Stabilisation of 2,6-diarylpyridinium cation by through-space polar-π interactions

Article abstract of DOI:10.1002/chem.201304462

The through-space polar-π interactions between pyridinium ion and the adjacent aromatic rings in 2,6-diarylpyridines affect the pK_avalues. Hammett analysis illustrates that the basicity of pyridines correlates well with the sigma values of the

Full text of DOI:10.1002/chem.201304462

DOI: 10.1002/chem.201304462
Communication
&
Aromatic Interactions
Stabilisation of 2,6-Diarylpyridinium Cation by Through-Space
Polar–p Interactions
Joan Simꢀ Padial,[a] Renꢁ de Gelder,[a] Cꢁlia Fonseca Guerra,[b] F. Matthias Bickelhaupt,[a, b] and
[a]
´
Jasmin Mecinovic*
substituent effect on the pyridine system is based on through-
bond effect; either via induction or resonance effects, or a com-
bination of both. In contrast, stabilisation of pyridines (or pyri-
dinium cations) via through-space effect has not been com-
monly a subject of investigations employing rigorous physical-
organic chemistry approaches. Recently, it was determined
that (2,6-pyridino)paracyclophane that contains an aryl group
in front of the pyridine’s nitrogen is about two orders of mag-
nitude more basic than cyclophane that bears the tetrafluoro-
aryl group at the same position; this result can be in part at-
tributed to through-space effect.[5]
Abstract: The through-space polar–p interactions be-
tween pyridinium ion and the adjacent aromatic rings in
2,6-diarylpyridines affect the pKa values. Hammett analysis
illustrates that the basicity of pyridines correlates well
with the sigma values of the substituents at the para posi-
tion of the flanking aryl rings.
Weak non-covalent interactions dominate biological molecular
recognition events, including enzyme catalysis, DNA double
helix structure, protein folding, protein–protein interactions
and association of proteins and ligands.[1] Among functionali-
ties that often participate in molecular recognition are aromat-
ic rings, which provide a dominant stabilising effect for many
polar functional groups via polar–p interactions (i.e., energeti-
cally-favourable interaction between polar group and p system
of the aromatic ring).[2] In this regard, the most detailed struc-
tural and energetics studies were performed on small molecule
systems that involve interactions between aromatic rings and
diverse sets of functional groups, including alkyl, aryl,
perfluoroaryl, thiol, hydroxyl, carboxyl, silyl, borenium and vari-
ous cations and anions.[3] The mechanisms by which substitu-
ents on the aryl group affect polar groups include, depending
on the substitution pattern of the system, contributions from
through-bond and/or through-space effects.
We have envisaged that the basicity of pyridines that pos-
sess two flanking aryl groups might be perturbed by substitu-
ents positioned at the distinct para position of these rings via
the mechanism by which through-space polar–p interactions
provide a dominant contribution. 2,6-Diarylpyridines 1–6 were
synthesised from 2,6-dibromopyridine and meta- or para-sub-
stituted bromoxylenes under initial Grignard reaction, followed
by Kumada coupling (Scheme 1).
Early studies on the substituted pyridines demonstrated that
pyridine’s basicity can be significantly altered in the presence
of substituents located at the ortho, meta or para positions of
the aromatic ring.[4] In addition to causing different pKa values,
substituted pyridines also exhibit substantially different reactiv-
ities relative to unsubstituted pyridine towards electrophiles in
the nucleophilic substitution type reactions. Most, if not all, of
these studies suggested that the underlying mechanism of the
Scheme 1. Synthesis of 2,6-diarylpyridines 1–6 under Grignard–Kumada con-
ditions.
Potentiometric methods to measure pKa values in different
solvents were found inapplicable, because either starting 2,6-
diarylpyridines or their hydrochloride salts were insoluble in
these media. Another method, that was previously used for
the determination of differences of pKa values of pyridines and
imidazoles in DMSO by 1H NMR spectroscopy, was found to
work very well.[5,6] This NMR method utilizes the titration of
triflic acid into a mixture of two pyridines of interest and pro-
vides information about the differences in their pKa values (i.e.
DpKa). Our hypothesis that the substituents at the para posi-
tion of the flanking aryl rings influence the pKa values of pyri-
dines in a predictable manner can be tested in the presence of
the standard pyridine base. Measurements of pyridines 1–6 in
the presence of 2,6-dimethylpyridine (lutidine) as an internal
reference gave poor results, presumably because the differ-
ence in the pKa values between 2,6-diarylpyridines and lutidine
´
[a] J. S. Padial, Dr. R. de Gelder, Prof. Dr. F. M. Bickelhaupt, Dr. J. Mecinovic
Institute for Molecules and Materials
Radboud University Nijmegen
Heyendaalseweg 135, 6525 AJ Nijmegen (The Netherlands)
Fax: (+31) 24-365-3393
E-mail: j.mecinovic@science.ru.nl
[b] Dr. C. Fonseca Guerra, Prof. Dr. F. M. Bickelhaupt
Department of Theoretical Chemistry
and Amsterdam Center for Multiscale Modeling
VU University Amsterdam
De Boelelaan 1083, 1081 HV Amsterdam (The Netherlands)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304462.
Chem. Eur. J. 2014, 20, 6268 – 6271
6268
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
was too large for a good determination of DpKa (in competi-
tion experiments, shifts of 4-CH proton was only observed for
lutidine, but not for the set of pyridines 1–6). 2,6-Di-tert-butyl-
pyridine, a bulkier analogue of lutidine, is about two orders of
magnitude (DpKa =2.2) less basic than lutidine in 50% aque-
ous ethanol.[7] Thus, we have conceived that 2,6-di-tert-butyl-
pyridine might be a better reference for our set of pyridines,
due to smaller differences in their basicity/acidity constants.
Competitive titrations between an individual pyridine in the
series of 1–6 and 2,6-di-tert-butylpyridine gave excellent re-
sults, as exemplified by a linear correlation (R2 >0.97) between
the shift of the aromatic CH protons in pyridines 1–6 and the
Attempts to obtain a direct evidence for polar–p interactions
were then carried out using X-ray crystallography. Pyridinium
complex 3·HCl crystallised from methanol at room tempera-
ture, whereas crystals of 3·HClO4 and 3·HSO3(OEt) were ob-
tained from ethanol at 48C.[8] Crystallographic analyses of all
three salts illustrate that both flanking rings at the ortho posi-
tion of the pyridinium ion display dihedral angles between 618
and 828, and that both rings are approximately in ‘‘antiparallel’’
or ‘’staggered’’ positions (Figure 2A and Supporting Informa-
1
reference by H NMR (Table 1, see the Supporting Information).
The plot that shows a dependence of pKa of pyridines 1–5 on
the 2s illustrates that the acidity constant of the pyridinium
ion is in a strong correlation (R2 =0.99) with the Hammett
sigma values giving the 1 value of 1.1 (Figure 1). Overall these
Table 1. pKa values for pyridines 1–6.
[a]
Compound
X
s
pKa
1
2
3
4
5
6
H
0.00
2.40
3.04
2.83
2.26
1.99
3.02
p-OMe
p-Me
p-F
p-Cl
m-OMe
À0.27
À0.17
0.06
0.23
0.12
Figure 2. A) Crystal structure of 3·HCl. B) Calculated most stable conforma-
tion of pyridinium cation 3. Dihedral angle f is calculated from N-C2-Ca-Cb.
[a] Determined in [D6]DMSO.
tion). Average distances between the pyridinium proton and
carbon atoms on the neighbouring rings are: NH-Ca 2.5 ꢃ, NH-
Cb 3.3 ꢃ, NH-Cg 4.4 ꢃ, and NH-Cd 4.9 ꢃ. These data support the
view that the substituents at the para position (i.e., Cd) do not
directly interact with the pyridinium ion, and that the rings’
proximity and orientation stabilises the pyridinium ion. All
three crystal structures also suggested that in the crystal form
there is an interaction between pyridinium NH and the anion.
Distances between pyridinium NH proton and the anionic spe-
cies were 2.14 ꢃ for ClÀ, 1.98 ꢃ for oxygen in ClO4À, and 1.78 ꢃ
for oxygen in HSO3(OEt)À. These observations are in agreement
with the structure of the salt between cationic 3 and tetra-
chloro gallate anion.[9] Our crystal structures also revealed that
solvent molecules do not interact with the cation.
Computational studies on our pyridine system, using the
Amsterdam Density Functional (ADF) program, support the in-
terpretation of our experimental findings.[10] The optimisation
of the geometry for pyridinium ion 1, computed at BP86/TZ2P,
provided the conformation of the energetically most stable
form. The dihedral angle between the pyridinium C2 and the
adjacent aryl ring (i.e., N-C2-Ca-Cb) for the energetically most
stable structure was found to be 658 for a set of six pyridines
(Figure 2B). Our computed angles for the most stable confor-
mation are compatible with those derived from crystal struc-
tures. Importantly, the computations reveal that the potential
energy well, associated with varying the dihedral angle around
the absolute minimum at 658, is extremely shallow. Thus,
whereas conformers with angles below 308 are energetically
Figure 1. Dependence of pKa values on the Hammett sigma values of para-
substituted pyridines 1–5. 2s is the sum of the Hammett sigma values of
para substituents on both flanking rings.
results demonstrate that pyridines that contain electron-donat-
ing groups (e.g. OMe, Me) are significantly more basic than
pyridines that bear electron-withdrawing groups (e.g. Cl), and
that there is a linear relationship between the acidity and the
Hammett sigma values. Notably, the pyridine that possesses
a OMe substituent at the meta position (6) exhibits a similar
pKa value as the para analogue (2), despite the value of smeta
(0.12) being significantly different from the value of spara
(À0.27) for this substituent.
Chem. Eur. J. 2014, 20, 6268 – 6271
www.chemeurj.org
6269
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Figure 3. Relative energy DE of pyridinium cation 1 in the gas phase as
a function of the dihedral angle f, computed at BP86/TZ2P.
unfavourable, the difference in energy between conformers
that possess dihedral angles between 458 and 908 was calcu-
lated to be only 0.4 kcalmolÀ1 (Figure 3). Similarly, pyridinium
ions 4 and 5 were calculated to be energetically stable be-
tween 458 and 908; the difference in energy between conform-
ers in this range was found to be 0.6 kcalmolÀ1 (see the Sup-
porting Information).
Proton affinity energies (DEPA) for a set of pyridines were
then calculated in the gas phase and in DMSO, again at BP86/
TZ2P and using COSMO to simulate the effect of solvation
(Table 2). Dependence of proton affinity energies in the gas
Figure 4. Dependence of proton affinity energies (DE) in the gas phase (A)
and in DMSO (B) on the Hammett sigma values of para-substituted pyridines
1–5. 2s is the sum of the Hammett sigma values of para substituents on
both flanking rings.
Table 2. Proton affinity energies DEPA [in kcalmolÀ1] computed at BP86/
TZ2P.
not alter the structural feature of the pyridinium cation pro-
vides direct information about the impact of the electronic
effect, and eliminates the contribution on the basicity from the
solvent effect (we do, however, note that solvent has a substan-
tial effect on basicity of pyridines as exemplified by the large
difference in calculated proton affinities in the gas phase rela-
tive to DMSO; see above). A resonance effect is excluded, be-
cause the system is not planar as demonstrated by both, crys-
tallographic and computational studies. Furthermore, a reso-
nance effect is also eliminated by the observation that pyri-
dines 2 and 6 that bear meta and para-substituted methoxy
group, respectively, have the same basicity. Contributions from
the inductive effect are also excluded, because through-bond
effects diminish with the number of bonds (there are five
bonds between pyridinium NH and the aryl Cd). Similarly, in-
volvement of the steric effect is eliminated due to a substantial
distance (>5 ꢃ) between the NH and the substituent at the
para position whereas we nevertheless observe a significant
dependence of the basicity of the set of pyridines 1–5 on the
electronic properties of the para substituent. Our results, a de-
creasing basicity with increasing Hammett sigma values, are
compatible with the mechanism by which the field effect and
the polarisability effect via through-space polar–p interactions
determine the stability and basicity of 2,6-diarylpyridines.
Compound
X
DEPA in gas phase
DEPA in DMSO[a]
1
2
3
4
5
6
H
238.6
244.5
241.2
236.4
234.4
243.0
22.3
23.9
22.8
22.2
21.9
22.1[b]
p-OMe
p-Me
p-F
p-Cl
m-OMe
[a] Solvation in DMSO is simulated using COSMO. [b] Lower DEPA for 6
than for 2 may be ascribed to discontinuous change in COSMO cavity for
meta pyridine 6 as compared to para pyridines 1–5.
phase on the Hammett sigma values shows a strong linear re-
lationship with the slope of À11.0 (Figure 4A). In DMSO, the
slope is significantly shallower with a value of À1.8 (Figure 4B).
These results are compatible with the picture that the NH
cation strongly interacts with electron-rich aromatic rings at
the adjacent positions, and that this interaction is more pro-
found in the gas phase than in DMSO.
In principle, six mechanisms can be involved in the stabilisa-
tion of pyridinium cation by the adjacent aryl groups: 1) reso-
nance effect, 2) inductive effect, 3) field effect, 4) polarisability
effect, 5) steric effect and 6) solvation effect. Our analysis that
1) compares all sets of pyridines relative to a reference under
the same experimental conditions, and 2) the fact that a sub-
stituent at the distant para position of the flanking rings does
We have demonstrated that the acidity of 2,6-diarylpyridini-
um cations that bear two flanking aryl rings is substantially in-
fluenced by substituents positioned at the distant para posi-
Chem. Eur. J. 2014, 20, 6268 – 6271
www.chemeurj.org
6270
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Vinter, K. R. Lawson, C. J. Urch, C. A. Hunter, Org. Biomol. Chem. 2007, 5,
1062–1080; c) F. Cozzi, R. Annunziata, M. Benaglia, M. Cinquini, L. Rai-
mondi, K. K. Baldridge, J. S. Siegel, Org. Biomol. Chem. 2003, 1, 157–
162; d) K. S. Kim, P. Tarakeshwar, J. Y. Lett, Chem. Rev. 2000, 100, 4145–
4185; e) C. T. Chen, J. S. Siegel, J. Am. Chem. Soc. 1994, 116, 5959–5960;
f) S. Duttwyler, Q. Q. Do, A. Linden, K. K. Baldridge, J. S. Siegel, Angew.
Chem. 2008, 120, 1743–1746; Angew. Chem. Int. Ed. 2008, 47, 1719–
1722; g) S. L. Cockroft, C. A. Hunter, K. R. Lawson, J. Perkins, C. J. Urch, J.
Am. Chem. Soc. 2005, 127, 8594–8595; h) C. A. Hunter, C. M. Low, C.
Rotgers, J. G. Vinter, C. Zonta, Proc. Natl. Acad. Sci. USA 2002, 99, 4873–
4876; i) H. B. Mansaray, A. D. L. Rowe, N. Phillips, J. Niemeyer, M. Kelly,
D. A. Addy, J. I. Bates, S. Aldridge, Chem. Commun. 2011, 47, 12295–
12297; j) J. C. Ma, D. A. Dougherty, Chem. Rev. 1997, 97, 1303–1324;
k) A. Frontera, P. Gamez, M. Mascal, T. J. Mooibroek, J. Reedijk, Angew.
Chem. 2011, 123, 9736–9756; Angew. Chem. Int. Ed. 2011, 50, 9564–
9583.
tion of these two rings. In this set of pyridines, the pyridinium
NH cation is stabilised by aromatic rings via through-space
polar–p interactions caused by the neighbouring aryl groups.
In depth understanding of the molecular level mechanism by
which through-space interactions (either intramolecular or in-
termolecular) stabilise polar functionalities might be useful in
designing small-molecule ligands and inhibitors that bind to
proteins specifically and with high affinity. In this respect, our
study demonstrates that pyridinium ions favourably interact
with electron-rich aromatic rings that are located in close prox-
imity to pyridinium NH (<5 ꢃ) and possess a proper orienta-
tion (T-shape or distorted T-shape).[11] Because pyridine deriva-
tives construct the skeleton of many drugs, it is important to
consider that the pyridinium ion (as substituted pyridines can
partially exist in protonated form under physiological condi-
tions) can be stabilised by aromatic amino acids (i.e., Phe, Tyr,
Trp) that constitute the binding sites of several biologically im-
portant proteins.
[4] a) A. Fischer, W. J. Galloway, J. Waughan, J. Chem. Soc. 1964, 3591–
3596; b) D. H. Aue, H. M. Webb, M. T. Bowers, C. L. Liotta, C. J. Alexander,
H. P. Hopkins, J. Am. Chem. Soc. 1976, 98, 854–856; c) A. R. Katritzky,
D. E. Leahy, J. Chem. Res. (S) 1985, 28–29; d) C. D. Johnson, I. Roberts,
P. G. Taylor, J. Chem. Soc. Perkin Trans. 2 1981, 409–413.
[5] a) K. K. Baldridge, F. Cozzi, J. S. Siegel, Angew. Chem. 2012, 124, 2957–
2960; Angew. Chem. Int. Ed. 2012, 51, 2903–2906; b) R. Annunziata, M.
Benaglia, F. Cozzi, A. Mazzanti, Chem. Eur. J. 2009, 15, 4373–4381.
[6] C. L. Perrin, M. A. Fabian, K. B. Armstrong, J. Org. Chem. 1994, 59, 5246–
5253.
Acknowledgements
[7] H. C. Brown, B. Kanner, J. Am. Chem. Soc. 1953, 75, 3865–3865.
[8] CCDC 958095, 958096 and 958097 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of charge
We thank Radboud University Nijmegen, the Netherlands Or-
ganization for Scientific Research (NWO-CW), and the National
Research School Combination Catalysis (NRSC-C) for financial
support, Jan M. M. Smits for assistance with crystallography,
and Professor Alan E. Rowan for helpful discussions.
from
The
Cambridge
Crystallographic
Data
Centre
via
www.ccdc.cam.ac.uk/data_request/cif.
[9] T. Pell, D. P. Mills, A. J. Blake, W. Lewis, S. T. Liddle, Polyhedron 2010, 29,
120–125.
[10] a) G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca Guerra, S. J. A.
van Gisbergen, J. G. Snijders, T. Ziegler, J. Comput. Chem. 2001, 22, 931–
967; b) Computer code ADF 2013.01: see http://www.scm.com.; c) M.
Swart, F. M. Bickelhaupt, J. Chem. Theory Comput. 2006, 2, 281–287;
d) M. Swart, E. Rçsler, F. M. Bickelhaupt, J. Comput. Chem. 2006, 27,
1486–1493; e) A. Klamt, J. Phys. Chem. 1995, 99, 2224–2235; f) M.
Swart, E. Rçsler, F. M. Bickelhaupt, Eur. J. Inorg. Chem. 2007, 3646–3654.
[11] S. Tsuzuki, M. Mikami, S. Yamada, J. Am. Chem. Soc. 2007, 129, 8656–
8662.
Keywords: basicity
interactions · polar–p interactions · pyridines
·
Hammett analysis
·
non-covalent
[1] A. Karshikoff in Non-Covalent Interactions in Proteins, Imperial College
Press, London, 2006.
[2] a) E. A. Meyer, R. K. Castellano, F. Diederich, Angew. Chem. 2003, 115,
1244–1287; Angew. Chem. Int. Ed. 2003, 42, 1210–1250; b) L. M. Salo-
nen, M. Ellermann, F. Diederich, Angew. Chem. 2011, 123, 4908–4944;
Angew. Chem. Int. Ed. 2011, 50, 4808–4842; c) S. E. Wheeler, Acc. Chem.
Res. 2013, 46, 1029–1038.
Received: November 14, 2013
Revised: March 25, 2014
[3] a) I. K. Mati, S. L. Cockroft, Chem. Soc. Rev. 2010, 39, 4195–4205; b) S. L.
Cockroft, J. Perkins, C. Zonta, H. Adams, S. E. Spey, C. M. R. Low, J. G.
Published online on April 15, 2014
Chem. Eur. J. 2014, 20, 6268 – 6271
www.chemeurj.org
6271
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim