Solid-state effects of monofluorophenyl substitution in dithiadiazolyl radicals: Impact on S?S and S?N interactions and their classification via Hirshfeld surfaces and fingerprint plots

Article abstract of DOI:10.1016/j.poly.2012.07.066

Ortho- and para-fluorophenyl 1,2,3,5-dithiadiazolyl (DTDA) radicals were synthesised and structurally characterised in the solid-state. This enables the well-known molecular origins of the electrical conductivity properties in this series of molecules to be studied, specifically S?S conduction pathways. This work includes the first report of monofluorophenyl DTDA crystal structures, filling an important gap in the literature to complement the many structural records of di-, tri-, tetra- and pentafluorophenyl DTDA analogues. This report also overturns previous thinking that monofluoro-substitution in these phenyl DTDA compounds does not influence the supramolecular chemistry; indeed, we demonstrate that singular fluorine is indeed structurally (and therefore property) directing as per their di-, tri-, tetra- and pentafluorinated relatives. In particular, the S?S and S?N interactions that control the electrical conductivity in DTDAs are distinct in these mono-fluorophenyl DTDAs. Hirshfeld surfaces were employed to clarify the nature and extent of these interactions. Their ability to exploit the very sensitive features of surface topologies in order to identify S?S and S?N intermolecular interactions is important since these interactions are much more subtle than, say, classical hydrogen-bonding. Furthermore, we demonstrate that Hirshfeld surfaces can classify the entire set of intermolecular interactions for a compound via a fingerprint plot. This affords the instant recognition of a given type of DTDA supramolecular network. In turn, barcodes can be generated from these fingerprint plots which quantify the percentage contribution of atom pairs that are involved in intermolecular interactions within DTDAs. The predictive potential of such classification within the field of molecular design is shown via a comparison of our fingerprint plots with those of DTDAs from previous studies.

Full text of DOI:10.1016/j.poly.2012.07.066

Polyhedron 45 (2012) 61–70
Contents lists available at SciVerse ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Solid-state effects of monofluorophenyl substitution in dithiadiazolyl radicals:
Impact on SꢀꢀꢀS and SꢀꢀꢀN interactions and their classification via Hirshfeld surfaces
and fingerprint plots
c
Jacqueline M. Cole a,b, , Christine M. Aherne , Paul G. Waddell a,b, Arthur J. Banister c, Andrei S. Batsanov c,
⇑
Judith A.K. Howard c
a Cavendish Laboratory, University of Cambridge, J.J. Thomson Avenue, Cambridge CB3 0HE, UK
b Department of Chemistry, University of New Brunswick, P.O. Box 4400, Fredericton, NB, Canada E3B 5A3
c Department of Chemistry, University of Durham, South Road, Durham DH1 3LE, England, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Ortho- and para-fluorophenyl 1,2,3,5-dithiadiazolyl (DTDA) radicals were synthesised and structurally
characterised in the solid-state. This enables the well-known molecular origins of the electrical conduc-
tivity properties in this series of molecules to be studied, specifically SꢀꢀꢀS conduction pathways.
This work includes the first report of monofluorophenyl DTDA crystal structures, filling an important
gap in the literature to complement the many structural records of di-, tri-, tetra- and pentafluorophenyl
DTDA analogues. This report also overturns previous thinking that monofluoro-substitution in these phe-
nyl DTDA compounds does not influence the supramolecular chemistry; indeed, we demonstrate that
singular fluorine is indeed structurally (and therefore property) directing as per their di-, tri-, tetra-
and pentafluorinated relatives. In particular, the SꢀꢀꢀS and SꢀꢀꢀN interactions that control the electrical con-
ductivity in DTDAs are distinct in these mono-fluorophenyl DTDAs.
Received 26 April 2012
Accepted 19 July 2012
Available online 27 July 2012
Keywords:
Supramolecular
Sulfur–nitrogen
Structure–property relationships
Crystallography
Hirshfeld surfaces
Hirshfeld surfaces were employed to clarify the nature and extent of these interactions. Their ability to
exploit the very sensitive features of surface topologies in order to identify SꢀꢀꢀS and SꢀꢀꢀN intermolecular
interactions is important since these interactions are much more subtle than, say, classical hydrogen-
bonding.
Furthermore, we demonstrate that Hirshfeld surfaces can classify the entire set of intermolecular inter-
actions for a compound via a fingerprint plot. This affords the instant recognition of a given type of DTDA
supramolecular network. In turn, barcodes can be generated from these fingerprint plots which quantify
the percentage contribution of atom pairs that are involved in intermolecular interactions within DTDAs.
The predictive potential of such classification within the field of molecular design is shown via a compar-
ison of our fingerprint plots with those of DTDAs from previous studies.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
thought that the contacts between these dimers may take the form
of a 4-centre 2-electron interaction where an electron pair is delo-
Exciting prospects for sulfur–nitrogen chemistry have been
demonstrated by the unusual superconducting properties of
poly-sulfurnitride [1] and the potential of DTDA radicals (Fig. 1)
as ‘organic metal’ materials for magnetism and conduction. This
has led to a sustained interest in their solid-state structure–prop-
erty relationships [2–9]. It has been suggested that these useful
properties, in particular the electronic conduction properties, are
linked to the intermolecular, non-bonded, SꢀꢀꢀS contacts within
the structures of these compounds. A large number of these com-
pounds form co-planar dimers within their structures and it is
calised across the four sulfur atoms [10]. Stacking of DTDA rings
forming continuous SꢀꢀꢀS contact chains is thought to facilitate elec-
tronic conduction in these compounds [2].
Previous studies into halogen-substituted phenyl-DTDA com-
pounds have addressed the effect of the substituents on the reduc-
tion of dithiadiazolylium cations to form the corresponding radical
[11]. Associated structural studies have investigated the modes of
association (MA) of the S2N2C rings which form dimers (Fig. 2)
and the SꢀꢀꢀN close contact motifs (Fig. 3) in bi- and trifluorophenyl
DTDA radicals (SN) [12–16]. Intermolecular interactions in the
structures reported by these studies have been rationalised by con-
sidering non-spherical van der Waals radii [17] and electrostatic
potential maps [12].
⇑
Corresponding author at: Cavendish Laboratory, University of Cambridge, J.J.
Thomson Avenue, Cambridge CB3 0HE, UK. Tel.: +44 1223 337470.
Despite these studies, no structural investigation has been
reported for monofluorophenyl DTDA radicals; indeed, only two
E-mail address: jmc61@cam.ac.uk (J.M. Cole).
0277-5387/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.poly.2012.07.066

62
J.M. Cole et al. / Polyhedron 45 (2012) 61–70
of ortho- and para-fluorophenyl-DTDA radicals and identify a new
way to assess their intermolecular interactions by analysing Hirsh-
feld surfaces. Such an approach that has already proved most
useful in identifying the intermolecular interactions in hydrogen-
bonded structures [20–23]. Here, we extend this application to
the identification of the more subtle types of intermolecular con-
tacts that are observed in DTDA structures, with emphasis on
non-bonded SꢀꢀꢀS contacts.
Fig. 1. Phenyl 1,2,3,5-dithiadiazolyl radical.
2. Experimental
2.1. Synthesis
Unless stated otherwise, all manipulations were performed in
an inert atmosphere using standard Schlenk techniques. IR spectra
were measured as Nujol mulls using a Perkin-Elmer 577 spectro-
photometer. Microanalyses were carried out on a Carlo-Erba
1106 elemental analyser.
para-Fluorophenyl dithiadiazoyl radical (C7H4N2FS2, 1) was syn-
thesised via literature methods [24]. Single crystals suitable for
X-ray crystallography were grown via sublimation of the product
under vacuum at 373 K.
ortho-Fluorophenyl dithiadiazoyl radical (C7H4N2FS2, 2): Lithium
hexamethyldisilazane (1.39 g, 8 mmol) was dissolved in diethyl
ether (40 mL) to which ortho-fluorobenzonitrile (0.90 g, 7.5 mmol)
was added, producing a pale yellow solution. After stirring for 3
hours, sulfur dichloride (2.4 g, 23 mmol) was added to the now
dark orange solution, affording a yellow solution, to which more
solvent (25 mL) was added. The reaction was allowed to proceed
overnight and then evaporated to dryness affording a yellow resi-
due. The ortho-fluorophenyl dithiadiazoyl cation was removed
from the residue via sulfur dioxide extraction leaving lithium chlo-
ride. The dark purple radical was then formed upon reduction by a
Zn(Cu) redox couple (0.30 g, 4.61 mmol) in tetrahydrofuran. Single
crystals suitable for X-ray crystallography were grown via sublima-
tion of the product under vacuum at 373 K. Elemental Anal. (%, cal-
culated in brackets): C, 42.3 (42.2); H, 2.0 (2.0); N, 14.0 (14.1). IR:
mmax 1608m 1491w 1458s 1374s 1274w 1222w 1141m 1094m
1034w 908w 839w 804m 777s 763s 731m 652m 550m 511w.
Fig. 2. Modes of association observed in DTDA radicals.
2.2. X-ray crystallography
Crystal structure data for 1 and 2 (Tables 1 and 2) were col-
lected at 150 K on a Rigaku AFC6S diffractometer equipped with
Table 1
Experimental and refinement details for the X-ray structures of 1 and 2.
1
2
Fw
M
C7H4FN2S2
199.24
C7H4FN2S2
199.24
T (K)
150
150
Crystal system
Space group
a (Å)
b (Å)
c (Å)
monoclinic
P21/n
5.7941(12)
29.810(6)
9.2007(18)
90
monoclinic
P21/c
13.327(6)
18.183(6)
13.437(6)
90
a
(°)
b (°)
103.01(3)
90
1548.4(5)
8
0.71073
1.709
2022
107.37(3)
90
3108(2)
16
0.71073
1.703
5378
c
(°)
U (Å3)
Fig. 3. SꢀꢀꢀN close contact motifs observed in DTDA radicals [11].
Z
l
(Mo K
Dcalc (Mg mꢁ3
Total reflections
a
) (mmꢁ1
)
monohalogenated phenyl DTDA radicals (para-chloro- [18] and
para-iodophenyl DTDA radicals [19]) have been structurally char-
acterised and there is no such information for any ortho-monosub-
stituted derivative of the phenyl DTDA radical in the literature.
Here we report the synthesis and X-ray structural characterisation
)
Independent reflections (Rint
)
894 (0.079)
0.0484, 0.1255
0.384, ꢁ0.396
3578 (0.0631)
0.0995, 0.3084
1.334, ꢁ0.879
Final R, wR2
Largest difference: peak, hole (eÅꢁ3
)

J.M. Cole et al. / Polyhedron 45 (2012) 61–70
63
Table 2
Selected bond distances (Å) and angles (°) for 1 and 2.
Intramolecular S—S bond
distances
Intramolecular S—N bond
distances
Intradimer SꢀꢀꢀS contact
Interdimer SꢀꢀꢀS contact
Interplanar twist
angle
distances
distances
1
2
S1—S2 2.093(3)
S3—S4 2.108(3)
S1—N1 1.630(6)
S2—N2 1.628(6)
S3—N3 1.638(6)
S4—N4 1.619(6)
S1ꢀꢀꢀS3 3.055(3)
S2ꢀꢀꢀS4 3.051(3)
S1ꢀꢀꢀS3i 3.482(3)
11 6.8(4)
12 6.3(5)
S2ꢀꢀꢀS4i 3.354(3)
S1—S2 2.104(4)
S3—S4 2.100(4)
S5—S6 2.090(3)
S7—S8 2.097(4)
S1—N1 1.628(9)
S2—N2 1.616(8)
S3—N3 1.632(8)
S4—N4 1.638(9)
S5—N5 1.629(8)
S6—N6 1.613(7)
S7—N7 1.636(9)
S8—N8 1.616(9)
S1ꢀꢀꢀS3 3.054(4)
S2ꢀꢀꢀS4 3.078(4)
S5ꢀꢀꢀS7 3.121(4)
S6ꢀꢀꢀS8 3.032(3)
S3ꢀꢀꢀS6 3.403(3)
S4ꢀꢀꢀS8 3.560(4)
S1ꢀꢀꢀS6ii 3.758(4)
S2ꢀꢀꢀS8ii 3.702(4)
21 11.8(4)
22 8.0(5)
23 23.5(2)
24 5.2(3)
(i) 1/2 + x, 1/2 ꢁ y, 1/2 + z; (ii) x, 1.5 ꢁ y, ꢁ0.5 + z.
a molybdenum X-ray source (kMo
Cryosystems Cryostream open-flow nitrogen cooling device.
= 0.71073 Å) and an Oxford
a
3. Results and discussion
3.1. Molecular and dimeric geometry
K
The structures were solved by direct methods and refined by
full-matrix least-squares methods on F2 values of all data. Refine-
ments were performed using SHELXL [25]. Hydrogen atoms were
positioned geometrically and refined as riding on their parent phe-
nyl carbon atoms, with C–H = 0.950 Å and Uiso(H) = 1.2Ueq(C). The
structure of 2 shows substitutional disorder between opposing
ortho F and H atoms and a high Z0 value (Z0 = 4) results where
two out of four of these molecules are disordered. There was no
evidence for disorder in the other two asymmetric molecular units,
nor was there any sign of twinning despite the fact that a ꢂ c, as
can be a sign of twinning in a monoclinic crystal system. Its result-
ing R1 value is slightly high, as is a common observation in struc-
tures where Z0 > 1 [26]. Nonetheless, the model presents accurate
bond geometry and meets all relevant checks for crystallographic
integrity in Platon [27] such that we can be confident about its
reliability.
The crystallographic asymmetric unit (Fig. 4) of 1 comprises a
dimer formed from two symmetry-independent molecules
(Z0 = 2), that are related geometrically via a cis-cofacial arrange-
ment (MA-I). The asymmetric unit of 2 (Fig. 5) comprises two di-
mers, denoted here by d1 and d2; all molecules therein are
crystallographically independent (Z0 = 4). The dimer d1 lies perpen-
dicular to (100) while d2 lies parallel to (100) and both exhibit the
same MA-I mode of association. While d2 is ordered, the fluorine in
d1 exhibits substitutional disorder in both molecules about the two
possible ortho-positions in the phenyl ring. The fluorine atoms on
each molecule that correspond to the dominant disordered compo-
nent (70% for F1 and 85% for F3) lie on the same side of the dimer.
For both 1 and 2, the geometry of each crystallographically inde-
pendent molecule is similar to those in previously reported fluoro-
phenyl-substituted dithiadiazoyl (DTDA) compounds [12–16]. The
two crystallographically independent molecules in 1 exhibit inter-
planar twist angles between the DTDA and phenyl rings of 6.8(4)°
and 6.3(5)°. In 2, the observed interplanar twist angles within each
molecule of d1 are 8.0(5)° and 11.8(4)° while those observed in d2
are 5.2(3)° and 23.5(2)°. These geometries are also in line with pre-
vious findings.
2.3. Computational methods
Hirshfeld surfaces of the individual molecules of 1 and 2 have
been calculated using CRYSTALEXPLORER 2.1 [28] in order to identify
the important intermolecular interactions within the crystal
structures [21,20,22]. Hirshfeld surfaces are representative of
the electron distribution, calculated as the sum of the electron
densities of isotropic atoms. Identification of close contacts is
made possible via the normalised contact distance (dnorm) relative
to the distances from the surface to the nearest nucleus inside
and outside the surface (di and de respectively) given by Eq. (1).
These close intermolecular contacts i.e. those closer than van
der Waals, are indicated by the corresponding red areas on the
Hirshfeld surfaces of the molecules. In both 1 and 2, Z0 > 1 and
hence matching red spots are found on surfaces corresponding
to different molecules.
The dimers observed in the structures of 1 and 2 are stabilised
in the solid-state via intradimer SꢀꢀꢀS contacts (Table 2) and
di ꢁ rvi dw de ꢁ rvedw
dnorm
¼
þ
ð1Þ
rvedw
rvi dw
‘Fingerprint’ plots of each Hirshfeld surface were also derived
from CRYSTALEXPLORER 2.1 by plotting di against de. Such figures allow
one to classify a molecular set of intermolecular interactions
according to a single topological characteristic. Breaking the sur-
faces down into percentage areas of the surface associated with
specific interactions allows ‘barcode’ figures for each molecule to
be produced, facilitating further this characterisation.
Fig. 4. Asymmetric unit of
probability level.
1 with displacement ellipsoids drawn at the 50%

64
J.M. Cole et al. / Polyhedron 45 (2012) 61–70
Fig. 5. Asymmetric unit of 2 with displacement ellipsoids drawn at the 50% probability level.
intradimer
p
ꢀꢀꢀ
p
stacking interactions between the phenyl rings; in
compounds (3.361(3)–4.428(1) Å)12. The direction of the chains
alternates along (010) forming staggered layers of molecules.
Continuous chains of alternating intra- and interdimer SꢀꢀꢀS con-
tacts are also observed in the structure of 2, though these form in a
distinctly different way to those in 1. The chains comprise alternate
dimers lying perpendicular to each other, creating linear
[SꢀꢀꢀSꢀꢀꢀSꢀꢀꢀS] units (Fig. 7). The chain then continues at right angles
to this four-membered unit, much like the SꢀꢀꢀS contact chain ob-
served in 1, forming infinite chains of SꢀꢀꢀS contacts along [001].
These contacts are generally longer than those observed in 1 and
hence there is less potential for electron hopping between dimers
in 2. Viewing the structure down [100], parallels can be drawn be-
tween 2 and the ‘herringbone’ chains observed in the structure of
the monoclinic, dimeric polymorph of the 4-(20,60-diflurophenyl)-
DTDA radical [12]. The arrangement of d2 in 2 is similar to this
dithiadiazolyl radical with the exception that the d1 dimer acts
as a spacer oriented perpendicular to the d2 dimer lying parallel
to (100). In this previously published structure, the SꢀꢀꢀN contact
motif, identified as SN-IV, was attributed to close bifurcated SꢀꢀꢀN
contacts. In this case, the same could be said of the structure of 2
1, the centroidꢀꢀꢀcentroid distance is 3.74(1) Å and in 2 the cen-
troidꢀꢀꢀcentroid distance is 3.64(1) Å for d1 and 3.79(1) Å for d2.
Close contacts between the C atoms of the radical ring (in 1,
C7ꢀꢀꢀC14 3.281(9) Å; in 2, C7ꢀꢀꢀC14 3.31(1) Å and C21ꢀꢀꢀC28
3.35(1) Å) are also apparent.
3.2. Crystal packing
Considering the interdimer SꢀꢀꢀS contacts in 1, the dimeric units
pack together forming chains in a ‘herringbone’ arrangement along
[103] with the planes of the molecules lying parallel to (103)
(Fig. 6). This differs from the very common pinwheel configuration
seen in many of the fluorophenyl DTDA radicals [12,15] and the
herringbone structure of 2,6-fluorophenyl DTDA where the dimers
are arranged perpendicular to the direction of the motif [12]. These
chains consist of linear [SꢀꢀꢀSꢀꢀꢀS] units that repeat in an alternating
interdimer and intradimer fashion. The interdimer SꢀꢀꢀS distances
lie at the shorter end of the range observed for similar ‘in-plane’
contacts in the bi- and trifluorophenyl analogues of these
Fig. 6. Structure of 1 showing the chains formed via intermolecular SꢀꢀꢀS contacts (dashed lines) along [103].

J.M. Cole et al. / Polyhedron 45 (2012) 61–70
65
Fig. 7. Structure of 2 showing the chains formed via intermolecular SꢀꢀꢀS contacts (dashed lines) along [100].
Fig. 8. Hirshfeld surface of molecule 1 of the structure of 1.
Fig. 9. Hirshfeld surfaces of molecule 2 of the structure of 1.
since the SꢀꢀꢀN distances (S5ꢀꢀꢀN3 3.242(8) Å, S6ꢀꢀꢀN3 3.395(8) Å,
S7ꢀꢀꢀN4 3.460(9) Å and S8ꢀꢀꢀN4 3.56(1) Å) fulfil the criteria for close
contacts given the sums of the van der Waals radii (3.20–3.63 Å)
[12].
Examination of the Hirshfeld surfaces of the molecules in the
structure of 1 (Figs. 8 and 9) demonstrates the significance of the
SꢀꢀꢀS contacts both within and between dimers. This is despite

66
J.M. Cole et al. / Polyhedron 45 (2012) 61–70
Fig. 12. Hirshfeld surfaces of molecule 3 of the structure of 2.
These interactions are not continuous in the manner of the stack-
ing contacts in 1, which form infinite stacks, but are finite connect-
ing two of the planes parallel to (001). According to the Hirshfeld
surfaces, there is little evidence for SꢀꢀꢀN contacts since a red spot
on the surface (indicative of a closer than van der Waals interac-
tion) corresponding to this interaction is not observed. The close
approach of these molecules for one half of the dimer appears to
be due to SꢀꢀꢀC close contacts (S6ꢀꢀꢀC21 3.454(8) Å); the presence
of these interactions over those of the more usual SꢀꢀꢀN contacts
in this family of compounds is most likely a consequence of steric
effects rather than a more electrostatically favourable non-bonded
interaction. In general, regarding Hirshfeld surfaces, it is clear that
the intradimer contacts are much shorter than the interdimer con-
tacts but the latter can be seen to be also significant, as evidenced
by the corresponding red areas of the Hirshfeld surface. Though
significant in a number of reported structures of bi- and tri-fluoro-
phenyl radicals, SꢀꢀꢀN contacts are not as prominent in the monoflu-
orophenyl compounds. While it may be false to conclude that
single F atoms produce little structure directing influence, as they
play a role in many of the intermolecular interactions in the mono-
fluorophenyl structures, it can be suggested that combinations of F
atoms give rise specifically to structure-directing SꢀꢀꢀN contacts
whereas single F atoms do not.
Fig. 10. Hirshfeld surfaces of molecule 1 of the structure of 2.
the methodology of Clarke et al. [12] through which SꢀꢀꢀN contacts
within the range of the sum of the van der Waals radii can be
identified (S2ꢀꢀꢀN3 3.481(7) Å, S2ꢀꢀꢀN1 3.490(7) Å and S4ꢀꢀꢀN3
3.469(7) Å) forming chains down [100] with a SꢀꢀꢀN contact motif
similar to SN-III. Yet, these interactions are not ‘in-plane’ owing
to the staggered alignment of the chains of SꢀꢀꢀS contacts.
Hirshfeld surfaces show that SꢀꢀꢀN separations are not deemed
to form close contacts. As such, the SꢀꢀꢀS interactions dictate the
crystal packing since they seem to form at the expense of SꢀꢀꢀN con-
tacts. A potential Sꢀꢀꢀ
p interaction is also highlighted by the Hirsh-
feld surfaces (S4ꢀꢀꢀC9 3.302(6) Å, Fig. 9).
In 2, analysis of the Hirshfeld surfaces (Figs. 10–13) indicates
that the SꢀꢀꢀS contacts are generally stronger than SꢀꢀꢀN contacts
and are likely to be more important in terms of the mode of asso-
ciation of the two dimers. What appear to be SꢀꢀꢀN contacts are ob-
served between two d2 dimers in
a stacking arrangement
(S6ꢀꢀꢀN6 = 3.501(7) Å), interacting via the SN-II SꢀꢀꢀN contact motif.
Fig. 11. Hirshfeld surfaces of molecule 2 of the structure of 2. NꢀꢀꢀH and FꢀꢀꢀH
interactions (denoted by N/FꢀꢀꢀH) are difficult to delineate due to the manifest
disorder.
Fig. 13. Hirshfeld surfaces of molecule 4 of the structure of 2.

J.M. Cole et al. / Polyhedron 45 (2012) 61–70
67
appear that the F atom presumably has a significant structure-
directing influence. The chains of SꢀꢀꢀS contacts are similarly con-
nected in the structure of 2, in this case via bifurcated interactions
between the para-H atoms and N and F atoms on adjacent, in-plane
d1 dimers (N5ꢀꢀꢀH18 = 2.614(7) Å and N7ꢀꢀꢀH25 2.668(8) Å, F6ꢀꢀꢀH18
2.669(6) Å and F5ꢀꢀꢀH25 = 2.455(7) Å); these generate further chain
motifs along [001] (Fig. 14). Similar interactions are a common
feature of reported structures of this kind with ortho-fluorine
atoms [12] and they are readily identified in this structure by anal-
ysis of the Hirshfeld surfaces (Fig. 12). An interaction between F5
and the meta-H of an adjacent radical (F5ꢀꢀꢀH26 = 2.651(7) Å) is also
apparent from the Hirshfeld surfaces. However, the distance be-
tween these atoms in analogous positions on the other radical
comprising d1 is too long to be considered an interaction
(F6ꢀꢀꢀH19 = 3.169(6) Å), due to the greater interplanar twist angle
observed for that radical. In turn, this twist angle can be attributed
to repulsive HꢀꢀꢀH contacts between neighbouring d1 and d2 dimers.
The H atoms appear to be brought into closer than the van der
Waals range by a strong interaction between a N atom on d1 and
a para-H on d2 (N6ꢀꢀꢀH4 2.580(8) Å, Fig. 13).
Other potentially significant intermolecular interactions are
also apparent in these structures. In 1, close contacts occur as a re-
sult of the bifurcated interaction between a meta-H atom and S
(S1ꢀꢀꢀH10 3.002(2) Å) and N (N1ꢀꢀꢀH10 2.646(7) Å) atoms, linking
dimers along [010] (Fig. 9). Also present in the structure of 2 are
a series of HꢀꢀꢀC and SꢀꢀꢀC contacts, the latter of which include some
Fig. 14. Structure of 2 showing the chains formed via intermolecular FꢀꢀꢀH and NꢀꢀꢀH
contacts (dashed lines) along [001].
HꢀꢀꢀF close contacts are observed in both structures in line with
what was observed in other fluorophenyl DTDAs. In 1, there are
close contacts between meta-H and para-F atoms in opposing di-
meric molecules which form about a crystallographic inversion
centre (F1ꢀꢀꢀH10 2.754(5) Å and F2ꢀꢀꢀH3 2.793(4) Å). Since these
interactions serve to link the chains of [SꢀꢀꢀSꢀꢀꢀS] units, it would
potential Sꢀꢀꢀ
p interactions (Fig. 10). Most interactions involving
ortho-H and ortho-F atoms on d1 are not readily identifiable due
to the occupational disorder, though the F2ꢀꢀꢀH13 distance
(2.38(4) Å) is the shortest FꢀꢀꢀH contact (Fig. 10).
Fig. 15. Hirshfeld fingerprint plots for molecules 1 (top) and 2 (bottom) in 1 (showing all interactions (left), highlighting SꢀꢀꢀS interactions (middle) and highlighting NꢀꢀꢀH/
HꢀꢀꢀN interactions (right). Fingerprint plots for 2 are included in the supplementary material.

68
J.M. Cole et al. / Polyhedron 45 (2012) 61–70
Comparing 1 with the two previously reported para-monohalo-
of molecule 2. Very low percentages of NꢀꢀꢀN interactions are ob-
served in both molecules in spite of the cis-cofacial arrangement
of the molecules in the dimers. Though interestingly, despite there
being no readily identifiable close SꢀꢀꢀN contacts, SꢀꢀꢀN interactions
account for 6.7% and 12.4% of the Hirshfeld surface in molecule 1
and 2 respectively.
Fewer HꢀꢀꢀH interactions are apparent on the surfaces of mole-
cules in 2 (Fig. 17) but the bifurcated NꢀꢀꢀH interactions in mole-
cules 3 and 4 (which comprise d2) are evidenced by the increase
in percentage of NꢀꢀꢀH interactions for these molecules relative to
both 1 and 2. Again, the relatively low percentage of SꢀꢀꢀN contacts
is notable.
Generating analogous interaction ‘barcodes’ for the previously
reported fluorophenyl DTDA structures (provided in the supple-
mentary information) [12–16,29] allows a number of important
observations to be made. Firstly, regarding SꢀꢀꢀN interactions, their
Hirshfeld surface percentage contribution does not appear to fol-
low any trend as far as the presence of structure-directing close
contacts are concerned. For example, the mono-fluorinated com-
pounds 1 and 2, which exhibit no or few SꢀꢀꢀN close contacts con-
tain similar percentages of SꢀꢀꢀN interactions to those bi- and tri-
fluorinated compounds where these close contacts are observed.
Secondly, and more importantly as far as potential conducting
properties are concerned, the Hirshfeld surface area percentages
of SꢀꢀꢀS contacts observed in each molecule afford a direct indica-
tion of the mode of association and motif within the structure, as
higher percentages can be associated with continuous SꢀꢀꢀS contact
motifs. Those compounds which exhibit the ‘pinwheel’ motif as
well as those exhibiting direct stacking of dimers with the MA-I
mode of association [12–16,29], both of which result in infinite
chains of SꢀꢀꢀS contacts, have much higher percentages of SꢀꢀꢀS con-
tacts (ca. 9–15%) than those without these motifs or where other
modes of association are observed (ca. 3–8%). These percentages
could serve as a quantitative prediction of electrical conducting po-
tential. The high percentage barcode values observed in 1 and 2 (ca.
9–12%) are therefore indicative of their SꢀꢀꢀS contact motifs; though
the lower values observed in 2 compared to 1 may indicate a lower
potential for conduction properties, most likely a result of its long-
er inter-dimer SꢀꢀꢀS contacts.
genated phenyl DTDA radicals, the chloro- and iodo-analogues of 1,
a comparison is best made between 1 and the chloro-analogue [18]
since both exhibit the MA-I mode of association, whereas the iodo-
analogue exhibits the MA-II mode of association [19]. However,
the para-chlorophenyl DTDA radical does not form analogous SꢀꢀꢀS
chain motifs to those observed in 1 in the solid-state. Close SꢀꢀꢀCl
contacts (S3ꢀꢀꢀCl2 3.385(5) Å and S4ꢀꢀꢀCl2 3.521(5) Å) are instead
observed in the plane of the dimer containing the S atoms involved
in the interaction.
Comparing the structure of 2 to that of the non-fluorinated phe-
nyl DTDA radical [10,29], the impact of the ortho-fluorine substitu-
ent is clear. Both structures comprise four crystallographically
independent molecules (Z0 = 4) and hence two crystallographically
independent dimers, which lie on different mean planes to each
other. However, the discrete planes of molecules resulting from
the formation of chains of SꢀꢀꢀS contacts in 2 are absent in the
non-fluorinated form. It can be concluded that the aforementioned
bifurcated interactions between the para-H atoms and N and F
atoms along [001] (Fig. 14) exert their structure-directing influ-
ence by maintaining the discrete planes of molecules which assist
in the formation of the SꢀꢀꢀS close contact motif.
3.3. Fingerprint plots
An overall summary of intermolecular contacts can be given by
2D fingerprint plots of the Hirshfeld surfaces produced by plotting
di against de (e.g. Fig. 15) [20]. The most striking common charac-
teristic of the fingerprint plots of 1 and 2 are the intense red streaks
in the centre representing SꢀꢀꢀS contacts. Stubby ‘wings’ are visible
in each fingerprint plot of 1 (Fig. 15) starting at de ꢂ 1.0 and
di ꢂ 1.5 for molecule 1 where the N atom is participating and vice
versa in molecule 2. These wings represent NꢀꢀꢀH interactions be-
tween the two molecules; similar peaks can be seen in the finger-
print plots for the molecules of 2 as well as those corresponding to
FꢀꢀꢀH, CꢀꢀꢀH and SꢀꢀꢀH interactions.
Breaking down the Hirshfeld surface of 1 into percentage areas
associated with a particular atom-typeꢀꢀꢀatom-type interaction af-
fords interaction ‘barcodes’ (Fig. 16). These barcodes reveal that
SꢀꢀꢀS interactions account for ca. 12% of interactions for both mole-
cules of 1 and 9–10% of molecules of 2 highlighting the longer
interdimer SꢀꢀꢀS contacts observed in 2. The importance of HꢀꢀꢀH
interactions is readily identifiable here since they account for,
respectively, 17.7% and 15.5% of the Hirshfeld surface for molecules
1 and 2 (which comprise d1). The larger percentage of SꢀꢀꢀH and
SꢀꢀꢀC interactions in molecule 2 compared to molecule 1 reflect
In terms of structural prediction, the use of electrostatic isopo-
tentials, as per the methodology of Clarke et al. [12], has limited
usefulness for predicting the structures of monofluorinated DTDA
radicals, as evidenced by the structures of 1 and 2 reported herein.
However, it remains valid in terms of predicting structural motifs
in similar DTDA radicals. It also has the distinct advantage over
Hirshfeld surfaces that it can predict structural motifs simply from
a chemical sketch of the molecule. Indeed, Hirshfeld surfaces can
the end-on Sꢀꢀꢀ
p mode of association seen between two examples
Fig. 16. Barcode chart showing the breakdown of the intermolecular interactions for the two independent molecules of 1. Each atom-type/atom-type interaction given
includes reciprocal contacts.

J.M. Cole et al. / Polyhedron 45 (2012) 61–70
69
Fig. 17. Barcode chart showing the breakdown of the intermolecular interactions for the four independent molecules of 2. Each atom-type/atom-type interaction given
includes reciprocal contacts.
only be generated from a priori known crystal structures, where
the structural motif is of course already known, so rendering its
prediction moot. Instead, Hirshfeld surfaces act to rationalise
rather than predict structural motifs, as demonstrated herein. Nev-
ertheless, Hirshfeld surfaces represent a highly complementary
predictive tool to this electrostatic isopotential approach in that
they can predict properties of these structural motifs. It is in this
sense that Hirshfeld surfaces demonstrate potential use in the de-
sign of new DTDA materials with tailor-made functional
properties.
materials. This leads to the prospect of employing smart material
design strategies with this ‘training’ data, to predict DTDA materi-
als that have designer SꢀꢀꢀS contacts, in order to realise their poten-
tial as electrical conductors.
Acknowledgements
J.M.C. thanks the Royal Society for a University Research Fellow-
ship, the University of New Brunswick for the UNB Vice-Chancel-
lor’s Research Chair and NSERC for the Discovery Grant 355708
(for P.G.W.).
4. Concluding remarks
Appendix A. Supplementary data
Our analysis of two monofluorophenyl DTDA radicals refutes
the previous assertion that one fluorine atom alone does not pos-
sess much structure-directing ability [12]. Indeed, the positioning
of the ortho- or para-fluorine in 1 and 2 appears to dictate the
appearance of structural motifs. Considering intermolecular con-
tacts involving nitrogen atoms in DTDA radicals, however, it can
be said that more than one fluorine is necessary to produce a large
enough d- charge on N to allow for strong NꢀꢀꢀX contacts, as the
monofluorophenyl DTDA radicals do not exhibit as many close con-
tacts involving N atoms compared to their bi- and trifluorophenyl
counterparts.
CCDC 878564 and CCDC 878565 contain the supplementary
crystallographic data for 1 and 2, respectively. These data can be
obtained free of charge via http://www.ccdc.cam.ac.uk/conts/
retrieving.html, or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-
336-033; or e-mail: deposit@ccdc.cam.ac.uk. Supplementary data
associated with this article can be found, in the online version, at
http://dx.doi.org/10.1016/j.poly.2012.07.066.
We have also demonstrated that Hirshfeld surfaces are a very
useful tool for visualising intermolecular interactions in these mol-
ecules where such contacts appear subtle and yet, they are easily
identifiable by this method. Moreover, their relative strengths are
easily gauged allowing for simple rationalisation of the most
important, structure-directing interactions. Deriving fingerprint
plots and barcode graphs from these surfaces can also help to eval-
uate the potential of these compounds as electrical conductors
since they represent an ensemble classification of non-bonded con-
tacts via the partitioning of Hirshfeld surface areas into intermolec-
ular interaction percentage contributions; and higher Hirshfeld
surface area percentages of SꢀꢀꢀS contacts are indicative of infinite
SꢀꢀꢀS contact motifs.
References
[1] C.-H. Hsu, M.M. Labes, J. Chem. Phys. 61 (1974) 4640.
[2] J.M. Rawson, A.J. Banister, I. Lavender, Adv. Heterocycl. Chem. 62 (1995) 137.
[3] A.W. Cordes, R.C. Haddon, R.T. Oakley, Phosphorus Sulfur Silicon Relat. Elem.
179 (2004) 673.
[4] R.C. Haddon, Nature 256 (1975) 394.
[5] L. Ducasse, Chem. Phys. Lett. 352 (2002) 454.
[6] J.M. Rawson, A. Alberola, A. Whalley, J. Mater. Chem. 16 (2006) 2560.
[7] D.A. Haynes, CrystEngComm 13 (2011) 4793.
[8] J.M. Cole, C.M. Aherne, J.A.K. Howard, A.J. Banister, P.G. Waddell, Acta
Crystallogr., Sect. E Struct. Rep. Online 67 (2011) o2514.
[9] C.D. Martin, P.J. Ragogna, Annu. Rep. Prog. Chem. Sect. A Inorg. Chem. 107
(2011) 110.
[10] A. Vegas, A. Pérez-Salazar, A.J. Banister, R.G. Hey, Dalton Trans. (1979) 1812.
[11] C.M. Aherne, A.J. Banister, T.J. Hibbert, A.W. Luke, J.M. Rawson, Polyhedron 16
(1997) 4239.
[12] C.S. Clarke, D.A. Haynes, J.N.B. Smith, A.S. Batsanov, J.A.K. Howard, S.I. Pascux,
J.M. Rawson, CrystEngComm 12 (2010) 172.
Such classification affords the concept of performing systematic
large-scale comparisons of non-bonded contacts in DTDA

70
J.M. Cole et al. / Polyhedron 45 (2012) 61–70
[13] A.J. Banister, A.S. Batsanov, O.G. Dawe, P.L. Herbertson, J.A.K. Howard, S. Lynn,
I. May, J.N.B. Smith, J.M. Rawson, T.E. Rogers, B.K. Tanner, F. Palacio, Dalton
Trans. (1997) 2539.
[14] A.M.T. Bell, J.N.B. Smith, J.P. Attfield, J.M. Rawson, K. Shankland, W.I.F. David,
New J. Chem. 23 (1999) 565.
[21] F.P.A. Fabbiani, L.T. Byrne, J.J. McKinnon, M.A. Spackman, CrystEngComm 9
(2007) 728.
[22] H.F. Clausen, M.S. Chevallier, M.A. Spackman, B.B. Iversen, New J. Chem. 34
(2010) 193.
[23] S.K. Seth, G.C. Maity, T. Kar, J. Mol. Struct. 1000 (2011) 120.
[24] C.M. Aherne, A.J. Banister, I.B. Gorrell, M.I. Hansford, Z.V. Hauptman, A.W. Luke,
J.M. Rawson, Dalton Trans. (1993) 967.
[15] L. Beer, A. Wallace Cordes, D.J.T. Myles, R.T. Oakley, N.J. Taylor, CrystEngComm
2 (2000) 109.
[16] E.M. Fatila, M.C. Jennings, J. Goodreid, K.E. Preuss, Acta Crystallogr., Sect. C
Cryst. Struct. Comm. 66 (2010) o260.
[25] G.M. Sheldrick, Acta Crystallogr., Sect. A Found. Crystallogr. 64 (2008) 112.
[26] D.J. Watkin, Crystallogr. Rev. 16 (2010) 197.
[17] S.C. Nyburg, C.H. Faerman, Acta Crystallogr., Sect. B Struct. Sci. B41 (1985) 274.
[18] R.T. Boeré, K.H. Moock, M. Parvez, Z. Anorg. Allg. Chem. 620 (1994) 1589.
[19] N. Bricklebank, S. Hargreaves, S.E. Spey, Polyhedron 19 (2000) 1163.
[20] M.A. Spackman, D. Jayatilaka, CrystEngComm 11 (2009) 19.
[27] A.L. Spek, Acta Crystallogr., Sect. A Found. Crystallogr. A46 (1990) C34.
[28] S.K. Wollf, D.J. Grimwood, J.J. McKinnon, D. Jayatilaka, M.A. Spackman,
CRYSTALEXPLORER 2.1, University of Western Australia, Perth, Australia, 2007.
[29] C. Allen, D.A. Haynes, C.M. Pask, J.M. Rawson, CrystEngComm 11 (2009) 2048.