Importance of secondary interactions in twisted doubly hydrogen bonded complexes

Article abstract of DOI:10.1021/ol302803j

Three model hydrogen bond arrays that form complexes with large twist angles between their heterocyclic rings were synthesized differing only in the sequence of their hydrogen bond donors and acceptors. The complementary and self-complementary association

Full text of DOI:10.1021/ol302803j

ORGANIC
LETTERS
2012
Vol. 14, No. 22
5772–5775
Importance of Secondary Interactions
in Twisted Doubly Hydrogen Bonded
Complexes
Jiaxin Li,† Antonia T. Pandelieva,‡ Christopher N. Rowley,§ Tom K. Woo,*,‡ and
James A. Wisner*,†
The University of Western Ontario, Department of Chemistry, London, Ontario,
N6A 5B7 Canada, Department of Chemistry, Memorial University of Newfoundland,
St. John’s, NL, A1B 3X7, Canada, and Centre for Catalysis Research and Innovation,
Department of Chemistry, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
twoo@uottawa.ca; jwisner@uwo.ca
Received October 11, 2012
ABSTRACT
Three model hydrogen bond arrays that form complexes with large twist angles between their heterocyclic rings were synthesized differing only in
the sequence of their hydrogen bond donors and acceptors. The complementary and self-complementary association of the arrays to form
complexes was studied computationally and in solution. The analysis reveals the significant impact secondary interactions have on complex
stability in such an arrangement despite the very different topology in comparison to typical planar arrays.
The association of two molecular fragments through
hydrogen bonding between multiple donor/acceptor con-
tacts is a ubiquitous feature of molecular recognition and
assembly in naturally occurring and artificial systems.1
Hence, the strength of a binding interaction in this context
is very important, and the contribution of individual
hydrogen bond donor/acceptor pairs to complex stability
can be predicted with reasonable accuracy using both
empirical and purely theoretical techniques.2
space, the situation is more complex and not a simple
function of the number of primary hydrogen bonds pre-
sent. In particular, hydrogen bonding between nucleobases
or similar heterocycles has been studied in detail to under-
stand the relationship between the arrangement of the
donor(D)/acceptor(A) pairs that interact to form a com-
plex and the resulting stabilization.3 The influence of
secondary interactions in these cases was first introduced
by Jorgensen and Pranata,4 and empirical methods for
evaluating the stabilities of such complexes in nonpolar
solutions have been formulated with largely successful
results.5 These methods predict that nucleobase-like com-
plexes incorporating hydrogen bond arrays with favorable
However, in situations where adjacent hydrogen bond
donorꢀacceptor pairs may influence each other through
† The University of Western Ontario.
‡ Memorial University of Newfoundland.
§ University of Ottawa.
(3) (a) Zimmerman, S. C.; Corbin, P. S. Struct. Bonding (Berlin)
2000, 96, 63. (b) Cooke, G.; Rotello, V. M. Chem. Soc. Rev. 2002, 31, 275.
(c) Sessler, J. L.; Jayawickramarajah, J. Chem. Commun. 2005, 1939. (d)
Sivakova, S.; Rowan, S. J. Chem. Soc. Rev. 2005, 34, 9. (e) Wilson, A. J.
Soft Matter 2007, 3, 409.
(4) Jorgensen, W. L.; Pranata, J. J. Am. Chem. Soc. 1990, 112, 2008.
(5) (a) Pranata, J.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem.
Soc. 1991, 113, 2810–2819. (b) Sartorius, J.; Schneider, H.-J. Chem.;
Eur. J. 1996, 2, 1446. (c) Quinn, J. R.; Zimmerman, S. C.; Del Bene, J. E.;
Shavitt, I. J. Am. Chem. Soc. 2007, 129, 934.
(1) (a) Jeffrey, G. A.; Saenger, W. Hydrogen Bonding in Biological
Structures; Springer-Verlag: Berlin, 1991. (b) Jeffrey, G. A. An Introduction
to Hydrogen Bonding; Oxford University Press: New York, 1997. (c)
Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural
Chemistry and Biology; Oxford University Press Inc.: New York, 1999.
(2) (a) Hydrogen Bonding - New Insights; Grabowski, S. J., Ed. In
Series Challenges and Advances in Computational Chemistry and Physics;
Leszczynski, J., Ed.; Springer: New York, 2006. (b) Abraham, M. H. J. Org.
Chem. 2001, 66, 3484. (c) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43,
5310.
r
10.1021/ol302803j
Published on Web 11/06/2012
2012 American Chemical Society

secondary interactions (e.g contiguous (AA)n and (DD)n
sequences)6 will be more stable than those containing the
opposite arrangement (e.g., alternating (AD)n sequences).
Complexes formed from the planar heterocycles de-
scribed above generally contain an approximately copla-
nar array of hydrogen bond donor and acceptor groups.
We haverecently described the formation ofdouble-helical
complexes from oligomers of pyridine, thiazine-1,1-dioxide,
and indole-based heterocycles that self-assemble through
complementary hydrogen bonding arrays.7 In contrast to
planar arrays like the nucleobases, the adjacent hydrogen
bond donorꢀacceptor pairs in our double helical examples
are oriented at an approximately 90° angle to one another
when viewed down the axis of the double helix (i.e., they
have a pitch of approximately 4 heterocycles/helical turn).
This lead to an assumption that secondary hydrogen bond
interactions would likely be mimimal or nonexistent due to
this orthogonality. Much to our surprise, even though the
two examples pictured in Figure 1 are both assembled
through four primary hydrogen bonds, the overall
stabilities of theirdimersinsolutionareremarkablydifferent
can potentially interact in solution to form a complemen-
tary DD/AA complex (2•3) and a self-complementary AD/
DA dimer (7•7). If secondary interactions between adjacent
donorꢀacceptor pairs influence the stability of the resulting
duplexes, then complex 2•3 (containing two attractive second-
ary interactions; blue arrows) should be more stable than 7•7
(containing two repulsive secondary interactions; red arrows).
The model compounds were designed specifically so that
a direct comparison could be made between the complex
2•3 and the dimer 7•7. The structures of 2, 3, and 7 are
therefore isomeric only in the connectivity between their
heterocyclic rings. Methyl groups were installed in posi-
tions ortho- to the connection between the pyrzidyl and
thiazine rings of 7 to strictly prevent any attractive intra-
molecular interactions via NꢀH N hydrogen bonding
3 3 3
that might affect the dimerization equilibrium.7d
Scheme 1. Synthesis of the DD (2), AA (3), and AD (7) Model
Compounds and Their Intended Hydrogen Bonded Complexes
(Kdimer = 5 and 5700 Mꢀ1;ΔG=ꢀ4.0 and ꢀ21.4 kJ molꢀ1
)
and not simply accounted for by the relatively small differ-
ences in donor/acceptor character of their hydrogen bond-
ing subunits.7a,d However, this disparity could be a result of
their different sequences of donors and acceptors (ADADA
versus AADD). The accumulation of these and other similar
results have led us to postulate that secondary hydrogen
bond interactions may have a significant influence on com-
plex stability in this type of twisted arrangement too. Herein,
we examine the extent of this putative influence using
identical hydrogen bond donor/acceptor heterocycles in
the comparator molecules.
Figure 1. Two double-helical complexes stabilized by four pri-
mary hydrogen bonds (Kdimer measured in CDCl3 at rt). Arrows
indicate potentially attractive (blue) and repulsive (red) second-
ary interactions.
(6) (a) Murray, T. J.; Zimmerman, S. C. J. Am. Chem. Soc. 1992, 114,
4010. (b) Bell, D. A.; Anslyn, E. V. Tetrahedron 1995, 51, 7161. (c) Djurdjevic,
S.; Leigh, D. A.; McNab, H.; Parsons, S.; Teobaldi, G.; Zerbetto, F. J. Am.
Chem. Soc. 2007, 129, 476. (d) Blight, B. A.; Camara-Campos, A.; Djurdjevic,
S.; Kaller, M.; Leigh, D. A.; McMillan, F. M.; McNab, H.; Slawin, A. M. Z.
J. Am. Chem. Soc. 2009, 131, 14116. (e) Blight, B. A.; Hunter, C. A.; Leigh,
D. A.; McNab, H.; Thomson, P. I. T. Nat. Chem. 2011, 3, 244.
The simplest approach to establishing the impact of
secondary interactions in this system is to examine the sta-
bility of minimal doubly hydrogen bonded analogues. We
therefore synthesized three model compounds (Scheme 1) that
Org. Lett., Vol. 14, No. 22, 2012
5773

1H NMR spectroscopy was used to examine both com-
plex and dimer formation in CDCl3 solution at 298 K. The
DD array 2 was titrated with AA array 3, and the changes
in the chemical shift of the NH proton resonances were
monitored during the addition (Figure 2). The data were
fit to a 1:1 binding model8 using nonlinear least-squares
regression to provide an association constant Ka = 375 Mꢀ1
(ΔG = ꢀ3.51 kcal molꢀ1). In a similar manner, a solution of
7 was diluted from an initial concentration of 4 M, and the
change in the chemical shift of the NH resonance was
observed. The data was fit to a 1:1 dimerization model8
to give a dimerization constant Kdimer = 0.4 Mꢀ1 (ΔG =
þ0.55 kcal molꢀ1). The extreme solubility of 7 in chloro-
form permits measurements that describe the majority
of the complexation isotherm, even at this low value of
stabilized by hydrogen bonds between the two N acceptors
˚
of 3 and the NH donors of 2 (NꢀH N = 2.86 A,
3 3 3
NꢀH N = 139°). The hydrogen bondgeometryis likely
3 3 3
distorted away from the linear ideal by the steric influence
of the methyl substituents on the two molecules. This
inference is supported by the interheterocyclic dihedral
angles observed in 2 and 3 (HNꢀCꢀCꢀNH = 103° and
NꢀCꢀCꢀN = 79° respectively) that preclude a more
linear arrangement of the donor/acceptor groups while
maintaining both hydrogen bond interactions.
To investigate the origin of this difference computation-
ally, we constructed dimer and monomer structures of 2•3
and 7•7 based on the crystallographic structure. Disper-
sion corrected density functional theory (DFT-D3) using
the B3LYP functional9 and the def2-TZVPP basis set10
was used to calculate the interaction energy of these dimers
in their minimum energy geometries (see Supporting
Information). The calculated 2•3 dimerization energy
was ꢀ22.8 kcal molꢀ1 while the 7•7 dimerization energy
was ꢀ18.5 kcal molꢀ1. The difference in dimerization
energies ΔΔE = 4.34 kcal molꢀ1 is remarkably close to
the experimental ΔΔG of 4.06 kcal molꢀ1. Although
solvent, entropic, and zero-point energy corrections are
not included in the calculated energies, we would expect
these corrections to be similar for the two dimers since they
only differ in the connectivity of the heterocyclic rings.
Further the dispersion interactions for the two dimers were
calculated to be within 0.56 kcal molꢀ1 of one another
(ꢀ19.35 and ꢀ18.79 kcal molꢀ1 for 2•3and 7•7, respectivley).
Thus, the correlation of this dimerization energy to the
relative dimerization free energies supports the hypothesis
that the difference in dimerization behavior of these com-
pounds corresponds directly to the molecular interaction
energies.
Kdimer, allowing a confident evaluation of the dimerization
equilibrium.
Figure 2. Calculated (curves) and experimental (b) isotherms
measured for the complexation of 2 with 3 (top) and the
dimerization of 7 (bottom).
The structure of complex 2•3 was confirmed in the solid
state using X-ray crystallography (Figure 3). Single crys-
tals were grown by the slow diffusion of isopropyl ether
into a chloroform solution of a 1:1 mixture of 2 and 3. The
complex displays C2 symmetry along an axis bisecting the
bonds connecting the heterocycles in both molecules and is
Figure 3. Stick representation of the X-ray crystal structure of
complex 2•3 looking down the C2 axis with intermolecular
hydrogen bonds indicated as dashed orange lines. All CꢀH
hydrogen and methyl carbon atoms have been removed for
clarity.
(7) (a) Li, J.; Wisner, J. A.; Jennings, M. C. Org. Lett. 2007, 9, 3267.
(b) Wang, H.-B.; Mudraboyina, B. P.; Li, J.; Wisner, J. A. Chem.
Commun. 2010, 46, 7343. (c) Wang, H.-B.; Mudraboyina, B. P.; Wisner,
J. A. Chem.;Eur. J. 2012, 18, 1322. (d) Mudraboyina, B. P.; Wisner,
J. A. Chem.;Eur. J. 201210.1002/chem.201201668.
(8) Bisson, A. P.; Carver, F. J.; Eggleston, D. S.; Haltiwanger, R. C.;
Hunter, C. A.; Livingstone, D. L.; McCabe, J. F.; Rotger, C.; Rowan,
A. E. J. Am. Chem. Soc. 2000, 122, 8856.
(9) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(10) Weigend, F.; Furche, F.; Ahlrichs, R. J. Chem. Phys. 2003, 119,
12753.
5774
Org. Lett., Vol. 14, No. 22, 2012

Secondary interactions in hydrogen bonded donar/
acceptor systems have primarily been attributed to electro-
static interactions between neighboring hydrogen bond
pairs.4 To analyze the origin of the difference in interaction
energiesobserved between2•3 and7•7 weassignedNatural
Population Analysis (NPA) charges in structures 2, 3,
and 7. The charges of the N and H atoms of the hydrogen
bond donor NꢀH bonds were approximately ꢀ0.52 eꢀ1
and 0.42 eꢀ1 respectively in 2 and 7, while the hydrogen
bond accepting pyridine nitrogen atoms were approxi-
mately ꢀ0.42 eꢀ1 for both structures 3 and 7, indicating
that the electronic characteristics of the hydrogen bonding
sites of these compounds are constant and differences in
bond polarity or anionicity cannot explain the differences
in dimerization energy.
dimerization energies of 2•3 and 7•7, although the trend
and origin of these differences are apparent through this
decomposition.
Table 1. Decomposition of the Interaction Energies of
Complexes 2•3 and 7•7 in Terms of Primary, Secondary,
Net Hydrogen Bond and Total Electrostatic Energies
complexation energy component
(kcal molꢀ1
)
2•3
7•7
primary hydrogen bond
interactions
ꢀ8.31
ꢀ8.00
secondary hydrogen bond
interactions
6.00
16.2
net hydrogen bond interactions
total electrostatic interactions
ꢀ2.31
ꢀ15.8
8.20
ꢀ11.0
We then calculated the Coulombic interaction energy
between the fragments in the two dimers using the assigned
NPA charges (Table 1). The primary H-bonded interac-
tions, defined as those involving only the NꢀH bond and
the N atom of the acceptor, are similar in the two dimers
(ꢀ8.31 and ꢀ8.00 kcal molꢀ1 in 2•3 and 7•7, respectively).
This reflects that the charges and geometries of the hydro-
gen bonds in these two dimers are very similar. The
secondary H-bonding electrostatic interactions between
the atoms of adjacent hydrogen bonding sites show much
sharper differences. The secondary interactions of the
7•7 dimer are 10.2 kcal molꢀ1 more repulsive than in the
2•3 dimer. This reflects that there is a strongly replusive
interaction between the NꢀH bonds of the opposing
monomer of 7 that is not present in the 2•3 dimer. The
sum of the primary and secondary hydrogen bonding
interactions shows that the net strength of the electrostatic
interactions at the hydrogen bonding centers is much
larger in 2•3 than in 7•7. This difference is partially can-
celed by electrostatic interactions involving other atoms, as
the difference in the total electrostatic interactions of the
twodimersis only4.80kcal molꢀ1. This purelyelectrostatic
model of the dimerization overestimates the difference in
Our analysis is consistent with the results presented by
Jorgensen and Pranata,4 which attributed the difference in
dimerization energies of nucleic acid pairs to secondary
interactions between neighboring hydrogen bond acceptor
and donor sites of the opposing monomer. The sharp
difference observed in the dimerization energies of AA/
DD complex 2•3 and the AD/AD complex 7•7 is in
accordance with this secondary interaction hypothesis,
indicating that this model can be used as a design principle
in hydrogen bonded supramolecular dimers with large
twist angles, such as these.
Acknowledgment. J.A.W. and J.L. would like to thank
NSERC for funding this research.
Supporting Information Available. Details of synthetic
procedures, characterization, spectroscopic data, theore-
tical methods, and calculated complex structures. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 22, 2012
5775