Structural Variations in the Dithiadiazolyl Radicals p-ROC6F4CNSSN (R = Me, Et, nPr, nBu): A Case Study of Reversible and Irreversible Phase Transitions in p-EtOC6F4CNSSN

Article abstract of DOI:10.1021/acs.cgd.7b01065

The 4′-alkoxy-tetrafluorophenyl dithiadiazolyls, ROC₆F₄CNSSN [R = Me (1), Et (2), ⁿPr (3), ⁿBu(4)] all adopt cis-oid dimers in the solid state. The methoxy derivative 1 adopts a π-stacked AA'AA' motif, whereas p

Full text of DOI:10.1021/acs.cgd.7b01065

Article
pubs.acs.org/crystal
Cite This: Cryst. Growth Des. XXXX, XXX, XXX−XXX
Structural Variations in the Dithiadiazolyl Radicals p‑ROC6F4CNSSN (R
= Me, Et, nPr, nBu): A Case Study of Reversible and Irreversible Phase
Transitions in p‑EtOC6F4CNSSN
Published as part of a Crystal Growth and Design virtual special issue Honoring Prof. William Jones and His
Contributions to Organic Solid-State Chemistry
Yassine Beldjoudi,† Rui Sun,†,# Ana Arauzo,‡ Javier Campo,‡ Robert J. Less,§
and Jeremy M. Rawson*,†,§
†Department of Chemistry & Biochemistry, University of Windsor, 401 Sunset Avenue, Windsor, Ontario N9B 3P4, Canada
‡Aragon
́
Materials Science Institute (CSIC-University of Zaragoza) and Condensed Matter Physics Department, Facultad de Ciencias,
E-50009 Zaragoza, Spain
§Department of Chemistry, The University of Cambridge, Lensfield Road, Cambridge, U.K. CB2 1EW
S
* Supporting Information
ABSTRACT: The 4′-alkoxy-tetrafluorophenyl dithiadiazolyls,
n
n
ROC6F4CNSSN [R = Me (1), Et (2), Pr (3), Bu(4)] all adopt cis-oid
dimers in the solid state. The methoxy derivative 1 adopts a π-stacked
AA’AA’ motif, whereas propoxy (3) and butoxy (4) derivatives exhibit an
AA’BB’ stacking. The ethoxy derivative (2) is polymorphic. The α-phase
(2α) adopts an AA’BB’ motif comparable with 3 and 4, whereas 2β and 2γ
are reminiscent of 1 but combine a mixture of both monomers and dimers
in the solid state. The structure of 2β exhibits Z’ = 6 with two dimers and
two monomers in the asymmetric unit but undergoes a thermally induced
phase transition upon cooling below −25 °C to form 2γ (Z’ = 14) with six
dimers and two monomers in the asymmetric unit. The transition is associated with both rotation and translation of the
dithiadiazolyl ring. Detailed differential scanning calorimetry and variable temperature powder X-ray diffraction studies coupled
with SQUID magnetometry have been used to show that 2α converts irreversibly to 2β upon heating and that 2β and 2γ
interconvert through a reversible phase transition with a small thermal hysteresis in its magnetic response.
Scheme 1. Common Favorable Electrostatic in Plane Sδ+···
INTRODUCTION
■
Nδ− Contacts Identified in DTDA Radicals
Over the last 30 years, 1,2,3,5-dithiadiazolyl (DTDA) radicals
have been explored as building blocks for neutral radical
conductors,1 as building blocks in the design of some of the
highest TC organic magnets2 and as components in photo-
conducting devices.3 They have also been used as ligands in
coordination chemistry4 and have been shown to undergo
oxidative addition to low oxidation state metals.5 More recent
work has shown that polyaromatic hydrocarbons bearing a
DTDA moiety exhibit photoluminescence.6
The electronic properties of DTDA radicals such as
conductivity and magnetism are heavily dependent on the
solid-state structure.7 Work by Rawson and Haynes identified
common packing motifs between DTDA radicals in the absence
of additional functionality (Scheme 1),8 and the implementa-
tion of other strong structure-directing groups to dictate the
structure has been reviewed.9 The strong tendency of these
radicals to dimerize (ΔHdim ≈ 35 kJ·mol−1)10 has recently been
used by Preuss as a supramolecular synthon for the self-
organization of lanthanide-DTDA complexes.11
retain their paramagnetic behavior.7 Several strategies have
been explored to potentially inhibit the dimerization process.
While the presence of sterically demanding substituents can
disrupt the common cis-oid dimerization (Scheme 2),
alternative dimerization modes have been observed. The fine
balance between monomer and dimer structures is exemplified
by 2,4,6-(F3C)3C6H2CNSSN, which was found to exist in both
trans-antarafacial and monomeric forms.12
Received: August 2, 2017
Revised: October 17, 2017
Published: November 17, 2017
The utilization of DTDA radicals as building blocks for the
design of organic magnetic materials necessitates that molecules
© XXXX American Chemical Society
A
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
isolated when the “cold” finger temperature was maintained at
+30 °C. Polymorph 2γ was formed through a solid state phase
transition by cooling 2β below −25 °C.
Scheme 2. Common Dimerization Modes of DTDA
Radicals: (i) Cis-oid, (ii) Twisted, (iii) Trans-antarafacial,
(iv) Trans-cofacial and (v) Orthogonal Configurations
Crystal Structures. Crystallographic data for 1−4 are
summarized in Table 1. Radicals 1, 3, and 4 adopt a single
crystalline phase, whereas 2 crystallizes in three different
phases: 2α, 2β, and 2γ.
Table 1. Crystallographic Data for 1−4
1
3
4
The most commonly implemented approach to suppress
dimerization has been the use of perfluoroaryl substituents
coupled with structure-directing groups. In these cases, the
electrostatic repulsion between the fluorines in the ortho
position of the perfluoroaryl ring and the nitrogen of the
heterocyclic ring leads to a large twist angle (energy minimum
at θ ≈ 50°)13 between the two rings. This makes the
perfluoroaryl ring sterically demanding, destabilizing the
dimer. In the simplest case, 2,6-F2C6H3CNSSN, the compound
is found to be trimorphic with two cis dimers formed
(tetragonal and monoclinic phases)13 and a third monoclinic
phase formed which comprises a trans-antarafacial dimer
(Scheme 2) and a monomer,14 reflecting the fine energetic
balance between monomer and dimer in this system. In such
cases, the inclusion of additional strong structure-directing
groups may overcome the tendency for dimerization,
exemplified by the presence of CN···S interactions in both α-
and β-phases of p-NCC6F4CNSSN15,16 as well as p-
NCC6F4C6F4CNSSN.17 Similarly NO2···S interactions in p-
O2NC6F4CNSSN18 and Br···N sigma-hole interactions in the
case of p-BrC6F4CNSSN19 appear adequate to overcome
dimerization. In the current paper, we examine how subtle
tuning of the 4′-substituent in the series p-ROC6F4CNSSN (R
formula
C8H3OF4N2S2
283.24
C10H7OF4N2S2
311.3
C11H9OF4N2S2
325.3
FW
temp (K)
150(2)
173(2)
173(2)
crystal system
monoclinic
P21/c
monoclinic
P21/c
monoclinic
P21/c
space group
a/Å
7.3635(4)
31.7443(17)
8.3990(4)
90.00
16.2391(10)
18.0475(9)
8.1473(5)
90.00
16.9314(7)
8.0672(3)
22.8387(8)
90.00
b/Å
c/Å
α/°
β/°
γ/°
V/Å3
101.436(2)
90.00
100.045(2)
90.00
124.995(2)
90.00
1922.28(17)
8
2351.2(2)
8
2555.51(17)
8
Z
Dc/g·cm−3
1.955
1.759
1.691
Rint
0.040
0.055
0.046
R1
0.051
0.047
0.043
wR2
0.123
0.103
0.099
S
1.15
1.06
1.05
Δρmax, Δρmin
0.42, −0.54
0.45, −0.38
0.49, −0.39
(e Å−3
)
2α
2β
2γ
formula
FW
C9H5OF4N2S2
297.27
C9H5OF4N2S2
297.27
C9H5OF4N2S2
297.27
n
n
= Me, Et, Pr, and Bu, Scheme 3) affects the solid-state
structure and probe the polymorphic nature of 2.
temp (K)
173(2)
253(2)
173(2)
crystal system
monoclinic
P21/c
triclinic
triclinic
space group
P1
̅
P1
̅
Scheme 3
a/Å
15.6073(15)
17.3379(15)
8.1387(7)
90.00
8.3271(8)
11.9249(12)
35.154(4)
94.5571(7)
91.851(6)
106.655(6)
3328.1(6)
12
8.4511(15)
25.774(4)
34.955(6)
90.013(4)
90.207(5)
94.726(5)
7588(2)
28
b/Å
c/Å
α/°
β/°
γ/°
V/Å3
99.208(5)
90.00
2173.9(3)
8
Z
Dc/g·cm−3
1.817
1.780
1.822
RESULTS
■
Rint
0.073
0.049
0.096
All four 4-alkoxy perfluoroaryl carbonitrile derivatives,
ROC6F4CN, were synthesized from a mixture of pentafluor-
obenzonitrile and sodium alkoxide in the corresponding alcohol
based on the literature procedure (see ESI-2).20 The 4′-alkoxy-
functionalized perfluoroaryl DTDA radicals 1−4 were then
prepared using standard synthetic methodologies21 using
triphenyl antimony as the reducing agent due to the low
melting point of the radicals (60−80 °C), which avoided any
contamination with Ph3SbCl2 (mp = 142 °C) during
sublimation.22 In this manner radicals 1−4 were obtained in
crystalline form in 37−70% recovered yield. Radical 2 was
found to be polymorphic, and careful monitoring of the
temperature of the coldfinger was required to selectively
crystallize either 2α or 2β, a procedure that has been already
utilized to selectively prepare α or β polymorphs of p-
NCC6F4CNSSN.15 In this fashion 2α was selectively formed
when collected on a coldfinger at −15 °C, whereas 2β was
R1
0.077
0.087
0.130
wR2
0.170
0.245
0.341
S
1.11
1.02
1.04
Δρmax, Δρmin (e Å−3
)
0.50, −0.45
1.04, −0.45
1.21, −0.80
Selected structure parameters for 1−4 are reported in Table
Crystal Structure of 1. Radical 1 crystallizes in the space
2.
group P21/c with one cis-oid dimer (two crystallographically
independent molecules) in the asymmetric unit. The twist
angles between the heterocyclic ring plane and the aryl ring
plane for the two molecules are 31.2(5)° and 30.8(5)°,
comparable with other perfluorinated aryl DTDA radicals
(14.5−57.8°, mean 37.8°).16−19,23 Previous computational
studies revealed an energy minimum at ∼50° but a relatively
shallow potential well for deviation away from this minimum
B
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Table 2. Selected Structural Parameters for Dimers of 1−4
angle between C6F4 and
DTDA ring planes (deg)
angle between DTDA in plane S···N contacts in plane S···F contacts
intradimer dS···S (Å)
intradimer dO···O (Å)
ring planes (deg)
(Å)
(Å)
1
3.179(2)−3.198(2)
mean: 3.188
3.751(4)
30.8(5)−31.2(5)
mean: 31.0
5.6(2)
3.295(3)−3.387(3)
3.018(2)−3.232(2)
2α
2β
2γ
3
3.066(2)−3.090(2)
mean: 3.078
4.198(4)
29.2(7)−35.3(7)
mean: 32.2
8.7(3)
8.7(4)
6.3(2)
10.0(6)
13.3(5)
3.102(4)−3.235(4)
3.139(2)−3.372(2)
2.964(3)−3.324(3)
3.095(2)−3.247(2)
2.998(1)−3.236(1)
2.990(3)−3.235(3)
3.023(9)−3.220(9)
2.82(3)−3.17(3)
3.162(8)−3.204(9)
mean: 3.183
4.087(4)−4.204(2)
mean: 4.145
27(3)−39(3)
mean: 33
3.124(6)−3.177(6)
mean: 3.150
3.877(2)−4.056(2)
mean: 3.966
30(2)−36(2)
mean: 33
3.077(1)−3.137(1)
mean: 3.107
4.460(3)
40.0(3)−35.1(3)
mean: 37.5
3.091(1)−3.189(1)
3.021(1)−3.200(1)
4
3.099(1)−3.049(1)
mean: 3.074
5.379(2)
44.4(2)−35.2(2)
mean: 39.7
such that angles in the range 30−90° fall within 3 kJ·mol−1 of
this minimum.13 A comparison of the experimentally observed
torsion angles for fluorinated and nonfluorinated aryl groups in
relation to the computed energies are presented in Figure S.4.
The intradimer S···S contacts of 3.179(2) and 3.198(2) Å are
significantly less than the van der Waals radii (3.60 Å), whereas
the intradimer O···O contact, at 3.751(4) Å, is much larger than
twice the sum of the van der Waals radius of O (3.04 Å). The
nature of these intradimer contacts afford a slightly “wedge-
shaped” dimer. The angle between the two DTDA planes
within the dimer is ∼5.6(2)°. The dimers of 1 adopt a π-
stacked structure along the a-axis (Figure 1a) in which the
Figure 2. Crystal structure of 2α. (a) Antiparallel arrangement of
dimers parallel to the crystallographic a axis; (b) propagation of
interstack contacts via SN-IV type interactions in the bc-plane.
dimer S···S contacts (3.049(1)−3.206(1) Å) remain in the
typical range for DTDA radicals (Table 2), the intradimer O···
O separation increases steadily with increasing chain length
from 3.751(4) to 5.379(2) Å, revealing a linear relationship
between the length of the alkoxy group and O···O separation
(Figure S3 (left), Supporting Information). This increasing O···
O separation reflects additional strain through “hinging” at the
intradimer DTDA S···S contacts with the angle between DTDA
ring planes increasing form 5.6(2) to 13.3(5)° (Table 2).
A consequence of this increased “wedging” of the dimer
geometry is to generate a fundamental change in the π-stacking,
while 1, 2α, 3, and 4 all crystallize in the P21/c space group, the
inversion center is located between stacks in 1 (affording an
AA’AA’ π-stacking motif) but is positioned within stacks in 2α,
3, and 4, generating an AA’BB’ packing. The latter appears
better able to accommodate the more “wedge-shaped” dimers
associated with 2α, 3, and 4. The structure of 2α is shown in
Figure 2. Perpendicular to the stacking direction 2α, 3, and 4
form similar interstack contacts to 1 (compare Figure 1b and
Figure 2b).
In all cases the in-plane S···N contacts are supplemented by
S···F contacts linking dimers together in the bc plane. These fall
in the range 2.82−3.26 Å (Table 2), cf. sum of the van der
Waals radii of S and F (3.27 Å).
Crystals of 2β were collected at +30 °C but underwent a
phase transition below −25 °C to form 2γ upon cooling,
confirmed by DSC measurements (Supporting Information,
Figure S6). Both 2β and 2γ showed a propensity for twinning
Figure 1. Crystal structure of 1. (a) π-Stacking parallel to the
crystallographic a-axis; (b) propagation of interstack contacts via SN-
IV type interactions in the bc-plane.
interdimer S···S contacts (4.156(2)−4.180(2) Å) are markedly
longer than the intradimer S···S distances, like many π-stacked
DTDA radicals.6,13,24−26 Conversely the interdimer O···O
distance of 3.615(4) Å is comparable with the intradimer
contact (3.751(4) Å). Each molecule within a dimer forms
close contacts to two neighboring molecules perpendicular to
the stacking direction through electrostatically favorable δ+S···
Nδ− contacts of the SN-IV type (Scheme 1),9 leading to a
chain-like motif in the bc-plane (Figure 1b).
Crystal Structures of 2α, 3, and 4. Radicals 2α, 3, and 4
crystallize as cis-oid dimers in the space group P21/c, each with
two molecules (one dimer) in the asymmetric unit (Figure 2;
Figures S1 and S2, Supporting Information). The torsion angles
between heterocyclic and perfluoroaryl rings in 2α, 3, and 4
span the range 29.2−44.4(2)°, comparable with 1 and other
perfluoroaryl DTDAs (vide supra).16,17,23 Although the intra-
C
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
and cooling through the phase transition led to significant
degradation in crystal quality. The structures described here are
the best of multiple crystals examined.
adjacent to each other in neighboring stacks offering an
orthogonal π−π interaction between DTDA rings (Figure 3b).
The closest S···S contacts in this arrangement are in excess of
3.8 Å, consistent with a van der Waals interaction rather than a
π−π interaction between SOMOs.
Crystal Structure of 2β. At −20 °C a crystal of 2β was
indexed on a triclinic cell (P1) with six molecules in the
̅
asymmetric unit. These six crystallographically independent
molecules comprise two cis-oid dimers and two monomeric
radicals. The heterocyclic ring planes of the monomers are
markedly twisted with respect to the dimer DTDA ring planes
(Figure 3). Structures of DTDA radicals comprising mixtures of
Crystal Structure of 2γ. On cooling below −25 °C, 2β
undergoes a reversible phase transition to 2γ which is evident in
variable temperature powder X-ray diffraction (VT-PXRD) and
DSC studies (vide infra). Although samples cooled through the
phase transition retained crystallinity, the diffraction pattern
persistently reflected some degree of twinning in all crystals
examined and was reflected in a poorer internal R value (Rint)
compared to the high temperature phase (Table 1). The
emergence of new reflections was consistent with the formation
of a supercell. An examination of the unit cell parameters for 2β
and 2γ reveal similar a and c parameters but an increase in the
crystallographic b-axis from 11.9249(12) to 25.774(4) Å (Table
1). A full data set was measured at 173 K (75 K below the
phase transition) on a crystal cooled through the phase
transition. The new data initially indexed on a higher symmetry
monoclinic supercell. Structure solution in the P21 space group
provided an initial solution with Z’ = 14, but the refinement
stalled at R1 = 27% with evidence for twinning (systematic
observation of Fo > Fc for the most disagreeable reflections). A
possible two component twin, close to merohedral was
identified but failed to provide a significant improvement in
R1. When the data were reprocessed in the lower symmetry
Figure 3. Crystal structure of 2β. (a) One of the two crystallo-
graphically independent stacks of dimers and monomers in 2β. (b)
Propagation of interstack contacts via SN-IV and SN-I type
interactions.
triclinic setting, a similar solution in P1 (Z’ = 14) was identified.
̅
In this case application of the merohedral twin law (−1 0 0 0
−1 0 0 0 1) led to a marked decrease in R1 and refinement
proceeded smoothly to afford a final R1 of 13%. Although the R
value is high it is in line with the intrinsic quality of the data
(Rint ≈ 9%) and linked to the inherent degradation in crystal
quality on passing through the phase transition. Attempts to
grow single crystals of this low temperature phase directly by
using a low temperature coldfinger proved unsuccessful due to
the thermodynamic preference to crystallize as the form 2α at
these lower temperatures.
both dimers and monomers are rare but not without
precedent.12,14,27 The intradimer S···S contacts fall in the
range 3.162(8)−3.204(9) Å (Table 2), while the corresponding
intradimer O···O separations are 4.087(4) and 4.204(2) Å,
comparable with 2α (4.198(4) Å). The hinge angles between
DTDA rings within these dimers are in range 27−39°.
The unit cell parameters of 2γ are similar to those of 2β but
with an expansion of the crystallographic b axis and some
modification of the unit cell angles. The structure of 2γ
contains 14 molecules in the asymmetric unit, three of which
show a small degree of disorder (<10%) in the orientation of
the DTDA ring. Considering only the major component of the
disorder then the low temperature structure comprises six cis-
oid dimers and two monomers which can be split into two
crystallographically independent stacks, each containing three
dimers and a monomer (Figure 4a). While many DTDA
radicals exhibit large Z’ structures,9b the current example has
the largest Z’ value of reported DTDA radicals to date.
In 2β the structure comprised an alternating dimer−
monomer repeat along the stacking direction (Figure 3a),
whereas 2γ now exhibits a repeat unit of three dimers and a
monomer along the stacking direction (Figure 4a). The
combination of the three molecule repeat for 2β and seven
molecule repeat for 2γ requires a common supercell containing
21 molecules along the stacking direction. As a consequence,
the transition from 2β to 2γ not only requires disruption of the
existing dimers but also monomers becoming involved in dimer
formation. A representation of the displacement of the radicals
in 2β and 2γ along the stacking direction in relation to the
supercell is presented in Figure 5. Given the disruption along
the stacking direction associated with the phase transition, it is
As with other structures in this series, 2β adopts a π-stacked
structure parallel to the crystallographic b-axis. The two dimers
and two monomers in the asymmetric unit form two
crystallographically independent stacks with each stack
containing one dimer and one monomer (Figure 3a). Although
the perfluoroaryl rings associated with the two monomeric
radicals are well-located, their DTDA rings appear thermally
disordered (vide infra) via rotation of the CNSSN ring about
the CAr−CDTDA bond. In each case this ring was modeled over
two orientations. This disorder reveals that there is adequate
space between dimers along the stacking direction to
accommodate significant DTDA ring rotation in this structure.
Notably the major component of disorder in both independent
monomers (60% and 85%) reveals the DTDA ring is close to
perpendicular to the DTDA rings in the dimers (Figure 3a).
This behavior is reminiscent of (o-ClC6H4CNSSN)2, although
the latter shows no structural disorder.27b Notably the
interstack contacts still adopt a similar motif to that observed
in the other structures in this series (Figure 3b). In 2β rotation
of 33% of the DTDA rings leads to a mixture of SN-I (two S···
N contacts with coplanarity unnecessary) and SN-IV (S···N and
S···S contacts between near coplanar DTDA heterocycles).
Although the closest centroid···centroid distances between
monomers along the stacking direction is 12.4 Å, the two
independent rotationally disordered radicals are located
D
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
dispersion and weak interdimer bonding interactions, whereas
the latter is stabilized through dipole−dipole interactions. The
increase of the alkyl chain length affords a linear increase of the
unit cell volume (Figure S4 (right)) and a monotonic increase
in O···O separation within dimers, affording a more “wedge-
shaped” motif for larger alkyl groups. This subtle change in
geometry of the cis-oid dimers leads to marked changes in the
solid-state packing. While the smaller methoxy derivative 1
forms a π-stacked structure, the larger alkoxy derivatives (3 and
4) which exhibit more wedge-shaped geometries adopt AA′BB′
motifs as a more efficient packing mode (Figure 6). In the case
Figure 4. Crystal structure of 2γ. (a) One of the two crystallo-
graphically independent stacks of dimers and monomers; (b)
propagation of interstack contacts via SN-IV and SN-I type
interactions (molecules colored by crystallographic independence).
Figure 6. Packing arrangements for 1−4. Trapezoids represent dimers
and rectangles represent rotationally disordered monomers.
of 2 there is a fine balance between these two scenarios with
polymorphs 2β and 2γ exhibiting the AA′AA′ motif and 2α
revealing an AA′BB′ stacking pattern. In 2β and 2γ the AA′AA′
packing of the wedge-shaped motif leads to some degree of
inefficient packing. At low temperature (2γ) there is just one
monomer observed for every three dimers, but thermal
expansion of the lattice leads to additional monomers at
elevated temperatures (2β exhibits one monomer for each
dimer) enhancing the number of paramagnetic centers in the
lattice on warming.
Figure 5. Comparison of the common supercell associated with π-
stacking in 2β (top) and 2γ (bottom).
Thermal Studies on 1−4. A summary of the thermody-
namic data determined from DSC studies for 1−4 is presented
in Table S2, and values quoted in the text refer to onset
temperatures (Tonset). DSC studies on 1 and 4 revealed the
presence of a single endothermic process at 78 and 58 °C for 1
and 4 respectively associated with their melting points. The
absence of any additional features on thermal cycling coupled
with VT-PXRD studies (see Supporting Information) showed
no evidence for polymorphism in these two materials.
The selective preparation of both polymorphs 2α and 2β
permitted a series of studies on the relative stabilities and phase
transitions of 2. For 2α variable temperature single crystal and
PXRD studies showed no significant structural change in 2α
from −170 to +65 °C (Figure 7a). At 75(2) °C PXRD studies
revealed 2α had converted to 2β (Figure 6). DSC studies on a
pure sample of 2α revealed the presence of two successive
endothermic transitions on warming at +76 and +80 °C with a
total enthalpy change of ΔH = +20 kJ·mol−1 on the heating
cycle (Supporting Information, Figure S6). The first transition
is attributed to the conversion of 2α to 2β, whereas the second
is attributed to the melt of 2β. Upon slow cooling,
recrystallization of liquid 2 occurs with formation of 2β,
confirmed by VT-PXRD studies (Figure 7). A comparison of
the structures of 2α and 2β reveal that the transformation of 2α
to 2β requires a major lattice reorganization (50% of molecules
to undergo a 180° rotation or 100% of molecules to undergo a
90° rotation). Previous studies have identified such major
perhaps unsurprising that there was an inherent decrease in
crystal quality on passing through this phase transition. As with
2β, the monomeric radicals in 2γ exhibit some rotational
disorder, although this is much reduced when compared to 2β.
In 2γ monomers are not only well isolated along the stacking
direction (monomer···monomer separation ∼28 Å) but also
between stacks (monomer···monomer separation ∼13 Å).
Between stacks the radicals are once again linked via a
combination of SN-I and SN-IV type contacts (Figure 4b).
Comparison of Structures of 1−4. DTDA radicals
bearing simple phenyl ring substituents often adopt herring-
bone motifs through δ+S···π interactions as exemplified by
(PhCNSSN)2.28 The inclusion of the perfluoroaryl group
reduces π-spin density evidenced by Haynes’ charge density
studies on (PhCNSSN)2, and (C6F5CNSSN)2,29 suppressing
herringbone motifs in perfluoro-aryl DTDAs.16−19,23 Their
heavier Se analogues30 also tend to adopt π-stacked/layer-like
structures. The structures of 1−4 all comprise layer-like
architectures which maximize interstack S···N contacts of the
type SN-I and SN-IV (Figures 1b, 2b, 3b, 4b and Supporting
Information, Figures S1b, S2b). The arrangement of molecules
along the stacking direction appears sensitive to the functional
group size. While compounds 1, 2β, and 2γ adopt π-stacked
structures (of the AA′AA′ variety), 2α, 3, and 4 adopt layered
structures in which discrete dimers align antiparallel adopting
an AA′BB′ motif. The former appears stabilized through
E
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
more intense feature near 22°/2θ on cooling to −25 °C. More
subtle changes in peak position and intensities were evident on
cooling from −25 °C to −170 °C, suggestive of a dynamic
process in this regime. In addition the re-emergence of the
original PXRD profile on warming back to room temperature
confirmed the reversible nature of the 2β ↔ 2γ transition seen
in DSC studies. Powder patterns for 2β and 2γ were in good
agreement with those predicted from the single crystal X-ray
data.
Application of the density rule (2β = 1.781 g·cm−3, cf. 2γ =
1.821 g·cm−3) suggests that the lattice enthalpy of 2γ is greater
than that of 2β, and therefore the conversion of 2γ to 2β on
warming is entropically driven, consistent with the increasing
number of disordered monomers. In this context, 2β can be
considered as the most entropically favorable phase, whereas
2α is the most enthalpically favorable phase with 2γ metastable
in the low temperature region. This is entirely consistent with
experimental observations that sublimation at high temperature
(“cold” finger temperature +30 °C) affords selectively the most
entropically favorable phase 2β, whereas sublimation using low
substrate temperatures (coldfinger set to −15 °C) affords the
most enthalpically stable phase (2α) with 2γ only isolated
through the solid-state transformation 2β → 2γ on cooling.
Radical 3 reveals two endothermic transitions on heating
(Tonset = 64 and 91 °C, Figure S7). VT-PXRD studies on 3
reveal no significant change in the profile of the powder pattern
across the temperature range 20−65 °C, agreeing well with that
predicted based on the single crystal structure determination
(Figure S9, Supporting Information). At 80 °C, the pattern
changed significantly to afford just a few broad low intensity
features. We tentatively assign this to the formation of a glassy
phase with some degree of structural order (possibly a liquid
crystalline phase). The second transition at 91 °C has a much
lower enthalpy change (ΔH = 1.54 kJ mol−1) consistent with
the breakdown of the final components of short-range ordering
and formation of an isotropic liquid. The integrity of the radical
over this temperature range was confirmed by the VT-PXRD
studies which revealed reformation of crystalline 3 upon
cooling (see Figure S9, Supporting Information).
Figure 7. Variable temperature PXRD profiles observed on heating
2α.
lattice reorganizations occur via a melt/recrystallization
process.15
The endothermic nature of the conversion of 2α to 2β
indicates this spontaneous process is entropically driven such
that 2α is the enthalpically preferred of the two polymorphs
and 2β is the entropically more stable phase. On the basis of
the density rule,31 (2α = 1.842 g·cm−3 cf. 2β = 1.781 g·cm−3)
the lattice enthalpy of 2α is greater than 2β in agreement with
these experimental observations. DSC studies on a pure sample
of 2β heated from −60 to +150 °C revealed two endothermic
transitions occurring at −25 °C and +78 °C. The transition at
−25 °C is associated with the phase transition from the low
temperature phase 2γ to the high temperature phase 2β,
corroborated by single crystal and PXRD studies (Figure 8).
Magnetic Studies on 2. DC magnetic studies on 2α and
2β were measured on a Quantum Design SQUID magneto-
meter. Data for 2α (69 mg) were measured from 5 to 300 K
with the sample sealed in a gelatin capsule with an applied field
of 10 000 Oe. The sample of 2β (47 mg) was mounted in a
quartz tube and measured from 150 to 375 K with an applied
field of 10 000 Oe. Additional heating and cooling cycles on 2β
(46 mg) mounted in a gelatin capsule were undertaken in
applied fields of 500 Oe to probe the solid-state transition
between 2β and 2γ. Data were corrected for diamagnetism of
both the sample and holder.
For 2α, the value of χT remains approximately constant at
0.003 emu·K·mol−1 on warming up to 250 K (Figure 9 (top)),
consistent with a small number of S = 1/2 defects in the crystal
lattice (0.6%). Above 250 K the susceptibility of 2α begins to
increase. The onset of significant sample paramagnetism in
such DTDA dimers has been reported for a number of DTDA
radicals and has previously been shown to be consistent with
thermal population of a low-lying excited state triplet.25,26,32 In
the current case, modeling the magnetic data to the Bleaney−
Bowers model33 [H = −2JS1·S2] failed to reproduce the
sample’s magnetic behavior (dashed line in Figure 9 (top)).
Alternatively paramagnetism may arise from some degree of
dissociation of the dimers to form monomers in the solid state.
Figure 8. Variable temperature PXRD profiles of 2β on cooling,
converting to 2γ at −25 °C and subsequent conversion back to 2β on
warming to room temperature.
The enthalpy change associated with this process (+0.6 kJ·
mol−1) is extremely small. The transition at +78 °C (ΔHfus
=
+19.2 kJ·mol−1) is assigned to the melting of 2β. On cooling a
small transition at −30 °C was recorded reflecting the
reversible nature of the 2γ to 2β transition. This behavior
was confirmed by VT-PXRD studies (Figure 8) which revealed
the disappearance of weak peaks around 7° and 16° as well as a
F
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Neglecting the possibility of radical−radical exchange
coupling, the magnetic data of 2α can be well fitted to the
equilibrium model with just two parameters. The values of
ΔHdim (−30 kJ·mol−1) and ΔSdim (−71 J·K−1·mol−1) are
gratifyingly comparable with experimentally determined ΔHdim
and ΔSdim values for (PhCNSSN)2 in solution (ΔHdim = −35
kJ·mol−1, ΔS = −121 J·K−1·mol−1).10b A surge in the value of
χT commences around 340 K and levels off at 350 K and is
associated with the melting of 2α to form liquid 2.
For 2β/2γ the low temperature value of χT remains constant
at 0.057 emu·K·mol−1 up to 250 K (Figure 9 (bottom)), close
to the value expected for 1-in-7 radicals being unpaired (0.054
emu·K·mol−1), consistent with the structural data for 2γ. From
246 to 264 K there is a step in the value of χT, increasing to
0.089 emu·K·mol−1, a little less than that predicted for one-
third of the molecules contributing to sample paramagnetism
(0.125 emu·K·mol−1) anticipated from the structure of 2β.
From 330 to 350 K there is a further increase in paramagnetism
up to a maximum of 0.180 emu·K·mol−1 as the sample
approaches the melting point, corresponding to ca. 50%
unpaired spins. This behavior is comparable with other studies
on DTDA radicals in the liquid phase where the value of χT
remains below the value expected for a pure S = 1/2 system and
has been attributed to aggregation in the melt.34 The final
decrease in χT above the melting point is associated with loss of
sample from the cavity space through sublimation and reflected
in coating of the quartz sample holder with radical when
removed from the magnetometer. Additional studies of the
phase transition 2β ↔ 2γ in heating and cooling modes confirm
the reversible nature of the transition and presence of a small
thermal hysteresis loop. If the onset temperatures for the phase
transition on heating/cooling are considered then there is
negligible hysteresis (TC = 257(1) K), but if the transition
temperatures are defined by the maximum change in gradient
of χT(T) then this affords (TC(↑) = 261(1) K, TC(↓) = 256(1) K
(scan rate = 1 K·min−1) (Figure 10). Additional heat capacity
measurements (scan rate = 0.1 K·min−1) agree well with the
SQUID data, affording a peak maximum at 259 K.
Figure 9. Plots of χT vs T for 2α and 2β/2γ. For 2α (top) the dashed
line corresponds to the fit to the Bleaney−Bowers model, whereas the
solid line corresponds to the equilibrium model. For 2β/2γ (bottom)
the vertical dashed line at 259 K corresponds to the maximum
observed in the heat capacity, and the vertical dashed line at 350 K
corresponds to the melting point. The dashed line through the data
merely acts as a guide to the eye.
Similar arguments have been implemented to model the weak
paramagnetism observed in the liquid phase of F3CCNSSN.34
In this context, we applied a simple monomer−dimer
equilibrium model (eq 1) based on a two parameter fit
(ΔHdim and ΔSdim) which define the percentage monomer
present (see ESI-9).
(EtOC6F4CNSSN)2 ⇌ 2EtOC6F4CNSSN·
(1)
This behavior is not unreasonable within the context of the
recent pair exchange dynamics model proposed for related
thiazyl radicals (vide infra).35 This dimer dissociation model
assumes that both exchange coupling (i) between the two
radicals generated and (ii) to neighboring radicals is negligible.
At low temperature when the equilibrium lies far to the left (eq
1) then the small numbers of monomer pairs generated are well
separated from each other. The latter is certainly a good
assumption. The assumption that exchange coupling between
radical pairs generated by dimer dissociation is negligible
should be treated more cautiously, although computed
exchange couplings between radicals at/near the sum of the
van der Waals radii are typically an order of magnitude or more
less (<102 K) than the intradimer exchange coupling.36
Figure 10. Plot of χT vs T on heating and cooling for 2β/2γ and the
heat capacity measured over the same temperature range. The dashed
lines merely act as a guide to the eye.
G
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
associated with a lanthanide-bridging DTDA complex described
by Preuss11b comprise two sequential structural phase
transitions at 160 and 310 K, which involve rupture of 50%
(160 K) and 100% (310 K) of the DTDA π*−π* dimer
interactions.
DISCUSSION
■
Polymorphism in DTDAs is typically manifested in different
modes of dimerization and/or different solid-state packing
motifs. Even “simple” radicals such as HCNSSN37 and
ClCNSSN8,38 where there is no torsional flexibility have been
found to be polymorphic. For example ClCNSSN has been
isolated as both cis-oid dimers (α- and γ-polymorphs) and
twisted dimers (β, δ and ε-polymorphs) reflecting small
energetic differences between cis-oid and twisted dimer
conformations.8,38 Several examples of structures comprising
both monomers and dimers in the solid state have been
reported,12,14 highlighting that their dimerization enthalpies are
comparable with other packing forces.
CONCLUSIONS
■
The length of the alkoxy chain influences the solid-state
structures of the alkoxy-tetrafluorophenyl DTDA radicals 1−4
in a systematic fashion. The ethoxy derivative 2 appears to be at
the crossing point in terms of structure stability and was found
to be polymorphic; 2α adopts the AA’BB’ dimer motif observed
for 3 and 4, whereas 2β and 2γ adopt a cofacial packing
arrangement akin to 1. Unlike 1, the structure of 2β is more
complex with the structure seemingly unable to support
formation of a pure dimer phase. At low temperature, the
structure of 2γ comprises a 3:1 ratio of dimers to monomers,
whereas in the high temperature phase 2β exhibits a
dimer:monomer ratio of 1:1. The rich polymorphism and
phase transition behavior associated with 2 was probed by
DSC, VT-PXRD, and SQUID magnetometry. These reveal an
irreversible transition from 2α to 2β upon warming and a
reversible transition between the high temperature phase 2β
and the low temperature form (2γ), which are associated with a
unique dynamic process along the stacking direction involving
both displacive and rotational motion of the DTDA radicals.
There was no evidence for polymorphism in 1 and 4 based
on DSC and VT-PXRD studies. For 3 the variable temperature
studies showed some evidence for an intermediate glassy or
liquid crystalline phase occurring between the crystalline and
liquid phases (64−91 °C), but again no direct evidence for
additional polymorphs was observed during our thermal studies
on this material. The polymorphism in 2 appears to arise as a
result of its structurally intermediate nature between 1 (which
adopts an AA’AA’ stacked structure) and 3 and 4 (which adopt
an AA’BB’ stacking). The significant structural differences
between 2α and 2β are reflected in a melt−recrystallization
process for the irreversible endothermic 2α → 2β trans-
formation, comparable to the phase transition between α and β
phases of p-NCC6F4CNSSN.15 The instability of 2β on cooling
cannot be alleviated by a 2β → 2α transformation, but adopts
an alternative metastable state 2γ which is more enthalpically
favorable than 2β. The reversible solid-state phase transition
between 2β and 2γ necessitates a combination of small
molecular displacements and ring rotations along the stacking
direction (Figure 5). The differing number of monomers in 2β
(33%) and 2γ (14%) leads to an associated change in magnetic
response. Such single crystal to single crystal phase transitions
are not without precedence in thiazyl radical chemistry. In the
current case, the aryl substituents provide a well-defined
framework of near-regular displacements along the stacking
direction which are mismatched with the spin-Peierls driven
dimerization of the DTDA groups. This mismatch between
regular and irregular displacements appears to support dynamic
behavior along the stacking direction. Similar displacive
transitions have been reported within the family of dithiazolyls
such as TTTA,39 PDTA,40 and NCBDTA (Scheme 4).41
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.cgd.7b01065.
■
S
Experimental details of the preparation of the starting
nitriles, ROC6F4CN, and radicals 1−4; crystallographic
details of the data collection, processing, and refinement
for 1−4; DSC studies on 1−4; VT-PXRD studies on 3;
computations of the exchange coupling in dimers of 1−
4; details of the magnetic measurements on 2 (PDF)
Accession Codes
CCDC 1565879−1565882 and 1565884−1565888 contain the
supplementary crystallographic data for this paper (crystal
structures of 1−4 including variable temperature structural
studies on 2α (103−268 K)). These data can be obtained free
of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
Scheme 4. (a) 1,3,5-Trithia-2,4,6-triazapentalenyl (TTTA),
(b) 1,3,2-Dithiazolo[4,5-b]pyrazin-2-yl (PDTA), (c) 3-
Cyanobenzo-1,3,2-dithiazolyl (NCBDTA)
AUTHOR INFORMATION
Corresponding Author
*E-mail: jmrawson@uwindsor.ca.
■
ORCID
Robert J. Less: 0000-0002-5796-3110
Jeremy M. Rawson: 0000-0003-0480-5386
Present Address
In the case of TTTA and PDTA these transitions show
significant thermal hysteresis associated with the cleavage of
interstack contacts.42 Ab initio molecular dynamics simulations
on TTTA35 indicate fast intrastack pair-exchange dynamics in
which displacive motion of the radicals along the stacking
direction plays a key role in driving the phase transition.
Structural transitions in DTDA chemistry which resemble the
2β ↔ 2γ transformation are few.27a,43 The first order transitions
#(R.S.) Department of Chemistry and Chemical Biology,
Harvard University, 12 Oxford Street, Cambridge, MA 02138,
United States.
Author Contributions
Synthetic work was carried out by Y.B., R.S., and R.J.L.;
magnetic measurements were made by A.A. and J.C. Structural
studies and magnetic analysis by Y.B. and J.M.R. Manuscript
H
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
(17) Alberola, A.; Less, R. J.; Palacio, F.; Pask, C. M.; Rawson, J. M.
Molecules 2004, 9, 771.
(18) Alberola, A.; Less, R. J.; Pask, C. M.; Rawson, J. M.; Palacio, F.;
Oliete, P.; Paulsen, C.; Yamaguchi, A.; Farley, R. D.; Murphy, D. M.
Angew. Chem., Int. Ed. 2003, 42, 4782.
(19) Antorrena, G.; Davies, J. E.; Hartley, M.; Palacio, F.; Rawson, J.
M.; Smith, J. N. B.; Steiner, A. Chem. Commun. 1999, 1393.
(20) (a) Kyoichi, T.; Hideki, S. Patent No. JP 06145129, May 24,
1994. (b) Birchall, J. M.; Haszeldine, R. N.; Jones, M. E. J. Chem. Soc. C
1971, 1343.
(21) Robinson, S. W.; Haynes, D. A.; Rawson, J. M. CrystEngComm
2013, 15, 10205.
́
(22) Boere, R. T. CrystEngComm 2016, 18, 2748. (b) Mills, M. B.;
Hollingshead, A. G.; Maahs, A. C.; Soldatov, D. V.; Preuss, K. E.
CrystEngComm 2015, 17, 7816.
(23) (a) Allen, C.; Haynes, D. A.; Pask, C. M.; Rawson, J. M.
CrystEngComm 2009, 11, 2048. (b) Domagała, S.; Haynes, D. A.
CrystEngComm 2016, 18, 7116.
(24) (a) Banister, A. J.; Batsanov, A. S.; Dawe, O. G.; Herbertson, P.
L.; Howard, J. A. K.; Lynn, S.; May, I.; Smith, J. N. B.; Rawson, J. M.;
Rogers, T. E.; Tanner, B. K.; Antorrena, G.; Palacio, F. J. Chem. Soc.,
Dalton Trans. 1997, 2539. (c) Bell, A. M. T.; Smith, J. N. B.; Attfield, J.
P.; Rawson, J. M.; Shankland, K.; David, W. I. F. New J. Chem. 1999,
23, 565.
(25) (a) Constantinides, C. P.; Eisler, D. J.; Alberola, A.; Carter, E.;
Murphy, D. M.; Rawson, J. M. CrystEngComm 2014, 16, 7298.
(26) Constantinides, C. P.; Carter, E.; Eisler, D.; Beldjoudi, Y.;
Murphy, D. M.; Rawson, J. M. Cryst. Growth Des. 2017, 17, 3017.
(27) (a) Barclay, T. M.; Cordes, A. W.; George, N. A.; Haddon, R.
C.; Itkis, M. E.; Oakley, R. T. Chem. Commun. 1999, 2269.
(b) Alberola, A.; Carter, E.; Constantinides, C. P.; Eisler, D. J.;
Murphy, D. M.; Rawson, J. M. Chem. Commun. 2011, 47, 2532.
drafted by Y.B. and J.M.R. with additional contributions from
all authors. All authors have given approval to the final version
of the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We would like to thank the Canada Research Chairs Program
for financial support (J.M.R.), the University of Windsor for a
scholarship (Y.B.) and C.F.I./O.R.F. support for infrastructure.
We thank the Inorganic Chemistry Exchange (I.C.E.) Program
for the opportunity for R.S. to visit U. Windsor. We also
́
acknowledge support from the Ministerio de Economia y
Competividad of Spain (Grant No. MAT2015-68200-C2-2-P).
Additional support from Diputacion General de Aragon (DGA-
́ ́
M4) is also acknowledged. (J.C. and A.A.)
REFERENCES
■
(1) Bryan, C. D.; Fleming, R. M.; Glarum, S. H.; Haddon, R. C.;
Oakley, R. T.; Palstra, T. T. M.; Perel, A. S.; Schneemeyer, L. F.;
Waszczak, J. V.; Cordes, A. W.; George, N. A. Nature 1993, 365, 821.
(2) Thomson, R. I.; Pask, C. M.; Lloyd, G. O.; Mito, M.; Rawson, J.
M. Chem. - Eur. J. 2012, 18, 8629.
(3) Iwasaki, A.; Hu, L.; Suizu, R.; Nomura, K.; Yoshikawa, H.; Awaga,
K.; Noda, Y.; Kanai, K.; Ouchi, Y.; Seki, K.; Ito, H. Angew. Chem., Int.
Ed. 2009, 48, 4022.
(4) (a) Preuss, K. E. Dalton Trans. 2007, 2357. (b) Preuss, K. E.
Coord. Chem. Rev. 2015, 289, 45. (c) Lau, H. F.; Ng, V. W. L.; Koh, L.
L.; Tan, G. K.; Goh, L. Y.; Roemmele, T. L.; Seagrave, S. D.; Boere,
́
R.
T. Angew. Chem., Int. Ed. 2006, 45, 4498.
́
(28) Vegas, A.; Perez-Salazar, A.; Banister, A. J.; Hey, R. G. J. Chem.
Soc., Dalton Trans. 1980, 1812.
(29) Domagała, S.; Kosc, K.; Robinson, S. W.; Haynes, D. A.;
(5) (a) Banister, A. J.; May, I.; Rawson, J. M.; Smith, J. N. B. J.
Organomet. Chem. 1998, 550, 241. (b) Banister, A. J.; Gorrell, I. B.;
Clegg, W.; Jørgensen, K. A. J. Chem. Soc., Dalton Trans. 1989, 2229.
(6) (a) Beldjoudi, Y.; Osorio-Roman, I.; Nascimento, V.; Rawson, J.
M. J. Mater. Chem. C 2017, 5, 2794. (b) Beldjoudi, Y. Ph.D. Thesis,
University of Windsor, 2016.
(7) Rawson, J. M.; Alberola, A.; Whalley, A. J. Mater. Chem. 2006, 16,
2560.
(8) Bond, A. D.; Haynes, D. A.; Pask, C. M.; Rawson, J. M. J. Chem.
Soc., Dalton Trans. 2002, 2522.
Wozniak, K. Cryst. Growth Des. 2014, 14, 4834.
́
(30) (a) Feeder, N.; Less, R. J.; Rawson, J. M.; Oliete, P.; Palacio, F.
Chem. Commun. 2000, 2449. (b) Melen, R. L.; Less, R. J.; Pask, C. M.;
Rawson, J. M. Inorg. Chem. 2016, 55, 11747.
(31) (a) Burger, A.; Ramberger, R. Microchim. Acta 1979, 72, 273.
(b) Burger, A.; Ramberger, R. Microchim. Acta 1979, 72, 259.
(32) Shuvaev, K. V.; Decken, A.; Grein, F.; Abedin, T. S. M.;
Thompson, L. K.; Passmore, J. Dalton Trans. 2008, 4029.
(33) Bleaney, B.; Bowers, K. D. Proc. R. Soc. London, Ser. A 1952, 214,
451.
(9) (a) Haynes, D. A. CrystEngComm 2011, 13, 4793. (b) Boere,
́
R.
T.; Hill, N. D. D. CrystEngComm 2017, 19, 3698.
(10) (a) Brooks, W. V. F.; Burford, N.; Passmore, J.; Schriver, M. J.;
Sutcliffe, L. H. J. Chem. Soc., Chem. Commun. 1987, 0, 69. (b) Fairhurst,
S. A.; Johnson, K. M.; Sutcliffe, L. H.; Preston, K. F.; Banister, A. J.;
Hauptman, Z. V.; Passmore, J. J. Chem. Soc., Dalton Trans. 1986, 1465.
(34) Du, H.; Haddon, R. C.; Krossing, I.; Passmore, J.; Rawson, J. M.;
Schriver, M. J. Chem. Commun. 2002, 1836.
(35) Vela, S.; Mota, F.; Deumal, M.; Suizu, R.; Shuku, Y.; Mizuno, A.;
Awaga, K.; Shiga, M.; Novoa, J. J.; Ribas-Arino, J. Nat. Commun. 2014,
5, 4411.
(36) Luzon, J.; Campo, J.; Palacio, F.; McIntyre, G. J.; Rawson, J. M.
Polyhedron 2005, 24, 2579. (b) Deumal, M.; Rawson, J. M.; Goeta, A.
E.; Howard, J. A. K.; Copley, R. C.; Robb, M. A.; Novoa, J. J. Chem. -
Eur. J. 2010, 16, 2741.
(11) (a) Fatila, E. M.; Maahs, A. C.; Mills, M. B.; Rouzier
Soldatov, D. V.; Clerac, R.; Preuss, K. E. Chem. Commun. 2016, 52,
5414. (b) Fatila, E. M.; Mayo, R. A.; Rouzieres, M.; Jennings, M. C.;
Dechambenoit, P.; Soldatov, D. V.; Mathoniere, C.; Clerac, R.;
Coulon, C.; Preuss, K. E. Chem. Mater. 2015, 27, 4023.
̀
es, M.;
́
̀
̀
́
(12) Alberola, A.; Clarke, C. S.; Haynes, D. A.; Pascu, S. I.; Rawson, J.
(37) (a) Bryan, C. D.; Cordes, A. W.; Haddon, R. C.; Hicks, R. G.;
Kennepohl, D. K.; MacKinnon, C. D.; Oakley, R. T.; Palstra, T. T. M.;
Perel, A. S.; Scott, S. R.; Schneemeyer, L. F.; Waszczak, J. V. J. Am.
Chem. Soc. 1994, 116, 1205. (b) Cordes, A. W.; Bryan, C. D.; Davis,
W. M.; de Laat, R. H.; Glarum, S. H.; Goddard, J. D.; Haddon, R. C.;
Hicks, R. G.; Kennepohl, D. K.; Oakley, R. T.; Scott, S. R.; Westwood,
N. P. C. J. Am. Chem. Soc. 1993, 115, 7232.
(38) (a) Knapp, C.; Lork, E.; Gupta, K.; Mews, R. Z. Anorg. Allg.
Chem. 2005, 631, 1640. (b) Clarke, C. S.; Pascu, S. I.; Rawson, J. M.
CrystEngComm 2004, 6, 79.
M. Chem. Commun. 2005, 4726.
(13) Clarke, C. S.; Haynes, D. A.; Smith, J. N. B.; Batsanov, A. S.;
Howard, J. A. K.; Pascu, S. I.; Rawson, J. M. CrystEngComm 2010, 12,
172.
(14) Fatila, E. M.; Jennings, M. C.; Goodreid, J.; Preuss, K. E. Acta
Crystallogr., Sect. C: Cryst. Struct. Commun. 2010, 66, o260.
(15) Beldjoudi, Y.; Arauzo, A.; Palacio, F.; Pilkington, M.; Rawson, J.
M. J. Am. Chem. Soc. 2016, 138, 16779.
(16) (a) Banister, A. J.; Bricklebank, N.; Clegg, W.; Elsegood, M. R.
J.; Gregory, C. I.; Lavender, I.; Rawson, J. M.; Tanner, B. K. J. Chem.
Soc., Chem. Commun. 1995, 679. (b) Banister, A. J.; Bricklebank, N.;
Lavender, I.; Rawson, J. M.; Gregory, C. I.; Tanner, B. K.; Clegg, W.;
Elsegood, M. R. J.; Palacio, F. Angew. Chem., Int. Ed. Engl. 1996, 35,
2533.
(39) Fujita, W.; Awaga, K. Science 1999, 286, 261.
(40) Brusso, J. L.; Clements, O. P.; Haddon, R. C.; Itkis, M. E.;
Leitch, A. A.; Oakley, R. T.; Reed, R. W.; Richardson, J. F. J. Am. Chem.
Soc. 2004, 126, 14692.
I
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
(41) Alberola, A.; Collis, R. J.; Humphrey, S. M.; Less, R. J.; Rawson,
J. M. Inorg. Chem. 2006, 45, 1903.
(42) Clarke, C. S.; Jornet-Somoza, J.; Mota, F.; Novoa, J. J.; Deumal,
M. J. Am. Chem. Soc. 2010, 132, 17817.
(43) Suizu, R.; Iwasaki, A.; Shuku, Y.; Awaga, K. J. Mater. Chem. C
2015, 3, 7968.
J
DOI: 10.1021/acs.cgd.7b01065
Cryst. Growth Des. XXXX, XXX, XXX−XXX