Structural diversities of charge transfer organic complexes. Focus on benzenoid hydrocarbons and 7,7,8,8-tetracyanoquinodimethane

Article abstract of DOI:10.1039/c3ce41703d

We have obtained three-component systems: the complexes comprising of two different benzenoid hydrocarbons together with one molecule of 7,7,8,8-tetracyanoquinodimethane (TCNQ). The X-ray single-crystal structures of naphthalene-perylene-TCNQ and pyrene-perylene-TCNQ revealed that they form face-to-face stacking between perylene and TCNQ molecules containing another hydrocarbon as a guest in the structure. We also present a pyrene-TCNQ complex with two pyrene moieties acting in similar way. In this system, charge transfer is far more efficient than in any other complexes of TCNQ with benzenoid hydrocarbons. Additionally, we have also obtained as references pyrene-TCNQ, chrysene-TCNQ, phenanthrene-TCNQ and naphthalene-TCNQ with 1:1 molecular ratios. These systems were also used to estimate the degrees of charge transfer. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3ce41703d

CrystEngComm
View Article Online
View Journal | View Issue
PAPER
Structural diversities of charge transfer organic
complexes. Focus on benzenoid hydrocarbons
and 7,7,8,8-tetracyanoquinodimethane†
a
Cite this: CrystEngComm, 2014, 16,
415
Michał A. Dobrowolski, Gaston Garbarino,b Mohamed Mezouar,b
*
Arkadiusz Ciesielskiac and Michał K. Cyrańskia
We have obtained three-component systems: the complexes comprising of two different benzenoid
hydrocarbons together with one molecule of 7,7,8,8-tetracyanoquinodimethane (TCNQ). The X-ray
single-crystal structures of naphthalene–perylene–TCNQ and pyrene–perylene–TCNQ revealed that
they form face-to-face stacking between perylene and TCNQ molecules containing another hydrocarbon
as a guest in the structure. We also present a pyrene–TCNQ complex with two pyrene moieties acting
in similar way. In this system, charge transfer is far more efficient than in any other complexes of
TCNQ with benzenoid hydrocarbons. Additionally, we have also obtained as references pyrene–TCNQ,
chrysene–TCNQ, phenanthrene–TCNQ and naphthalene–TCNQ with 1 : 1 molecular ratios. These systems
were also used to estimate the degrees of charge transfer.
Received 26th August 2013,
Accepted 15th October 2013
DOI: 10.1039/c3ce41703d
www.rsc.org/crystengcomm
and acceptor (A) molecules are formed (⋯DADADADA⋯),
whereas in the second, the molecules segregate into parallel
Introduction
Organic charge-transfer complexes show a large variety of phys-
ical properties such as charge ordering, spin density waves
(SDW), spin-Peierls ground state and superconductivity.1 Poly-
cyclic aromatic hydrocarbons (PAHs) are typical examples of
π-electron systems able to donate electrons depending on
their ionization potential I. 7,7,8,8-tetracyanoquinodimethane
(TCNQ) is the archetypal example of an acceptor molecule
which can be easily reduced to form the open shell electron
radical anion TCNQ˙− when it is exposed to contact with elec-
tron donors. Benzenoid hydrocarbons containing in their crys-
tal structures molecules of TCNQ exhibit semiconducting
properties with a relatively small band gap and moderate
charge-transfer between the counterions.2 Moreover, the elec-
tron transfer can be tuned by composition, stoichiometry,
temperature or pressure. The salts of closed-shell organic elec-
tron donors with the acceptor TCNQ and stoichiometry 1 : 1
form crystal structures that usually fall into one of two broad
categories.3 In the first, alternating mixed stacks of donor (D)
stacks of donors (⋯DDDD⋯) and acceptors (⋯AAAA⋯).
Concerning electrical properties, materials of the first cate-
gory are limited by simple band theory to semiconducting
behavior, while the second type includes a broad range from
insulators to metals. Calculations on the electronic structure
showed that the tetracene–TCNQ complex is an indirect band
gap semiconductor with a valence band maximum at Y and
the charge transfer between the two constituent molecules is
0.1.4 Increasing the number of benzene rings in the PAH can
lead to better conducting properties. For instance, the degree
of charge transfer in coronene–TCNQ was estimated to be the
order of 0.3 and the structure showed semiconductor behav-
iour with an optical gap of 1.55 eV and a transport gap of
0.49 eV.2 A picene single crystal field-effect transistor with an
HfO2 gate dielectric and TCNQ-coated electrodes shows p-channel
characteristics with a smooth hole injection and a field effect
mobility of more than 1 cm2 V−1 s−1 in two-terminal measure-
ment.5 Recently, it has been shown that PAHs doped with
alkali metals revealed superconductivity.6 In the case of
phenanthrene changing the alkali metal to lanthanum or
samarium and application of high pressure even enhances
this effect.7 In BCS theory,8 the suppression effect on Tc is
usually expected as an external pressure is applied because
it compresses the crystal lattice and consequently reduces
the density of states at the Fermi level. Tc in those complexes
exhibits positive pressure dependence, strongly suggesting
unconventional superconductivity in the doped phenan-
threne. The investigations of angular PAHs doped with
a Faculty of Chemistry, University of Warsaw, Pasteura 1, 02-093 Warsaw, Poland.
E-mail: miked@chem.uw.edu.pl; Fax: +48 22 8222892; Tel: +48 22 8220211 ext. 360
b European Synchrotron Radiation Facility, 6 Rue Jules Horowitz, BP 220, 38043
Grenoble, France
c Institute of Biochemistry and Biophysics Polish Academy of Sciences,
Pawinskiego 5a, 02-106 Warsaw, Poland
† Electronic supplementary information (ESI) available. It contains labeling of
atoms, additional fingerprint plots, a full list of bond lengths, bond angles
and torsion angles of the experimentally studies systems (1)–(7). CCDC
957179–957185. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c3ce41703d
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 415

View Article Online
Paper
CrystEngComm
potassium revealed that superconductivity increases with the
number of benzene rings.9 For 1,2:8,9-dibenzopentacene the
transition temperature is as high as 33 K.
The aim of this paper is to show how different hydrocar-
bons could modify the packing of PAH–TCNQ complexes and
hence the degree of calculated charge transfer.
quantities of perylene and TCNQ were dissolved in molten naph-
thalene, then benzene was added and the crystals were grown.
All other crystals were obtained by dissolving stoichiometric
quantities of hydrocarbons and TCNQ in benzene [(2), (3)]
and toluene [(4), (5), (6), (7)]. The solutions of (4), (6) and (7)
were refluxed with stirring for 2 hours, then left to allow slow
evaporation of the solvent. In most cases single crystals suit-
able for further X-ray studies were obtained by slow evapora-
tion of benzene or toluene.
Results and discussion
Synthesis
Several benzenoid hydrocarbons (naphthalene, anthracene,
phenanthrene, chrysene, pyrene, and perylene) and/or their
stoichiometric (1 : 1) mixtures were used as a starting material
to form complexes with TCNQ. In the case of (1), equimolar
Structures
Structural details for all the complexes are given in
Tables 2, 3 and 4 (see later).
Fig. 1 The crystal packing along [001] direction in complexes (1) and (2).
Fig. 2 The crystal packing along [100] direction in complexes (1) and (2) showing overlap between perylene and TCNQ.
416 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Three-component systems
pyrene and TCNQ are similar with the angle between the
mean planes 2.1°. The six-membered TCNQ ring lies exactly
above and beneath the central C–C bond of pyrene with a
shift of only 1.3 Å. It is shown in Fig. 4b. The shift is defined
as the square root of the sum of the distance between two
centroids (which always intersect with the hydrocarbon cen-
troid) squared and the distance between two mean planes of
TCNQ squared (from the pythagorean theorem).
The intra-stack separation is 3.45 Å. Interestingly there are
short contacts between the stacks, TCNQ interacts with a
pyrene from a second “sandwich complex” with C–H⋯N dis-
tance 2.596 Å (see Fig. 5). It packs very closely to another
Naphthalene–perylene–7,7,8,8-tetracyanoquinodimethane (1) and
pyrene–perylene–7,7,8,8-tetracyanoquinodimethane (2) complexes
both crystallize in a triclinic system, space group P1, with half
a molecule of each hydrocarbon and half a molecule of TCNQ
forming the asymmetric part of the unit cell. In both struc-
tures, molecules form stacking between the perylene and
TCNQ along the [100] direction, whereas other compounds
(e.g. naphthalene or pyrene, see Fig. 1) act as guests. All the
molecules lie at inversion centers.
¯
As a mean distance between the molecules in stacks we
chose half of the distance between two best planes covering
TCNQ moieties. The intra-stack separation is comparable in
both cases. It is 3.31 Å for (1) and 3.32 Å for (2). This is very
close in value to the equivalent distance in the semiconduct-
ing coronene–TCNQ complex, which has been estimated as
3.33 Å.2 At the same time it is much shorter than in the par-
ent perylene–TCNQ complex (3.45 Å),10 suggesting that addi-
tional hydrocarbons may enhance the charge transfer of
these complexes (vide infra). Along the stacks the tilt angles
for perylene and TCNQ are very small, the angles between the
mean planes are 1.0° and 1.4°, respectively. Moreover, the
central ring of perylene is shifted with respect to the TCNQ
ring by ca. 2.6 Å and 2.9 Å (for explanation of the shift see
Fig. 4b). The naphthalene and pyrene molecules fit perpen-
dicularly between the stacks with angles to the mean plane of
perylene 87.8° and 89.3°, respectively, as shown in Fig. 2.
Besides π⋯π stacking, the crystal structures of (1) and (2)
are stabilized by weak C–H⋯N interactions between TCNQ
and perylene molecules, with the distances in the range of
2.602–2.614 Å and 2.533–2.695 Å for (1) and (2), respectively.
These interactions are shown in Fig. 3a and c. Additionally,
in (1) naphthalene interacts with TCNQ with a C–H⋯N
distance of 2.654 Å. Notably in the case of (2) interactions of
this kind are not present. Both structures are also stabilized
by C–H⋯π interactions with perylene molecules with the
distances between the centroid of naphthalene, pyrene rings
and H atom from perylene equal to 2.966 Å and 2.837 Å,
respectively. In both cases, infinite zigzag motifs are formed,
as shown in Fig. 3b and d.
Two-component systems
Pyrene–TCNQ complex 2 : 1 (3). This is an interesting case,
where there are two hydrocarbon molecules per TCNQ. One
is forming face-to-face π–π stacking with TCNQ along the
[100] direction and the second acts as a guest in the struc-
ture, interacting via C–H⋯π contacts (see Fig. 4a). This situa-
tion resembles the sandwich herringbone structures of
hydrocarbons classified by Desiraju & Gavezotti.11 The com-
¯
pound crystallizes in a triclinic system, space group P1 with
two halves of pyrene and half of the TCNQ forming the asym-
metric part of the unit cell. All the molecules lie at the inver-
sion centers, as in (1) and (2). The guest pyrene molecule is
somewhat tilted with the angle to the mean plane of stacked
hydrocarbon equal to 69.2°. Along the stacks, the tilts for
Fig. 3 (a) Short interactions in crystal of (1); (b) zigzag motif in (1).
(c) Short interactions in crystal of (2); (d) zigzag motif in (2).
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 417

View Article Online
Paper
CrystEngComm
Fig. 4 (a) The crystal packing along [010] direction in pyrene–TCNQ complex (3); (b) definition of the shift between hydrocarbon and TCNQ in the stack.
TCNQ molecule (the contact between C and N atom from
cyano groups is 3.242 Å), forming a kind of “ladder” of
stacks. This kind of motif is shown in Fig. 6. The structure is
stabilized by C–H⋯π interactions. The distances between
centroids of guest pyrene rings and C–H from TCNQ and
stacked pyrene are 3.083 Å and 2.675 Å, respectively.
and half of a TCNQ molecule as the asymmetric part of the
unit cell. The chrysene is slightly tilted with respect to the
TCNQ along the stacks, the angle between mean planes is
3.1° (see Fig. 9a). The shift is equal to 2.4 Å (see Fig. 9b).
The intra-stack separation is 3.37 Å. In this structure there
are two types of stabilizing interactions, C–H⋯N from
chrysene and delocalized charge between two cyano groups
from TCNQ in the range of 2.871–2.876 Å; and C–H⋯N from
TCNQ rings to chrysene with the distance equal to 2.729 Å
(see Fig. 10). The stacks form infinite columns of alternating
chrysene and TCNQ molecules as shown in Fig. 11.
Pyrene–TCNQ complex 1 : 1 (4). For comparison we have
also synthesized a pyrene–TCNQ complex with the molecular
ratio 1 : 1, (4). The compound crystallizes in a monoclinic
system, space group P21/n with half of the molecule of pyrene
and TCNQ as the asymmetric part of the unit cell. The
pyrene is tilted with respect to the TCNQ along the stacks,
with an angle between mean planes of 6.0° (see Fig. 7a). The
six-membered TCNQ ring lies above and beneath the pyrene
moiety as shown in Fig. 7b. This is very similar packing to
the case of (3), however the shift is larger and equal to 1.8 Å.
The intra-stack separation is 3.38 Å, which is smaller
than in (3) and the structure is stabilized by C–H⋯N contacts
with the distance equal to 2.656 Å (see Fig. 8a). Moreover,
TCNQ molecules pack even closer than in the case of (3),
the contact between the C and N atoms from cyano groups
is 3.226 Å. The arrangement of the stacks in the crystal forms
a “herringbone” motif along the [010] direction as shown
in Fig. 8b.
Phenanthrene–TCNQ complex 1 : 1 (6). The structure of
phenanthrene and TCNQ is disordered. It crystallizes in a
¯
triclinic system, space group P1 with one phenanthrene and
half a TCNQ molecule forming the asymmetric part of the
unit cell. TCNQ lies at the center of inversion, whereas
phenanthrene is slightly shifted. This situation causes
Chrysene–TCNQ complex 1 : 1 (5). Complex (5) crystallizes
¯
in a triclinic system, space group P1 with half of a chrysene
Fig. 5 Short interactions in crystal of pyrene–TCNQ complex (3).
418 | CrystEngComm, 2014, 16, 415–429
Fig. 6 “Ladder” stack motif in pyrene–TCNQ complex (3).
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Fig. 7 (a) The crystal packing along [010] direction in pyrene–TCNQ complex (4); (b) along [100] direction showing overlap between pyrene
and TCNQ.
Fig. 8 (a) Short interactions in crystal of (4); (b) herringbone motif between the stacks in (4).
disorder, which in turn makes the molecule resemble pyrene
(see Fig. 12). Despite this kind of similarity, the packing and
hence interactions in the crystal lattice are completely different
as compared with pyrene–TCNQ complex (4).
differently than in the other complexes (see Fig. 13a). There
are two molecules of hydrocarbon per one TCNQ in the
“sandwich”, whereas for all the synthesized complexes there
were two TCNQ and one hydrocarbon moieties.
The phenanthrene is somewhat tilted with respect to
the TCNQ along the stacks. The angle between mean planes
is 2.0°. The stacks in the unit cell are formed somewhat
The intra-stack separation is 3.35 Å. The structure is stabi-
lized by C–H⋯N contacts with the nitrogens from TCNQ and
atoms from adjacent phenanthrene molecules with contact
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 419

View Article Online
Paper
CrystEngComm
Fig. 9 (a) The crystal packing along [001] direction in chrysene–TCNQ complex (5); (b) along [100] direction showing overlap between chrysene
and TCNQ.
Fig. 10 Short interactions in crystal of chrysene–TCNQ complex (5).
Fig. 12 Molecule of phenanthrene disordered over the center of
inversion.
Fig. 13 (a) The crystal packing along [010] direction in phenanthrene–
TCNQ complex (6).
two cyano groups from TCNQ in the range of 2.792–2.990 Å
(see Fig. 14). The stacks are arranged in a similar way to (5),
forming infinite columns of alternating phenanthrene and
TCNQ molecules (see Fig. 15).
Fig. 11 Stacked columns of chrysene–TCNQ complex (5).
Naphthalene–TCNQ complex 1 : 1 (7). The structure of
naphthalene–TCNQ is significantly different from the others.
¯
distances in the range 2.654–2.748 Å, and there are also C–H
Although it crystallizes in a triclinic system, space group P1
contacts from phenanthrene and delocalized charge between
with half of the naphthalene and TCNQ molecules forming
420 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Fig. 14 Short interactions in crystal of phenanthrene–TCNQ complex (6).
Fig. 16 The crystal packing in naphthalene–TCNQ complex (7).
the asymmetric part of the unit cell, it forms very weak
stacking as shown in Fig. 16. Interestingly, one naphthalene
is surrounded by four TCNQ molecules, effectively forming
channels. Fig. 17 presents this kind of motif. The naphtha-
lene is tilted with respect to the TCNQ with an angle between
mean planes of 14.8° and the molecule is shifted by 4.8 Å.
The intra-stack separation is 3.09 Å. The structure is stabi-
lized by C–H⋯N interactions with adjacent TCNQ molecules
with the distance 2.679 Å and C–H⋯N interactions with
naphthalene equal to 2.677 Å (see Fig. 18). Interestingly,
there are also very short contacts between hydrogens from
naphthalenes (2.380 Å).
transition in the phenanthrene crystal at 0.7 GPa. Investiga-
tions of Hirshfeld surfaces and fingerprint plots indicate
that high-pressure polymorph packing is significantly differ-
ent from the ambient pressure form, which is dominated by
π⋯π stacking with a limited contribution from C–H⋯π
interactions.
In all cases except (7) there are strong π⋯π stacking inter-
actions between TCNQ and hydrocarbons illustrated by the
green-yellow colour around de = di ≈ 1.8 Å (in Fig. 19, it is
even red in (6) (see ESI†), but due to disorder this structure
should be treated with some care).
When the structure contains additional hydrocarbon,
TCNQ acts only as C–H⋯N acceptor, as shown by the charac-
teristic spike where de < di (Fig. 20a). If the molecular ratio
is 1 : 1 there are two spikes, upper (de > di) corresponds to
the HB donor and the lower (de < di) corresponds to the
HB acceptor (see Fig. 20b).
Concerning the C⋯H interactions: in complex (1) TCNQ is
mostly a donor, whereas in (2) the perylene moiety is slightly
more shifted and it gives rise to extra C⋯H contacts in the
Analysis of fingerprints based on the Hirshfeld surfaces
It has been shown that Hirshfeld surfaces12 are a good way to
compare similar crystal structures. External (de) and internal
(di) distances between atoms create a surface, which is able
to illustrate differences between different packing motifs
based on 2-D fingerprint plots of selected information.
McKinnon et al. revealed that it is a very effective way to visu-
alize intermolecular interactions for large variety of benzenoid
hydrocarbons.12 Fabbiani et al.13 showed that there is a phase
Fig. 15 Stacked columns of phenanthrene–TCNQ complex (6).
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 421

View Article Online
Paper
CrystEngComm
Fig. 17 The naphthalene channels in complex (7).
stacks, causing TCNQ to act as an acceptor as shown in
Fig. 21a and b. The situation is somewhat different in (3)
where TCNQ is a strong acceptor/donor of C⋯H interactions
to the “guest” pyrene molecule (Fig. 21c). In other cases
TCNQ acts like an acceptor (see Fig. 21d, tails where de < di,
see also ESI† for details).
fingerprint plot, where de = di ≈ 1.1 Å (see Fig. 23c). These
kinds of interactions are also present in other structures (e.g.
(5) as shown in Fig. 23a).
Degree of charge transfer
In all the structures with “guest” hydrocarbons one can
see a large contribution of C–H⋯π interactions. It is clear
from the appearance of “wings” in the fingerprint plots,
where de > di (see Fig. 22). Perylene in (1) and (2) and pyrene
(3) act as a donors to “guest” PAHs. They also accept C⋯H
interactions from TCNQ.
Interestingly, in complexes (5), (6) and (7) the molecules
are arranged in chains, where the hydrocarbon is a C–H
donor to delocalized charge at two cyano groups in TCNQ
moiety, which is manifested by a tail, where de > di (in
Fig. 23a, the appearance of the green colour at the end of
the tails suggest interaction with delocalized π electrons).
TCNQ acts as an acceptor, de < di (see Fig. 23b). Moreover,
naphthalene in (7) acts as a C–H⋯π acceptor from two TCNQ
molecules (“wing” de < di, see Fig. 23c). There are also
very short H⋯H contacts between two hydrocarbons in the
structure, they show up as a characteristic spike in the
The extent of the charge transfer can be estimated using
Raman spectroscopy.14 However, in our case, most of the
complexes were not stable enough and decomposed using the
confocal laser beam. The exciting laser line was 488 nm with
50 swd mitutoyo objectives. The argon laser power was kept at
10 mW on the DAC to avoid heating of the sample, but the
crystals had been destroyed. It has also been shown that the
degree of charge transfer in the complex can be assessed very
well from the geometry of TCNQ, reasoning from the nodal
structure of the frontier b2g molecular orbital.15 The popula-
tion of an electron at the LUMO of TCNQ should force the
two double bonds (a and c) to increase their bond length, at
the same time shortening the b and d (see Scheme 1). In the
method suggested by Kistenmacher15b the degree of charge
transfer can be estimated using the ratio c/(b + d).
Another way of estimating the extent of charge transfer in
this kind of system was proposed by Woźniak and Krygowski,
based on three canonical forms of TCNQ (two Kekule struc-
tures and a quinoidal one).16 Their relative contributions were
determined from the molecular geometry using the Harmonic
Oscillator Stabilization Energy approach (HOSE).17 It has been
shown that this approach is superior to the above-presented
method of Flandrois, Chasseau and Kistenmacher15 and leads
to more reliable values of charge transfer.16,18,19 The slightly
higher precision of the HOSE model in estimating the charge
of TCNQ species may result from better chemical grounds.
There is no argument that the lengths of b, c and d bonds in
the radical anion must be equal. No such assumption is made
in the HOSE model.16
Fig. 18 Short interactions in crystal of naphthalene–TCNQ complex (7).
422 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Fig. 19 Fingerprint plots visualizing de and di for C⋯C contacts (including reciprocal ones) generated for TCNQ and hydrocarbon molecules in
complexes (1) and (7).
Fig. 20 Fingerprint plots visualizing de and di for C–H⋯N contacts (including reciprocal ones) generated for TCNQ molecule in complexes
(a) (1) and (b) (4).
Table 1 presents the estimated values of the charge trans-
fer using both methods based on the molecular geometry of
TCNQ along with the values of mean plane separation
between TCNQ and benzenoid hydrocarbons, their shift and
tilt, as well as the degree of charge transfer estimated by the
HOSE model16,17 and directly from TCNQ geometry.15 Since
the susceptibility of charge transfer may depend on benze-
noid hydrocarbon, their characteristics such as aromatic
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 423

View Article Online
Paper
CrystEngComm
Fig. 21 Fingerprint plots visualizing de and di for C⋯H contacts (including reciprocal ones) generated for TCNQ molecule in complexes (a) (1),
(b) (2), (c) (3) and (d) (4).
stabilization energies (ASE)20 or ASE per π electron (ASE/π)
are also given.
Based on the values presented in Table 1, in the case of
the pure perylene–TCNQ complex there is no charge transfer.
The value of c/(b + d) equals 0.475(3), which is smaller than
the value for neutral TCNQ. As one can see, the estimation of
the degree of charge transfer in this way is not very accurate.
Moreover, the b and d bond lengths are the mean value
between two distances in most cases. In our calculations, we
also take into account the canonical forms of TCNQ and we
assess the charge using the HOSE model.
For neutral TCNQ q = 0.06 ē, which is almost the same as
for perylene–TCNQ q = 0.05 ē (see Table 1); but when the sec-
ond hydrocarbon is added to the complex, the conducting
properties are larger i.e. 0.21 ē, 0.22 ē in (1) and (2), respec-
tively. It means that the separation of stacks enhances the effi-
ciency of electron transfer in these kind of complexes. The
same is true for (3) and (4), where the second molecule of
pyrene stabilizes the complex by additional C–H⋯π interac-
tions. The degree of charge transfer increases from 0.13 ē (4) to
0.52 ē (3) and is even better than semiconducting coronene–
TCNQ complex, where the charge transfer is 0.32 ē.2
Fig. 22 Fingerprint plot visualizing de and di for C–H⋯π and C⋯H
contacts (including reciprocal ones) generated for perylene molecule in
complex (1).
424 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Fig. 23 Fingerprint plots visualizing de and di for C–H⋯π and C⋯H contacts (including reciprocal ones) generated for chrysene and TCNQ
molecules in complex (5) (a), (b) and naphthalene in (7) (c).
There are three important factors that may govern the
effectiveness of charge transfer in PAH–TCNQ complexes:
1. Mean plane separation. It has been suggested2 that the
closer the planes covering the TCNQ and PAH moieties, the
more effective the charge transfer. From our analysis it is
clear that this is not always the case. For example, in (3) the
distance is 3.45 Å and for coronene–TCNQ it is 3.33 Å, but
the charge transfer is evidently more efficient for pyrene–
TCNQ (0.52 ē vs. 0.32 ē). It is even more striking for (3), the
tetracene–TCNQ and perylene–TCNQ complexes, where the
mean plane separations are equal, but the differences in q
are huge: 0.52 ē, 0.18 ē, 0.05 ē, respectively.
2. Shift between the molecules in the stacks. It seems to be a
very important factor, even more important than 1. To ensure
efficient charge transfer the molecules should not be shifted
far from each other. For example the shift between pyrene and
TCNQ is only 1.3 Å in (3) and 1.8 Å in (4), but the difference in
charge transfer is dramatic 0.52 ē vs. 0.13 ē! In the cases where
the hydrocarbons overlap less efficiently i.e. (1) and (2) (2.6 Å
and 2.9 Å), the charge of TCNQ is the almost same, q = 0.21 ē
vs. 0.22 ē, so the shift plays a less important role.
3. Stability of the hydrocarbon. It might be expected that
the electronic nature of benzenoid hydrocarbons plays a cru-
cial role in the charge transfer. However, based on the ASE or
ASE/π values it can be concluded that there is no clear rela-
tionship. For example chrysene–TCNQ and tetracene–TCNQ
have similar mean plane separations (3.37 Å vs. 3.44 Å) and
shifts (2.4 Å vs. 2.0 Å), but the difference in ASE is substantial
(56.4 kcal mol−1 vs. 39.2 kcal mol−1). The degrees of charge
transfer in those compounds are also comparable (0.10 ē vs.
0.18 ē).
Scheme 1
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 425

View Article Online
Paper
CrystEngComm
426 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
Table
2 Crystal data and structure refinement for naphthalene–perylene–7,7,8,8-tetracyanoquinodimethane (1), pyrene–perylene–7,7,8,8-
tetracyanoquinodimethane (2), pyrene–TCNQ (2 : 1) (3) complexes
Compound
(1)
(2)
(3)
Empirical formula
Formula weight
Space group
Unit cell dimensions
a [Å]
b [Å]
c [Å]
C20H12, C12H4N4, C10H8
584.65
P1
C20H12, C12H4N4, C16H10
658.73
P1
2(C16H10), C12H4N4
608.67
P1
¯
¯
¯
7.126(1)
10.736(2)
10.971(2)
60.99(3)
89.03(3)
86.86(3)
732.9(3)
1
7.236(2)
7.004(2)
10.090(5)
11.783(5)
107.42(5)
106.46(3)
93.44(4)
752.4(6)
1
11.072(3)
11.308(3)
62.06(2)
86.353(2)
81.766(2)
792.1(3)
α [°]
β [°]
γ [°]
Volume V [Å3]
Z [molecules per cell]
1
Dcalculated [Mg m−3
]
1.325
0.079
1.381
0.082
1.343
0.080
Absorption coefficient μ/mm−1
θ range for data collection [°]
Limiting indices
3.55–25.5
−6 ≤ h ≥ 8
−12 ≤ k ≥ 12
−13 ≤ l ≥ 13
5694/2704
2704/256
0.896
3.49–25.75
−8 ≤ h ≥ 8
−13 ≤ k ≥ 13
−13 ≤ l ≥ 13
10239/2913
2913/287
1.077
3.65–25.1
−8 ≤ h ≥ 8
−12 ≤ k ≥ 12
−14 ≤ l ≥ 14
10917/2667
2667/265
0.977
Reflections collected/unique
Data/parameters
Goodness of fit
Final R index (I > 2σ)
wR2
0.0509
0.1182
0.209 and −0.219
0.0903
0.2492
0.429 and −0.389
0.0945
0.2292
0.353 and −0.350
Largest diff. peak and hole [Å−3
]
Table 3 Crystal data and structure refinement for: pyrene–TCNQ (1 : 1) (4), chrysene–TCNQ (5) and phenanthrene–TCNQ (6) complexes
Compound
(4)
(5)
(6)
Empirical formula
Formula weight
Space group
Unit cell dimensions
a [Å]
b [Å]
c [Å]
C16H10, C12H4N4
406.43
P21/n
C18H12, C12H4N4
432.47
P1
C14H10, C12H4N4
382.41
P1
¯
¯
6.992(1)
10.069(2)
14.671(3)
90.00
103.52(3)
90.00
1004.2(3)
2
1.344
7.1533(5)
7.9721(5)
9.2258(5)
80.11(5)
89.79(5)
86.67(5)
517.4(6)
1
6.7083(9)
8.0897(1)
9.118(2)
101.21(2)
106.46(3)
99.52(1)
471.6(1)
1
α [°]
β [°]
γ [°]
Volume V [Å3]
Z [molecules per cell]
Dcalculated [Mg m−3
]
1.388
0.084
1.347
0.082
Absorption coefficient μ/mm−1
θ range for data collection [°]
Limiting indices
0.081
2.48–25.05
−8 ≤ h ≥ 8
−11 ≤ k ≥ 11
−17 ≤ l ≥ 17
9996/1779
1779/173
1.036
3.62–25.2
−8 ≤ h ≥ 8
−9 ≤ k ≥ 9
−11 ≤ l ≥ 10
7076/1860
1860/186
1.088
2.31–25.05
−7 ≤ h ≥ 7
−8 ≤ k ≥ 9
−10 ≤ l ≥ 9
4006/1799
1799/165
0.969
Reflections collected/unique
Data/parameters
Goodness of fit
Final R index (I > 2σ)
wR2
0.0337
0.0792
0.257 and −0.152
0.0330
0.0887
0.161 and −0.225
0.0761
0.1763
0.555 and −0.443
Largest diff. peak and hole [Å−3
]
(0.71073 Å). The crystals were positioned 62.25 mm from
the KM4CCD camera; 187 frames were measured at 0.8°
intervals with a counting time of 60 s, 150 frames were
measured at 1° intervals with a counting time of 35 s,
150 frames were measured at 0.8° intervals with a counting
time of 45 s, 187 frames were measured at 0.8° intervals
with a counting time of 25 s and 136 frames were mea-
sured at 0.7° intervals with a counting time of 50 s, respec-
tively for (1), (2), (3), (5) and (6).
Experimental
X-ray diffraction
The X-ray measurements of naphthalene–perylene–7,7,8,8-
tetracyanoquinodimethane (1), pyrene–perylene–7,7,8,8-
tetracyanoquinodimethane (2), pyrene–TCNQ (2 : 1) (3),
chrysene–TCNQ (5) and phenanthrene–TCNQ (6) complexes
were performed at 100 (2) K on a KUMA CCD k-axis diffrac-
tometer with graphite-monochromated Mo Kα radiation
This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 427

View Article Online
Paper
CrystEngComm
Table 4 Crystal data and structure refinement for naphthalene–TCNQ
(7) complex
Conclusions
We have synthesized five new complexes of benzenoid
hydrocarbons: naphthalene–perylene, pyrene–perylene, pyrene,
phenanthrene and TCNQ and redetermined the structures of
the chrysene–TCNQ and naphthalene–TCNQ complexes,
which were previously solved with inadequate quality. Within
these compounds we have obtained structures of three
completely novel three-component complexes (naphthalene–
perylene–TCNQ, pyrene–perylene–TCNQ and pyrene–TCNQ 2 : 1).
Two of them, (1) and (2), comprise of different hydrocarbons
and TCNQ in the unit cell. Pyrene–TCNQ (3) contains two
pyrene molecules. One of them interacts in the stacking,
whereas another one acts as a guest molecule in the crystal
structure. We have analyzed the crystal structures, interac-
tions and different structural motifs. They were illustrated
and described by fingerprint plots based on Hirshfeld sur-
faces. In all cases there are strong π⋯π stacking interactions
between TCNQ and hydrocarbons (except naphthalene–
TCNQ). For the 2 : 1 complexes, characteristic spikes for an
acceptor C–H⋯N interactions are present, whereas for 1 : 1
complexes one can see both spikes (hydrogen bond acceptor/
donor). There are large contributions from C–H⋯π interac-
tions in the structures with an additional “guest” hydrocar-
bon, manifested by “wings” in the plots. In complexes (5), (6)
and (7) hydrocarbons act as donors of C–H to delocalized
charge at cyano groups. Importantly, we have assessed the
degree of charge transfer in the complexes and analyzed fac-
tors that may play a significant role. It turns out that addi-
tional hydrocarbon in the structure significantly increases the
charge transfer. Mean plane separation is an important factor,
but it seems that the shift is even more crucial. It is not yet
clear whether the effectiveness of charge transfer depends on
the nature of PAH. Along the stacks the tilts for hydrocarbons
and TCNQ are different, but it does not appear to correlate
with charge transfer.
Compound
(7)
Empirical formula
Formula weight
Space group
Unit cell dimensions
a [Å]
b [Å]
c [Å]
C10H8, C12H4N4
332.36
¯
P1
7.493(1)
7.800(1)
8.103(1)
84.30(6)
81.12(6)
61.34(5)
410.4(9)
1
α [°]
β [°]
γ [°]
Volume V [Å3]
Z [molecules/cell]
Dcalculated [Mg m−3
]
1.345
0.083
Absorption coefficient μ/mm−1
θ range for data collection [°]
Limiting indices
2.98–25.24
−8 ≤ h ≥ 8
−9 ≤ k ≥ 9
−9 ≤ l ≥ 9
8231/1487
1487/142
1.054
0.0292
0.0795
0.207 and −0.151
Reflections collected/unique
Data/parameters
Goodness of fit
Final R index (I > 2σ)
wR2
Largest diff. peak and hole [Å−3
]
Pyrene–TCNQ (1 : 1) (4) and naphthalene–TCNQ (7) com-
plexes were measured at 100 (2) K on a Bruker D8 VENTURE
diffractometer with TRIUMPH monochromator and MoKα
radiation (0.71073 Å). The crystals were positioned 40 mm (4)
and 55 mm (7) from the CCD camera; 230 frames were
measured at 0.5° intervals with a counting time of 25 s and
560 frames were measured at 0.25° intervals with a counting
time of 10 s for (4) and (7).
Data collection, cell refinement and data reduction for (1),
(2), (3), (5) and (6) were carried out with the KUMA Diffrac-
tion programs: CrysAlis CCD and CrysAlis RED22 and Bruker
SAINT software package for (4), (7). The data were corrected
for Lorentz and polarization effects, but no absorption cor-
rection was applied (except (4) and (7), where data were
corrected for absorption effects using the multi-scan method
(SADABS)). The structures were solved by direct methods23
and refined using SHELXL.24 The refinement was based on
F2 for all reflections except for those with very negative F2.
The weighted R factor, wR and all goodness-of-fit S values
are based on F2. The non-hydrogen atoms were refined
anisotropically. The hydrogen atoms were located from a dif-
ference map and were refined isotropically. The atomic scat-
tering factors were taken from the International Tables.25
Crystal structures and refinement are specified in Tables 2–4.
The structure of (6) is disordered with phenanthrene mole-
cule occupying two alternative sites with 50% occupancy
each. A full list of bond lengths, bond angles, and torsion
angles is given in Tables 1–9 in the ESI.† Analysis of finger-
prints based on the Hirshfeld surfaces was performed with
the CrystalExplorer program (ver. 3.0).26 The figures were
generated with Mercury (ver. 3.1).27
The estimated charge at TCNQ in synthesized pyrene–TCNQ
complex, where molecular ratio is 2:1, is much larger than pre-
viously reported coronene–TCNQ semiconductor. It could there-
fore be a good candidate for high-pressure, low-temperature
studies of superconductivity (the charge at TCNQ is 0.5 ē!).
Acknowledgements
M. A. D. gratefully acknowledges the Polish Ministry of Higher
Education and Science (grant 649/MOB/2011/0) and Dr Łukasz
Dobrzycki (University of Warsaw) for his help. The authors
also thank Dr Sian Howard for proof-reading the manuscript.
References
1 D. Jérome, Chem. Rev., 2004, 104, 5565–5591.
2 X. Chi, C. Besnard, V. K. Thorsmølle, V. Y. Butko,
A. J. Taylor, T. Siegrist and A. P. Ramirez, Chem. Mater.,
2004, 16, 5751–5755.
3 F. H. Herbstein, Perspect. Struct. Chem., 1971, 4, 166–395.
428 | CrystEngComm, 2014, 16, 415–429
This journal is © The Royal Society of Chemistry 2014

View Article Online
CrystEngComm
Paper
4 A. J. C. Buurma, O. D. Jurchescu, I. Shokaryev, J. Baas,
A. Meetsma, G. A. de Wijs, R. A. de Groot and
T. T. M. Palstra, J. Phys. Chem. C, 2007, 111, 3486–3489.
5 N. Kawai, R. Eguchi, H. Goto, K. Akaike, Y. Kaji, T. Kambe,
A. Fujiwara and Y. Kubozono, J. Phys. Chem. C, 2012, 116,
7983–7988.
6 (a) R. Mitsuhashi, Y. Suzuki, Y. Yamanari, H. Mitamura,
T. Kambe, N. Ikeda, H. Okamoto, A. Fujiwara, M. Yamaji,
N. Kawasaki, Y. Maniwa and Y. Kubozono, Nature, 2010,
464, 76–79; (b) X. F. Wang, Y. J. Yan, Z. Gui, R. H. Liu,
J. J. Ying, X. G. Luo and X. H. Chen, Phys. Rev. B: Condens.
Matter Mater. Phys., 2011, 84, 214523; (c) X. F. Wang,
R. H. Liu, Z. Gui, Y. L. Xie, Y. J. Yan, J. J. Ying, X. G. Luo and
X. H. Chen, Nat. Commun., 2011, 2, 507–513; (d)
Y. Kasahara, Y. Takeuchi and Y. Iwasa, Phys. Rev. B:
Condens. Matter Mater. Phys., 2012, 85, 214520.
7 X. F. Wang, X. G. Luo, J. J. Ying, Z. J. Xiang, S. L. Zhang,
R. R. Zhang, Y. H. Zhang, Y. J. Yan, A. F. Wang, P. Cheng,
G. J. Ye and X. H. Chen, J. Phys.: Condens. Matter, 2012,
24, 345701.
8 (a) J. Bardeen, L. N. Cooper and J. R. Schrieffer, Phys. Rev.,
1957, 106, 162–164; (b) J. Bardeen, L. N. Cooper and
J. R. Schrieffer, Phys. Rev., 1957, 108, 1175–1204.
R′r and R′′r stand for the bond lengths of a molecule; Rso and
Rdo stand for formal single and double bond in the i-th
canonical structure, n1 and n2 are the numbers of the
corresponding formal single and double bonds in the i-th
canonical structure, respectively. kr stands for the force
constants. In the process of deformation the n1 bonds
corresponding to the single bonds in the i-th canonical
structure are lengthened, whereas the n2 bonds corresponding
to the double bonds in the i-th canonical structure are
shortened. Each canonical structure may have a different
HOSE value. The weight of the i-th canonical structure, wi, in
the description of geometry of the real molecule is inversely
proportional to its destabilization energy (i.e. HOSEi), the
energy by which the i-th canonical structure is less stable than
1
HOSE

i 
w 
i100%
i
N
the real molecule:
. The summation
1
HOSE

i 

i1
runs over all structures. For neutral TCNQ q = 0 and the
contribution of the quinoidal structure accounted by the
HOSE model is 91.3%, whereas for singly charged TCNQ
species q = −1 and the quinoidal structure contribution is
50.6% (see ref. 16 for details).
18 T. M. Krygowski and M. K. Cyrański, Advances in Molecular
Structure Research, ed. M. Hargittai and I. Hargittai,
JAI Press, London, 1997.
9 M. Q. Xue, T. B. Cao, D. M. Wang, Y. Wu, H. X. Yang,
X. L. Dong, J. B. He, F. W. Li and G. F. Chen, Sci. Rep., 2012,
2, 389.
10 I. J. Tickle and C. K. Prout, J. Chem. Soc., Perkin Trans. 2,
1973, 720–723.
19 T. M. Krygowski, A. Ciesielski, B. Świrska and P. Leszczyński,
Pol. J. Chem., 1994, 10, 2097–2107.
20 A. Ciesielski, D. K. Stepień, M. A. Dobrowolski,
L. Dobrzycki and M. K. Cyrański, Chem. Commun., 2012,
48, 10129–10131.
11 G. R. Desiraju and A. Gavezzotti, Acta Crystallogr., Sect. B:
Struct. Sci., 1989, 45, 473–482.
12 J. J. McKinnon, M. A. Spackman and A. S. Mitchell, Acta
Crystallogr., Sect. B: Struct. Sci., 2004, 60, 627–668.
13 F. P. A. Fabbiani, D. R. Allan, S. Parsons and C. R. Pulham,
Acta Crystallogr., Sect. B: Struct. Sci., 2006, 62, 826–842.
14 See e.g. R. Voggu, B. Das, C. S. Rout and C. N. R. Rao,
J. Phys.: Condens. Matter, 2008, 20, 472204.
15 (a) S. Flandrois and D. Chasseau, Acta Crystallogr., Sect. B:
Struct. Crystallogr. Cryst. Chem., 1977, 33, 2744–2750; (b)
T. J. Kistenmacher, T. J. Emge, A. N. Bloch and D. O. Cowan,
Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem.,
1982, 38, 1193–1199.
21 R. E. Long, R. A. Sparks and K. N. Trueblood, Acta
Crystallogr., 1965, 18, 931–939.
22 Oxford Diffraction (2001), CrysAlis CCD and CrysAlis RED,
Version 171.33.66, Oxford Diffraction Ltd., Wrocław, Poland.
23 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
1990, 46, 467–473.
24 G. M. Sheldrick, SHELXL93. Program for the Refinement of
Crystal Structures, University of Göttingen, Germany.
25 International Tables for Crystallography, ed. Wilson A. J. C,
Kluwer, Dordrecht, 1992, vol. C.
26 S. K. Wolff, D. J. Grimwood, J. J. McKinnon, M. J. Turner,
D. Jayatilaka and M. A. Spackman, Crystal Explorer 3.0,
University of Western Australia, 2005–2009.
27 C. F. Macrae, P. R. Edgington, P. McCabe, E. Pidcock,
G. P. Shields, R. Taylor, M. Towler and J. van de Streek,
J. Appl. Crystallogr., 2006, 39, 453–457.
16 K. Woźniak and T. M. Krygowski, J. Mol. Struct., 1989, 193,
81–92.
17 T. M. Krygowski, R. Anulewicz and J. Kruszewski, Acta Crystallogr.,
Sect. B: Struct. Sci., 1983, 39, 732–739. HOSE is defined as follows:
n1
n2


s
2
d
2




HOSE  301.15 i
R  R ik 
R  R ik
where

o 

o 


r 

i
r
r
r
r1
r1

This journal is © The Royal Society of Chemistry 2014
CrystEngComm, 2014, 16, 415–429 | 429