Electronically tunable anion-π interactions in pyrylium complexes: Experimental and theoretical studies

Article abstract of DOI:10.1039/c4cp01977f

Noncovalent interactions of anions with electron-deficient aromatic rings that have been studied so far involve non-heteroaromatic or nitrogen-based heteroaromatic systems. Here we report the first case of an organic oxygenated aromatic system, in particular the tri-aryl-pyrylium tetrafluoroborate system, for which noncovalent anion-π interactions of the pyrylium cation with the tetrafluoroborate anion have been experimentally detected and demonstrated by means of ¹⁹F NMR spectroscopy in solution. A series of pyrylium tetrafluoroborate salts were synthesized in the presence of BF ₃·Et₂O, by direct reaction of 4-substituted benzaldehydes with 4-substituted acetophenones or via the previously obtained chalcone of the less reactive ketone. Correlations of ¹⁹F NMR chemical shifts of tetrafluoroborate anion for most of the synthesized tri-arylpyrylium tetrafluoroborate complexes with both the pyrylium cation molecular weight and the standard substituent Hammett constants, demonstrate anion-π⁺ interaction to act between the polyatomic anion BF ₄^- and the pyrylium aromatic system. DFT calculations reveal that an additional (C-H)⁺-anion hydrogen bond involving the H(5) of pyrylium ring exists for these fluorescent dyes that show a tunable cup-to-cap shape cavity. The strong fluorescence emission observed for some representative pyrylium tetrafluoroborates described herein, makes them a promising class of tunable emission wavelength dyes for laser technology applications. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cp01977f

PCCP
View Article Online
View Journal | View Issue
PAPER
Electronically tunable anionꢀp interactions in
pyrylium complexes: experimental and
theoretical studies†
Cite this: Phys.Chem.Chem.Phys.,
2014, 16, 18442
Antonio Franconetti,a Lidia Contreras-Bernal,a Sorel Jatunov,a Manuel Gomez-
´
Guillen,a Manuel Angulo,b Rafael Prado-Gotorc and Francisca Cabrera-Escribano*a
´
Noncovalent interactions of anions with electron-deficient aromatic rings that have been studied so far
involve non-heteroaromatic or nitrogen-based heteroaromatic systems. Here we report the first case of
an organic oxygenated aromatic system, in particular the tri-aryl-pyrylium tetrafluoroborate system, for
which noncovalent anion–p interactions of the pyrylium cation with the tetrafluoroborate anion have
been experimentally detected and demonstrated by means of 19F NMR spectroscopy in solution. A series
of pyrylium tetrafluoroborate salts were synthesized in the presence of BF3ꢁEt2O, by direct reaction of
4-substituted benzaldehydes with 4-substituted acetophenones or via the previously obtained chalcone
of the less reactive ketone. Correlations of 19F NMR chemical shifts of tetrafluoroborate anion for most
of the synthesized tri-arylpyrylium tetrafluoroborate complexes with both the pyrylium cation molecular
weight and the standard substituent Hammett constants, demonstrate anion–p+ interaction to act
ꢀ
between the polyatomic anion BF4 and the pyrylium aromatic system. DFT calculations reveal that an
Received 7th May 2014,
Accepted 16th July 2014
additional (C–H)+–anion hydrogen bond involving the H(5) of pyrylium ring exists for these fluorescent
dyes that show a tunable cup-to-cap shape cavity. The strong fluorescence emission observed for some
representative pyrylium tetrafluoroborates described herein, makes them a promising class of tunable
emission wavelength dyes for laser technology applications.
DOI: 10.1039/c4cp01977f
www.rsc.org/pccp
including cation–p interactions are widely accepted and classical
noncovalent bond interactions. Less common are the anion
1. Introduction
Noncovalent interactions involving aromatic systems play coordination6 and the anion–p interactions. Nevertheless, since
important roles in chemistry and biology.1 In particular, 2002, when the term anion–p interaction was introduced on the
host–guest and supramolecular systems,2 design of advanced basis of theoretical calculations7 for designating an attractive force
functional materials,3,4 and specific biological recognition,5 (B20 to 70 kJ molꢀ1) between an electron-deficient aromatic
including drug-receptor exchanges and protein folding, are p system and an anion, more than three hundred articles about
the areas of their pivotal significance.
it have been published.8 Currently, the field of anion–p research is
There are different types of reported noncovalent bond inter- highly promising and relevant in the scientific community due to
action in the literature involving both neutral and charged species. the potential application of anion–p interactions for the design of
Electrostatic, hydrogen bonding, p–p stacking, cation coordination anion receptors9,10 and anion-templated supramolecular architec-
tures,11 selective hosts for anions recognition,8a,12 anion carriers,13
catalysts,14 and their presumed involvement in biological func-
a
tions.5a,15,16a In comparison with a number of theoretical calcula-
tions on anion–p interactions,7,16,17 experimental studies of this
exciting noncovalent bond interaction are limited,9 particularly in
solution.18 It is generally accepted that anion–p interactions can be
characterized by anion–ring distances lower than the sum of the
van der Waals radii of the participating atoms and at such a
position that the anion-ring centroid vector is perpendicular to the
plane of the aromatic ring.19,20
´
´
´
Departamento de Quımica Organica, Facultad de Quımica, Universidad de Sevilla,
´
´
C/ Profesor Garcıa Gonzalez 1, 41012 Sevilla, Spain. E-mail: fcabrera@us.es;
Fax: +34 954624960; Tel: +34 954556868
b
c
´
Servicio de Resonancia Magnetica Nuclear, CITIUS, Universidad de Sevilla,
Avda. Reina Mercedes 4, 41012 Sevilla, Spain
´
´
´
Departamento de Quımica Fısica, Facultad de Quımica, Universidad de Sevilla,
´
´
C/ Profesor Garcıa Gonzalez 1, 41012 Sevilla, Spain
† Electronic supplementary information (ESI) available: Experimental details,
chemical shifts and Hammett correlations, UV-visible and fluorescence spectra,
characterization data including 1H and 13C NMR spectra of all new compounds,
19F NMR spectra of studied compounds and cartesian coordinates for some
representative complexes. See DOI: 10.1039/c4cp01977f
A comparison of analogous cation–p and anion–p complexes
generally shows the anion–p distance to be longer and the
18442 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
interaction to be energetically weaker.21 However, a different ring, but it also depends upon the stabilizing effect of the
picture appears when charged aromatic rings participate in anion. In this communication, the behaviour of the tetrafluor-
anion–p bonding. Positive charges can be easily generated in oborate anion in a number of previously synthesized pyrylium
azine rings by simply adjusting the pH of the medium, and this salts has been analyzed by 19F NMR spectroscopy; we have
can be used to increment the anion binding ability of the ring. found small but significant variations on chemical shifts from
Several examples of anion–p complexes between aromatic cations light to heavily modified tri-arylpyrylium cations. From our
(tropylium, quinolizinylium, protonated 2-aminopyrimidine, results, regardless of the classical vision of host–guest inter-
protonated adenine) and various anions have been reported by action, pyrylium tetrafluoroborates can be considered to be
Frontera and co-workers along with crystallographic structures ‘‘symbiotic’’ relationships between cations and tetrafluorobo-
that support the theoretical findings.22 As expected, in these rate anions.
complexes the interaction energies are large (480 kcal molꢀ1
)
and electrostatic effects dominate these denoted as anion–p+
interactions.8e On the other hand, (C–H)+ꢁꢁꢁXꢀ hydrogen bonds
have been reported in several 1,3-disubstituted imidazolium and a
bis(triazolium)-based anion receptors for guest binding.23 This is
a novel type of charged hydrogen bonding wherein the charge–
charge electrostatic interaction dominates, in contrast to the well-
known neutral hydrogen bonds that occur in the most positively
charged anion receptors which form N–Hꢁꢁꢁanion hydrogen
bonds as binding forces.24 Very recently, taking advantage of both
noncovalent anion–p+/(C–H)+ interactions of imidazolium rings
with anions as stabilizing forces for crystal packing, Singh and
co-workers4 have designed and reported some self-assembled
arrays of cone-shaped calyx[n]imidazolium systems (n = 4, 5) as
a new class of positively charged homo-calix compounds for
different applications. Moreover, Caballero et al.23a have described
an unprecedented bidentate triazolium receptor for anion recog-
nition, for which NMR spectroscopic data in solution indicate that
one triazolium ring is acting as a hydrogen bond donor, whereas
the second triazolium ring behaves as an anion–p receptor.
However, despite the progress made in the field in the last few
years, all studies about the participation of charged aromatic rings
in anion–p bonding reported so far involve non-heteroaromatic or
nitrogen-based heteroaromatic systems. Here we report the first
case of an organic oxygenated aromatic system,22a in particular a
pyrylium-based tri-arylated series of compounds, for which non-
covalent anion–p+/(C–H)+ interactions of the pyrylium ring with
the tetrafluoroborate anion have been experimentally detected
and demonstrated.
2. Experimental section
19F NMR spectroscopy
Spectra were acquired on a Bruker Avance 500 MHz spectro-
meter equipped with a 5 mm BBFO broadband probehead
incorporating a z-gradient coil. Spectra were recorded at
300 K in triplicate, and sample concentrations were 20 mM in
all cases. DMSO-d6 containing 0.1% TMS was used as the
solvent. Spectra were acquired using a high-power 901 pulse
of 14.8 ms and a recycle delay of 5 s. The 19F chemical shifts
were referenced indirectly32 to CCl3F (X = 94.094011).
Synthesis of pyrylium tetrafluoroborate salts
Route I
General procedure from an aldehyde and a ketone. BF3ꢁEt2O
(3.1 equiv.) was added dropwise under a N2 atmosphere to a
solution of p-substituted benzaldehyde* (1 equiv.) and p-sub-
stituted acetophenone (3 equiv.). The mixture was heated to
80 1C until the reaction was complete, monitored by TLC (1 : 4
EtOAc–hexane). The residue was eventually treated with EtOH
or Et2O. Crystallization from 1,2-dichloroethane or acetone
gave the corresponding pure pyrylium tetrafluoroborate.
*Note: solid p-substituted benzaldehydes were previously
dissolved in toluene.
Route II
General procedure from a chalcone. BF3ꢁEt2O (3.1 equiv.) was
added dropwise under a N2 atmosphere to a solution of
chalcone (1 equiv.) and p-substituted acetophenone (3 equiv.)
in anhydrous dichloromethane. The mixture was heated to
reflux (45 1C) until the reaction was complete, monitored by
TLC (1 : 4 EtOAc–hexane). Solvents were evaporated to dryness
and the residue was eventually washed with H2O and EtOH.
Crystallization from 1,2-dichloroethane or acetone gave the
corresponding pure pyrylium tetrafluoroborate.
Pyrylium salts are a very important class of cationic organic
molecules containing a trivalent oxygen atom in a six-membered
aromatic ring.25 The high electronegativity of oxygen is the reason
for two experimental observations: (a) pyrylium salts represent the
strongest single modification of the electron distribution caused
by one heteroatom in a benzene ring, and (b) pyrylium ions have
the highest conductivity among the compounds with benzene
rings. A number of references are available for this kind of
compounds such as crystallographic and structural studies26 and
specially their good absorption, fluorescence, and electron accep-
3. Results and discussion
tor behaviour.27 Pyrylium salts can function as intermediates for a A series of pyrylium salts (10–28) with tunable electronic effects
variety of syntheses,28 and have been exploited to design sensors have been synthesized following two approaches (Scheme 1,
for anions,29 amines, amino acids and chameleon labels for routes I and II, Table 1), between their different feasible syn-
quantifying proteins,30 as well as a kind of ionic liquid crystal thetic methods.33 In the case of symmetrically substituted
materials which can be used in the display industry.31
pyrylium salts, these compounds can be directly synthesized
Pyrylium ions are less reactive than other oxonium ions. from 4-substituted aldehydes and ketones (route I). Thus, treat-
Their improved stability is influenced by the aromaticity of the ment of biphenyl-4-carboxaldehyde and 4-fluoroacetophenone
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18443

View Article Online
PCCP
Paper
are observed for these ortho protons in the range of 8.24–8.79 ppm,
the most deshielded one corresponding to two ortho protons of the
phenyl ring A and a second doublet for four protons assigned to
ortho protons on phenyl rings at pyrylium positions 2 and 6 (rings B
and C). When a fluorine atom is present in the phenyl ring
(compounds 15, 17, 18, 23, 25, 27), a double doublet is observed
instead, due to the coupling with fluorine.
Related to non-symmetrical pyrylium derivatives (14, 16, 20,
21, and 24), it was observed that the 1H NMR spectrum in
DMSO-d6 of compound 14 shows a strange lack of peaks as
expected for a non-symmetrical compound. A signal for protons
H-3 and H-5 of pyrylium cation was observed at 9.16 ppm
(a singlet for two protons). Moreover, another main feature of
this spectrum is the presence of one doublet for four protons at
8.60 ppm assigned to the ortho protons in phenyl rings at
positions 4 and 6 (rings A and C) of the pyrylium cation.
Likewise, compound 16 exhibits a multiplet at 8.74–8.66 ppm
for these four ortho protons. Substitution patterns of compounds
14 and 16 were confirmed by 1H–13C HMBC experiments. It was
first suggested that these features could be caused by solvation
effects, but this explanation was discarded when acetone-d6 was
used; in this solvent a singlet for H-3 and H-5 was still observed
as well as two signals for the ortho protons on phenyl rings A and
Scheme 1
Table 1 Pyrylium salts with tunable electronic effects synthesized following
two approaches (Scheme 1, routes I and II)
Compound
Route
R1
R2
R3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
I, II
I
II
II
II
I
II
I
II
I
II
II
I, II
I, II
II
I
II
H
H
Ph
Ph
H
F
F
H
F
NO2
NO2
NO2
NO2
NO2
Ph
Ph
OMe
OMe
OBz
H
Me
H
Me
H
H
F
F
F
H
H
Me
H
Me
F
H
H
F
F
H
H
F
Me
F
F
F
1
F
C, showing virtual H–1H coupling (ESI,† Fig. S11). We propose
that this effect, in solution, may be due to tetrafluoroborate–p+
interaction, in such a way that being a fluorine atom of the anion
closer to H-3 (or H-5) than H-5 (or H-3) of the pyrylium ring, a
change occurs in the expected chemical shift for one of them.
A similar experimental effect for H-3 and H-5 (a singlet at
Me
Me
F
Me
F
Me
F
Me
Me
F
I, II
I
Me 9.14 ppm) is observed for compound 16 but not for the rest of
unsymmetrically substituted pyrylium salts 20, 21, and 24.
13C NMR corroborated the presence of the pyrylium moiety.
with BF3ꢁEt2O in toluene gives 4-biphenyl-2,6-di-(4-fluoro- Thus, on the spectra of symmetrically substituted derivatives a
phenyl)pyrylium tetrafluoroborate 25 in 74% yield. For unsym- signal in the range of 167.9–172.7 ppm is observed for C-2 and
metrically substituted systems, the synthesis of a chalcone C-6; C-4 gives rise to a signal at 164.7–167.1 ppm, and C-3 and
from the less reactive ketone, as a previous step, is required C-5 generate one signal between 113.0 and 117.7 ppm. As
(Scheme 1, route II). In their turn, E-chalcones 1–9 were obtained expected, unsymmetrical derivatives give separated signals for
by Claisen–Schmidt condensation between commercially avail- each of the pairs C-2/C-6 and C-3/C-5 of the pyrylium ring.
able aldehydes and ketones.
To test our hypothesis, anion exchange was induced by
In both strategies, the Lewis acid BF3ꢁEt2O was used to addition of increasing amounts of acid. When compound 14
mediate cyclization and dehydration steps of the 1,5-dicarbonyl was treated in a NMR tube with H2SO4 or HClO4 (4500 equiv.),
intermediate. For example, treatment of chalcone 2 (R1 = Ph, R2 = a downshift of the signal for H-3 and H-5 from 9.16 ppm to give
H) and acetophenone with BF3ꢁEt2O in anhydrous CH2Cl2 gives separate signals at 8.51 and 8.50 ppm for each proton, was
4-biphenyl-2,6-di-phenylpyrylium tetrafluoroborate 12 in 70% observed (Fig. 1). For the pyrylium perchlorate salts, the new
yield. The preparation of pyrylium salts from chalcones is a chemical shifts were in the range described in the literature.41
simple process, but the yields are unequal.
If changes in the aromatic protons of the pyrylium cation
Symmetrically substituted pyrylium salts (10,34 11,35 12,36 can be detected as a consequence of the anion–pyrylium system
13, 15,37 17,35 18,38 19,39 22, 23, 25, 26,40 27, and 28) show in interaction, the corresponding changes must be exhibited in
1
their H NMR spectrum in DMSO-d6 a singlet in the range of the tetrahedral tetrafluoroborate anion. Using 19F NMR spectro-
8.93–9.30 ppm for the two typical equivalent protons H-3 and scopy as a tool for this purpose, we have found a relationship
H-5 of the pyrylium ring. For compound 10, a doublet at between 19F chemical shifts and the molecular weight of the
8.60 ppm (J = 7.5 Hz) for 6 protons is observed due to the ortho pyrylium cation (Fig. 2). From the extensive studies7a,8d,42 of the
protons of three phenyl rings (A, B and C) attached to positions physical nature of the anion–p interactions it was concluded
4, 2 and 6 of the pyrylium ring, respectively. For all the rest of that electrostatic forces, specifically the quadrupole moment of
symmetrically substituted derivatives (compounds 11, 12, 13, the arene and ion-induced polarization, are the main energetic
15, 17, 18, 19, 22, 23, 25, 26, 27, and 28) two different doublets contributors to the anion–p complex. Accordingly, a molecular
18444 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
Table 2 Series of pyrylium salts synthesized and Dd(Hz) observed for each
compound relative to the unsubstituted pyrylium dye 10
Series
a
Compound (X)
DdX/10/dX ꢀ d10(Hz)a
0
0.9
4.1
4.3
10
11
12
13
b
c
14
15
16
17
18
0.9
1.9
3.6
3.9
7.3
19
22
23
3.5
9.4
8.7
Fig. 1 1H NMR titration of compound 14 with H2SO4 in acetone-d6.
d
25
26
27
3.9
5.3
3.7
a
Corrected values for compounds 19, 22 and 23 were 4.3, 9.4 and 10.2,
respectively.
pyrylium oxygen, which will undergo a higher interaction with
the anion. As compared with compound 10, larger chemical
shifts of BF4ꢀ are observed. Indeed, a maximized interaction for
compound 18 containing three F atoms in the para position of
each aromatic ring can be achieved.
The position of each substituent group proves to be significant
for the interaction. The essential feature was the dependence of
the chemical shift value, dF, as a function of fluorine position,
which gives rise to small variations among regioisomers, i.e.,
compounds with the same number of F atoms and Mw. Chemical
shift deviations for symmetrically rather than unsymmetrically
substituted pyrylium compounds take place, as shown for pairs
17/16 and 15/14: (Dd17/16 0.3 Hz) and (Dd15/14 1.0 Hz).
Fig. 2 Tetrafluoroborate anion chemical shift in 19F NMR for different
pyrylium salts in DMSO-d6 (20 mM) for coupling with 11B vs. molecular
weight of the pyrylium cation. In this graph, triangles correspond to
compounds of series (a); circles to series (b); hexagons for series (c); squares
to series (d). For coupling with 10B, see ESI,† Fig. S1.
For the first series (a), an also positive, but lower slope than
those observed for series (b) with increasing Mw is observed.
A weaker contact of the guest with the pyrylium cation is evidenced
weight dependence of the quadrupolar molecular moment43 from the reference state (compound 10, Table 2). Compounds 25
and/or the ion-induced polarization might be expected to and 27 from series (d) correlate acceptably well with this straight
justify our experimental findings.
line, but not 26 (compounds 24 and 28 not studied).
We realized that 19F NMR chemical shifts of tetrafluoroborate
The trend for nitro-compounds (c) was not satisfactory at first
anion in DMSO-d6, for most of the synthesized tri-arylpyrylium because a dF, lower than expected, was found as a consequence of
tetrafluoroborate salts, undergo small but significant changes the spontaneous decrease of the tri-arylpyrylium salt concentration
due to the effect of substituents on the pyrylium cation. To in the sample. These nitro-containing structures evolve to an
illustrate the magnitude of the effect, which proves significant equilibrium state between pyrylium compounds (Scheme 2), for
concentration dependence, we show in Table 2 the Dd(Hz) example 23 and the corresponding substituted pent-2-en-1,5-dione
observed for each case relative to the unsubstituted pyrylium
dye 10. For our study, four series of pyrylium salts synthesized
can be differentiated: (a) compounds 10, 11, 12 and 13, with only
alkyl or aryl substituents; (b) compounds 14, 15, 16, 17 and 18
containing only fluorine atoms as substituent in aromatic rings;
(c) nitro-compounds (19, 20, 21, 22 and 23) with or without F
atoms and (d) hybrid-pyrylium compounds or pyrylium salts 24,
25, 26, 27 and 28 with hybrid electronic effects.
In the second series, (b), due to the electronegativity of the
Scheme 2
fluorine atom, there is an increased positive charge over the
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18445

View Article Online
PCCP
Paper
Table 3 Equilibrium ratio between nitro-containing pyrylium dyes and
the corresponding substituted pent-2-en-1,5-diones (by 1H NMR)
Compounds
R2
R3
Pyrylium/dione ratio
19/19a
20/20a
21/21a
22/22a
23/23a
H
F
Me
Me
F
H
H
F
Me
F
1 : 0.4
1 : 1.4
1 : 0.8
1 : 0.3
1 : 1.3
1
1
23a (23/23a, 1:1.3 by H NMR), so that the H NMR spectrum
in DMSO-d6 now exhibits a singlet for the olefinic proton (H-2) at
7.74 ppm and a singlet at 4.86 ppm assigned to methylene protons
in an a position to the carbonyl group (H-4 and H-40). This ring-
opening process is not complete after several days. According to
the substituent that stabilizes the pyrylium cation, the formation
of the corresponding pent-2-ene-1,5-dione (19a) from compound
19 is less acute (19/19a, 1 : 0.4) than that observed for fluorine-
containing phenyl B and C rings (Table 3). In the case of pent-2-
ene-1,5-dione 22a derived from compound 22, the ratio (22/22a)
Fig. 3 Plot of tetrafluoroborate anion chemical shift in 19F NMR for
different pyrylium salts at a constant concentration of 20 mM in DMSO-
P
d6 (for coupling with 11B) vs. substituent Hammett constants ( sp). For
coupling with 10B, see ESI,† Fig. S2.
was 1 : 0.3, 1H NMR spectrum showing singlets at 7.69 and of the pyrylium ring, in such a way that the anion–p+ interaction
4.85 ppm for H-2 and H-4/H-40, respectively. For all 1,5-diones, is relatively small. In other words, a minor anion contribution is
no perceptible changes in the aromatic regions are observed. required to stabilize the pyrylium system. As a consequence, the
A predictable decrease of dF based on the calculated ratio of sensitivity of the anion–cation interaction to the substituent
pyrylium/dione was observed (Table 3). For these compounds, dF group nature will be limited, as it is deduced in fact from the plot.
exclusively due to the pyrilium contribution has been calculated
However, for only fluorine-containing pyrylium salts [series
taking into account the reported pyrilium/dione ratios and these (b), compounds 14–18] a higher and positive slope results. A
resulting corrected dF values have been employed for being strong electron deficiency is caused now by the substituents on
represented in Fig. 2. The described behaviour of nitro- the pyrylium cation. The electrostatic interaction Oꢁ ꢁ ꢁB is not
compounds can be considered as a strong evidence of ‘‘symbiotic strong enough for maintaining the complex anion–cation in a
relationship’’ of both ions in the pyrylium–tetrafluoroborate solvent like DMSO, thus requiring a fluorine atom of the anion
complex. To further highlight this special relationship, when to interact more effectively with the p-system. The higher
ꢀ
pyrylium cation is separated from BF4 by solvent, it prefers to the electron deficiency caused at the pyrylium system by the
decompose itself rather than being disconnected from its anion. substituents, the more significant is this anion–p+ interaction.
DMSO-d6, whose capability of solvating cations and splitting ionic The limit situation was found in nitro-pyrylium salts where
aggregates is well known, was used as a solvent in this NMR study. extreme electron-withdrawing (–I, –R) effects induced consider-
Relevant information can be obtained using linear free energy able instability, and the consequent correlation failure related
relationships. Thus, assigned tetrafluoroborate anion 19F chemical to the pyrylium salt concentration decrease in the sample.
shifts (d, ppm) for different pyrylium salts in DMSO-d6 have been
These results confirm those inferred from the tetrafluoro-
correlated with the standard Hammett constants44 for summed borate anion 19F chemical shifts (d, ppm) vs. pyrylium cation
P
para position substituents ( sp). The plot obtained using the molecular weight, described above.
classical Hammett equation, type: d = rs + h, is shown in Fig. 3.
Additionally, on the basis of our experimental results
Two straight lines are clearly observed in the plot, with an abrupt (Table 3) and Hammett correlations we propose a stability range
change of slope. While a poor correlation coefficient (r2 = 0.714) for the studied pyrylium tetrafluoroborate complexes between
was obtained for compounds 12, 13 and 25–27, from series (a) and ꢀ0.61 and 0.40 of the corresponding s value. Compound 22
(d), respectively, a satisfactory correlation (r2 = 0.911) was found proved to be more stable (spara 0.44) than 19 (spara 0.78) thanks
for pyrylium salts having only fluorine as the substituent in the to donating methyl groups and, of course more stable than 21,
phenyl rings [series (b), compounds 14–18]. Compound 11 and 23 and 20.
nitro-compounds [series (c), compounds 19, 22 and 23] failed to
P
From these results, the lowest value of sp for anion–p+
correlate.
interaction seems to be a value of at least 0.0. Values under 0.0
These results reveal quantitative correlation between the s can be understood as a sole electrostatic interaction and for
constants of substituent groups in the pyrylium cation and the values higher than the stated range of stability (ꢀ0.61 to 0.4 for
P
tetrafluoroborate–pyrylium cation interaction. Compounds that
show this correlation giving a small negative r value have an
electron-donating (+R) substituent group at the phenyl ring A
sp) an enhancement of decomposition products is expected.
3.1 Photophysical properties
(R1 = Ph, OMe). Substituent groups of this kind stabilize the The UV-vis absorption and fluorescence properties of some
cation since they disperse the positive charge of the oxygen atom representative synthesized pyrylium tetrafluoroborates (10, 13,
18446 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
14, 15, 22, 23, 26, 27 and 28) were studied. All studied compounds
showed one UV-vis absorption band in the range 406–439 nm in
acetone (ESI,† Table S2). Some of them showed a second absorp-
tion band (10, 14, 15 and 28 at 357, 356, 358 and 368 nm,
respectively) in acetone. Previously, Suresh and co-workers
reported theoretical studies of absorption bands denoted as ‘‘x’’
and ‘‘y’’ bands27 for pyrylium salts. Two independent chromo-
phores assigned to the x-band (2,6-diarylpyrylium moieties,
longer wavelength) and the y-band (4-arylpyrylium moiety,
shorter wavelength) act on pyrylium salts. A hypsochromic shift
of the absorption bands to shorter wavelengths occurs when
solvent polarity is increased. Thus, the absorption spectrum of
compound 10 in CH2Cl2 shows the same two bands (417 and
369 nm) red-shifted reported in the literature,45 whereas the
absorption spectrum of compound 14 in MeOH shows two bands
at 254 and 322 nm. When compound 14 was treated with
increasing amounts (0.2, 0.6, 1.0, 1.5 and 4.0 equiv.) of HClO4,
AcOH and TFAA, (ESI,† Fig. S3) a bathochromic shift of
the absorption bands (357 and 409 nm) resulted. Addition of
0.2 equiv. of HClO4 was sufficient for the completion of this
displacement. Adding up to 4 equiv. any changes with respect to
0.2 equiv. were observed. The same procedure was applied for
TFAA and AcOH. TFAA and HClO4 cause a hypochromic effect on
UV spectra while the presence of AcOH has the opposite effect.
Unfortunately, in our opinion, any evidence of the nature of
anion–cation interactions can be extracted from the just exposed
absorption experiments.
Nevertheless, interesting photophysical changes in different
solvents were observed, in particular related to the stability of
synthesized complexes. For example, in compound 10, three
absorption bands with time-dependence intensity were attri-
buted to the x-band (412 nm), y-band (359 nm) and another one
at 298 nm located in the carbonyl absorption region (ESI,†
Fig. S4). Furthermore, treatment of compound 10 in DMSO-d6
with 10 cycles of UV radiation showed an increase of the blue
shifted band (298 nm) and respectively, loss of intensity in the
pyrylium bands. Subsequently, the 1H NMR spectrum reveals a
ring-opening compound (1,5-di-ketone) (ESI,† Fig. S5).
Fig. 4 Colour perception and brightness of pyrylium tetrafluoroborates
10, 12, 13, 15 and 22 in acetone solution.
In its turn, replacing an electron-withdrawing group like NO2 in
the phenyl ring A by an electron-donating group, like OMe (pairs
22/26 and 23/27) or OBz (pair 22/28), also provides a blue shift in
maximum emission to shorter wavelengths (Dle = 22 to 28 nm)
as a consequence of electronic reorganization and subsequent
displacement of the positive charge to the methoxy or benzyloxy
group, and then tetrafluoroborate should move towards the
phenyl ring A during electronic excitation.
All complexes are stable with the exception of 15 (R1 = F, R2 =
H, R3 = H) and 26 (R1 = OMe, R2 = Me, R3 = Me), fluorescence
intensities of which slowly decay with time (t1/2 = 214.2 min and
272.2 min, respectively), probably due to some photoinstability
of the complex, and specially, pyrylium salt 22 (R1 = NO2, R2 =
Me, R3 = Me), which is highly unstable. In this case, the
maximum fluorescence emission peak becomes null (t1/2
166.3 min) (ESI,† Fig. S6).
=
On the other hand, most of the investigated pyrylium tetra-
fluoroborate salts showed blue to orange emission colours, as it
is shown in Fig. 4 for compounds 10, 12, 13, 15 and 22.
Furthermore, the strong fluorescence emission of these pyrylium
tetrafluoroborate salts makes them a promising class of tunable
emission wavelength dyes for laser technology applications.46
All complexes showed fluorescence characteristics in acetone
solution due to the extended p-conjugation. The wavelength
emission band of the investigated pyrylium tetrafluoroborates
3.2 Theoretical calculations
(10, 13, 15, 22, 23, 26, 27 and 28) was in the range 464–533 nm To establish the theoretical disposition of BF4ꢀ, density func-
depending on the substituent in the para position of phenyl tional theory (DFT) calculations with the Becke 3 Lee-Yang-Park
rings A, B and C attached to the pyrylium core (see Table S2 in (B3LYP) exchange correlation functional at the 6-31G(d) level
the ESI†). From these results we concluded that, in general, of theory were employed. During the optimization process no
incorporation of substituent groups at these para positions symmetry constraints were enforced.47 This basis set proved to
causes a red-shift in the emission band. Furthermore, changing be more in agreement with our experimental results as com-
the conjugation length, i.e. the length of the oligophenyl moiety pared with B3LYP/6-31+G(d,p). Other levels of this theory, such
as in compound 13 (le 533 nm, R1 = Ph, R2 = Me, R3 = Me), or as MP2/6-31++G**, B3LYP/DZVP or/and MP2/aug-cc-pVDZ, have
introducing different substituent groups in the aromatic rings, been previously reported for this kind of interactions.11,12
the maximum emission can be controlled. Thus, it is noteworthy TD-DFT calculations to support the UV-vis results to substantiate
that replacing two methyl groups by fluorine atoms in rings B the mode of binding of the anion has been applied for other
and C (pairs 22/23 and 26/27) causes shifts in the maximum systems.48 Particularly, for pyrylium compounds B3LYP/
emission to shorter wavelengths (14 and 20 nm, respectively) LanLsDZ or ZINDO/S methods have been also applied.27,49 These
while introducing a fluorine atom the in phenyl ring A (pair 10/15) theoretical calculations by B3LYP/6-31G(d) are in excellent agree-
does not even remotely affect the emission band wavelength. ment with the correlation and experimental results observed.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18447

View Article Online
PCCP
Paper
Interaction energies of the pyrylium complexes have been
also calculated at the M06-2X/6-31G(d)50 level, which has been
successfully employed in different molecular systems51 and
provides more accurate values in our case. From the inspection
of the results shown in Table 4, several interesting points arise.
First, the unsubstituted compound 10 shows a theoretical
distance RBꢁ ꢁ ꢁO of 3.023 Å. For all the other investigated pyrylium
salts this distance increases (DRBꢁ ꢁ ꢁO = 0.214 to 0.543 Å, being
maximum for 27), with the exception of compound 12 (R1 = Ph,
R2 = H, R3 = H where RBꢁ ꢁ ꢁO = 3.005 Å). In addition, the anion
distance to the aromatic pyrylium centroid (Re) diminished with
respect to 10 (Re = 3.453 Å) in a range of 0.132–0.239 Å (minimum
value for 28), with the exception of compound 12 (Re = 3.467 Å).
That indicates that the incorporation of a sole phenyl group at
the para position of phenyl ring A (e.g. pairs 10/12 and 11/13 and
17/25) [DRBꢁ ꢁ ꢁO = 0.018, ꢀ0.113 and 0.033 Å, respectively; DRe =
ꢀ0.014, 0.018 and ꢀ0.010 Å, respectively; DRFꢁ ꢁ ꢁC(5) = ꢀ0.038,
0.002 and ꢀ0.018 Å, respectively; DRFꢁ ꢁ ꢁC(3) = 0.018, 0.035 and
ꢀ0.002 Å, respectively; DRFꢁ ꢁ ꢁH(5) = ꢀ0.044, 0.037 and ꢀ0.045 Å,
respectively; and DRFꢁ ꢁ ꢁH(3) = 0.028, 0.161 and ꢀ0.030 Å, respec-
tively] has a poor influence on the anion–p+ interaction, while
the dispersion of charge through the cation, in general, favours
the anion–cation interaction. Thus, a maximum dispersion of
charge in 27 (R1 = OMe, R2 = R3 = F) could be correlated with the
lowest electrostatic interaction and justify the maximum RBꢁ ꢁ ꢁO
distance observed for this compound.
Fig. 5 A superimposition of structures based on DFT calculations for
compounds of series (a).
Fig. 6 A superimposition of structures based on DFT calculations for
compounds of series (b) in comparison with compound 10.
From compound 10 to compound 13, series (a), a slight
reduction of RFꢁ ꢁ ꢁH(5) (3.511 and 3.162 Å, respectively) and
RFꢁ ꢁ ꢁH(3) (3.722 and 2.911 Å, respectively) was observed
(Table 4). For this series, the anion is practically oveꢀr the
trivalent oxygen atom and two fluorine atoms of BF4 are interaction anion–p+. In fluorine compounds (14 to 18) a
almost equidistant to electropositive C-2 and C-6 atoms maximized contact with a distance shorter than DRvdw for
(2.714 and 2.543 Å, respectively for compound 10) (Fig. 5).
Fꢁ ꢁ ꢁC(5) in comparison with C-5 in series (a), is observed. In
When fluorine atoms were incorporated into the molecule, the case of compound 18, the lowest value for RFꢁ ꢁ ꢁH(5) (2.821 Å)
the RBꢁ ꢁ ꢁO distance increased up to a maximum value of 3.480 Å is indicative of the higher anion–p+ interaction suggested. It is
and the anion was directly positioned over C-6 (Fig. 6).
noteworthy that taking for comparison the RFꢁ ꢁ ꢁH(5) and RFꢁ ꢁ ꢁH(3)
An increasingly shorter distance for Fꢁ ꢁ ꢁH(5) in the pyrylium values for compounds 14 and 16, these distances are identical
framework was calculated. This reinforcement of interaction for compound 16, while there is a difference of 0.195 Å in the
with the positive charge of the cation provides a non-covalent case of 14, all of that in accordance with the observations in the
Table 4 Theoretical calculations by B3LYP/6-31G(d) method of distances and dihedral angles for some pyrylium tetrafluoroboratesa
DEint (kcal molꢀ1
)
b
Compound
RBꢁ ꢁ ꢁO (Å)
RFꢁ ꢁ ꢁC(5) (Å)
RFꢁ ꢁ ꢁC(3) (Å)
RFꢁ ꢁ ꢁH(5) (Å)
RFꢁ ꢁ ꢁH(3) (Å)
Re (Å)
f2 (1)
f4 (1)
f6 (1)
10
11
12
13
14
15
16
17
18
25
26
27
28
3.023
3.287
3.005
3.400
3.247
3.346
3.310
3.246
3.480
3.279
3.549
3.566
3.488
3.048
2.824
3.086
2.822
2.972
2.891
2.812
2.821
2.874
2.803
2.863
2.860
2.781
3.297
2.881
3.279
2.846
2.847
2.882
2.849
2.870
2.979
2.868
3.042
3.058
2.845
3.511
3.199
3.555
3.162
3.122
2.907
3.086
3.144
2.821
3.099
2.768
2.753
2.777
3.722
3.072
3.694
2.911
3.317
3.353
3.098
3.132
3.395
3.102
3.417
3.442
3.120
3.453
3.258
3.467
3.240
3.321
3.301
3.255
3.264
3.272
3.254
3.277
3.281
3.214
11.1
ꢀ26.5
15.8
ꢀ4.7
7.2
ꢀ15.2
ꢀ16.9
ꢀ16.6
24.0
ꢀ72.1 (ꢀ78.8)
ꢀ72.9 (ꢀ79.0)
ꢀ67.8 (ꢀ75.2)
ꢀ71.1 (ꢀ77.9)
ꢀ77.8 (ꢀ84.5)
ꢀ76.9 (ꢀ83.6)
ꢀ77.6 (ꢀ83.8)
ꢀ78.2 (ꢀ84.4)
ꢀ79.1 (ꢀ86.1)
ꢀ74.9 (ꢀ81.8)
ꢀ70.6 (ꢀ77.6)
ꢀ74.9 (ꢀ82.2)
ꢀ67.4 (ꢀ75.3)
ꢀ1.0
ꢀ26.8
ꢀ27.2
ꢀ30.0
ꢀ27.0
ꢀ25.0
ꢀ28.7
ꢀ23.5
ꢀ27.5
ꢀ28.1
ꢀ25.8
0.8
9.0
5.4
7.0
6.8
8.4
5.9
5.8
5.8
ꢀ4.5
17.1
ꢀ16.4
ꢀ17.7
24.7
ꢀ17.2
28.7
26.9
23.1
ꢀ5.2
a
All calculated contacts between aromatic rings and protons have been qualitatively confirmed by 2D NOESY experiments where NOE
b
enhancements for compounds 10, 14, 16 and 28 were observed. Interaction energies (without BSSE correction) were calculated from DEint
=
Ecomplex ꢀ E(BF4ꢀ) ꢀ E(Py+). In parentheses the interaction energies at M06-2X/6-31G(d) level of theory are shown.
18448 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
1H NMR spectrum of 14 described above. In general, the
incorporation of a fluorine atom at the para position of the
phenyl ring A (pair 10/15) causes, apart from an increase of
RBꢁ ꢁ ꢁO (DRBꢁ ꢁ ꢁO = 0.323 Å), also a decrease of Re (DRe = ꢀ0.152 Å),
RFꢁ ꢁ ꢁC(5) (DRFꢁ ꢁ ꢁC(5) = ꢀ0.157 Å), RFꢁ ꢁ ꢁC(3) (DRFꢁ ꢁ ꢁC(3) = ꢀ0.415 Å),
RFꢁꢁꢁH(5) (DRFꢁꢁꢁH(5) = ꢀ0.604 Å), and RFꢁꢁꢁH(3) (DRFꢁꢁꢁH(3) = ꢀ0.369 Å);
analogously, the incorporation of two fluorine atoms at the para
position of phenyl rings B and C (e.g. pairs 10/17 and 12/25)
provokes not only a diminution of Re (DRe = ꢀ0.189 and ꢀ0.213 Å,
respectively), but also of RFꢁꢁꢁC(5) (DRFꢁꢁꢁC(5) = ꢀ0.227 and, ꢀ0.283 Å,
respectively), RFꢁꢁꢁC(3) (DRFꢁꢁꢁC(3) = ꢀ0.427 and ꢀ0.411 Å, respec-
Fig. 8 A perspective view of compounds 10 and 27 showing cup-to-cap
shape cavity.
tively), RFꢁꢁꢁH(5) (DRFꢁꢁꢁH(5) = ꢀ0.367 and, ꢀ0.456 Å, respectively), at ortho positions (rings B and C) present dihedral angles of
and specially RFꢁꢁꢁH(3) (DRFꢁꢁꢁH(3) = ꢀ0.590 and, ꢀ0.592 Å, respec- 11.11 and ꢀ15.21 with the pyrylium core. The phenyl ring A is
tively). As a conclusion, the incorporation of fluorine atoms at the also rotated at ꢀ4.71. The aromatic rings are rotated with
para position of phenyl rings A, B and C, apart from diminishing respect to the pyrylium cation, forming a simulated cavity in
Re, strongly decreases the distance between fluorine atoms of the the attempt to accommodate the anion. Nevertheless, the cavity
tetrafluoroborate anion and C(3)/C(5) as well as H(3)/H(5) of the shape changing from the cup-to-cap form (Fig. 8) is determined
pyrylium ring. That can be interpreted in terms of effective anion– by the substituent groups nature and electronic effects as it is
p+, and (C–H)+ꢁꢁꢁFꢀ hydrogen bonds interactions, respectively. It is shown for compound 27, having an electron-donating methoxy
worth mentioning that the incorporation of two methyl groups at group at the para position of the phenyl ring A.
the para position of both phenyl rings B and C has a similar effect
A comparison of the interaction energies calculated using
the B3LYP/6-31G(d) method or at the M06-2X/6-31G(d) level
to that shown by the presence of fluorine atoms.
ꢀ
A different disposition of the BF4 anion in series (d) of (Table 4) shows that, although there is a parallelism between
compounds (26, 27 and 28) resulted. The incorporation of both groups of data, the B3LYP/6-31G(d) values underestimate
electron donating groups such as OMe or OBz promotes the the noncovalent interaction, as has been described for other
migration of guest tetrafluoroborate over C-4 (Fig. 7) with the cases previously.51b,c An analysis of these interaction energies
boron atom positioned in the centroid (Re 3.214–3.281 Å), one for our different pyrylium tetrafluoroborate complexes reveals
of its F atoms over C-4, and other F atoms perpendicular to that compound 18 has the highest anion–cation binding energy
C(5)–C(6) bonds. From the data in Table 4, the incorporation of with calculated B3LYP/6-31G(d) [M06-2X/6-31G(d)] values of
a methoxy group at the para position of the phenyl ring A (pairs ꢀ79.1 kcal molꢀ1 [ꢀ86.1 kcal molꢀ1], as expected due to its
11/26 and 17/27) causes, apart from an increase of RBꢁ ꢁ ꢁO three strongly electron-withdrawing fluorine substituents. In
(DRBꢁ ꢁ ꢁO = 0.262 and 0.320 Å, respectively), a slight increase of contrast, an electron-donating (+R effect) methoxy group as the
Re (DRe = 0.019 and 0.017 Å), an increasing of RFꢁ ꢁ ꢁC(5) (DRFꢁ ꢁ ꢁC(5)
=
substituent, to lead a larger electron density at the aromatic
0.039 Å for both cases), RFꢁ ꢁ ꢁC(3) (DRFꢁ ꢁ ꢁC(3) = 0.161 and 0.188 Å), system, it should result in a weaker binding energy of the
RFꢁ ꢁ ꢁH(3) (DRFꢁ ꢁ ꢁH(3) = 0.345 and 0.310 Å) and a decrease of RFꢁ ꢁ ꢁH(5) pyrylium cation with the tetrafluoroborate anion. Nevertheless,
(DRFꢁ ꢁ ꢁH(5) = ꢀ0.431 and ꢀ0.391 Å).
we realize that compound 26 (R1 = OMe, R2 = R3 = Me) has an
On the other hand, the dihedral angles on geometry- interaction energy, ꢀ70.6 kcal molꢀ1 [ꢀ77.6 kcal molꢀ1], simi-
optimized structures by B3LYP/6-31G(d) calculations are depen- lar to that calculated for compound 13, (R1 = Ph, R2 = R3 = Me)
dent on the substituent groups at the para position of the ꢀ71.1 kcal molꢀ1 [ꢀ77.9 kcal molꢀ1], and higher than those
aromatic ring attached to C-2 and C-6 (rings B and C). Thus, for for compound 12, (R1 = Ph, R2 = R3 = H) ꢀ67.8 kcal molꢀ1
the unsubstituted compound 10, the phenyl rings of pyrylium [ꢀ75.2 kcal molꢀ1]. Moreover, the interaction energy for com-
pound 27 is the same [B3LYP (ꢀ74.9 kcal molꢀ1)] or slightly
higher [M06-2X (ꢀ82.2 kcal molꢀ1)] than for compound 25
[B3LYP (ꢀ74.9 kcal molꢀ1) and M0-62X (ꢀ81.8 kcal molꢀ1)].
This unexpected behaviour can be explained in terms of the so
called cooperativity, a feature observed previously in structures
stabilized by weak noncovalent interactions.51a Thus, the
presence of a methyl group in the methoxy group attached to
ring A in compound 26, leads to a cooperative stabilization of the
otherwise weakly stable anion–p complex by bending of the side
group toward the anion to facilitate (C–H)+–anion interactions.
Additionally, we have found a relationship between 19F chemical
shifts and the interaction energies calculated at the M06-2X/
6-31G(d) level, (ESI,† Fig. S7). Two straight lines are observed in
the plot. While a satisfactory correlation (r2 = 0.938) with a
Fig. 7 A superimposition of structures based on DFT calculations for
compounds of series (d) in comparison with compound 10.
higher and negative slope results for only fluorine-containing
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18449

View Article Online
PCCP
Paper
pyrylium salts [series (b), compounds 14–18], a poor correlation pyrylium cation (for which the Mu¨lliken charge shows to be the
line (r2 = 0.409) with a near zero slope was obtained for more electropositive of the pyrylium salt protons) is inferred.
compounds 12, 13 and 25–27, from series (a) and (d), respec- In most cases, C-3 and H-3 in host pyrylium are not involved in
tively. Thus, the higher the electron deficiency is at the pyrylium any interaction.
system provoked by the substituents, the greater is the inter-
action energy of the complex, while an interplay of cooperativity
Conclusions
between anion–p and (C–H)+–anion interactions could justify the
results obtained for compounds from series (a) and (d).
We have found that 19F NMR chemical shifts of tetrafluoroborate
anions at a constant concentration in DMSO-d6 for most of the
here described tri-arylpyrylium tetrafluoroborate salts undergo
small but significant changes caused by the effect of substituents
on the pyrylium cation. Correlations of these parameters with
both the pyrylium cation molecular weight and the standard
substituent Hammett constants, demonstrate that an anion–p+
interaction between the polyatomic anion BF4ꢀ and the pyrylium
aromatic system does exist. Two trends are shown by the com-
plexes depending on the substituents: electron-withdrawing
substituent groups like fluorine enhance the magnitude of the
anion–p+ non-covalent interaction, while electron-donating
groups, such as OMe or Ph, have a poor global effect.
DFT calculations are in excellent agreement with the observed
correlations and experimental results. Furthermore, apart from
corroborating the anion–p+ interaction by means of calculated Re
distances, from the boron atom to the centroid of the pyrylium
moiety, shorter than RBꢁ ꢁ ꢁO distances, suggests the existence of:
(1) an additional non covalent anion–p+ contact involving a
fluorine atom of the anion with acidic C(5), in practically all of
the substituted pyrylium salts described here (exceptions are
compounds 10ꢀand 12), (2) a (C–H)+–anion interaction between a
F atom of BF4 and H(5) of the pyrylium ring in some cases
(compounds 15, 18 and 26–28), and (3) an arrangement of the
aromatic rings A, B and C, which are rotated with respect to the
pyrylium ring, to conform to a cup-to-cap shape cavity, depending
on the nature and electronic effects of the substituent groups.
On the other hand, although the study on photophysical
properties gave no evidence of the investigated anion–p+ inter-
action, the strong fluorescence emission of some representative
pyrylium tetrafluoroborates described herein makes them a
promising class of tunable emission wavelength dyes for laser
technology applications.
The effect of the solvent (DMSO) on the tetrafluoroborate–
pyrylium cation interaction was also investigated by using
Polarizable Continuum Method (PCM)52 coupled to B3LYP/
6-31G(d) by means of Gaussian 09.53 In general, DMSO causes
ꢀ
higher lengths between pyrylium cation and BF4 anion as
expected of the intrinsic effects on the cation by this solvent
and very small interaction energies. Thus, by using the B3LYP/
6-31G(d) level, the calculated DEint for compounds 10, 14 and
18 were ꢀ7.5, ꢀ6.7 and ꢀ8.1 kcal molꢀ1, respectively. For
compound 10 the predicted Bꢁ ꢁ ꢁO+ distance increased B0.35 Å
(B0.36 Å for Re) in the PCM model. A second point observed in
this model was an enhanced rotation of A, B and C aryl rings
with respect to the cation core (ꢀ31.4, ꢀ29.6 and 30.71, respec-
tively). For fluorinated compound 14 a major distance for Bꢁ ꢁ ꢁO+
(3.390 Å) was also calculated in the PCM model. A similar
rotation to that described for compound 10 was observed
(ꢀ30.91 in the C-2 ring). However, the phenyl ring at the C-6
position, which contains the fluorine atom, is rotaꢀted over
ꢀ15.81 less than the phenyl ring in 10. In general, BF4 in this
model is situated over charged oxygen but is much closer to
ortho protons in phenyl rings B and C with a distance shorter
than DRvdw Fꢁ ꢁ ꢁH (2.242 and 2.245 Å for B and C, respectively)
(Fig. 9). Certainly, the aforementioned model brings to light a
(C–H)+–anion interaction between H-6b of the aromatic ring in B
and a fluorine atom of the anion (and H-6c for the aromatic
ring C). By incorporating fluorine atoms into pyrylium salts, the
calculated Mu¨lliken charges over H-3, H-5, H-6b and H-6c
change significantly. Thus, an increase of positive charge from
0.190 (compound 10) to 0.200 (compound 18) is observed for
H-6b that indicates in this last case an enhanced (C–H)+–anion
interaction involving this proton. It is worthy of note that by
using this procedure, compound 14 shows the same charge
(0.212) in fictitious equivalent H-3 and H-5 protons.
On the basis of DFT calculations, two situations can be
observed for this interaction. First, an anion–p+ contact invol- Acknowledgements
ving F atoms of the anion with acidic C-5, and in the second
The authors thank the AECID (Projects A/023577/09 and
A/030422/10, and scholarship to one of them, S. Jatunov) and
place, (C–H)+–anion between the same F atoms and H-5 of the
´
the Junta de Andalucıa (FQM 142 and Project P09-AGR-4597) for
financial support. The authors also thank the Servicio de
´
Resonancia Magnetica Nuclear, CITIUS, Universidad de Sevilla
for facilities on NMR experiment performance.
Notes and references
1 (a) M. L. Waters, Acc. Chem. Res., 2013, 46, 873; (b) E. A. Meyer,
R. K. Castellano and F. Diederich, Angew. Chem., Int. Ed., 2003,
42, 1210.
Fig. 9 Structures based on DFT calculations including solvent effects
(PCM) for compounds 10, 14 and 18.
18450 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
2 (a) J.-M. Lehn, Supramolecular Chemistry. Concepts and
Perspectives, Wiley-VCH, Weinheim, 1995; (b) J.-M. Lehn,
Science, 2002, 295, 2400; (c) J.-M. Lehn, Chem. Soc. Rev.,
2007, 36, 151; (d) H.-J. Schneider and A. Yatsimirski,
Principles and Methods in Supramolecular Chemistry, Wiley,
(c) O. B. Berryman and D. W. Johnson, Chem. Commun.,
2009, 3143; (d) O. B. Berryman, A. C. Sather, B. P. Hay,
J. S. Meisner and D. W. Johnson, J. Am. Chem. Soc., 2008,
130, 10895; (e) O. B. Berryman, F. Hof, M. J. Hynes and
D. W. Johnson, Chem. Commun., 2006, 506.
Chichester, 2000; (e) J. W. Steed and J. L. Atwood, Supramol. 11 H. T. Chifotides, I. D. Gles and K. R. Dunbar, J. Am. Chem.
Chem., Wiley, London, 2009; ( f ) F. P. Schmidtchen and Soc., 2013, 135, 3039, and references therein.
M. Berger, Chem. Rev., 1997, 97, 1609; (g) H.-J. Schneider, 12 (a) D. Quinonero, A. Frontera, A. Costa and P. M. Deya, in
˜
`
Acc. Chem. Res., 2013, 46, 1010.
Anion–p Interactions in Molecular Recognition in Anion Coor-
dination Chemistry, ed. K. Bowman-James, A. Bianchi and
3 M. M. J. Smulders, I. A. Riddell, C. Browne and
J. R. Nitschke, Chem. Soc. Rev., 2013, 42, 1728.
4 Y. Chun, N. J. Singh, I.-C. Hwang, J. W. Lee, S. U. Yu and
K. S. Kim, Nat. Commun., 2013, 4, 1797, DOI: 10.1038/
ncoms2758.
´
˜
E. Garcıa-Espana, Wiley-VCH, Weinheim, 2012, ch. 6,
p. 321, and references therein; (b) S. Chakravarty,
Z.-Z. Sheng, B. Iverson and B. Moore, FEBS Lett., 2012,
586, 4180; (c) O. Perraud, V. Robert, H. Gornitzka,
A. Martinez and J.-P. Dutasta, Angew. Chem., Int. Ed., 2012,
5 (a) N. J. Zondlo, Acc. Chem. Res., 2013, 46, 1039;
´
˜
´
´
´
(b) J. L. Asensio, A. Arda, F. J. Canada and J. Jimenez-
Barbero, Acc. Chem. Res., 2013, 46, 946; (c) L. M. Salonen,
M. Ellermann and F. Diederich, Angew. Chem., Int. Ed., 2011,
51, 504; (d) G. Gil-Ramırez, E. C. Escudero-Adan, J. Benet-
Buchholtz and P. Ballester, Angew. Chem., Int. Ed., 2008,
47, 4114.
50, 4808; (d) J. L. Seganish, P. V. Santacroce, K. J. Salimian, 13 (a) R. E. Dawson, A. Hennig, D. P. Weimann, D. Emery,
J. C. Fettinger, P. Zavalij and J. T. Davis, Angew. Chem., Int.
Ed., 2006, 45, 3334; (e) C. J. Fowler, T. J. Haverlock,
B. A. Moyer, J. A. Shriver, D. E. Gross, M. Marquez,
J. L. Sessler, M. A. Hossain and K. Bowman-James, J. Am.
V. Ravikumar, J. Montenegro, T. Takeuchi, S. Gabutti,
M. Mayor, J. Mareda, C. A. Schalley and S. Matile, Nat. Chem.,
2010, 2, 533; (b) J. Mareda and S. Matile, Chem. – Eur. J., 2009,
15, 28.
Chem. Soc., 2008, 130, 14386; ( f ) K. Mu¨ller-Dethlefs and 14 (a) E. H. Krenske and K. N. Houk, Acc. Chem. Res., 2013,
P. Hobza, Chem. Rev., 2000, 100, 143.
6 (a) Supramolecular Chemistry of Anions, ed. A. Bianchi,
46, 979; (b) Y. Zhao, C. Beuchat, Y. Domoto, J. Gajewy,
A. Wilson, J. Mareda, N. Sakai and S. Matile, J. Am. Chem.
Soc., 2014, 136, 2101.
´
˜
K. Bowman-James and E. Garcıa-Espana, Wiley-VCH, New
´
˜
´
˜
`
York, 1997; (b) E. Garcıa-Espana, P. Dıaz, J. M. Llinares and 15 (a) C. Estarellas, A. Frontera, D. Quinonero and P. M. Deya,
A. Bianchi, Coord. Chem. Rev., 2006, 250, 2952.
7 (a) I. Alkorta, I. Rozas and J. Elguero, J. Am. Chem. Soc., 2002,
Chem. – Asian J., 2011, 6, 2316; (b) A. Robertazzi, F. Krull,
E.-W. Knapp and P. Gamez, CrystEngComm., 2011, 13, 3293.
˜
˜
`
124, 8593; (b) D. Quinonero, C. Garau, C. Rotger, 16 (a) C. Estarellas, A. Frontera, D. Quinonero and P. M. Deya,
`
A. Frontera, P. Ballester, A. Costa and P. M. Deya, Angew.
Angew. Chem., Int. Ed., 2011, 50, 415; (b) C. Garau,
˜
Chem., Int. Ed., 2002, 41, 3389; (c) M. Mascal, A. Armstrong
and M. D. Bartberger, J. Am. Chem. Soc., 2002, 124, 6274.
8 For recent reviews on anion–p interactions see:
(a) P. Ballester, Acc. Chem. Res., 2013, 46, 874; (b) M. M.
D. Quinonero, A. Frontera, P. Ballester, A. Costa and
P. M. Deya, J. Phys. Chem. A, 2005, 109, 9341; (c) D. Kim,
P. Tarakeshwar and K. S. Kim, J. Phys. Chem. A, 2004,
108, 1250.
Watt, M. S. Collins and D. W. P. Johnson, Acc. Chem. Res., 17 (a) S. E. Wheeler, Acc. Chem. Res., 2013, 46, 1029;
2013, 46, 955; (c) H. T. Chifotides and K. R. Dunbar, Acc.
Chem. Res., 2013, 46, 894; (d) A. Frontera, Coord. Chem. Rev.,
2013, 257, 1716; (e) A. Frontera, P. Gamez, M. Mascal,
(b) S. E. Wheeler, J. Am. Chem. Soc., 2011, 133, 10262;
(c) S. E. Wheeler and K. N. Houk, J. Phys. Chem. A, 2010,
114, 8658.
T. J. Mooibroek and J. Reedijk, Angew. Chem., Int. Ed., 18 When this work was in preparation for its publication,
2011, 50, 9564; ( f ) B. L. Schottel, H. T. Chifotides and K. R.
Dunbar, Chem. Soc. Rev., 2008, 37, 68; (g) C. Caltagirone and
P. A. Gale, Chem. Soc. Rev., 2009, 38, 520; (h) P. Gamez,
T. J. Mooibroek, S. J. Teat and J. Reedijk, Acc. Chem. Res.,
2007, 40, 435; (i) B. P. Hay and V. S. Bryantsev, Chem.
Commun., 2008, 2417; ( j) P. Ballester, Anions and p–Aromatic
unsuccessful solution studies of anion–p interactions in
N-(pentafluorobenzyl)pyridinium salts next to the experi-
mental evidence of anion–p interactions in crystals of these
salts were reported. M. Giese, M. Albrecht, T. Repenko,
J. Sackmann, A. Valkonen and K. Rissanen, Eur. J. Org.
Chem., 2014, 2435.
Systems. Do they interact attractively? Recognition of Anions, 19 (a) B. L. Schottel, H. T. Chifotides and K. R. Dunbar, Chem.
Structure#Bonding Series, Springer Verlag, Berlin, 2008, vol.
129, p. 127.
9 D.-X. Wang and M.-X. Wang, J. Am. Chem. Soc., 2013,
135, 892, and references therein.
10 (a) A. V. Jentzsch, D. Emery, J. Mareda, P. Metrangolo,
G. Resnati and S. Matile, Angew. Chem., Int. Ed., 2011,
50, 11675; (b) M. G. Chudzinski, C. A. McClary and
M. S. Taylor, J. Am. Chem. Soc., 2011, 133, 10559;
Soc. Rev., 2008, 37, 68; (b) O. B. Berryman, V. S. Bryantsev,
D. P. Stay, D. W. Johnson and B. P. Hay, J. Am. Chem. Soc.,
2007, 129, 48.
20 According to anion–p Berryman’s definition19b the anion
must be positioned over the centroid of the aromatic ring
and the interaction must be electrostatic with negligible
charge-transfer. A broader operative definition used by
Dunbar’s group, among other research groups, considers
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18451

View Article Online
PCCP
Paper
that any interaction between the anion and the p-system of 30 B. K. Wetzl, S. M. Yarmoluk, D. M. Craig and O. S. Wolfbeis,
an aromatic ring is an anion–p interaction, regardless of Angew. Chem., Int. Ed., 2004, 43, 5400.
position of the anion over the p–face of the ring and with or 31 (a) F. A. Scaramuzzo, A. Gonzalez-Campo, C. C. Wu,
´
without charge-transfer interactions. This less unrestrictive
criterion is supported by a recent theoretical study on
the directionality of anion–p interactions [C. Estarellas,
A. H. Velders, V. Subramaniam, G. Doddi, P. Mencarelli,
M. Barteri, P. Jonkheijm and J. Huskens, Chem. Commun.,
2010, 46, 4193; (b) K. Binnemans, Chem. Rev., 2005, 105, 4148.
´
˜
`
A. Bauza, A. Frontera, D. Quinonero and P. M. Deya, Phys. 32 (a) R. K. Harris, E. D. Becker, S. M. Cabral de Menezes,
Chem. Chem. Phys., 2011, 13, 5696].
R. Goodfellow, P. Granger, R. E. Hoffman and K. W. Zilm,
Pure Appl. Chem., 2008, 80, 59; (b) R. K. Harris, E. D. Becker,
S. M. Cabral de Menezes, R. Goodfellow and P. Granger,
Pure Appl. Chem., 2001, 73, 1795.
˜
21 C. Garau, A. Frontera, D. Quinonero, P. Ballester, A. Costa
`
and P. M. Deya, J. Phys. Chem. A, 2004, 108, 9423.
´
˜
`
22 (a) A. Bauza, D. Quinonero, P. M. Deya and A. Frontera,
´
Chem. Phys. Lett., 2012, 530, 145; (b) A. Garcıa-Raso, 33 (a) A. T. Balaban, W. Schroth and G. Fischer, in Adv.
F. M. Albert, J. J. Fiol, A. Tasada, M. Barcelo-Oliver,
Heterocycl. Chem., ed. A. R. Katritzky and A. J. Boulton,
Academic Press, New York, 1969, vol. 10, p. 241;
(b) A. T. Balaban, in Names Reactions in Heterocyclic Chem-
istry II, ed. J. J. Li and E. J. Corey, John Wiley & Sons, Inc.,
2011, p. 402.
˜
E. Molins, C. Estarellas, A. Frontera, D. Quinonero and
`
P. M. Deya, Cryst. Growth Des., 2009, 9, 2363;
˜
`
(c) C. Estarellas, A. Frontera, D. Quinonero and P. M. Deya,
´
J. Chem. Theory Comput., 2008, 4, 1981; (d) A. Garcıa-Raso,
¨
F. M. Albert, J. J. Fiol, A. Tasada, M. Barcelo-Oliver, E. Molins, 34 (a) H. Johansson, T. Cailly, A. R. B. Thomsen, H. Brauner-
˜
`
D. Escudero, A. Frontera, D. Quinonero and P. M. Deya, Inorg.
Osborne and D. S. Pedersen, Beilstein J. Org. Chem., 2013,
9, 1383; (b) M. Martiny, E. Steckhan and T. Esch, Chem. Ber.,
1993, 126, 1671.
˜
Chem., 2007, 46, 10724; (e) D. Quinonero, A. Frontera,
D. Escudero, P. Ballester, A. Costa and P. M. Deya, Chem-
`
´
PhysChem, 2007, 8, 1182; ( f ) A. Garcıa-Raso, F. M. Albert, 35 L. E. E. Broeckx, S. Gueven, F. J. L. Heutz, M. Lutz, D. Vogt
J. J. Fiol, A. Tasada, M. Barcelo-Oliver, E. Molins, D. Escudero, and C. Mueller, Chem. – Eur. J., 2013, 19, 13087.
A. Frontera, D. Quinonero and P. M. Deya, Eur. J. Org. Chem., 36 C. Reichardt and R. Mueller, Justus Liebigs Ann. Chem., 1976,
2007, 5821. 11, 1937.
23 (a) A. Caballero, F. Zapata, L. Gonzalez, P. Molina, I. Alkorta 37 A. R. Katritzky, J. Adamson, E. M. Elisseou, G. Musumarra,
˜
`
´
and J. Elguero, Chem. Commun., 2014, 50, 4680; (b) H. Ihm,
S. Yun, H. G. Kim, J. K. Kim and K. S. Kim, Org. Lett., 2002,
R. C. Patel, K. Sakizadeh and W. Yeung, J. Chem. Soc., Perkin
Trans. 2, 1982, 1041.
4, 2897; (c) K. Sato, S. Arai and T. Yamagishi, Tetrahedron 38 A. R. Katritzky, R. T. C. Brownlee and G. Musumarra,
´
Lett., 1999, 40, 5219; (d) E. Alcalde, C. Alvarez-Ru´a, S. Garcıa-
Tetrahedron, 1980, 36, 1643.
´
´
Granda, E. Garcıa-Rodriguez, N. Mesquida and L. Perez- 39 P. P. Laine, F. Bedioui, F. Loiseau, C. Chiorboli and
´
Garcıa, Chem. Commun., 1999, 295.
S. Campagna, J. Am. Chem. Soc., 2006, 128, 7510.
24 (a) M. Haj-Zaroubi, N. W. Mitzel and F. P. Schmidtchen, 40 M. Song and C. Fan, Lett. Org. Chem., 2013, 10, 27.
Angew. Chem., Int. Ed., 2002, 41, 104; (b) H. Miyaji, W. Sato 41 (a) Y. Liu, M. Han, H.-Y. Zhang, L.-X. Yang and W. Yiang,
and J. L. Sessler, Angew. Chem., Int. Ed., 2000, 39, 1777;
(c) K. Niikura, A. P. Bisson and E. V. Anslyn, J. Chem. Soc.,
Perkin Trans. 2, 1999, 1111; (d) K. S. Kim, P. Tarakeshwar
and J. Y. Lee, Chem. Rev., 2000, 100, 4145.
Org. Lett., 2008, 10, 2873; (b) A. Moghimi, M. F. Rastegar and
M. Ghandi, Can. J. Chem., 2004, 82, 1716; (c) A. Moghimi,
M. F. Rastegar, M. Ghandi, M. Taghizadeh, A. Yari,
M. Shamsipur, G. P. A. Yap and H. Rahbarnoohi, J. Org.
Chem., 2002, 67, 2065.
25 (a) R. G. Gieling, M. Babur, L. Mamnani, N. Burrows,
B. A. Telfer, F. Carta, J.-Y. Winum, A. Scozzafava, 42 (a) D. Y. Kim, N. J. Singh, J. W. Lee and K. S. Kim, J. Chem.
C. T. Supuran and K. J. Williams, J. Med. Chem., 2012,
55, 5591; (b) S. Farid, J. P. Dinnocenzo, P. B. Merkel,
R. H. Young and D. Shukla, J. Am. Chem. Soc., 2011,
133, 4791.
Theory Comput., 2008, 4, 1162; (b) D. Y. Kim, N. J. Singh and
K. S. Kim, J. Chem. Theory Comput., 2008, 4, 1401; (c) D. Kim,
E. C. Lee, K. S. Kim and P. Tarakeshwar, J. Phys. Chem. A,
2007, 111, 7980; (d) D. Kim, P. Tarakeshwar and K. S. Kim,
J. Phys. Chem. A, 2004, 108, 1250; (e) M. Mascal, A. Armstrong
and M. D. Bartberger, J. Am. Chem. Soc., 2002, 124, 6274.
26 I. Turowska-Tyrk and T. M. Krygowski, J. Mol. Struct., 1991,
263, 235.
27 N. Manoj, G. Ajayakumar, K. R. Gopidas and C. H. Suresh, 43 (a) J. Vrbancich and L. D. Ritchie, J. Chem. Soc., Faraday
J. Phys. Chem. A, 2006, 110, 11338.
28 (a) A. Mouradzadegun and F. Abadast, Synlett, 2014, 448;
Trans. 2, 1980, 76, 648; (b) K. Yamaoka and K. Fukudome,
J. Phys. Chem., 1990, 94, 6896.
(b) A. T. Balaban, A. Dinculescu, G. N. Dorofeenko, 44 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165.
´
G. W. Fischer, A. V. Koblik, V. V. Mezheritskii and 45 M. A. Miranda and H. Garcıa, Chem. Rev., 1994, 94, 1063.
W. Schroth, in Advances in Heterocyclic Chemistry, Supplement 46 M. Fakis, J. Polyzos, G. Tsigaridas, J. Parthenios, A. Fragos,
2, ed. A. R. Katritzky, Academic Press, New York, 1982, p. 85.
´˜
V. Giannetas, P. Persephonis and J. Mikroyannidis, Chem.
Phys. Lett., 2000, 323, 111.
´
´
´
´
29 B. Garcıa-Acosta, F. Garcıa, J. M. Garcıa, R. Martınez-Manez,
´
´
F. Sancenon, N. San-Jose and J. Soto, Org. Lett., 2007, 47 (a) M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert,
9, 2429.
M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga,
18452 | Phys. Chem. Chem. Phys., 2014, 16, 18442--18453
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis and 52 J. Tomasi, B. Menucci and R. Cammi, Chem. Rev., 2005,
J. A. Montgomery, J. Comput. Chem., 1993, 14, 1347; 105, 2999.
(b) M. S. Gordon and M. W. Schmidt, Advances in Electronic 53 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
Structure Theory: GAMESS a Decade Later, in Theory and
Applications of Computational Chemistry, the First 40 years,
ed. C. E. Dykstra, G. Frenking, K. S. Lim and G. E. Scuseria,
Elsevier, Amsterdam, 2005.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
48 Y. Zhang, X.-Y. Chen, H.-J. Wang, K.-S. Diao and J.-M. Chen,
J. Mol. Struct.: THEOCHEM, 2010, 952, 16.
49 M. Kondo, M. Uchikawa, K. Namiki, W.-W. Zhang, S. Kume,
E. Nishibori, H. Suwa, S. Aoyagi, M. Sakata, M. Murata,
Y. Kobayashi and H. Nishihara, J. Am. Chem. Soc., 2009,
131, 12112.
50 (a) Y. Zhao and D. G. Thrular, Acc. Chem. Res., 2008, 41, 157;
(b) Y. Zhao, N. E. Schultz and D. G. Thrular, J. Chem. Theory
Comput., 2006, 2, 364.
51 (a) T. K. Maldal, S. Samanta, S. Chakraborty and A. Datta,
ChemPhysChem, 2013, 14, 1149; (b) A. Nijamudheen, D. Jose,
A. Shine and A. Datta, J. Phys. Chem Lett., 2012, 3, 1493;
(c) S. A. Abraham, D. Jose and A. Datta, ChemPhysChem,
2012, 13, 695.
¨
J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox,
Gaussian 09, Revision C.01, Gaussian, Inc., Wallingford, CT,
2010.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 18442--18453 | 18453