X-ray diffraction and theoretical studies for the quantitative assessment of intermolecular arene-perfluoroarene stacking interactions

Article abstract of DOI:10.1002/chem.200501248

The arene-perfluoroarene stacking interaction was studied by experimental and theoretical methods. A series of compounds with different possibilities for formation of this recognition motif in the solid state were synthesized, and their crystal structures

Full text of DOI:10.1002/chem.200501248

DOI: 10.1002/chem.200501248
X-ray Diffraction and Theoretical Studies for the Quantitative Assessment of
Intermolecular Arene–Perfluoroarene Stacking Interactions
Sergio Bacchi,[a, b] Maurizio Benaglia,[a] Franco Cozzi,[a] Francesco Demartin,[c]
Giuseppe Filippini,[d] and Angelo Gavezzotti*[c]
Abstract: The arene–perfluoroarene
stacking interaction was studied by ex-
perimental and theoretical methods. A
series of compounds with different pos-
sibilities for formation of this recogni-
tion motif in the solid state were syn-
thesized, and their crystal structures
determined by single-crystal X-ray dif-
fraction. The crystal packing of these
compounds, as well as the packing of
related compounds retrieved from crys-
tallographic databases, were analyzed
with quantitative crystal potentials:
total lattice energies and the cohesive
energies of closest molecular pairs in
the crystals were calculated. The
arene–perfluoroarene
recognition
respectively. Pixel energy partitioning
reveals that whenever aromatic rings
stack, the largest cohesive energy con-
tribution comes from dispersion, which
roughly amounts to 20 kJmolꢀ1 per
phenyl ring, while the coulombic term
is minor but significant enough to
make a difference between the arene–
arene or perfluoroarene–perfluoroar-
ene interactions on the one hand, and
arene–perfluoroarene interactions on
the other, whereby the latter are fa-
vored by about 10 kJmolꢀ1 per phenyl
ring. No evidence of special interaction
which can be attributed to H···F con-
frontation was recognizable.
motif emerges as a dominant interac-
tion in the non-hydrogen-bonding com-
pounds studied here, to the point that
asymmetric dimers formed over the
stacking motif carry over to asymmet-
ric units made of two molecules in the
crystal both for pure compounds and
for molecular complexes; however,
inter-ring distances and angles range
from 3.70 to 4.85 Š and from 5 to 218,
Keywords: crystal engineering
molecular recognition pixel
calculations · stacking interactions
·
·
Introduction
plex,[1] many investigations have been devoted to the geo-
metric and energetic features of the interaction between
these two molecules.[2] Structure analysis has demonstrated[3]
that in the solid state the benzene and hexafluorobenzene
molecules form infinite columns of alternating hydrogenated
and fluorinated rings. It was later realized that a parallel
stacked structure is widespread among the arene–perfluo-
roarene complexes, consistent with a simplified model in
which the interaction is favored by electrostatic attraction
when the charge distributions are approximated by central
quadrupoles of opposite phase.[4–6] The discovery of analo-
gous structural motifs in a number of crystallographic stud-
ies[7] showed that such alternating columnar arrangements
are invariably adopted whenever perfluorinated and hydro-
genated aromatic rings are present, either in crystals of pure
compounds or of molecular complexes formed by any kind
of planar aromatic molecules. The arene–perfluoroarene
face-to-face stacking interaction[8] thus emerges as an ubiq-
uitous noncovalent interaction[9,10] that deserves the status of
a reliable and robust “synthon” which can be classified and
exploited as a structural driving force in supramolecular
chemistry and crystal engineering.[11,12]
After the discovery by Patrick and Prosser in 1960 that ben-
zene and hexafluorobenzene form a 1:1 molecular com-
[a] Dr. S. Bacchi, Dr. M. Benaglia, Prof. F. Cozzi
Dipartimento di Chimica Organica e Industriale
Università degli Studi di Milano
via Venezian 21, 20133 Milano (Italy)
[b] Dr. S. Bacchi
GlaxoSmithKline - Research Center
via Fleming 4, 35131 Verona (Italy)
[c] Prof. F. Demartin, Prof. A. Gavezzotti
Dipartimento di Chimica Strutturale e Stereochimica Inorganica
Università degli Studi di Milano
via Venezian 21, 20133 Milano (Italy)
Fax : (+39)02-5031-4454
E-mail: angelo.gavezzotti@unimi.it
[d] Dr. G. Filippini
ISTM-CNR, c/o Dipartimento di Chimica Fisica ed Elettrochimica
Università degli Studi di Milano
via Golgi 19, 20133 Milano (Italy)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
3538
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3538 – 3546

FULL PAPER
The crystallographic studies reported so far[3,7,9] have
dealt almost exclusively with cocrystals of aromatic hydro-
allows neither complete ring coplanarity nor intramolecular
ring stacking. Henceforth, the mnemonic symbols H-HH, H-
FF, H-HF, and F-HH will be used for compounds 1, 2, 3, 4
respectively, to indicate the hydrogenated or fluorinated
nature of the carrier and the two prongs in that order.
ꢀ
carbons and perfluorohydrocarbons having C C triple and
double bonds but no other functional groups. To the best of
our knowledge, only very few studies have presented a de-
tailed description of arene–perfluoroarene interaction in
cocrystals of functionalized aromatic molecules. In one of
these,[13] it was shown that phenol and pentafluorophenol
form columnar face-to-face stacks, and this arrangement was
interpreted as mainly due to aryl–perfluoroaryl interactions,
which were considered more important than hydrogen-bond
formation in determining the structure of the cocrystal. An-
other study[14] described the solid-state structures of three
N-pentafluorophenyl benzamides having a nitro, a hydrogen,
and a dimethylamino substituent in the para position of the
benzoyl moiety. In all of the crystal structures, the molecules
are linked by infinite H-bonded chains; a face-to-face
stacked arrangement between aromatic moieties was ob-
served only in the case of N-pentafluorophenyl 4-dimethyla-
minobenzamide, in which the electron-poor perfluoroaryl
residue of one molecule can interact favorably with the elec-
tron-rich dimethylaminophenyl residue of another molecule,
in agreement with the above-mentioned interpretation of
the arene–perfluoroarene interaction in terms of central
multipoles.
Computational and Theoretical Methods
Qualitative analyses of crystal packing use geometrical crite-
ria for the assessment of the relative importance of energet-
ic factors, and as such, are often affected by subjective judg-
ment. For the quantitative analysis of crystal packing, we
use either overall lattice energies, or the concept of neighbor
molecular pairs, also called structure determinants:[16] a ref-
erence molecule in a crystal structure is picked (the choice
is arbitrary), and each molecular pair formed by the refer-
ence and one surrounding molecule is characterized by the
distance between centers of mass, the symmetry operator
connecting the two molecules, and the molecule–molecule
interaction energy. In the simplest approach, intermolecular
energies are computed by atom–atom potentials supple-
mented by a separately optimized F···F interaction func-
tion;[17] C···F and H···F interaction curves were obtained by
the usual averaging procedures. The whole parameter set is
collected in Table 1. This formulation is especially attractive
With a view to a better understanding and a wider practi-
cal application of the arene–perfluoroarene interaction as
structural determinant in supramolecular chemistry and
crystal engineering, it would be highly desirable to acquire
more systematic information 1) on the crystal packing of
flexible molecular systems without classical hydrogen-bond-
ing functional groups; 2) on the packing modes of molecules
carrying both an aromatic and a perfluoroaromatic residue,
and thus possibly able to undergo also an intramolecular
stacking interaction; and 3) on the competition between the
aromatic interaction and other packing forces, such as classi-
cal hydrogen bonding, the other prime driving force in crys-
tal packing,[15] and other less energetic recognition patterns.
We start here by determining the packing arrangement of
molecules whose structure can
Table 1. UNI force-field parameters[17] for atom–atom energy in the form
E=Aexp
No coulombic Rꢀ1 terms need be added. R8 is the minimum-energy dis-
tance, and e the well depth in kJmolꢀ1
A
.
.
A
B
C
R8
e
ꢀ
H F
ꢀ
F F
64257.8
170916.4
24158.4
226145.2
120792.1
196600.9
4.110
4.220
4.010
3.470
4.100
3.840
248.36
564.84
109.20
2418.35
472.79
1168.75
3.29
3.20
3.36
3.89
3.29
3.50
ꢀ0.110
ꢀ0.293
ꢀ0.042
ꢀ0.387
ꢀ0.205
ꢀ0.350
ꢀ
H H
ꢀ
C C
ꢀ
H C
ꢀ
C F
offer, in the solid state, a choice
of interactions among arene–
arene, fluoroarene–arene, and
fluoroarene–fluoroarene rings:
the phenyl and pentafluoro-
phenyl esters of benzene-1,2-di-
carboxylic acid and tetrafluoro-
benzene-1,2-dicarboxylic acid
because it gives the complete lattice energy without re-
course to any separate coulombic sums and thus does not re-
quire the derivation of atomic charge parameters. This ad-
vantage is obtained at the price of poor selectivity among
structures with small differences in coulombic energy. Point-
charge coulombic energies were separately calculated by
summations over atomic charge parameters obtained from
the electrostatic potential fit procedure (“pop=ESP” com-
mand) embedded in the Gaussian program suite.[18]
(Scheme 1). The molecules are
composed of a “carrier” ring
(the phthalic or perfluoroph-
thalic acid ring) and two hydro-
A more reliable evaluation of intermolecular interaction
energies is obtained by the Pixel method.[19] In this ap-
proach, the molecular electron density is first calculated by
standard quantum chemical methods to give a delocalized
description of the electron distribution by a large number
(ca. 10000) of negative-charge pixels. The coulombic energy
is then calculated by sums over pixel–pixel, pixel–nucleus,
and nucleus–nucleus coulombic terms. A local polarizability
genated
or
fluorinated
“prongs”. The two prongs have
torsional freedom with respect
to the carrier ring, but the over-
all constitution of the molecule
Scheme 1.
Chem. Eur. J. 2006, 12, 3538 – 3546
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3539

A. Gavezzotti et al.
is then assigned to each pixel. The electric field generated
by pixels and nuclei in surrounding molecules is calculated,
and the linear, static polarization energy is evaluated; an
empirical damping function, using one disposable parame-
ter, is introduced to avoid singularities. The intermolecular
overlap between molecular electron densities is calculated,
and the exchange repulsion energy is evaluated as propor-
tional to the overlap integral, with a correction due to the
electronegativity differences; the formulation requires two
more disposable parameters.[15]
A London-type formula is used to evaluate dispersion en-
ergies. The London approach[20] requires an estimate of the
“oscillator strength” of the interacting electrons, which is
usually approximated by the molecular ionization potential.
In the original formulation of the Pixel method, this quanti-
ty was taken as the energy of the HOMO, under the as-
sumption that the interacting electrons would be the periph-
eral ones and hence roughly at the energy level of the
HOMO. This assumption works reasonably well with small
molecules containing C, H, N, or O atoms, but runs into
trouble in heavily fluorinated aromatic compounds, where
the interacting electrons belong in fact, at least in good part,
to the fluorine atoms, whose ionization potential is much
lower than the energy of the HOMO, which is usually a p-
type molecular orbital. In an attempt to correct this defi-
ciency, the London oscillator strength Lp was calculated sep-
arately for each pair of interacting molecules, taking into ac-
count the different nature of the interacting electrons
[Eq. (1)],
X
Lp ¼
ijðIio þ IjoÞ=2 Sij
ð1Þ
where the summation runs on all atomic species in the mole-
cules, Iio is a rescaled atomic ionization potential for atomic
species i, and Sij the percentage of the total overlap between
the charge densities due to species i and j. The rescaled ioni-
zation potentials were taken as 0.300, 0.362, and 0.463 har-
tree for carbon, hydrogen, and fluorine, respectively. The
procedure amounts in fact to taking different dispersion-
energy coefficients according to the different kinds of ap-
proaching atomic basins. The overall performance of this
modification was checked by calculating the lattice energy
of a few crystals for which the heat of sublimation is known
(see Table 2).
The Pixel method has been successfully applied to the cal-
culation of energies of gas-phase dimers, where it has been
demonstrated[15] that the quality of the Pixel results is often
similar to that of quantum chemical calculations, at a frac-
tion of the computational cost. The Pixel method allows the
calculation of lattice energies in good agreement with crystal
sublimation enthalpies for a wide selection of organic com-
pounds, and also performs well in energy ranking for poly-
morphs of organic crystal structures.[21]
Table 2. Aryl–perfluoroaryl crystals: lattice energies and cohesion energies of the molecular stacked dimer (kJmolꢀ1).
[b]
Compound
Molecular complex with
Crystal: CSD refcode[a]
DHsubl
ꢀE
UNI
N
ꢀE
G
UNI
perfluoronaphthalene
perfluorobiphenyl
4,4’-difluorobiphenyl
hexafluorobenzene
1,2,3,4,5-pentafluorobiphenyl
1-phenyl-2-perfluorophenylacetylene
1-phenyl-2-perfluorophenylethylene
hexafluorobenzene
–
–
–
–
–
–
–
OFNAPH01
DECFDP01
ZZZAOS02
computer-generated[d]
PFBIPH
79*
81
92
93
54
93
101
110
99
124
162
146
153
173
177
189
207
157
174
173
180
161
259
199
73
81
98
43
97
-
-
-
86*
91*
49*
81
92
100
90
118
145
143
145
154
171
173
187
146
165
158
162
154
228
184
41
48
51
18
28
38
32
21
39
46
52
55
36
38
42
44
27
75
47
36
45
ASIJIW
SERQEL
101
benzene
naphthalene
pyrene
anthracene
trans-stilbene
biphenyl
anthracene
pyrene
triphenylene
naphthalene
diphenylacetylene
acenaphthene
biphenyl
naphthalene
triphenylene
diphenylacetylene
BICVUE01
IVOBOK
ZZZGKE01
ZZZGMW01
TIJTUB
ASAKIO
ECUTUR
ECUVIH
ECUVON
NPOFNP
OCAYIA
XUNJAR
BPPFBP
CEKYUM
CUKXIP
20
perfluoronaphthalene
47
perfluorobiphenyl
52
30
perfluorotriphenylene
perfluorodiphenylacetylene
ASIJER
46
[a] CSD: see reference [23]. Table S1 (Supporting Information) contains full literature references. [b] Values with asterisk: experimental heats of sublima-
tion; other values: sum of the sublimation enthalpies of the corresponding hydrocarbons (benzene 45, naphthalene 73, anthracene 98, pyrene 100, stil-
bene 100, biphenyl 81, triphenylene 114, diphenylacetylene 92, acenaphthene 85 kJmolꢀ1). All data from reference [24]. [c] For one-component crystals,
stacking energy of the parallel dimer showing aryl–perfluoroaryl interaction; for binary crystals, stacking energy of the complex dimer. Unoptimized geo-
metries as found in the crystal. [d] The experimental crystal structure of hexafluorobenzene is rather poor; a computer-generated monoclinic polymorph
with one molecule in the asymmetric unit was used instead.
3540
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3538 – 3546

Arene–Perfluoroarene Stacking Interactions
FULL PAPER
Quantum mechanical molecular energies and electron
densities were calculated by Gaussian[18] at the MP2/6-
31G** level. All molecular geometries were fixed as extract-
ed from the corresponding crystal structures, except for the
Figures S2 and S3 (Supporting Information) show an impres-
sive gallery of packing diagrams for the crystal structures in
Table 2 in which the parallel-stacked interaction motif be-
tween an aryl and a perfluoroaryl moiety appears. This
motif is ubiquitous for any kind of aromatic pair. For an as-
sessment of the implied energies, we first calculated the in-
teraction energies of isolated benzene–benzene, benzene–
hexafluorobenzene, and hexafluorobenzene–hexafluoroben-
zene dimers (Table 3). Complete fluorination brings about a
ꢀ
ꢀ
usual renormalization of C H geometries with a C H dis-
tance of 1.08 Š. The lattice energies were calculated by in-
cluding in the crystal model all molecules up to a separation
between molecular centers of 18 Š. All crystal-packing cal-
culations were carried out using the OPiX program pack-
age,[22] which includes a packing analysis and atom–atom lat-
tice energy calculation module (ZipOpec), a polymorph
generation module including a lattice energy minimizer
(Prom-Minop modules), and a module for the calculation of
the dimer and lattice energies by the Pixel method.
Table 3. Pixel results for the binding energies [kJmolꢀ1
stacked dimers at an inter-ring distance of 3.6 Š.
]
of parallel
[a]
[b]
[c]
[d]
[e]
Compound
Ecoul
Epol
Edisp
Erep
Etot
C6H6–C6F6
C6F6–C6F6
C6H6–C6H6
ꢀ7.5
2.7
2.7
ꢀ2.2
ꢀ2.1
ꢀ2.1
ꢀ22.2
ꢀ22.4
ꢀ22.0
11.7
10.7
12.9
ꢀ20.2
ꢀ11.2
ꢀ8.5
The Cambridge Structural Database[23] was searched for
crystal structures of compounds which could undergo aryl–
perfluoroaryl interactions, and the NIST database[24] was
searched for thermodynamic functions of aromatic hydrocar-
bons and aromatic fluorinated compounds. The results are
collected in Table 2.
[a] Pixel coulombic energy. [b] Pixel polarization energy. [c] Pixel disper-
sion energy. [d] Pixel repulsion energy. [e] Total Pixel energy for the mo-
lecular pair.
20% stabilization with respect to the hydrocarbon; mixed
interaction doubles the dimerization energy. While the bulk
of stabilization comes from dispersion, as is easily predicta-
ble in aromatic systems with polarizable p-electrons, the
energy difference between homodimers and the heterodimer
is entirely due to the coulombic-polarization term. In view
of these results, it is not correct to state that the arene–per-
fluoroarene dimerization energy is essentially coulombic;
rather, the coulombic component makes the difference in
stabilization.
The UNI force field arene–perfluoroarene stacking ener-
gies (Table 2) are strikingly similar to those obtained by
much more sophisticated methods; however, this is probably
due to a cancellation of errors, since the UNI force field
overestimates the aromatic stacking energy and misses the
extra coulombic stabilization in the heterodimer. The
arene–perfluoroarene energy advantage appears clearly
from the data in Table 2: on average, the lattice energies of
mixed compounds are 10–20% larger than the sum of the
sublimation enthalpies of the corresponding hydrocarbons.
The dimerization energies are additive over the number of
rings: for example, benzene–hexafluorobenzene 18, naph-
thalene–perfluoronaphthalene 36, triphenylene–perfluorotri-
phenylene 75 kJmolꢀ1; for the same perfluorinated moiety,
the dimerization energy increases smoothly with increasing
number of rings in the hydrogenated partner.
Results and Discussion
Single-crystal X-ray diffraction of 1–4 reveals that all intra-
molecular bond lengths and angles are within normal
ranges: C F distances are 1.332–1.340 Š in H-HF and F-
ꢀ
HH, and range from 1.32 to 1.35 Š in H-FF. Molecules in
the crystal (Figure S1, Supporting Information) exhibit a
wide range of conformational flexibility. Inter-ring angles
have been calculated as the angle (<908) formed by the unit
vectors along the direction of maximum inertia of the ring
formed by the six carbon atoms. They are (carrier–prong 1,
carrier–prong 2, prong 1–prong 2, respectively): 42, 30, and
578 in H-HH; 21, 55, and 578 in H-HF; 60, 65, and 708 in H-
FF, molecule A; 20, 87 and 88 in H-FF, molecule B; and 5,
75 and 728 in F-HH. These angles thus span the full range
between coplanar and orthogonal, and are different even in
the two molecules in the asymmetric unit of H-FF. This is
clear evidence that molecular conformation in the crystal is
driven by intermolecular forces.
There are practically no short atom–atom contacts involv-
ing fluorine atoms: just one H···F distance of 2.41 Š (sum of
atomic radii 2.56 Š),[25] and one C···F distance of 2.95 Š
(sum of atomic radii 3.23 Š), both in H-FF. We thus see no
need for a discussion of special atom–atom bonding interac-
tions involving fluorine.[10] Other short atom–atom contacts
involve the ester groups: O···H (sum of average atomic radii
2.68 Š): 2.34, 2.41, and 2.45 Š in H-FF; 2.46 Š in H-HH;
and 2.52 Š in H-HF; C···O (sum of atomic radii 3.35 Š):
3.13 Š in H-FF and a strikingly short 3.09 Š in F-HH. Dis-
cussing short contacts in terms of atom–atom bonding
hardly seems advisable; inasmuch as a short contact implies
a bond, then one should also discuss an incipient intermolec-
The Pixel partitioned lattice energies are listed in Table 4
along with the atom–atom energies. The Pixel method
cannot be applied to the H-FF crystal, because of a techni-
cal problem with two molecules in the asymmetric unit.
Pixel and UNI total lattice energies agree in the relative
ranking. In the crystals of these aromatic compounds, the
dispersion contributions mostly come from interactions be-
tween the electron densities of the aromatic ring, and the
most effective arrangement is a parallel-stack structure. Sta-
bilizing coulombic interactions may come from contacts be-
ꢀ
ular F C···O=C bond in the last-named cases above.
For comparison, the Cambridge Database was searched
for characteristic arene–perfluoroarene stacking patterns.
Chem. Eur. J. 2006, 12, 3538 – 3546
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3541

A. Gavezzotti et al.
Table 4. Pixel partitioned lattice energies [kJmolꢀ1].
[a]
[b]
[c]
[d]
[e]
[f]
[g]
Compound
Ecoul
Eqq
Epol
Edisp
Erep
Etot
EUNI
H-HH
H-HF
H-FF
F-HH
ꢀ48
ꢀ38
–
ꢀ32
ꢀ24
ꢀ18
–
ꢀ16
ꢀ18
ꢀ15
–
ꢀ13
ꢀ179
ꢀ170
–
ꢀ155
87
70
–
ꢀ157
ꢀ153
–
ꢀ137
ꢀ161
ꢀ158
ꢀ158
ꢀ149
63
[a] Pixel coulombic energy. [b] Coulombic energy estimated by the point-
charge method. [c] Pixel polarization energy. [d] Pixel dispersion energy.
[e] Pixel repulsion energy. [f] Total Pixel lattice energy. [g] Lattice energy
estimated by the UNI atom–atom force field.[17]
tween positively charged rims and negatively charged cores
of the benzene rings in a T-shaped arrangement, or from
contact between positively charged aromatic hydrogen re-
gions and negatively charged oxygen regions at the ester
linkage, mostly at or around the carbonyl oxygen atoms
ꢀ
(what is usually denoted briefly as C H···O interactions).
Destabilizing coulombic interactions are expected between
parallel-stacked benzene rings, while stabilizing coulombic
interactions are expected between benzene and fluoroben-
zene moieties. In fact, both the H-HH and the H-HF crystal
structures are largely built on dispersive interactions. How-
ever, the H-HH crystal has a larger coulombic component,
perhaps contrary to expectation: the reason is that, being
unable to take advantage of aryl–perfluoroaryl interactions,
this structure relies more heavily on stabilizing contact be-
ꢀ
tween C H···O interactions (see below). Interactions be-
ꢀ
tween C H hydrogen atoms and the benzene p clouds are
significantly present only in the F-HH crystal. Taken togeth-
er, these results outline a broad ranking of packing effects
Figure 1. The H-HH crystal structure: molecular dimers A, B, and C
(a to c).
ꢀ
in these crystals, as aryl–perfluoroaryl>C H···O coulom-
ꢀ
bic>C H···p.
The F-HH crystal structure (cell volume 433 Š per mole-
cule) is less densely packed than the H-HF structure (cell
volume slightly smaller, 424 Š per molecule, in spite of
having one fluorine atom more than F-HH): this is clearly
reflected in the smaller lattice energy of F-HH, as calculated
by both UNI and Pixel; the latter calculation reveals a sub-
stantial loss in dispersion energy. The reasons for this must
be somehow related to differences in electron density of
and dispersion contributions. This pair shows efficient inter-
locking of the aromatic rings, in which the two prongs of
one molecule embrace one prong in the other, but without
arene stacking; at the same time, one of the prongs is able
to form a favorable contact with the oxygen-rich region of
ꢀ
the carrier, with a short C=O···H C distance of 2.46 Š. The
second pair (B) shows a contact between an almost parallel
arrangement of the carrier and one prong in the two mole-
cules, and the second prong is turned away from contact.
This arrangement provides another source of substantial
coulombic stabilizing interaction, with O···H distances of
around 2.6 Š; the dispersion contribution is smaller, due to
smaller ring overlap, but the repulsion is also smaller, for
the same reason, and the two effects balance out. The third
pair (C) shows a significant decrease in coulombic stabiliza-
tion and a resisting amount of dispersion stabilization, in
compliance with a more parallel stack-type arrangement.
The next pair (D) is essentially a stacking of the two carrier
rings, with a small coulombic contribution and a still signifi-
cant dispersion contribution. While the first pairs have in-
version-center symmetry, the next three pairs consist of rib-
bons along glide plane or screw axes. Consistent with the
complex shape of the molecule, these string symmetries are
less effective in crystal stabilization.
CF5 and CF4(COO)2 rings.
G
In the H-HH crystal (Table 5 and Figure 1) the most sta-
bilizing structural pair (A) also has the largest coulombic
Table 5. Interacting pairs (structure determinants) in the H-HH crystal
(see Figure 1).
[b]
[c]
[d]
[e]
[f]
[g]
Symmetry operator[a] Ecoul
Epol
Edisp
Erep
EPixel
EUNI
A (I; 0, 1, 1)
ꢀ21
ꢀ19
ꢀ13
ꢀ9
ꢀ8
ꢀ6
ꢀ4
ꢀ3
ꢀ4
ꢀ46
ꢀ36
ꢀ40
ꢀ31
ꢀ27
26
16
23
14
14
ꢀ49
ꢀ45
ꢀ35
ꢀ30
ꢀ25
ꢀ40
ꢀ36
ꢀ38
ꢀ31
ꢀ27
B (I; ꢀ1, 1, 1)
C (I; 0, 0, 1)
D (I; ꢀ1, 0, 1)
E (G; 0, 1/2, ꢁ1/2)
ꢀ8
[a] Symmetry labels: I, inversion center; G, glide plane. [b] Pixel coulom-
bic energy. [c] Pixel polarization energy. [d] Pixel dispersion energy.
[e] Pixel repulsion energy. [f] Total Pixel energy for the molecular pair.
[g] Energy of the molecular pair estimated by the UNI atom–atom force
field.[17] Energies in kJmolꢀ1
.
3542
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3538 – 3546

Arene–Perfluoroarene Stacking Interactions
FULL PAPER
Contrary to the crystal structure of H-HH, which shows
almost a continuum of cohesive energies of the molecular
pairs, the crystal structure of H-HF (Table 6) shows a largely
Table 6. Interacting pairs (structure determinants) in the H-HF and F-
HH crystals (see Figures 2–4).
Symmetry operator[a]
Ecoul
Epol
Edisp
Erep
Etot,Pixel
Etot,UNI
H-HF
A (T; 0, 1, 0)
B (S; 0, ꢁ1/2, 1/2))
C (G; 0, ꢀ1/2, ꢁ1/2)
D (I; 1, 0, 1)
F-HH
ꢀ8
ꢀ13
ꢀ7
ꢀ4
ꢀ5
ꢀ3
ꢀ1
ꢀ58
ꢀ28
ꢀ24
ꢀ28
20
19
10
8
ꢀ50
ꢀ27
ꢀ25
ꢀ24
ꢀ61
ꢀ23
ꢀ23
ꢀ23
Figure 3. H-HF crystal: molecular pair D.
ꢀ2
A (I; 1,1,0)
B (I; 1,1,1)
C (I; 0,0,0)
ꢀ12
ꢀ10
ꢀ13
ꢀ5
ꢀ3
ꢀ4
ꢀ75
ꢀ29
ꢀ24
35
10
14
ꢀ57
ꢀ32
ꢀ28
ꢀ74
ꢀ28
ꢀ23
no special features in terms of structure and hence are less
attractive to the chemistꢁs eye. The result is that these im-
portant pairs are often neglected in cursory crystal-packing
analyses based only on short atom–atom distances or other
conspicuous structural features.
[a] Symmetry labels: I, inversion center; G, glide plane; S, screw axis; T,
identity. See footnotes to Table 5 for other symbols.
The F-HH structure determinants (Table 6) are somewhat
similar in energy and configuration to those of the H-HF
crystal: an arene–perfluoroarene stacking interaction again
dominates the packing (Figure 4), while the second best rec-
ognition mode is a double perpendicular arene–arene motif,
ꢀ
classifiable as a C H···p interaction, although the coulombic
Figure 2. The H-HF crystal: molecular dimers A (a) and B (b).
predominant pair (A, Figure 2) whose cohesive energy is
more than twice that of any other pair. This pair is clearly
formed over a strongly predominant energetic driving force,
that is, the simultaneous stacking of offset arene and arene–
perfluoroarene rings, easily obtained by pure translation.
The nature of this interaction is almost completely disper-
ꢀ
sive, and coulombic terms contribute very little. The C
H···O-type interactions appear only in the second best pair
(B, Figure 2), with a herringbone structure reaching a mod-
erate coulombic stabilization at the expense of a large de-
crease in dispersion stabilization. The C pair has another
arene–perfluoroarene stacking mode, while further pairs
(Figure 3) are less interpretable, good examples of what
may be called “anonymous” structural pairs, which contrib-
ute in a substantial way to the crystal stabilization but have
Figure 4. The F-HH crystal: molecular pairs A, B, and C (a to c).
Chem. Eur. J. 2006, 12, 3538 – 3546
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3543

A. Gavezzotti et al.
contribution is not particularly large. The third determinant
shown in Figure 4 is another example of “anonymous”
dimer: if a particular motif is to be recognized at all, it
would be a sort of incipient “nucleophilic attack” of a car-
bonyl oxygen atom on a fluorine-carrying, positively charg-
sults, with columns of stacked single rings and alternating
ribbons of fluorinated and hydrogenated rings. Since nearly
all cohesive forces in this crystal come from aromatic-ring
stacking, the most relevant energy contribution is the disper-
sion contribution. The smaller coulombic contribution in
dimers A and B results from contact between ester oxygen
ꢀ
ed carbon atom (very short F C···C=O distance of 3.09 Š).
ꢀ
The largely dominant driving force in the packing of the
atom and benzene rim (C H) regions, which involves some
H-FF crystal (Table 7) is again an arene–perfluoroarene
short O···H distances of 2.41 (A), 2.54 (B), and 2.45 Š (C).
The shortest O···H distance (2.33), in pair F, which is not
among the highest ranking in the crystal, confirms that
O···H distances are not reliable indicators of cohesive
strength.
Table 7. Interacting pairs (structure determinants) in the H-FF crystal
(see Figure 5).
Symmetry operator[a]
Ecoul Epol Edisp Erep EPixel EUNI
A (Z; 0, 0, 0)
ꢀ14
ꢀ13
ꢀ6
ꢀ5 ꢀ51
ꢀ4 ꢀ31
ꢀ4 ꢀ40
ꢀ4 ꢀ34
ꢀ2 ꢀ33
22
11
21
17
15
ꢀ49
ꢀ37
ꢀ28
ꢀ28
ꢀ21
ꢀ55
ꢀ39
ꢀ38
ꢀ21
ꢀ33
B (Z; a translation, 0, ꢀ1, 0)
Conclusion
C (Ia; 1, 2, 1)
D (Tb; 1, 0, 0)
E (Z; a translation, ꢀ1, 0, 0)
ꢀ7
ꢀ1
Our theoretical and experimental analysis quantitatively
confirms the qualitative idea that the stacked arene–per-
fluoroarene recognition pattern is a stable and reproducible
one. The crystal structures of isomorphic compounds in
which benzene and perfluorobenzene rings alternate show a
preference for organization in heterodimers (H···F rings)
versus homodimers (H···H or F···F rings). A database study
of arene–perfluoroarene molecular complexes reveals that
the dimer forms the asymmetric unit in the crystal and in-
variably has the closest intermolecular contact. Quantitative
analysis shows that it is a significant driving force in crystal
packing; with a cohesive energy of 20–25 kJmolꢀ1 per
phenyl ring, it is second only to hydrogen bonding in car-
boxylic acids and amides (30–35 kJmolꢀ1 per bond), it com-
petes with weaker hydrogen bonds such as those found in al-
cohols (25 kJmolꢀ1 per bond), and it largely overcomes all
kinds of weaker binding modes commonly invoked in crys-
[a] Symmetry labels: Z, symmetry-unrelated a and b molecules in the
asymmetric unit; I, inversion center; T, identity. See footnotes to Table 5
for other symbols.
stacking interaction, which is evident in all the highest rank-
ing molecular pairs (Figure 5). So strong is the drive to
arene–perfluoroarene interaction that the crystal includes
ꢀ
tal-packing analyses, such as C H···O hydrogen bonding,
which cannot exceed 10 kJmolꢀ1.[15] This energetic stability
does not, however, entail a stringent structural form: the dis-
tances between ring centroids in aromatic–perfluoroaromat-
ic contacts and the corresponding inter-ring angles found in
the crystal structures of the compounds we have examined
are 3.70 (5), 3.95 (6), 4.17 (20), 4.75 (20), and 4.85 Š (218).
The arene–perfluoroarene “synthon” can thus bear a stretch
of as much as 1 Š and angular flexion of about 208.
Figure 5. The H-FF crystal: molecular pairs A and B. Arene–perfluoroar-
ene stacking occurs in two directions.
two molecules in the asymmetric unit, so that both fluorinat-
ed prongs have the possibility to interact with the hydrogen-
ated carrier. Figure 6 shows the impressive structure that re-
Calculations of partitioned energies by the Pixel method
clearly show that all stacked-arene cohesive interactions are
mainly dispersive in nature, and arise from electron correla-
tion between p-electron clouds. However, coulombic ener-
gies obviously differ, since those between homodimers are
generally repulsive and destabilizing by some 3 kJmolꢀ1,
while that in the heterodimer is stabilizing by about
8 kJmolꢀ1. Thus, cohesion is mainly generated by dispersion,
but structural selectivity depends on coulombic terms.
Our results in terms of Pixel partitioned energies produce
a clear quantitative picture of the energetic ranking of mo-
lecular recognition modes in organic crystals. While the ab-
solute values may vary by a few kilojoules per mole, in
more accurate (e.g., fully ab initio) treatments, the relative
orderings do not.[15] Even rough estimates based on empiri-
Figure 6. Crystal packing of H-FF. Horizontal: b translation, vertical: a
translation. The columns of alternating arene–perfluoroarene rings seen
edge-on are clearly visible. The ribbons of perfluoroarene rings sit on top
of corresponding ribbons of arene rings. This view is perpendicular to the
view in Figure 5.
3544
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3538 – 3546

Arene–Perfluoroarene Stacking Interactions
FULL PAPER
Diphenyl 3,4,5,6-tetrafluoro-1,2-benzenedicarboxylate (4, F-HH): A cata-
lytic amount of DMF (1 drop) was added to a stirred solution of tetra-
fluorophthalic acid (0.5 g, 2.10 mmol) in thionyl chloride (20 mL), and
the resulting mixture was refluxed under nitrogen for 18 h. The reaction
mixture was then evaporated under vacuum and the crude oil was used
as such in the next step. A solution of 4-dimethylaminopyridine (0.10 g,
0.84 mmol) and phenol (0.37 g, 3.73 mmol) in pyridine (1.4 mL) was
added to a stirred solution of the dichloride (0.46 g, 1.69 mmol) in dry di-
chloromethane (5 mL). The yellow solution was stirred under nitrogen
for 24 h; the reaction was quenched by addition of water (5 mL), and the
resulting mixture was extracted with EtOAc (3”10 mL). The combined
organic phases were washed with water (4”10 mL), dried over sodium
sulfate, filtered, and concentrated under vacuum to afford a red oil. This
was purified by flash chromatography with a cyclohexane:EtOAc (95:5)
as eluant to afford the product as a white solid (0.305 g, 0.781 mmol,
46% yield). This was crystallized from pentane to give large crystals
(m.p. 778C). IR (Nujol): n˜ =1748.3, 1519.0, 1482.8, 1458.7, 1213.2,
cal atom–atom force fields can give acceptable quantitative
insights. This analysis further[26,27] shows that when all inter-
acting neighbors in a crystal structure are properly consid-
ered with their relative energies, conclusions based on
simple analysis of some short distances and preconceived
“synthons” are easily overthrown.
Experimental Section
Synthesis: Diphenyl 1,2-benzenedicarboxylate[28] (1, H-HH) and bis(pen-
tafluorophenyl) 1,2-benzenedicarboxylate[29] (2, H-FF) were prepared ac-
cording to literature procedures. They were purified by recrystallization
from hexane to afford analytically pure, sharply melting materials (1:
m.p. 768C; lit.:[28] 74–768C. 2: m.p. 968C; lit.:[29] 96–978C).
1184.3 cmꢀ1 1H NMR (400 MHz, [D6]DMSO): d=7.51 (t, J=8.0 Hz,
.
Pentafluorophenyl phenyl 1,2-benzenedicarboxylate (3, H-HF): Triethyl-
amine (0.155 mL, 5.41 mmol) and then a solution of pentafluorophenol
(0.907 g, 4.92 mmol) in dry THF (2 mL) were added to a stirred solution
of phthaloyl dichloride (1.0 g, 4.92 mmol) in dry THF (8 mL). A white
precipitate formed. The suspension was stirred at room temperature for
3 h, and the reaction was quenched by addition of saturated aqueous am-
monium chloride (15 mL). The organic phase was separated, and the
aqueous phase was extracted with dichloromethane (2”15 mL). The
combined organic phases were dried over sodium sulfate, filtered, and
concentrated under vacuum to afford a yellow oil (1.73 g). The crude
product, which was contaminated with about 50% of phthaloyl dichloride
by HPLC analysis, was purified by flash chromatography with cyclohexa-
ne:dichloromethane (9:1) as eluant. The obtained product (0.8 g), which
was still contaminated with phthaloyl dichloride (ca. 20% by HPLC),
was dissolved in EtOAc and subjected to flash chromatography on a
short column with cyclohexane:ethyl acetate (9:1) as eluant. On standing,
the obtained colorless oil (0.703 g),
4H), 7.38 (t, J=8.0 Hz, 2H), 7.23 (d, J=7.6 Hz, 4H), 7.59 (dt, J=1.2,
7.6 Hz, 1H), 7.45 (t, J=8.0 Hz, 1H), 7.25–7.32 ppm (m, 3H); 13C NMR
(100 MHz, [D6]DMSO): d=160.5, 150.2, 146.2 (JCꢀF =250.0 Hz), 143.0
(JCꢀF =252.0 Hz), 130.5, 127.4, 121.6, 116.3 ppm; 19F NMR (376 MHz,
[D6]DMSO): d=ꢀ147.4 (d, J=15.0 Hz, 2F), ꢀ136.3 ppm (d, J=15.0 Hz,
2 F); elemental analysis (%) calcd for C20H10F4O4: C 61.55, H 2.58;
found: C 61.87, H 2.43.
X-ray crystallography: Data were collected at room temperature on a
CAD4 diffractometer. No absorption corrections were applied. No inten-
sity decay was detected during data collection. The structures were
solved by direct methods (SIR2000)[30] and refined by full-matrix least-
squares technique on F2 (SHELXL-97).[31] Anisotropic thermal parame-
ters were assigned to all non-hydrogen atoms, while hydrogen atoms
were placed in calculated positions and refined using a riding model.
Crystallographic data are collected in Table 8. Further details: room tem-
which was still contaminated with
about 5% phthaloyl chloride (by
Table 8. Crystal data.[a]
HPLC), became a white waxy solid. It
was used without further purification.
Compound Space group, Z
a
b
c
a
b
g
1exptl [gcmꢀ3
]
V[b] [Š3]
A solution of triethylamine (0.11 mL,
0.784 mmol) and phenol (0.067 g,
0.713 mmol) in dry THF (2 mL) was
H-HH
H-FF
H-HF
F-HH
P21/c, 4
P1, 4
8.884 12.580 14.507
7.532 12.784 20.287 100.10
18.590 5.975 17.073
7.709 10.560 11.945 105.09
–
98.32
97.27
–
1.318
401.1
476.4
424.0
433.0
¯
91.40 1.737
1.599
P21/c, 4
–
116.57
99.75 106.92 1.497
–
added to
product (0.250 g, 0.713 mmol) in dry
THF (5 mL). white precipitate
a stirred solution of this
¯
P1, 2
[a] Cell edges in angstroms, cell angles in degrees. Typical standard deviations are 0.002 Š for cell edges, and
0.028 for cell angles. [b] Cell volume per mole.
A
formed. The suspension was stirred at
room temperature for 3 h, and the re-
action was quenched by addition of sa-
turated aqueous ammonium chloride
perature; least squares on reflections with I>2s(I); 1: C20H14O4, M=
318.31, m=0.09 mmꢀ1, 1825 reflections and 217 parameters, final R1=
(15 mL). The organic phase was separated, and the aqueous phase was
extracted with ethyl acetate (2”15 mL). The combined organic phases
were dried over sodium sulfate, filtered, and concentrated under vacuum
to afford a colorless oil (0.281 g). This was purified by flash chromatogra-
phy with cyclohexane:EtOAc (95:5) as eluant to afford three fractions.
The first (0.114 g) was unconverted starting material, which was recycled;
the second (0.043 g) was a mixture of starting material and product,
which was purified by flash chromatography; and the third fraction
(0.025 g) was the product, which was >90% pure by HPLC. The overall
yield of product was 15% (0.047 g). White needles (m.p. 107.5–1088C)
were obtained from dichloromethane:hexane. IR (Nujol): n˜ =1764.8,
0.045, wR2=0.147, GOF=1.074; 2: C20H4O4F10
, M=498.23, m=
0.18 mmꢀ1, 3358 reflections and 613 parameters, final R1=0.048, wR2=
0.207, GOF=0.872; 3: C20H9O4F5, M=408.27, m=0.15 mmꢀ1, 1580 reflec-
tions and 262 parameters, final R1=0.058, wR2=0.118, GOF=0.994; 4:
C20H10O4F4, M=390.28, m=0.13 mmꢀ1, 2533 reflections and 253 parame-
ters, final R1=0.039, wR2=0.114, GOF=1.012.
CCDC 285979–285982 contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from the Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
1754.5, 1519.0, 1464.3, 1262.4, 1092.8, 1043.1 cmꢀ1 1H NMR (400 MHz,
;
CDCl3): d=10 (dd, J=1.2, 7.6 Hz, 1H), 8.03 (dd, J=1.2, 7.6 Hz, 1H),
7.80 (dt, J=1.2, 7.6 Hz, 1H), 7.59 (dt, J=1.2, 7.6 Hz, 1H), 7.45 (t, J=
8.0 Hz, 1H), 7.25–7.32 ppm (m, 3H); 13C NMR (100 MHz, CDCl3): d=
165.5, 162.7, 150.7 (2 C), 141.3 (JCꢀF =251.0 Hz), 141.0 (JCꢀF =252.5 Hz),
138.0 (JCꢀF =254 Hz), 133.0, 132.8, 131.7, 130.0, 129.8, 129.5, 128.5, 126.2,
121.3 ppm; 19F NMR (376 MHz, CDCl3): d=ꢀ151.7 (d, J=21.1 Hz, 2F),
ꢀ157.3 (t, J=22.6 Hz, 1F), ꢀ161.9 ppm (d, J=19.9 Hz, 2F); elemental
analysis (%) calcd for C20H9F5O4: C 58.84, H 2.22; found: C 59.48, H
2.05.
Acknowledgements
The pictures were drawn with the help of the program SCHAKAL.[32] Fi-
nancial support by MIUR (FIRB 2001 grant number RBAU018XA7 to
F.C.J. is gratefully acknowledged.
Chem. Eur. J. 2006, 12, 3538 – 3546
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3545

A. Gavezzotti et al.
[13] S. Meejoo, B. M. Kariuki, K. D. M. Harris, ChemPhysChem 2003, 4,
766–769.
[14] H. Adams, J.-J. Jimenez Blanco, G. Chessari, C. A. Hunter, C. M. R.
Low, J. M. Sanderson, J. G. Vinter, Chem. Eur. J. 2001, 7, 3494–
3503.
[1] C. R. Patrick, G. S. Prosser, Nature 1960, 187, 1021.
[2] For an early review, see: J. H. Williams, Acc. Chem. Res. 1993, 26,
593–598; for quantum-chemical theoretical calculations on aromat-
ic–fluoroaromatic systems, see: M. O. Sinnokrot, E. F. Valeev, C. D.
Sherrill, J. Am. Chem. Soc. 2002, 124, 10887–10893; P. Hobza, H. L.
Selzle, E. W. Schlag, J. Phys. Chem. 1996, 100, 18790–18794; A. P.
West, S. Mecozzi, D. A. Dougherty, J. Phys. Org. Chem. 1997, 10,
347–350; J. Hernandez-Trujillo, F. Colmenares, G. Cuevas, M.
Costas, Chem. Phys. Lett. 1997, 265, 503–507; Y. Zhao, D. G. Truh-
lar, J. Phys. Chem. A J. Phys. Chem. B 2005, 109, 4209–4212; W. B.
Schweizer, J. D. Dunitz, J. Chem. Theor. Comp. 2005, 1, 834–840
and references therein; force-field treatment of aromatic interac-
tions: G. Chessari, C. A. Hunter, C. M. R. Low, M. J. Packer, J. G.
Vinter, C. Zonta, Chem. Eur. J. 2002, 8, 2860–2867. See also
ref. [15].
[3] a) J. Vrbancich, G. L. D. Ritchie, J. Chem. Soc. Faraday Trans. 2
1980, 648–659; b) M. Luhmer, K. Bartik, A. Dejaegere, P. Bovy, J.
Reisse, Bull. Soc. Chim. Fr. 1994, 131, 603–606; J. C. Collings, P. S.
Smith, D. S. Yufit, A. S. Batsanov, J.A.K. Howard, T. B. Marder,
CrystEngComm 2004, 6, 25–28, and other papers by the same group
in Table S1 (see Supporting Information).
[4] For a quantitative evaluation of the effect on arene–arene interac-
tion exerted by increasing fluorine substitution on one of two paral-
lel-stacked aromatic rings, see: F. Cozzi, F. Ponzini, R. Annunziata,
M. Cinquini, J. S. Siegel, Angew. Chem. 1995, 107, 1092–1093;
Angew. Chem. Int. Ed. Engl. 1995, 34, 1019–1020.
[5] a) C. A. Hunter, J.K. M. Sanders, J. Am. Chem. Soc. 1990, 112,
5525–5534; b) F. Cozzi, M. Cinquini, R. Annunziata, J. S. Siegel, J.
Am. Chem. Soc. 1993, 115, 5330–5331, and references therein;
c) F. J. Carver, C. A. Hunter, E. M. Seward, Chem. Commun. 1998,
775–776.
[15] The presence of a hydrogen bond can influence a crystal structure
so strongly that other, less energetic interactions, such as the arene–
perfluoroarene interaction, can be overwhelmed. For an assessment
of the relative energies associated with aromatic stacking and with
hydrogen-bond formation in crystals, see: A. Gavezzotti, J. Chem.
Theor. Comp. 2005, 1, 834–840, and references therein.
[16] A. Gavezzotti, G. Filippini, J. Am. Chem. Soc. 1995, 117, 12299–
12305.
[17] For the hydrocarbon part, see G. Filippini, A. Gavezzotti, Acta Crys-
tallogr. Sect. B 1993, 49, 868–880; the F···F interaction-energy curve
was optimized essentially by the same procedure, that is, by setting
the energy minimum at a distance slightly shorter than the preferred
contact distance in crystals of fluorinated compounds, and adjusting
the well depth to reproduce a few crystal heats of sublimation (see
Table 2).
[18] Gaussian03, RevisionA.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M.A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, M. Kiene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K.
Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakr-
zewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. J. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G.
Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox,
T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challa-
combe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonza-
lez, J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 2003.
[6] J. H. Williams, J.K. Cockroft, A. N. Fitch, Angew. Chem. 1992, 104,
1666–1668; Angew. Chem. Int. Ed. Engl. 1992, 31, 1655–1657.
[7] Review: K. Reichenbächer, H. I. Süss, J. Hulliger, Chem. Soc. Rev.
2005, 34, 22–30.
[8] For a review on the proteiform aspects of aromatic–aromatic inter-
actions and their implications in many areas of chemistry, see E. A.
Meyer, R. K. Castellano, F. Diederich, Angew. Chem. 2003, 115,
1244–1287; Angew. Chem. Int. Ed. 2003, 42, 1210–1250.
[19] a) A. Gavezzotti, J. Phys. Chem. 2002, B106, 4145–4154; b) A. Gav-
ezzotti, J. Phys. Chem. 2003, B107, 2344–2353; c) A. Gavezzotti,
CrystEngComm 2003, 5, 429–438.
[9] a) G. W. Coates, A. R. Dunn, L. M. Henling, D. A. Dougherty, R. H.
Grubbs, Angew. Chem. 1997, 109, 290–293; Angew. Chem. Int. Ed.
Engl. 1997, 36, 248–251; b) G. W. Coates, A. R. Dunn, L. M. Hen-
ling, J. W. Ziller, E. B. Lobkovsky, R. H. Grubbs, J. Am. Chem. Soc.
1998, 120, 3641–3649; c) C. Dai, P. Nguyen, T. B. Marder, A. J.
Scott, W. Clegg, C. Viney, Chem. Commun. 1999, 2493–2494;
d) M. L. Renak, G. P. Bartholomew, S. Wang, P. J. Ricatto, R. J. La-
chicotte, G. C. Bazan, J. Am. Chem. Soc. 1999, 121, 7787–7799; e) F.
Ponzini, R. Zagha, K. Hardcastle, J. S. Siegel, Angew. Chem. 2000,
112, 2413–2415; Angew. Chem. Int. Ed. 2000, 39, 2323–2325;
f) W. J. Feast, P. W. Lçvenich, H. Puschmann, C. Taliani, Chem.
Commun. 2001, 505–506; g) M. R. Haneline, M. Tsunoda, F. P.
Gabbaꢂ, J. Am. Chem. Soc. 2002, 124, 3737–3742.
[10] The occurrence of a favorable interaction between aromatic H and
aromatic F was sometimes suggested as a possible stabilizing factor
in these structures (e.g., see refs. [9d] and [9f]). However, it has
been noted that organofluorine almost never accepts hydrogen
bonds: J. D. Dunitz, R. Taylor, Chem. Eur. J. 1997, 3, 89–98.
[11] For examples dealing with DNA-like molecules, see: a) K.A. Frey,
S. A. Woski, Chem. Commun. 2002, 2206–2207; b) G. Mathis, J.
Hunziker, Angew. Chem. 2002, 114, 3335–3338; Angew. Chem. Int.
Ed. 2002, 41, 3203–3205; c) R. Faraoni, R. K. Castellano, V. Gram-
lich, F. Diederich, Chem. Commun. 2004, 370–371.
[20] F. London, Trans. Farady Soc. 1937, 33, 8–26.
[21] A. Gavezzotti, CrystEngComm 2003, 5, 439–446.
[22] A. Gavezzotti, Z. Kristallogr. 2005, 220, 499–510.
[23] F. H. Allen, O. Kennard, Chem. Des. Autom. News 1993, 8, 31–37.
[24] H. Y. Afeefy, J. F. Liebman and S. E. Stein, Neutral Thermochemical
Data; and J. S. Chickos, Heat of Sublimation Data, in NIST Chemis-
try WebBook, NIST Standard Reference Database Number 69 (Eds.:
P. J. Linstrom, W. G. Mallard), March 2003, National Institute of
Standards and Technology, Gaithersburg MD, 20899 (http://webbook.
nist.gov).
[25] R. S. Rowland, R. Taylor, J. Phys. Chem. 1996, 100, 7384–7391.
[26] A. Gavezzotti, Struct. Chem. 2005, 16, 177–185.
[27] J. D. Dunitz, A. Gavezzotti, Cryst. Growth Des. 2005, 5, 2180–2189.
[28] F. F. Blicke, O. J. Weinkauff, J. Am. Chem. Soc. 1932, 54, 330–334.
[29] N. Ito, H. Kudo, A. Kameyama, T. Nishikubo, T. Anadia, J. Polym.
Sci. Part A 2002, 41, 213–222.
[30] A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C. Giacovaz-
zo, A. Guagliardi, A. G. Moliterni, G. Polidori and R. Spagna, J.
Appl. Crystallogr. 1999, 32, 115–119.
[31] G. M. Sheldrick, SHELX-97, University of Goettingen, Germany
1997.
[32] E. Keller, SCHAKAL92, A Program for the Graphic Representa-
tion of Molecular and Crystallographic Models, University of Frei-
burg, 1993.
[12] For an example dealing with a catenane molecule, see: R. E. Gil-
lard, J. F. Stoddardt, A. J. P. White, B. J. Williams, D. J. Williams, J.
Org. Chem. 1996, 61, 4504–4505.
Received: October 10, 2005
Published online: February 28, 2006
3546
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3538 – 3546