. ReV. 1996, 255. (c) Palmer, L. C.; Shivanyuk, A.;
Yamanaka, M.; Rebek, J., Jr. Chem. Commun. 2005, 857. (d) MacGillivray,
L. R.; Atwood, J. L. Angew. Chem., Int. Ed. 1999, 38, 1018. (e) Middel,
O.; Verboom, W.; Reinhoudt, D. N. J. Org. Chem. 2001, 66, 3998. (f) Philip,
I. E.; Kaifer, A. E. J. Am. Chem. Soc. 2002, 124, 12678. (g) Cave, G. W.
V.; Hardie, M. J.; Roberts, B. A.; Raston, C. L. Eur. J. Org. Chem. 2001,
3227. (h) Shivanyuk, A.; Paulus, E. F.; Bo¨hmer, V.; Angew. Chem., Int.
Ed. 1999, 38, 2906. (i) Reek, J. N. H.; Schenning, A. P. H. J.; Bosman, A.
W.; Meijer, E. W.; Crossley, M. J. Chem. Commun. 1998, 11. (j) Umemoto,
K.; Tsukui, H.; Kusukawa, T.; Biradha, K.; Fujita, M. Angew. Chem., Int.
Ed. 2001, 40, 2620.
(11) (a) Kang, J.; Rebek, J., Jr. Nature 1997, 385, 50. (b) Heinz, T.;
Rudkevich, D. M.; Rebek, J., Jr. Nature 1998, 394, 764. (c) Castellano, R.
K.; Clark, R.; Craig, S. L.; Nuckolls, C.; Rebek, J., Jr. Proc. Natl. Acad.
Sci. U.S.A. 2000, 97, 12418. (d) Johnson, D. W.; Hof, F.; Iovine, P. M.;
Nuckolls, C.; Rebek, J., Jr. Angew. Chem. 2002, 114, 3947. (e) Corbellini,
F.; di Costanzo, L.; Crego-Calama, M.; Geremia, S.; Reinhoudt, D. N. J.
Am. Chem. Soc. 2003, 125, 9946. (f) Corbellini, F.; van Leeuwen, F. W.
B.; Beijleveld, H.; Kooijman, H.; Spek, A. L.; Verboom, W.; Crego-Calama,
M.; Reinhoudt, D. N. New J. Chem. 2005, 29, 243. (g) Zadmard, R.;
Schrader, T.; Grawe, T.; Kraft, A. Org. Lett. 2002, 4, 1687. (h) Choi, H.-
J.; Park, Y. S.; Cho, C. S.; Koh, K.; Kim, S.-H.; Paek, K. Org. Lett. 2004,
6, 4431. (i) Kobayashi, K.; Shirasaka, T.; Yamaguchi, K.; Sakamoto, S.;
Horn, E.; Furukawa, N. Chem. Commun. 2000, 41.
(12) (a) Inabe, T. New J. Chem. 1991, 15, 129. (b) Hadjoudis, E. Mol.
Eng. 1995, 5, 301. (c) Harada, J.; Uekusa, H.; Ohashi, Y. J. Am. Chem.
Soc. 1999, 121, 5809.
(13) (a) Ogawa, K.; Fujiwara, T. Chem. Lett. 1999, 657. (b) Pizzala, H.;
Carles, M.; Stone, W. E. E.; Thevand, A. J. Chem. Soc., Perkin Trans. 2
2000, 935. (c) Alarco´n, S. H.; Pagani, D.; Bacigalupo, J.; Olivieri, A. C. J.
Mol. Struct. 2003, 647. (d) Ogawa, K.; Harada, J.; Fujiwara, T.; Yoshida,
S. J. Phys. Chem. A 2001, 105, 3425. (e) Vargas, V.; Amigo, L. J. Chem.
Soc., Perkin Trans. 2 2001, 1124. (f) Ogawa, K.; Harada, J. J. Mol. Struct.
2003, 647, 211. (g) Amimoto, K.; Kanatomi, H.; Nagakari, A.; Fukuda,
H.; Koyama, H.; Kawato, T. Chem. Commun. 2003, 870.
(9) (a) Sijbesma, R. P.; Beijer, F. H.; Brunsveld, L.; Folmer, B. J. B.;
Hirschberg, J. H. K. K.; Lange, R. F. M.; Lowe, J. K. L.; Meijer, E. W.
Science 1997, 278, 1601. (b) Zhao, X., Wang, X.-Z.; Jiang, X.-K.; Chen,
Y.-Q.; Li, Z.-T.; Chen, G. J. J. Am. Chem. Soc. 2003, 125, 15128. (c) Corbin,
P. S.; Zimmerman, S. C. J. Am. Chem. Soc. 1998, 120, 9710. (d) Ma, Y.;
Kolotuchin, S. V.; Zimmerman, S. C. J. Am. Chem. Soc. 2002, 124, 13757.
(e) Park, T.; Zimmerman, S. C.; Nakashima, S. J. Am. Chem. Soc. 2005,
127, 6520. (f) Quinn, J. R.; Zimmerman, S. C. Org. Lett. 2004, 6, 1649.
(g) Choi, I. S.; Li, X.; Simanek, E. E.; Akaba, R.; Whitesides, G. M. Chem.
Mater. 1999, 11, 684. (h) Mathias, J. P.; Seto, C. T.; Simanek, E. E.;
Whitesides, G. M. J. Am. Chem. Soc. 1994, 116, 1725. (i) Lin, C. H.; Tour,
J. J. Org. Chem. 2002, 67, 7761. (j) Guo, D.; Sijbesma, R. P.; Zuilhof, H.
Org. Lett. 2004, 6, 3667. (k) Moralez, J. G.; Raez, J.; Yamakazi, T.; Motkuri,
K.; Kovalenko, A.; Fenniri, H. J. Am. Chem. Soc. 2005, 127, 8307.
(14) (a) Feringa, B. L.; Jager, W. F.; de Lange, B. Tetrahedron 1993,
49, 8267. (b) Delaire, J. A.; Nakatani, K. Chem. ReV. 2000, 100, 1817. (c)
Lo, D. S. Appl. Optics 1974, 13, 861. (d) Nakatani, K.; Delaire, J. A. Chem.
Mater. 1997, 9, 2682.
(15) (a) Suzuki, T.; Arai, T. Chem. Lett. 2001, 105, 3425. (b) Charette,
J. J.; de Hoffmann, E. J. Org. Chem. 1979, 44, 2256. (c) Hansen, P. E.;
Duus, F.; Schmitt, P. Org. Magn. Reson. 1982, 18, 58.
(16) Chong, J. H.; Sauer, M.; Patrick, B. O.; MacLachlan, M. J. Org.
Lett. 2003, 5, 3823.
(17) Ogawa, K.; Harada, J.; Tamura, I.; Noda, Y. Chem. Lett. 2000, 528.
(18) Ogawa, K.; Kasahara, Y.; Ohtani, Y.; Harada, J. J. Am. Chem. Soc.
1998, 120, 7107.
(19) Krygowski, T. M.; Woz´niak, K.; Anulewicz, R.; Pawlak, D.;
Kolodziejski, W.; Grech, E.; Szady, A. J. Phys. Chem. A 1997, 101, 9399.
776 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
SCHEME 2. Synthesis of Keto-Enamine Tris(N-salicylideneaniline)s
acceptor chromophores. In most cases hexasubstituted aromatics,
such as 1,3,5-benzenetriamides or 1,3,5-benzenetriols, are used
as suitable cores, which upon further functionalization self-
assemble into well-defined secondary conformations.20
We recently communicated a convenient synthesis of tri-
formylphloroglucinol 1. Upon reaction with aniline derivatives,
this sterically encumbered molecule does not form tris(imine)s
but rather affords tris(N-salicylideneaniline)s (TSANs) (e.g., 3)
that are exclusively found in the keto-enamine (NH) form. They
are formed as a mixture of two geometrical isomers with
approximate Cs and C3h symmetry (Scheme 2). In the C3h
isomer, all of the carbonyl groups on the central ring are
involved in intramolecular hydrogen-bonding.16 Yelamaggad et
al. recently demonstrated that when the aromatic substituents
are highly branched, TSANs form discotic liquid crystalline
phases.21 These molecules possess multiple hydrogen-bonding
groups around a rigid central core, suggesting their use in
developing hydrogen-bonded capsules and assemblies.
Here, we report our studies on TSANs and bis(N-salicylide-
neaniline)s, BSANs, as well as their assembly into hydrogen-
bonded structures. Their complementary hydrogen-bonding
motifs lead them to form capsule-like dimers and clefts in
solution. We have performed the first calculations of tautomers
of TSANs and BSANs to determine their relative stabilities and
to compare their calculated geometries with those determined
from single-crystal X-ray diffraction (SCXRD). These molecules
are new, modular motifs for the formation of supramolecular
structures.
in a single step from phloroglucinol under Duff formylation23
conditions.16
The reaction of 1 with 2 in refluxing ethanol afforded the
TSAN 3 in 52% yield. The IR spectrum of this yellow-orange,
fluorescent compound confirmed the disappearance of an
aldehyde, an intense absorption band observed at 1641 cm-1 in
1. As we initially expected the enol-imine form for 3 to be the
1
only product formed, the H NMR spectrum of 3 (Figure 1)
FIGURE 1. 1H NMR spectra of compounds 3 and 5 (CDCl3, 300
MHz).
was surprisingly complicated. The enol-imine form for 3 would
be expected to show singlets for the imine and phenol in the
OH form, but instead the spectrum showed multiple peaks
between δ 8.5-9.0 and δ 12.8-13.5 ppm. 2-D NMR experi-
ments showed that only the keto-enamine (NH) form of 3 is
present, in a mixture of two geometrical isomers with C3h and
Cs symmetries, as shown in Scheme 2. The carbonyl group of
the central ring was observed in the 13C NMR spectrum at ∼185
ppm, characteristic of a carbonyl group rather than a phenol
(this resonance is observed at 174 ppm in 1). Interestingly,
J-coupling (3JHH ∼13.5 Hz) was observed between the vinylic
protons and the NH protons of 3. This strong coupling is
consistent with localization of the proton on the nitrogen atom
and slow exchange. Indeed, the 100% keto-enamine form for
Results and Discussion
TSANs. To initiate this work, 1,3,5-triformylphloroglucinol
1 was required. Although this molecule was previously prepared
in seven steps from 1,3,5-trichlorobenzene,22 with an overall
yield of 21%, the product was not characterized. We discovered
that compound 1 could be readily prepared in nearly pure form
(20) (a) Corbin, P. S.; Zimmerman, S. C.; Thiessen, P. A.; Hawryluk,
N. A.; Murray, T. J. J. Am. Chem. Soc. 2001, 123, 10475. (b) Li, H.; Homan,
E. A.; Lampkins, A. J.; Ghiviriga, I.; Castellano, R. K. Org. Lett. 2005, 7,
443. (c) Bushey, M. L.; Nguyen, T.-Q.; Zhang, W.; Horoszewski, D.;
Nuckolls, C. Angew. Chem., Int. Ed. 2004, 43, 5446. (d) Brunsveld, L.;
Zhang, H.; Glasbeek, M.; Vekemans, J. A. J. M.; Meijer, E. W. J. Am.
Chem. Soc. 2000, 122, 6175. (e) Bushey, M. L.; Hwang, A.; Stephens, P.
W.; Nuckolls, C. J. Am. Chem. Soc. 2001, 123, 8157. (f) Miao, Q.;
Lefenfeld, M.; Nguyen, T.-Q.; Siegrist, T.; Kloc, C.; Nuckolls, C. AdV.
Mater. 2005, 17, 407.
(21) Yelamaggad, C. V.; Achalkumar, A. S.; Shankar Rao, D. S.; Krishna
Prasad, S. J. Am. Chem. Soc. 2004, 126, 6506.
(22) Himmelman, W.; Roos, E.; Sobel, J. German Patent 2002063, 1971;
Chem. Abstr. 1971, 30, 3227.
3
an N-salicylaldimine is expected to show a JCH-NH coupling
of this magnitude.24
The two geometrical isomers of 3 are present in a ratio of
ca. 1:1.4 (C3h:Cs) in CDCl3 and 1:1.5 in DMSO-d6. Statistically,
a random orientation of the three enamines would generate a
(23) Anderson, A. A.; Goetzen, T.; Shackelford, S. A.; Tsank, S. Synth.
Commun. 2000, 30, 3227.
(24) (a) Salman, S. R.; Saleh, N. A. I. Can. J. Anal. Sci. Spectrosc. 1997,
42, 9. (b) Salman, S. R.; Lindon, J. C.; Farrant, R. D.; Carpenter, T. A.
Magn. Reson. Chem. 1993, 31, 991.
J. Org. Chem, Vol. 71, No. 2, 2006 777

Sauer et al.
SCHEME 3. Synthesis of TSAN 5
ratio of 1:3 for these isomers. This indicates that the molecule
favors the more symmetrical isomer that leads to hydrogen-
bonding to each ketone. Also, the difference in the two solvents
is minor. The two geometrical isomers do not interconvert on
the NMR time scale even when heated to 120 °C.
t
We reacted 1 with the BOC-protected o-phenylenediamine
4 (Scheme 3) to design TSANs equipped with an additional
functional group available for hydrogen-bonding. The 1H NMR
spectrum of 5 shows only broad singlets for the enamine and
vinylic groups, in contrast to the spectra of other derivatives
(e.g., 3) that show multiple peaks corresponding to the two
geometrical isomers at 12.5-13.5 and 8.0-9.0 ppm (Figure 1).
The electrospray ionization mass spectrum (ESI-MS) of
compound 5 (Supporting Information) shows the presence of
not only the [5 + H]+ ion but also significant quantities of the
[52 + H]+ species, suggesting the formation of a dimer in
solution. To confirm the dimeric structure of 5, crystals suitable
for SCXRD were obtained from slow evaporation of a solution
of 5 in 1:1 dichloromethane/hexanes. The molecular structure
of 5 is shown in Figure 2 clearly indicating the formation of a
capsule-type dimer.
FIGURE 2. Structure of 5 as determined by SCXRD. The view of
the dimer is shown from the top in (a), and the tBOC protecting groups
have been removed from the side view in (b) for clarity.
of 90 ( 50 M-1 25
This order of magnitude is realistic for a
.
multiply hydrogen-bonded dimeric aggregate.9f,26 As the Cs
isomer is never observed in CDCl3, we can only provide a lower
bounds on Kassoc in CDCl3. Assuming that the isomerization
between the two isomers is facile in CDCl3, that any trace of
monomeric 5 is present in the same ratio in CDCl3 as in DMSO-
d6 (i.e., 1.7:1 for C3h:Cs isomers), and that there is less than 2%
of the Cs isomer present at 5 mM in CDCl3, the equilibrium
In the structure, two molecules of 5 form a self-complemen-
tary hydrogen-bonded dimer. Six hydrogen bonds between the
NHtBoc groups and the carbonyl groups of the central ring hold
the dimer together. The presence of just one isomer in solution
can be explained by the fact that only isomers with C3h
symmetry are able to form the thermodynamically favored
constant for association (Kassoc) must be >3 × 104 M-1
.
Variable-temperature 1H NMR experiments corroborated the
existence of a dimer in solution (see Supporting Information
1
for spectra). The H NMR spectrum of 5 is broad at room
temperature. When the sample was heated to >60 °C (in C2D2-
Cl4), the peaks became sharp and a mixture of the two isomers
(C3h and Cs) was apparent. When the same sample was cooled
(in CD2Cl2), the peaks became sharp at -20 °C and the major
species present (>95%) was the C3h isomer as expected for the
dimer. The Cs isomer was not present, but a new product, which
may be a larger aggregate, was also visible at low temperatures.
These results indicate that below room temperature, only the
C3h isomer is present, whereas at high temperature, the dimer
dissociates to give a mixture of C3h and Cs isomers. This
dynamic monomer-dimer equilibrium is responsible for the
broad peaks observed at room temperature.
1
dimeric aggregates. The H NMR spectrum is consistent with
the high (C3) symmetry afforded by a dimer rather than a larger
aggregate.
In CDCl3, there is no evidence of monomeric compound 5;
only a single geometric isomer is present and the NH peaks are
1
broad in the H NMR spectrum. In DMSO-d6, the molecule
has fully dissociated and only sharp peaks are present, showing
both the C3h and Cs isomers, with clear J-coupling between the
vinylic CH and the NH groups. Interestingly, the ratio of C3h:
Cs isomers in DMSO-d6 is 1.7:1, showing that the C3h isomer
is strongly biased over the Cs isomer compared with statistical
expectation. The C3h geometry minimizes steric interactions
between the bulky tBOC-o-phenylenediamine substituents present
in compound 5.
In addition to the intermolecular hydrogen bonds, further
stabilization of the dimer may originate from cofacial interac-
tions of the central cyclohexanetrione cores. These rings are
When 5 in CDCl3 (ca. 6.6 mM) was titrated with DMSO-d6,
dissociation of the dimer was only evident when the solvent
contained ca. 15% (v/v) DMSO-d6 or greater. Using a solvent
mixture of 1:5 DMSO-d6/CDCl3 and using the NH peaks to
distinguish the Cs isomer (monomeric) from the C3h isomer
(monomeric and dimeric), we estimated an association constant
(25) To obtain this number, samples were prepared in 1:5 (v/v) DMSO-
d6/CDCl3 at various concentrations. 1H NMR spectra were obtained and
the regions of interest (NH protons, 11.5-13.0 ppm) were integrated to
determine the ratio of species present. This calculation included several
assumptions, which are discussed in Supporting Information.
(26) Murray, T. J.; Zimmerman, S. C. Tetrahedron Lett. 1995, 36, 7627.
778 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
SCHEME 4. Synthesis of Compound 10a
SCHEME 5. Synthesis of Compounds 11 and 12
The reaction of 10 with aniline derivatives (e.g., 2, Scheme
5) afforded products that are predominantly in the keto-enamine
tautomer and were also obtained as a mixture of two geometrical
1
isomers (Figure 3). The H NMR spectrum of 11 shows the
presence of two geometrical isomers (isomers a and b), with
multiple peaks between δ 6.5-9.0 and 12.5-14.5 ppm. H-
1
a Reagents and conditions: (a) nBuLi/C2H4Br2, 65%; (b) NBS/MeCN,
n
86%; (c) BuLi/DMF/H+, 50%; (d) BBr3/CH2Cl2, 81%.
1H COSY NMR spectra elucidated coupling between doublets
in these two regions that could be assigned to the different
isomers. As with the TSANs, there is significant coupling (3JHH
≈12-13 Hz) between the vinylic and amine protons, indicating
that the protons are localized on the N atoms (keto-enamine
form) and exchange is slow on the NMR time scale. The 13C
NMR spectrum of 11 showed the carbonyl resonances at ca.
187 ppm, characteristic of the keto isomer.
coplanar and separated by only 3.35 Å in the solid state, close
enough for significant π-π interaction. The top ring is rotated
23.6° relative to the bottom ring as illustrated by the parameter
θ in Figure 2a. The peripheral aromatic phenylenediamine rings
may also participate in favorable π-π interactions. Although
this interaction is only weakly observed in the solid-state
structure (closest interatomic separation of 3.6 Å), it may be
more prominent in the solution state and enhance the overall
stability of the dimer.
BSANs. By reducing the number of functional groups around
the central ring, we felt that it might be possible to access
molecular clefts.5a,b,27 Compound 10 featuring two hydroxyl and
two formyl groups is a good candidate for building clefts.
Although there are reports of compound 10,28 there is only one
previous report of an aniline derivative of 10, but the authors
claimed that the bis(imine) form was present.29 Moreover, Schiff
base condensations of amines with 10 have been claimed to
afford imines in several patents without supporting evidence.30
DFT calculations of the product formed by reacting aniline with
10 indicated that it should also tautomerize to the keto-enamine
(NH) form. We established an improved synthetic procedure
to 10 in four steps, starting from 1,3-dimethoxybenzene (Scheme
4). Key to the procedure is the selective bromination of 7 and
the double metal-halogen exchange of 8. Compound 10 was
isolated in multigram quantities in 23% overall yield.
The two geometrical isomers of 11 are present in a ca. 3:1
ratio (isomer a:isomer b) in CDCl3. Assuming the two enamines
in 11 could be in any geometrical isomer that ensured both
amines were hydrogen-bonded to ketones, these isomers should
be present in a 1:1 ratio. The isomer that involves hydrogen-
bonding to both ketones (isomer a) is strongly favored in
solution. In DMSO-d6, the ratio of these two isomers is ca. 2.3:
1, suggesting less bias toward isomer a in this solvent. The
presence of water in the NMR solvent may also affect the
equilibrium by hydrogen-bonding to 11.
The reaction of 10 with monoprotected N-tBOC-o-phenylene-
diamine 4 afforded a product 12 with a significantly different
1H NMR spectrum, indicating the presence of a single geo-
metrical isomer in solution (Figure 4). Crystals of 12 suitable
for X-ray diffraction were obtained from CHCl3/MeCN. Al-
though a weak diffraction pattern was obtained, the structure
was solved to give reliable information about the extended
structure and some information about the bonding in the
structure. The molecular structure shown in Figure 5a (one of
the two molecules in the asymmetric unit) confirms that 12 is
locked exclusively in the keto-enamine (NH) form in the solid
state. The central ring is nearly flat, but highly distorted from
a typical benzene ring. One bond length (HCdCH; C45 to C46)
is 1.326(15) Å and typical of a vinylic group, whereas the other
four bond lengths range from 1.403(15) to 1.455(15) Å, more
typical of single bonds. Furthermore, the CdO length in the
central ring of this molecule is 1.271(12)-1.284(12) Å, which
is consistent with a ketone rather than a phenol. An elongated
C-N bond length relative to an imine further confirms the keto-
enamine form for the molecule in the solid state. A second
molecule of 12 in the asymmetric unit is similarly locked in
the keto-enamine form, though the molecule is slightly more
distorted about one amine. The molecular structure of 12 is
consistent with the calculated structure for the keto-enamine
form, which will be discussed later.
(27) (a) Rebek, J., Jr. Angew. Chem., Int. Ed. Engl. 1990, 29, 245. (b)
Klarner, F.-G.; Benkhoff, J.; Boese, R.; Burkert, U.; Kamieth, M.; Naatz,
U. Angew. Chem., Int. Ed. Engl. 1996, 35, 1130. (c) Bell, T. W.; Cragg, P.
J.; Firestone, A.; Kwok, A. D.-I.; Liu, J.; Ludwig, R.; Sodoma, A. J. Org.
Chem. 1998, 63, 2232. (d) Warrener, R. N.; Margetic, D.; Amarasekara,
A. S.; Butler, D. N.; Mahadevan, I. B.; Russell, R. A. Org. Lett. 1999, 1,
199. (e) MacGillivray, L. R.; Siebke, M. M.; Reid, J. L. Org. Lett. 2001, 3,
1257. (f) Goshe, A. J.; Steele, I. M.; Bosnich, B. J. Am. Chem. Soc. 2002,
125, 444. (g) Field, J. D.; Turner, P.; Harding, M. M.; Hatzikominos, T.;
Kim, L. New J. Chem. 2002, 26, 720. (h) Shao, X.-B.; Jiang, X.-K.; Zhao,
X.; Zhao, C.-X.; Chen, Y.; Li, Z.-T. J. Org. Chem. 2004, 69, 899. (i)
Stonicˇius, S.; Butkus, E.; Zˇilinkas, A.; Larsson, K.; O¨ hrstro¨m, L.; Berg,
U.; Wa¨rnmark, K. J. Org. Chem. 2004, 69, 5196. (j) Barboiu, M.; Petit, E.;
Vaughan, G. Chem. Eur. J. 2004, 10, 2263.
(28) (a) Pattekhan, H. H.; Divakar, S. J. Mol. Catal. A: Chem. 2001,
169, 185. (b) Geismann, T. A.; Schlatter, M. J.; Webb, I. D.; Roberts, J. D.
J. Org. Chem. 1946, 11, 741.
(29) Schiele, C.; Guenther, A. Z. Naturforsch. 1968, 23, 388.
(30) (a) Nishikawa, A. Japanese Patent JP 01085215 A21980330 Heisei,
1989. (b) Nishikawa, A.; Masaji, O.; Tokuyuki, K. Japanese Patent JP
62115754 A2 19870527 Showa, 1987.
The presence of a single geometrical isomer in BSAN 12
(1H NMR) is reminiscent of TSAN 5 and suggested the presence
J. Org. Chem, Vol. 71, No. 2, 2006 779

Sauer et al.
FIGURE 3. 1H-1H COSY NMR spectrum of 11 showing coupling between vinylic and NH protons. At the right are the two structures evident
1
in the H NMR spectrum.
from those of the monomer. In addition, when the molecule
dissociates, the equilibrium is affected by isomerization between
the two isomers (a and b). If we assume that the isomerization
between the two isomers is facile in CDCl3, that any trace of
monomeric 12 is present in the same ratio in CDCl3 as in
DMSO-d6 (i.e., 8.8:1 for isomers a:b), and that there is less than
2% of isomer b present at 560 µM in CDCl3, the equilibrium
constant for association (Kassoc) must be >104 M-1. We are
unable to provide an upper bound on the stability constant of
the dimer, but it could be much larger.
Although the molecule assembles into hydrogen-bonded
dimers in solution, the solid-state structure shows that the
molecules are organized into dimers that form an extended
FIGURE 4. 1H NMR spectra of compounds 11 and 12 (CDCl3, 300
MHz). The peaks at 12-15 ppm are the NH resonances.
hydrogen-bonded structure, Figure 6. Each dimer is formed by
a pair of nearly cofacial cyclohexenediketone rings of compound
12, rotated about 120° relative to one another. The rings are
3.6(1) Å apart, indicating a weak π-π interaction between the
rings. Two hydrogen bonds between the NHtBOC groups and
the carbonyl groups of the central ring hold the dimers together.
These are half of the hydrogen bonds that are likely present in
the dimer in solution, as modeled in Figure 5b,c. Moreover,
there are cofacial π-π interactions present between two of the
phenylenediamine substituents of the compound, with nearest
interatomic distances of only ca. 3.4 Å.
Hydrogen bonds formed between the NHtBOC groups not
involved in stabilizing the cofacial dimer, and ketones bridge
adjacent dimers in the solid state. This extended structure is
depicted in Figure 6a. The solubility of the crystals in CDCl3
or DMSO-d6 indicate that the extended structure is easily
disrupted.
of a dimer in solution. The simplicity of the NMR spectrum
also corroborates a dimer rather than a more complex, less
symmetrical aggregate. ESI-MS data provided additional evi-
dence for the presence of a dimer in solution, showing both
[12 + Na]+ and [122 + Na]+ species in the mass spectrum
(Supporting Information). When dissolved in DMSO-d6, the 1H
NMR spectrum indicates that 12 reverts to a mixture of isomers
a and b in a ratio of 8.8:1. With the bulky N-tBOC-o-
phenylenediamine substituents, the less sterically encumbered
geometry of isomer a is strongly preferred over isomer b. We
believe that in CDCl3, this molecule likely forms hydrogen-
bonded dimers similar to those observed in 5, but with only
four hydrogen bonds holding the cleft together. A molecular
model of the proposed dimer is shown in Figure 5b,c. In this
structure, there is a cleft evident between the two central rings.
The stability of the cleft was probed by dilution experiments
using NMR spectroscopy. Samples of 12 in CDCl3 ranging from
560 µM to 19 mM showed no change and no evidence of
dissociation. UV-vis spectra between 10-4 and 10-6 M showed
no shifts in the absorption spectrum. The cleft in CDCl3 (ca.
1.9 mM) was titrated with DMSO-d6. Dissociation of the dimer
was evident only at concentrations of >10% v/v DMSO-d6. We
wished to obtain association constants for the dimerization of
12 in solution, but peaks of the dimer could not be distinguished
Compound 10 is only one of five possible dihydroxydi-
formylbenzene isomers where the hydroxy/formyl groups are
arranged in pairs around a single benzene ring, as in salicyl-
aldehyde. To obtain information on how the position of the
hydroxyl/formyl groups within the central ring affects the
formation of the all keto-enamine (NH) form and possible
dimerization, we reacted compounds 13, 15, and 17 with 4
1
(Scheme 6). H NMR spectra clearly indicated that all three
products 14, 16, and 18 are present exclusively in the enol-
imine (OH) form. This is not at all surprising for 16 and 18,
780 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
FIGURE 5. (a) Molecular structure of one molecule of 12 in the solid state (thermal ellipsoids at 50% probability). (b, c) Molecular model of a
dimer of 12 as viewed from the top (b) and side (c) of the cleft showing hydrogen bonds (green). Hydrogen atoms have been omitted from the
model in (b) for clarity.
FIGURE 6. (a) SCXRD structure of ladder polymer formed by 12 with intermolecular hydrogen bonds shown in orange. (b, c) Face-to-face
stacking of 12 in the solid state (from SCXRD) with inter-dimer hydrogen bonds shown in green. Hydrogen atoms have been omitted for clarity.
since tautomerization of the two enol-imine groups would leave
free radicals on two of the CH groups of the central ring. The
lack of tautomerization observed for 14, however, was unex-
pected on the basis of our results with 10 and will be discussed
later.
10 with N-tBOC-m-phenylenediamine 19 in place of the ortho
derivative 4 to afford compounds 20 and 21, respectively
(Scheme 7). The 1H NMR spectra of 20 and 21 clearly indicate
that no aggregation occurs when 1 and 10 are reacted with the
meta-derivative instead of the ortho derivative. Both compounds
show the characteristic peak pattern that derives from the
presence of two geometrical isomers locked in the keto-enamine
form. Presumably this geometry does not permit effective
interaction between the NH groups and the carbonyl function-
In compounds 5 and 12, the presence of the tBOC-protected
amine is important for generating hydrogen-bonded dimers or
extended structures. To obtain information on how the position
of the amino substituents affects dimerization, we reacted 1 and
J. Org. Chem, Vol. 71, No. 2, 2006 781

Sauer et al.
SCHEME 6. Synthesis of Compounds 14, 16, and 18
steric bulk of accommodating four or six cyclohexane moieties
around the core may inhibit dimerization.
These results show that the geometry of the tBOC-protected
diamines is important for facilitating dimerization in these
molecules. If the tBOC-protected amine and the free amine could
be separated further, keeping the geometry appropriate for
dimerization, then capsules or clefts with adjustable interior
cavity sizes may be obtainable. In an effort to extend the space
between the central rings of the clefts, we synthesized N-tBOC-
1,8-diaminonaphthalene 25 and reacted it with compound 10
(Scheme 8). ESI-MS indicated the mass expected for the
condensation product, but the 1H NMR spectrum was inadequate
to confirm the product’s identity. Single crystals suitable for
X-ray diffraction were obtained from a mixture of dichloro-
methane and hexanes. The solid-state data (Figure 7) showed
that 26 does not form the anticipated keto-enamine (NH) form.
Surprisingly, the structure revealed the formation of a perimidine
derivative, arising presumably via imine formation followed by
nucleophilic substitution from the NHtBOC amide, and finally
proton transfer. The second aldehyde reacted with the diamine
to yield a regular imine-phenol (OH) form. Perimidine formation
at this imine may be hindered by steric crowding. In solution,
we only observe the ring-closed perimidine, not the ring-opened
chain structure.31
alities as the m-phenylenediamine ring must undergo excessive
twisting to position the NH groups in close proximity of the
carbonyl groups in its partner ring.
The central rings of compounds 12 and 26 are both derived
from compound 10. In compound 12, both hydroxyl/formyl pairs
have tautomerized to the keto-enamine form. This compound
shows C-C, C-O, and C-N bond lengths that are consistent
with the keto-enamine form. In compound 26, on the other hand,
the structure shows that neither hydroxyl/formyl pair have
tautomerized to the keto-enamine form. In this case, the
perimidine formed prevents the one hydroxy/formyl pair from
tautomerizing, and the second pair does not tautomerize either.
The C-O, C-C, and C-N bond lengths of the central ring in
26 are consistent with the enol-imine tautomer. It is interesting
that we have characterized the central core of 12 and 26 in both
the keto and enol forms.
To date, all of the keto-enamine tautomers we have investi-
gated have involved aniline derivatives. Compounds 1 and 10
were reacted with a racemic mixture of N-tBOC-trans-1,2-
diaminocyclohexane 22 to give compounds 23 and 24, respec-
tively (Scheme 7). These compounds with aliphatic substituents
are also locked in the keto-enamine form, demonstrating that
the aromaticity of the substituents is not a requirement for
stabilization of this tautomer. However, both 23 and 24 are found
as a mixture of two geometrical isomers (ca. 1.6:1 for C3h:Cs
isomer in 23, 1.3:1 for isomer a: isomer b in 24) and show no
evidence for aggregation. Since the starting compound 22 is
present as a racemic mixture, the products 23 and 24 contain
multiple isomers that may not interlock well. In addition, the
Compounds 3 and 5 may be classified as hetero[6]radialenes.32
These molecules possess six sp2 hybridized carbon atoms in a
SCHEME 7. Synthesis of Compounds 20, 21, 23, and 24
782 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
FIGURE 7. (a) Solid-state structure of compound 26 (thermal ellipsoids shown at 50% probability). (b) Schematic of the central ring in 26 and
bond lengths measured by SCXRD. ESDs are 0.003-0.004 Å for the bond lengths shown.
SCHEME 8. Synthesis of Compound 26
cyclic structure. Previous examples of 1,3,5-trisoxo[6]radialenes
have been limited to a few examples with sulfur substituents.33,34
Although compounds 11 and 12 are not technically radialenes,
they have a central 2,3-dimethylidenecyclohex-5-ene-1,3-one
that has only been observed in a few examples with ethylene-
dithiol substituents.33
Calculations. Computation of hydroxynaphthaldehyde anils,
azo compounds, and related systems has provided important
information about tautomerization in these compounds.35 In
general, enaminones (e.g., 4-methylamino-3-buten-2-one) are
found in their keto-enamine (NH) form.36 When they form part
of a benzene ring, as in N-salicylideneanilines, the molecules
are present in the enol-imine (OH) tautomeric form. In this case,
FIGURE 8. Structures of model compounds 27-29 used in the
calculations of the TSANs.
tautomerization from the enol-imine to the keto-enamine form
(31) Maloshitskaya, O.; Sinkkonen, J.; Ovcharenko, V. V.; Zelenin, K.
N.; Pihlaja, K. Tetrahedron 2004, 60, 6913.
(32) Hopf, H.; Mass, G. Angew. Chem., Int. Ed. Engl. 1992, 31, 931.
(33) (a) Kimura, M.; Watson, W. H.; Nakayama, J. J. Org. Chem. 1980,
45, 3719. (b) Coffin, M. A.; Bryce, M. R.; Clegg, W. J. Chem. Soc., Chem.
Commun. 1992, 401.
(34) (a) Mono, S.; Pritzkow, H.; Sundermeyer, W. Chem. Ber. 1993,
126, 2111. (b) Gompper, R.; Kutter, E. Chem. Ber. 1965, 1365.
(35) (a) Gilli, P.; Bertolasi, V,; Pretto, L.; Lycˇka, A.; Gilli, G. J. J. Am.
Chem. Soc. 2002, 124, 13554. (b) Alarco´n, S. H.; Olivieri, A. C.; Gonza´lez-
Sierra, M. J. Chem. Soc., Perkin Trans. 2 1994, 1067. (c) Rodr´ıguez-
Santiago, Sodupe, M.; Oliva, A.; Bertran, J. J. Am. Chem. Soc. 1999, 121,
8882. (d) Gilli, P.; Bertolasi, V.; Pretto, L.; Antonov, L.; Gilli, G. J. Am.
Chem. Soc. 2005, 127, 4943. (e) Fabian, W. M. F.; Antonov, L.;
Nedeltcheva, D.; Kamounah, F. S.; Taylor, P. J. J. Phys. Chem. A 2004,
108, 7603.
would result in the unfavorable loss of aromaticity. Only larger
aromatic rings (e.g., phenanthrene, naphthalene) can stabilize
the keto-enamine form. There have been reports of N-salicylide-
neanilines existing as mixtures of NH and OH forms, but the
structural characterization of a purely NH form has been
difficult.16-19 We have found that the condensation of amines
with both 1 and 10 afford compounds that are exclusively in
the keto-enamine (NH) form, whereas condensation of amines
with 13, 15, and 17 affords only the enol-imine (OH) tautomers.
In an effort to rationalize the stability of the keto-enamine form
and to compare the structural features of the two tautomers, we
have carried out the first ab initio and semiempirical calculations
on a series of model compounds 27-33 shown in Figures 8
and 9. In the model compounds, the group attached to the
(36) (a) von Auwers, K.; Wunderling, H. Ber. 1931, 64B, 2748. (b)
Dudedk, G. O.; Holm, R. H. J. Am. Chem. Soc. 1962, 84, 2691. (c) Zhuo,
J.-C. Magn. Reson. Chem. 1997, 35, 21.
J. Org. Chem, Vol. 71, No. 2, 2006 783

Sauer et al.
utilized all of the oxygen atoms available for hydrogen-bonding.
Although there are other geometrical isomers available where
one ketone is involved in two hydrogen bonds (as in isomer b
of compound 11), these were ignored for the purpose of
calculations. In each model compound, the minimum energy
geometries of both the keto and enol tautomers were determined
using B3LYP with the 6-31G* basis set. Calculations were
completed starting with different relative orientations of the
phenyl rings until the lowest energy structure was identified.
Single point calculations were then performed on the energy-
minimized structures using B3LYP with the 6-31G** basis set.
Geometry optimization calculations were also performed starting
with the geometries obtained from B3LYP/6-31G* using AM1
and PM3 to evaluate the abilities of these semiempirical methods
to predict the most stable tautomer. The results of the calcula-
tions are summarized in Tables 1 and 2.
To confirm that this level (B3LYP/6-31G*) was reasonable
for approximating the geometry of the tautomers, geometry
optimizations were performed on compounds 27 and 32 with
the 6-31G** basis set. The structures calculated with the larger
basis set were almost identical to those obtained with the 6-31G*
basis set, with bond lengths varying by only 0.001-0.002 Å.
TSANs. For the compounds containing three OH and three
imine groups on the benzene ring, compounds 27 and 29 are
predicted to exhibit keto-enamines that are more stable than
the enol-imine isomers in the gas phase. In these compounds,
three OH‚‚‚N interactions in the enol-imine form are replaced
with three NH‚‚‚O interactions in the keto-enamine tautomer.
In compound 28, however, where the three OH groups are
adjacent to one another, ab initio calculations predict that the
enol-imine isomer is most stable. In this case one OH‚‚‚O and
two OH‚‚‚N hydrogen-bonding interactions in the enol-imine
form are replaced by only two NH‚‚‚O interactions in the keto-
enamine form. This loss of hydrogen-bonding in the keto form,
as well as steric interactions of the three adjacent enamine
groups, is likely responsible for the destabilization of the keto-
enamine form. Semiempirical methods agree with the ab initio
results for 27 and 29 but predict the keto-enamine tautomer to
be most stable for 28.
FIGURE 9. Structures of model compounds 30-33 used in the
calculations of the BSANs.
nitrogen atoms was truncated to a phenyl group; this abbreviated
structure is a sensible analogue of the methoxyaniline derivatives
3 and 11. Compounds 27-29 (Figure 8) represent all of the
possible tris(N-salicylideneaniline) structural isomers generated
by the condensation of trihydroxytriformylbenzene and aniline.
Model compounds 30-33 (Figure 9) are the only possible
salicylaldimine isomers arising from the condensation of aniline
with dihydroxydiformylbenzene, where the two aromatic CH
substituents are arranged adjacent (ortho) to one another (i.e.,
1,2,3,4-tetrasubstitution). The tautomerization to the bis(keto-
enamine) form would only be plausible with these arrangements
since other arrangements (e.g., 1,2,4,5-tetrasubstitution) would
result in two free radicals on the central ring in the keto-enamine
form. Calculation of one of these isomers indicated that this
would be very high in energy and thus unlikely to form.
Conformations were selected that maximized O‚‚‚H‚‚‚N
hydrogen-bonding. In the cases of the keto-enamine forms of
compounds 27, 29, and 32, the conformation was selected that
We are particularly interested in the predicted structure for
compound 27, since it is similar to compound 5. The enol-imine
(OH) form of compound 27 is predicted to have a flat interior
shared by the imine groups and the central benzene ring, with
TABLE 1. Calculated Energies of Formation for TSANs 27-29
a
a
a
b
Eenol
Eketo
∆E(enol-keto)
(kcal/mol)
∆E(enol-keto)
(kcal/mol)
∆E (AM1)c
(kcal/mol)
∆E (PM3)c
(kcal/mol)
compd
(Hartree)
(Hartree)
27
28
29
-1431.513081
-1431.470838
-1431.486756
-1431.534675
-1431.452264
-1431.497803
13.55
-11.66
6.93
18.12
-7.5
11.00
22.94
9.29
17.00
5.97
3.41
4.67
a Single point determination using B3LYP with 6-31G** basis set on B3LYP/6-31G* geometry. b B3LYP/6-31G* basis set. c Geometry optimized beginning
from B3LYP/6-31G* geometry.
TABLE 2. Calculated Energies of Formation for BSANs 30-33
a
a
a
b
Eenol
Eketo
∆E(enol-keto)
(kcal/mol)
∆E(enol-keto)
(kcal/mol)
∆E (AM1)c
(kcal/mol)
∆E (PM3)c
(kcal/mol)
compd
(Hartree)
(Hartree)
30
31
32
33
-1031.744646
-1031.741870
-1031.757359
-1031.738162
-1031.739779
-1031.738826
-1031.761948
-1031.717439
-3.05
-1.91
2.88
-0.67
0.71
5.46
4.77
3.99
7.80
-4.37
-5.00
-2.52
-7.16
-13.00
-10.32
-4.51
a Single point determination using B3LYP with 6-31G** basis set on B3LYP/6-31G* geometry. b B3LYP/6-31G* basis set. c Geometry optimized beginning
from B3LYP/6-31G* geometry.
784 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
In compound 28, where the keto-enamine form is predicted to
be the least stable tautomer, steric interactions of the three
enamine groups lead to a large distortion of the central ring
from planarity in the calculated structure for the keto-enamine
form.
BSANs. Table 2 lists the energies of the structures calculated
for the BSANs 30-33. In TSANs 27 and 29, the stabilization
afforded from the formation of three keto-enamine tautomers
was sufficient to overcome the resonance stabilization of the
aromatic ring. In the BSANs, it is unclear whether the
stabilization of only two groups tautomerizing is sufficient to
overcome the stabilization of the aromatic ring. Both ab initio
and semiempirical methods predict that compound 33 will be
most stable in the enol-imine (OH) tautomer. In this case, steric
interactions of the aniline substituents as well as the loss of a
hydrogen bond in the keto form render the enol-imine tautomer
more stable. Steric interactions between the rigid enamine groups
in compound 31 also destabilize the keto form.
In our studies of BSANs, we were most surprised that
products formed by the condensation of aniline derivatives with
1,3-dihydroxy-2,4-diformylbenzene 10 (i.e., compounds 11, 12,
21) gave products locked in the keto-enamine (NH) form,
whereas products obtained from the condensation of aniline
derivatives with 1,2-dihydroxy-3,6-diformylbenzene 13 (i.e.,
compound 14) were locked in the enol-imine (OH) form. In
both cases, there is no decrease in the number of hydrogen bonds
between the keto-enamine and enol-imine forms, nor is there
significant steric interaction between the bulky peripheral phenyl
rings. Ab initio calculations agree with the experimental
observations and correctly predict the most stable tautomer for
compounds 30 and 32.
FIGURE 10. (a) Structure of the core of compound 5 as determined
by SCXRD. Thermal ellipsoids are shown at 50% probability. The
substituents have been truncated at the ipso-carbon atom of the phenyl
ring to emphasize the structure of the core. Only one position of N1
and N3 are shown, with disordered fragments omitted for clarity. (b)
Calculated energy-minimized structure (B3LYP/6-31G**) of the keto-
enamine form of compound 27. (c) Illustration of the core of 5 and 27
with key bond lengths identified for Table 3.
The enhanced stability of the keto-enamine tautomer in 32
versus 30 can be rationalized by considering the resonance
structures of the molecule. Stabilization by resonance-assisted
hydrogen-bonding is facilitated by ortho/para arrangement of
the keto groups to the enamines in 32.37 There are two additional
resonance structures that contribute in the case of the keto-form
of 32. This is illustrated in Figure 11.
TABLE 3. Bond Lengths Determined for 5 (from SCXRD) and 27
(Calculation)
5
27 keto
27 enol
bonda
(XRD)b (Å)
(DFT)c (Å)
(DFT)c (Å)
a
b
c
d
e
CdO
C-C
C-C
CdC
C-N
1.256(8)
1.45(2)
1.45(1)
1.38(1)
1.34(6)
1.257
1.466
1.460
1.386
1.337
1.327
1.419
1.422
1.440
1.299
We were hopeful that computation would allow us to
unequivocally assign the structures obtained for compounds 12
and 26 to the keto-enamine and enol-imine form, respectively.
The isolation and structural characterization of two different
tautomers is useful for determining the tautomeric form and
for evaluating the accuracy of our computations. The enol-imine
form of compound 32 is predicted to have a flat interior shared
by the imine groups and the central benzene ring, with the two
phenyl substituents twisted with a dihedral angle of ca. 34-
36° between imine and phenyl ring. In the more stable keto-
enamine form, however, the molecule is predicted to be flat,
enabling delocalization of the electrons on the nitrogen atoms
into the conjugated system containing the ketones. A structure
of the molecule and its relevant bond lengths are shown in
Figure 12. Although the figures of merit for the structure
determined from SCXRD are not outstanding due to disorder,
the bond lengths calculated for the keto-enamine structure
correspond quite well to those determined by X-ray diffraction
(there are two independent molecules of 12 in the asymmetric
unit), Table 4. Most importantly, the CdO bond lengths range
a See Figure 10. b T ) -100 °C. c B3LYP/6-31G** geometry optimized
structure.
the three phenyl substituents in a propeller orientation (dihedral
angle of ca. 31-34° between imine and phenyl ring). In the
more stable keto-enamine (NH) form, however, the molecule
is predicted to be flat, enabling delocalization of the electrons
on the nitrogen atoms into the conjugated system containing
the ketones (geometry optimization with 6-31G** basis set
revealed a slight (1-3°) torsion of the peripheral phenyl rings).
A structure of the molecule and its relevant bond lengths are
shown in Figure 10. The bond lengths calculated for the keto-
enamine structure agree with those determined by X-ray
diffraction (Table 3), verifying that TSAN 5 is in the keto-
enamine form.
The computed enol-imine form of compound 29 is similar
in geometry to compound 27, with a flat interior and propeller
orientation of the phenyl substituents. In the keto-enamine form,
however, steric interactions between the adjacent enamine
groups lead to twisting of the central ring and probably reduce
the degree of conjugation. As a consequence, the keto-enamine
form is only slightly more stable than the enol-imine tautomer.
(37) (a) Gilli, G.; Bellucci, F.; Ferretti, V.; Bertolasi, V. J. Am. Chem.
Soc. 1989, 111, 1023. (b) Bertolasi, V.; Nanni, L.; Gilli, P.; Ferretti, V.;
Gilli, G.; Issa, Y. M.; Sherif, O. E. New. J. Chem. 1994, 18, 251. (c) Gilli,
P.; Bertolasi, V.; Ferretti, V.; Gilli, G. J. Am. Chem. Soc. 2000, 122, 10405.
J. Org. Chem, Vol. 71, No. 2, 2006 785

Sauer et al.
TABLE 4. Bond Lengths Determined for 12 (SCXRD), 26
(SCXRD), and 32 (Calculation)
12
32 keto
32 enol
26
bonda
(XRD)b (Å) (DFT)c (Å)
(DFT)c (Å) (XRD)b (Å)
a
b
c
d
e
f
CdC
C-C
C-C
C-C
C-C
CdC
C-N
CdO
1.33(2)
1.44(1)
1.45(3)
1.42(2)
1.43(2)
1.39(2)
1.32(2)
1.27(1)
1.357
1.439
1.379
1.412
1.385(4)
1.401(4)
1.407(6)
1.418(3)
1.393(4)
1.463(3)
1.288(3)
1.360(4)
1.467, 1.469 1.424, 1.422
1.461
1.456
1.385, 1.387 1.446, 1.441
1.334, 1.342 1.296, 1.297
1.256, 1.259 1.330, 1.335
1.416
1.410
g
h
a See Figure 12. b T ) -100 °C. c B3LYP/6-31G** geometry optimized
structure.
bond lengths of the core in compound 26 are also included,
recognizing that the perimidine substituent is not an imine and
does not have the same extent of conjugation. Nonetheless, the
bond lengths determined for compound 26 agree closely with
those calculated for the enol-imine form of a salicylideneaniline.
The C-O bond lengths are characteristic of single bonds, while
the CdN bond length is consistent with an imine.
The DFT methods were useful to predict the most stable
tautomer of the BSANs and TSANs and to estimate their
expected geometries. The use of B3LYP/6-31G* with a 6-31G**
single point calculation is adequate to correctly predict the most
stable isomer for the BSANs and TSANs we prepared. The only
questionable result is that for compound 31. In this case, B3LYP/
6-31G* predicted that the keto-enamine tautomer would be most
stable, while a single point calculation with 6-31G** predicted
that the enol-imine tautomer was more stable. We believe the
6-31G** result as the tautomers of this compound should show
stability comparable to those of compound 30. Semiempirical
methods, which are substantially faster, were also evaluated.
PM3 did not accurately predict the most stable tautomer in most
cases, while AM1 methods usually agreed with the DFT/
experimental results. AM1 has previously shown good utility
in predicting the stability of salicylaldimine tautomers34b but
has not been used to examine BSANs or TSANs.
FIGURE 11. Resonance stabilization of the keto-enamine form of
compounds 30 and 32.
Our calculations indicate that compound 31, if it could be
synthesized, may be a good target for developing a bistable
molecular switch. In this compound, the energy difference
between the keto and enol tautomers is small so it may be
possible to utilize this compound or related BSANs for
photoswitching. We hope that others will be inspired to develop
synthetic routes to these novel compounds.
Conclusions
We have discovered new tris(N-salicylaldimine)s and bis(N-
salicylaldimine)s that undergo tautomerization to very stable
keto-enamine tautomers. We have structurally characterized the
keto-enamine form of two of these compounds, as well as an
enol-imine form of another. Calculations show that the tau-
tomerization is general and is stabilized by resonance between
the enamine and keto groups. The hydrogen-bonding abilities
of TSANs and BSANs make them good candidates for the
formation of dimeric capsules and clefts, respectively. We have
structurally characterized a dimeric capsule-type assembly and
an extended network formed by BSANs. Further studies of
aggregation in BSANs and TSANs are underway.
FIGURE 12. (a) Structure of the core of compound 12 (one of two
molecules in the asymmetric unit) as determined by SCXRD. Thermal
ellipsoids are shown at 50% probability. The substituents have been
truncated at the ipso-carbon atom of the phenyl ring to emphasize the
structure of the core. (b) Calculated energy-minimized structure
(B3LYP/6-31G**) of the keto-enamine form of compound 32. (c)
Illustration of the core of 12 and 32 with key bond lengths identified
for Table 4.
Experimental Section
from 1.258(12) to 1.284(12) Å, which are as expected for the
keto-enamine tautomer. Thus, BSAN 12 is indeed present in
the keto-enamine form, consistent with NMR solution data. The
Synthesis of 1,3-Dimethoxy-2,4-diformylbenzene (9). Under
a nitrogen atmosphere 1,3-dibromo-2,4-dimethoxybenzene (27.5 g,
786 J. Org. Chem., Vol. 71, No. 2, 2006

N-Salicylideneanilines
93 mmol) was dissolved in 600 mL of dry diethyl ether and cooled
to 0 °C. n-Butyllithium (175 mL of a 1.6 M solution in hexane,
280 mmol) was added over a period of 10 min and the resulting
yellow, milky mixture was stirred for 30 min. Anhydrous DMF
(15.5 mL, 200 mmol) was then added dropwise and the suspension
was stirred for an additional 30 min at 0 °C before being warmed
to room temperature. The reaction mixture was poured into 1 L of
water, followed by the addition of 400 mL of concentrated HCl.
The resulting two phases were separated and the aqueous layer was
extracted with diethyl ether (3 × 100 mL). The combined organic
layers were dried over MgSO4, filtered and concentrated under
vacuum. The resulting brown solid was recrystallized from toluene
yielding 9 as colorless needles (9 g, 46 mmol, 50%). 1H NMR (300
MHz, CDCl3) δ 10.44 (s, 1H), 10.26 (s, 1H), 8.05 (d, 1H, J ) 9
Hz), 6.85 (d, 1H, J ) 9 Hz), 3.98 (s, 6H) ppm; 13C NMR (100.6
MHz, CDCl3) δ 188.3, 187.9, 167.1, 165.9, 135.6, 123.4, 118.6,
108.1, 65.7, 56.8 ppm; MS (EI, 70 eV) m/z 194 (M+); IR ν (KBr)
) 3459, 2887, 1676, 1579, 1483, 1454, 1382, 1277, 1249, 1217,
1181, 1096, 947, 802, 734, 545 cm-1; mp ) 98-99 °C; HRMS-EI
(M+) calcd for C10H10O4 194.0579, found 194.0577. Anal. Calcd
for C10H10O4: C, 61.85, H, 5.19. Found: C, 61.87, H, 5.19.
Synthesis of 1,3-Dihydroxy-2,4-diformylbenzene (10). Under
a nitrogen atmosphere 1,3-dimethoxy-2,4-diformylbenzene (5 g,
25.8 mmol) was dissolved in 100 mL of dry CH2Cl2 and cooled to
0 °C. To this solution was added BBr3 (14.6 mL, 154 mmol),
resulting in a color change from pale yellow to red. After stirring
at 0 °C for 1 h the ice bath was removed and the solution was
stirred at room temperature overnight. The solution was poured
onto 400 mL of ice and the two phases were separated. The aqueous
layer was extracted with CH2Cl2 (2 × 50 mL) and the combined
organic layers were dried over MgSO4. Solvent was removed under
vacuum and the obtained red solid was purified by flash chroma-
tography (silica, CH2Cl2) yielding 10 as a white powder (3.5 g, 21
N-(tert-butyloxycarbonyl)-1,2-diaminobenzene 4 (0.415 g, 2 mmol).
Yield: 0.236 g (0.4 mmol, 72%) of 12 as a yellow solid. 1H NMR
(300 MHz, CDCl3) δ 14.05 (d, 1H, J ) 6.5 Hz, CdCNH), 13.67
(d, 1H, J ) 6.5 Hz, CdCNH), 8.87 (d, 1H, J ) 6 Hz, HC-N),
7.99 (d, 1H, J ) 6 Hz, HC-N), 7.95-7.83 (m, 2H), 7.25-7.01
(m, 4H), 6.21 (d, 1H, J ) 10 Hz), 1.47 (s, 18H) ppm; 13C NMR
(75.4 MHz, CDCl3) δ 181.4, 178.8, 157.8, 155.6, 154.9, 154.7,
142.1, 135.5, 135.2, 132.8, 131.9, 128.9, 127.9, 126.3, 126.0, 124.7,
123.8, 120.8, 119.7, 116.0, 111.8, 110.2, 82.6, 82.3, 29.6, 29.6 ppm;
MS (EI, 70 eV) m/z 546 (M+); IR ν (KBr) ) 3253, 2980, 2931,
1724, 1640, 1603, 1575, 1519, 1438, 1358, 1334, 1241, 1149, 1020,
742, 493 cm-1; mp ) 183-185 °C (dec); HRMS-EI (M+) calcd
for C30H34N4O6 546.2475, found 546.2478. Anal. Calcd for
C30H34N4O6: C, 65.92, H, 6.27, N, 10.25. Found: C, 65.52, H,
6.42, N, 10.51.
Synthesis of 14. Compound 14 was prepared by a procedure
analogous to the preparation of 11, using 13 (0.1 g, 0.6 mmol) and
N-(tert-butyloxycarbonyl)-1,2-diaminobenzene 4 (0.415 g, 2 mmol).
Yield: 0.252 g (0.46 mmol, 77%) of 14 as a yellow solid. 1H NMR
(300 MHz, CDCl3) δ 12.57 (s br, 2H), 8.61 (s, 2H), 8.19 (d, 2H,
J ) 6.5 Hz, 7.31-7.27 (m, 2H), 7.09-7.07 (m, 6H), 7.0 (s, 2H),
1.51 (s, 18H) ppm; 13C NMR (75.4 MHz, CDCl3) δ 164.9, 153.9,
151.1, 138.9, 134.1, 129.9, 124.6, 123.4, 122.7, 120.8, 119.8, 82.4,
29.7 ppm; MS (EI, 70 eV) m/z 546 (M+); IR ν (KBr) ) 3443,
3382, 2976, 2923, 1728, 1611, 1587, 1543, 1479, 1438, 1390, 1362,
1306, 1229, 1145, 1108, 1044, 1020, 879, 827, 750, 617, 565 cm-1
;
mp ) 218-220 °C (dec). Anal. Calcd for C30H34N4O6: C, 65.92,
H, 6.27, N, 10.25. Found: C, 65.68, H, 6.31, N, 10.44.
Synthesis of 16. Compound 16 was prepared by a procedure
analogous to the preparation of 11, using 15 (0.1 g, 0.6 mmol) and
4 (0.415 g, 2 mmol). Yield: 0.193 g (0.35 mmol, 69%) of 16 as
1
an orange solid. H NMR (300 MHz, CDCl3) δ 11.79 (s br, 2H),
8.57 (s, 2H), 8.18 (d, 2H, J ) 6.5 Hz), 7.36-7.28 (m, 2H), 7.10-
7.07 (m, 6H), 7.00 (s, 2H), 1.51 (s, 18H) ppm; 13C NMR (75.4
MHz, CDCl3) δ 163.1, 152.8, 152.5, 137.6, 132.6, 128.6, 123.3,
122.9, 119.7, 119.5, 118.3, 81.0, 28.3 ppm; MS (EI, 70 eV) m/z
546 (M+); IR ν (KBr) ) 3435, 3358, 2964, 2936, 1724, 1700, 1696,
1603, 1583, 1503, 1438, 1374, 1314, 1241, 1145, 1044, 1016, 963,
883, 843, 754, 589, cm-1; mp > 250 °C. Anal. Calcd for
C30H34N4O6: C, 65.92, H, 6.27, N, 10.25. Found: C, 66.14, H,
6.33, N, 10.49.
1
mmol, 81%). H NMR (300 MHz, CDCl3) δ 12.63 (s, 1H), 12.53
(s, 1H), 10.38 (s, 1 H), 9.69 (s, 1H), 7.63 (d, 1H, J ) 9 Hz), 6.54
(d, 1H, J ) 9 Hz) ppm; 13C NMR (100.6 MHz, CDCl3) δ 194.5,
193.6, 169.6, 167.3, 141.8, 113.6, 110.2, 109.4 ppm; MS (EI, 70
eV) m/z 166 (M+); IR ν (KBr) ) 3439, 3096, 2895, 1640, 1599,
1483, 1410, 1386, 1306, 1257, 1225, 1181, 1084, 939, 831, 750,
670, 617, 577, 497 cm-1; mp ) 127-128 °C; HRMS-EI (M+) calcd
for C8H6O4 166.0266, found 166.0266. Anal. Calcd for C8H6O4:
C, 57.84, H, 3.64. Found: C, 57.74, H, 3.64.
Synthesis of 18. Compound 18 was prepared by a procedure
analogous to the preparation of 11, using 17 (0.1 g, 0.6 mmol) and
4 (0.415 g, 2 mmol). Yield: 0.256 g (0.47 mmol, 78%) of 18 as a
yellow solid. 1H NMR (300 MHz, CDCl3) δ 13.26 (s br, 2H), 8.49
(s, 2H), 8.11 (d, 2H, J ) 6 Hz), 7.47 (s, 1H), 7.25-7.23 (m, 2H),
7.05-7.03 (m, 4H), 6.93 (s, 2H), 6.59 (s, 1H), 1.51 (s, 18H) ppm;
13C NMR (75.4 MHz, CDCl3) δ 165.9, 162.5, 152.6, 138.4, 137.8,
132.2, 127.8, 123.4, 119.6, 118.4, 113.6, 104.4, 80.9, 28.3 ppm;
MS (EI, 70 eV) m/z 546 (M+); IR ν (KBr) ) 3422, 2976, 2923,
1736, 1640, 1587, 1507, 1454, 1358, 1297, 1225, 1145, 1048, 1016,
984, 871, 746, 480 cm-1; mp ) 135-137 °C (dec). Anal. Calcd
for C30H34N4O6: C, 65.92, H, 6.27, N, 10.25. Found: C, 65.99, H,
6.20, N, 10.28.
Synthesis of 20. Compound 20 was prepared by a procedure
analogous to the preparation of 11, using 1 (0.137 g, 0.6 mmol)
and N-(tert-butyloxycarbonyl)-1,3-diaminobenzene 19 (0.642 g, 3.1
mmol). Yield: 0.337 g (0.4 mmol, 69%) of 20 as a yellow solid.
1H NMR (300 MHz, CDCl3) δ 13.29-13.25 and 12.97-12.89 (m,
3H), 8.70-8.63 (m, 3H), 7.46-6.90 (m, 15H), 6.67-6.61 (m, 3H),
1.26 (s, 27H) ppm; 13C NMR (75.4 MHz, CDCl3) δ 188.9, 186.2,
186.1, 154.6, 150.1, 142.9, 141.1, 140.9, 131.9, 116.9, 113.2, 108.6,
108.3, 108.1, 107.8, 81.3, 42.2, 41.9, 41.6, 41.3, 41.1, 40.8, 40.5,
30.0 ppm; MS (EI, 70 eV) m/z 781 (M+); IR ν (KBr) ) 3439,
2980, 2927, 1728, 1611, 1583, 1535, 1438, 1366, 1257, 1233, 1145,
883, 750, 674, 568 cm-1; mp ) 208-210 °C (dec); HRMS-EI (M+)
calcd for C42H48N6O9 781.3561, found 781.3549. Anal. Calcd for
C42H48N6O9: C, 64.60, H, 6.20, N, 10.76. Found: C, 64.00, H,
6.38, N, 10.76.
Synthesis of 11. To a solution of 10 (0.1 g, 0.6 mmol) in 50 mL
of ethanol was added 4-methoxyaniline (2) (0.246 g, 2 mmol) in
20 mL of ethanol and the solution was stirred at reflux overnight.
The solution was cooled to room temperature and the solvent was
removed under reduced pressure. The remaining yellow film was
dissolved in THF and precipitated from hexanes. The precipitate
was collected by filtration, washed with hexanes and dried under
vacuum to give 0.129 g (0.3 mmol, 57%) of compound 11 as a
yellow solid. 1H NMR (300 MHz, CDCl3) δ 14.33 (d, 1H, J ) 12
Hz, CdCNH of isomer a), 13.56 (d, 1H, J ) 13 Hz, CdCNH of
isomer b), 13.45 (d, 1H, J ) 12 Hz, CdCNH of isomer a), 12.67
(d, 1H, J ) 12.5 Hz, CdCNH of isomer b), 8.78 (d, 1H, J ) 12.5
Hz, HC-N of isomer a), 8.72 (d, 1H, J ) 13.5 Hz, HC-N of
isomer b), 7.61 (d, 1H, J ) 12 Hz, HC-N of isomer a), 7.51 (d,
1H, J ) 12.5 Hz, HC-N of isomer b), 7.24-6.83 (m, 18H), 5.96
(d, 2H, J ) 9.5 Hz), 3.76 (s, 6H), 3.73 (s, 6H) ppm; 13C NMR
(75.4 MHz, CDCl3) δ 187.5, 187.3, 184.7, 159.6, 159.5, 158.7,
158.6, 153.3, 152.7, 150.2, 148.7, 142.8, 134.0, 133.9, 133.8, 133.4,
121.1, 120.9, 120.1, 119.9, 119.5, 118.0, 116.5, 111.7, 107.9, 107.5,
56.9 ppm; MS (EI, 70 eV) m/z 376 (M+); IR ν (KBr) ) 3463,
3052, 2944, 2827, 1648, 1615, 1579, 1527, 1511, 1450, 1302, 1257,
1221, 1169, 1116, 1028, 815, 754, 706, 456, 509 cm-1; mp ) 178-
180 °C (dec); HRMS-EI (M+) calcd for C22H20N2O4 376.1423,
found 376.1418. Anal. Calcd for C22H20N2O4: C, 70.20, H, 5.36,
N, 7.44. Found: C, 69.87, H, 5.29, N, 7.57.
Synthesis of 12. Compound 12 was prepared by a procedure
analogous to the preparation of 11, using 10 (0.1 g, 0.6 mmol) and
J. Org. Chem, Vol. 71, No. 2, 2006 787

Sauer et al.
1
Synthesis of 21. Compound 21 was prepared by a procedure
analogous to the preparation of 11 from 10 (0.1 g, 0.6 mmol) and
19 (0.415 g, 2 mmol). Yield: 0.223 g (0.4 mmol, 68%) of 21 as a
yellow solid. 1H NMR (300 MHz, CDCl3) δ 14.10 (d, 1H, J ) 13
Hz), 13.56 (d, 1H, J ) 13 Hz), 13.28 (d, 1H, J ) 12 Hz), 12.75 (d,
1H, J ) 12 Hz), 8.91-8.82 (m, 4H), 7.76-6.89 (m, 18H), 6.03-
5.99 (m, 2H), 1.46 (s, 36H) ppm; 13C NMR (75.4 MHz, CDCl3) δ
188.3, 185.5, 161.3, 154.0, 153.9, 153.4, 153.0, 149.4, 143.4, 141.5,
141.3, 141.0, 140.7, 131.8, 131.6, 119.9, 118.9, 117.6, 116.2, 114.1,
113.6, 113.5, 112.9, 109.8, 108.3, 108.0, 82.4, 42.2, 29.7 ppm; MS
(EI, 70 eV) m/z 546 (M+); IR ν (KBr) ) 3282, 3080, 2972, 2919,
1728, 1644, 1591, 1527, 1495, 1442, 1366, 1293, 1233, 1149, 1048,
984, 867, 770, 678, 480 cm-1; mp ) 146-148 °C (dec); HRMS-
EI (M+) calcd for C30H34N4O6 546.2478, found 546.2468. Anal.
Calcd for C30H34N4O6: C, 65.92, H, 6.27, N, 10.25. Found: C,
65.60, H, 6.38, N, 10.34.
hexanes afforded 25 as red crystals (2.6 g, 10.2 mmol, 81%). H
NMR (300 MHz, CDCl3) δ 9.69 (br s, 1H), 8.03 (d, 1H, J ) 6
Hz), 7.47-7.17 (m, 4H), 6.78 (d, 1H, J ) 5.5 Hz), 4.02 (s br, 2H),
1.49 (s, 9H) ppm; 13C NMR (75.4 MHz, CDCl3) δ 153.1, 140.7,
136.2, 135.3, 126.0 125.4, 123.5, 122.6, 119.3, 118.8, 116.8, 116.7,
80.0, 28.4 ppm; MS (EI, 70 eV) m/z 258 (M+); IR ν (KBr) ) 3378,
3298, 3205, 3048, 2984, 2923, 2879, 1925, 1708, 1696, 1603, 1539,
1499, 1430, 1346, 1237, 1153, 1076, 976, 887, 819, 754, 658, 625,
468 cm-1; mp ) 89-90 °C. Anal. Calcd for C15H18N2O2: C, 69.74,
H, 7.02, N, 10.84. Found: C, 70.07, H, 7.20, N, 10.90.
Synthesis of 26. Compound 26 was prepared by a procedure
analogous to the preparation of 11, by reaction of 10 (0.05 g, 0.3
mmol) with 25 (0.155 g, 0.6 mmol) in ethanol. Yield: 0.120 g
1
(0.19 mmol, 62%) of 26 as a yellow solid. H NMR (300 MHz,
CDCl3) δ 11.76 (s, 1H), 10.80 (s, 1H), 9.00 (s, 1H), 8.98 (d, 2H,
J ) 8 Hz), 8.05 (d, 1H, J ) 6 Hz), 7.72 (d, 1H, J ) 6 Hz), 7.66
(d, 1H, J ) 6 Hz), 7.54-7.32 (m, 7H), 7.03-6.97 (m, 2H), 6.95
(d, 1H, J ) 3 Hz), 6.83 (d, 1H, J ) 5 Hz), 6.21 (d, 1H, 6.5 Hz),
4.93 (d, 1H, J ) 3 Hz); 1.54 (s, 9H), 1.12 (s, 9H) ppm; 13C NMR
(75.4 MHz, CDCl3) δ 161.4, 161.3, 159.6, 157.8, 153.0, 147.7,
135.8, 135.1, 135.0, 133.8, 132.6, 130.9, 128.3, 127.9, 127.5, 126.7,
126.6, 126.4, 126.4, 126.0, 125.5, 123.8, 123.7, 123.4, 123.1, 119.4,
119.4, 117.7, 117.4, 117.3, 109.9, 108.9, 107.6, 83.2, 79.9, 79.8,
79.5, 63.1, 53.3, 28.2, 27.8 ppm; ESI-MS m/z ) 647 (M+H+); IR
ν (KBr) ) 3410, 3358, 3060, 2968, 2923, 1688, 1607, 1563, 1523,
1487, 1426, 1354, 1318, 1229, 1145, 1072, 1040, 980, 815, 758,
529, 456 cm-1; mp ) 146-148 °C (dec). Anal. Calcd for
C38H38N4O6: C, 70.57, H, 5.92, N, 8.66. Found: C, 70.77, H, 6.10,
N, 9.06.
Computations. Ab initio DFT calculations were performed using
Spartan ‘04 for Windows (Wavefunction Inc., Irvine, CA). Both
the keto and enol isomers of 27-33 were geometry minimized using
the B3LYP method with a 6-31G* basis set. For each molecule,
the calculation was started with different possible orientations of
the peripheral benzene rings until the minimum energies (which
we are confident are at least very close to the global minimum)
were determined, then a B3LYP/6-31G** single point calculation
was performed. In the cases of 27 and 32, geometry minimized
structures with B3LYP/6-31G** were almost identical with respect
to bond lengths and bond angles to those determined by B3LYP/
6-31G*. The energies of the B3LYP/6-31G* structures are tabulated
in Tables 1 and 2. AM1 and PM3 geometry optimization calcula-
tions were performed beginning with the geometries obtained from
B3LYP/6-31G*.
Synthesis of 23. Compound 23 was prepared by a procedure
analogous to the preparation of 11, using 1 (0.1 g, 0.5 mmol) and
racemic N-(tert-butyloxycarbonyl)-trans-1,2-diaminocyclohexane 22
(0.336 g, 1.5 mmol). Yield: 0.283 g (0.35 mmol, 71%) of 23 as a
1
yellow solid. H NMR (300 MHz, CDCl3) δ 11.32 and 10.98-
10.79 (m, 3H), 8.17-8.08 (m, 3H), 4.66-4.47 (m, 2H), 3.44 (m,
2H), 2.96-2.92 (m, 2H), 2.26-2.01 (m, 6H), 1.76-1.70 (m), 1.47-
1.29 (m), 1.11-1.02 (m) ppm; 13C NMR (75.4 MHz, CDCl3) δ
187.4, 185.0, 182.8, 156.7, 155.9, 155.2, 104.5, 104.4, 79.4, 64.8,
63.8, 53.6, 32.3, 32.2, 32.0, 28.3, 28.2, 28.1, 24.8, 24.3 ppm; ESI-
MS m/z ) 799 (M + H+); IR ν (KBr) ) 3406, 3278, 2927, 2859,
2384, 2344, 2332, 1716, 1700, 1676, 1595, 1555, 1503, 1454, 1382,
1313, 1157, 1092, 1000, 959, 839, 694, 609, 569, 517 cm-1; mp )
94-95 °C. Anal. Calcd for C42H66N6O9: C, 63.13, H, 8.33, N,
10.52. Found: C, 62.74, H, 8.15, N, 10.66.
Synthesis of 24. Compound 24 was prepared by a procedure
analogous to the preparation of 11, using 10 (0.1 g, 0.6 mmol) and
22 (0.171 g, 1.2 mmol). Yield: 0.214 g (0.38 mmol, 64%) of 24
1
as a light yellow solid. H NMR (300 MHz, CDCl3) δ 12.40 (s),
11.68 (s), 11.48 (s), 10.92-10.86 (m), 8.39-8.33 (m), 7.23-7.11
(m), 6.78-6.74 (m), 5.80-5.52 (m), 5.06-4.65 (m), 3.43-3.08
(m), 1.99-195 (m), 1.72-1.70 (m), 1.50-1.41 (m), 1.37-1.17 (m),
0.82-0.76 (m) ppm; 13C NMR (75.4 MHz, CDCl3) δ 186.0, 185.4,
185.3, 183.1, 159.1, 158.2, 156.0, 155.2, 141.6, 141.1, 115.9, 115.8,
115.1, 109.3, 109.2, 104.5, 104.3, 79.5, 64.4, 62.8, 53.8, 32.7, 32.3,
32.2, 31.8, 28.9, 28.3, 28.1, 28.1, 28.1, 28.0, 24.5, 24.3 ppm; MS
(EI, 70 eV) m/z 558 (M+); IR ν (KBr) ) 3443, 3334, 2984, 2936,
2859, 1712, 1652, 1607, 1535, 1378, 1358, 1302, 1233, 1157, 1012,
996, 847, 766, 698, 629, 521 cm-1; mp ) 163-165 °C (dec). Anal.
Calcd for C30H46N4O6: C, 64.49, H, 8.30, N, 10.03. Found: C,
64.29, H, 8.15, N, 9.87.
Acknowledgment. M.S. thanks NATO for a postdoctoral
fellowship. J.H.C. is grateful to NSERC for a scholarship. We
thank NSERC of Canada (Discovery Grant) and UBC for
funding.
Synthesis of 25. Under a nitrogen atmosphere di-tert-butyl
dicarbonate (3.0 g, 13.7 mmol) was added to a mixture of 1,8-
diaminonaphthalene (2 g, 12.6 mmol) and triethylamine (2 mL, 14
mmol) in 50 mL of dry THF. The mixture was stirred for 12 h at
room temperature and the solvent removed under reduced pressure.
The obtained off-white solid was dissolved in 30 mL of toluene
and washed with NaOH (1 N, 10 mL), brine (10 mL) and H2O (10
mL). The organic layer was dried over MgSO4, filtered, and the
solvent removed by rotary evaporation. Recrystallization from
Supporting Information Available: X-ray crystallographic data
for 12 and 26 (CIF format); general experimental methods; synthesis
of 7, 8, and 19; NMR spectra of all new compounds; VT NMR
data; and atomic coordinates for calculated structures. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO052277T
788 J. Org. Chem., Vol. 71, No. 2, 2006