Fine tuning the performance of multiorbital radical conductors by substituent effects

Article abstract of DOI:10.1021/jacs.6b11779

A critical feature of the electronic structure of oxobenzene-bridged bisdithiazolyl radicals 2 is the presence of a low-lying LUMO which, in the solid state, improves charge transport by providing additional degrees of freedom for electron transfer. The magnitude of this multiorbital effect can be fine-tuned by variations in the π-electron releasing/accepting nature of the basal ligand. Here we demonstrate that incorporation of a nitro group significantly stabilizes the LUMO, and hence lowers U_eff, the effective Coulombic barrier to charge transfer. The effect is echoed, at the molecular level, in the observed trend in E_cell, the electrochemical cell potential for 2 with R = F, H and NO₂. The crystal structures of the MeCN and EtCN solvates of 2 with R = NO₂ have been determined. In the EtCN solvate the radicals are dimerized, but in the MeCN solvate the radicals form superimposed and evenly spaced πstacked arrays. This highly 1D material displays Pauli-like temperature independent paramagnetic behavior, with χ_TIP = 6 × 10-⁴ emu mol^-1, but its charge transport behavior, with σ_RT near 0.04 S cm^-1 and E_act = 0.05 eV, is more consistent with a Mott insulating ground state. High pressure crystallographic measurements confirm uniform compression of the π-stacked architecture with no phase change apparent up to 8 GPa. High pressure conductivity measurements indicate that the charge gap between the Mott insulator and metallic states can be closed near 6 GPa. These results are discussed in the light of DFT band structure calculations.

Full text of DOI:10.1021/jacs.6b11779

Article
pubs.acs.org/JACS
Fine Tuning the Performance of Multiorbital Radical Conductors by
Substituent Effects
Aaron Mailman,† Joanne W. L. Wong,† Stephen M. Winter,† Robert C. M. Claridge,†
Craig M. Robertson,‡ Abdeljalil Assoud,† Wenjun Yong,§ Eden Steven,∥ Paul A. Dube,⊥ John S. Tse,#
Serge Desgreniers,∇ Richard A. Secco,§ and Richard T. Oakley*,†
†Department of Chemistry, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada
‡Department of Chemistry, University of Liverpool, Liverpool L69 7ZD, United Kingdom
§Department of Earth Sciences, University of Western Ontario, London, Ontario N6A 5B7, Canada
∥Department of Physics, Florida State University, Tallahassee, Florida 32310, United States
⊥Brockhouse Institute for Materials Research, McMaster University, Hamilton, Ontario L8S 4M1, Canada
#Department of Physics, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5E2, Canada
∇Department of Physics, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
S
* Supporting Information
ABSTRACT: A critical feature of the electronic structure of
oxobenzene-bridged bisdithiazolyl radicals 2 is the presence of
a low-lying LUMO which, in the solid state, improves charge
transport by providing additional degrees of freedom for
electron transfer. The magnitude of this multiorbital effect can
be fine-tuned by variations in the π-electron releasing/
accepting nature of the basal ligand. Here we demonstrate
that incorporation of a nitro group significantly stabilizes the
LUMO, and hence lowers Ueff, the effective Coulombic barrier
to charge transfer. The effect is echoed, at the molecular level,
in the observed trend in Ecell, the electrochemical cell potential for 2 with R = F, H and NO2. The crystal structures of the MeCN
and EtCN solvates of 2 with R = NO2 have been determined. In the EtCN solvate the radicals are dimerized, but in the MeCN
solvate the radicals form superimposed and evenly spaced π-stacked arrays. This highly 1D material displays Pauli-like
temperature independent paramagnetic behavior, with χTIP = 6 × 10−4 emu mol−1, but its charge transport behavior, with σRT
near 0.04 S cm−1 and Eact = 0.05 eV, is more consistent with a Mott insulating ground state. High pressure crystallographic
measurements confirm uniform compression of the π-stacked architecture with no phase change apparent up to 8 GPa. High
pressure conductivity measurements indicate that the charge gap between the Mott insulator and metallic states can be closed
near 6 GPa. These results are discussed in the light of DFT band structure calculations.
However, for most organic π-radicals the reverse condition (W
< U) applies, and they possess Mott insulating ground states.5
Haddon’s proposed solution, which has inspired and guided
research on neutral radical conductors ever since, was to
employ odd-alternant hydrocarbons based on the phenalenyl
(PLY) framework, in which the highly delocalized, nonbonding
SOMO promotes a low U-value. Following his lead the last 40
years has witnessed extensive research into the structures and
properties of PLY derivatives,6 including the exploration of
substituent effects,7 complex polycyclic systems,8 mixed-valence
spiro-conjugated PLYs,9,10 and modes of association.11
However, although materials constructed from a purely
carbon-based framework displaying high conductivity have
been generated,12 metallic behavior has remained elusive.
INTRODUCTION
■
The idea that the unpaired electron in an organic radical might
serve as a carrier of charge, much like the valence electron in an
elemental metal like sodium, goes back over a century,1 but it
was not until the mid-1970s that Robert Haddon proposed
design criteria for a functional material based on the neutral
radical conductor concept.2 Recognizing the inherent weakness
of intermolecular interactions between organic molecules in the
solid state, Haddon concluded that charge transport required
reduction of the potential energy barrier to charge transfer (U)
to the point that the low kinetic energy of the unpaired
electrons, expressed in terms of the hopping integral (t)
between singly occupied molecular orbitals (SOMOs) on
neighboring radicals, would be sufficient to allow charge
migration. With respect to a hypothetical 1D-array of radicals,
these ideas imply formation of a metallic state with a half-filled
(f = 1/2) energy band when the bandwidth W ∼ 4|t| > U.3,4
Received: November 14, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
An alternative approach to organic radical-based conductors
involves the use of heavy heteroatoms.13,14 As in charge-transfer
conductors, where the presence of sulfur and/or its heavier
congener selenium imparts both softness (a lower U) and
increased orbital overlap (a larger t and W) to donors such as
tetrathiafulvalene,15 so too does their incorporation into the
framework of organic radicals. This greater overlap comes at a
cost, however, as it also increases the tendency for the radicals
to dimerize, with consequent loss of potential charge carriers.
Thus, in the absence of steric blockage, early generations of S/
Se-based heterocyclic radicals were strongly associated in the
solid state,16 and displayed insulating or weakly semiconducting
behavior.17 Improvements in conductivity could be induced by
p-type doping,18 which lowered U by changing the degree of
band filling, but the challenge to reduce U and to prevent
dimerization in purely neutral systems required greater spin
delocalization, as per Haddon’s original prescription.
Charge transport in oxobenzene-bridged radicals 2 is
improved relative to 1 by so-called multiorbital effects, which
arise from the interaction of the LUMO of the carbonyl group
with the π-system of the radical, and consequent introduction
of a low-lying π-LUMO.27 The presence of this extra orbital
increases the electronic and magnetic degrees of freedom
available for transfer of the unpaired electron, by reducing U to
U′ or U″, the latter two being equal to U − V + Δε K, where
Δε is the SOMO-LUMO gap, V represents the repulsion
between electrons in different orbitals, K is the electron
exchange term, and the sign of K reflects formation of a triplet
(−ve) or open shell singlet (+ve) anion. Density Functional
Theory (DFT) calculations suggest that U′ < U″ < U, that is,
the triplet anion is energetically favored, a feature which
enhances ferromagnetic exchange interactions and promotes
magnetic ordering at low temperatures.28 With respect to
charge transport, the effect of the low-lying LUMO is to reduce
U to U′ or U″, and thereby destabilize the Mott insulating state.
The resulting decrease in charge gap ΔC allows for higher
conductivities and lower thermal activation energies Eact than 1.
Further reduction and eventual closure of the charge gap ΔC
can be achieved by the application of relatively mild pressure,
with metallization for R = H, F, Ph observed at pressures
ranging from 3 to 12 GPa.29 In the case of R = F, Fermi liquid
behavior has been identified near 6 GPa.29 To a large extent the
variations in the solid state properties within this family of
radicals can be attributed to differences in electronic
dimensionality. For example, when R = H, the radicals are
aligned in relatively 1D slipped π-stack arrays,21e while for R =
F, the radicals adopt a classic brick wall architecture,21d and the
resulting 2D electronic structure29 is more conducive to
stabilization of a metallic state.
In addition to their role in influencing crystal packing, the
exocyclic ligand R in oxobenzene-bridged radicals 2 can also
play an important role in determining the magnitude of U. In
N-alkylpyridine-bridged radicals 1 this is not the case; U is
largely independent of nature of the ligands R1/R2, as the
SOMO is nodal at the sites of substitution.19 While the same
situation also applies in oxobenzene-bridged radicals 2, that is,
the SOMO is nodal at the sites of substitution, the LUMO is
not;27 to a first approximation its orbital energy ε1 should
therefore be raised (or lowered) depending on the π-electron
releasing (or accepting) power of the basal R-group. The
implication of this idea is that the SOMO-LUMO separation
Δε and hence the values of U′ and U″ should be tunable.
Notwithstanding the vagaries of crystal packing and resulting
bandwidth issues, improvements in performance at the solid
state level in oxobenzene-bridged radical 2 should therefore be
attainable simply by molecular modification, that is, by
judicious choice of R-group. Herein we demonstrate the
validity and utility of this concept by examining the molecular
and solid state transport properties of the nitro-substituted
radical 2a (2, R = NO2). The presence of this strong π-electron
withdrawing NO2 group leads to a reduction in the magnitude
of U′ (U″) relative to U, and the generation of a material with a
near metallic ground state.
Figure 1. Resonance-stabilized bisdithiazolyl radicals 1 and 2 (R1 = R2
= R = H), showing the Kohn−Sham frontier orbitals and Coulombic
barriers to intersite charge transfer along a chain of radicals, to
generate anions with closed shell singlet (U), triplet (U′) and open
shell singlet (U″) states. The latter two (for 2) are defined in terms of
the SOMO-LUMO energy separation Δε, the electron repulsion V
between electrons in different orbitals on the same site, and the
electron exchange term K.
Resonance stabilized heterocyclic radicals, in which spin
density is equally partitioned between heterocycles, were
developed to address the delocalization challenge.19 For
example, in the two bisdithiazolyls 1 and 2 shown in Figure
1, the radical SOMO and hence unpaired spin density spans the
entire molecule, regardless of the nature of the bridging unit, be
it an N-alkylpyridine ring in 120 or an oxobenzene moiety in
2.21 Initial work on derivatives of 1 confirmed that their
conductivity was improved (the Mott state was destabilized) by
virtue of a lowered Coulombic barrier (U) to charge transfer,
expressed in terms of the difference between the ionization
energy (IP) and electron affinity (EA) of the radical. At the
same time spin delocalization helped suppress pancake22
dimerization. While they were still Mott insulators, with charge
gaps ΔC = U − W between the Mott insulating and metal states
near 0.5 eV, replacement of sulfur by its heavier congener
selenium reduced the value of ΔC substantially,23 to the point
that “bad metal” states could be accessed under relatively mild
applied pressure.24 The selenium-based variants also displayed
strong isotropic and anisotropic magnetic exchange inter-
actions, which gave rise to magnetically ordered phases with
relatively high ordering temperatures and coercive fields.25,26
RESULTS AND DISCUSSION
■
Synthesis. Preparation of the nitro-substituted oxobenzene-
bridged bisdithiazolyl 2a (Scheme 1) begins with the stepwise
reduction of picric acid 3, first using ammonium polysulfide to
generate 2-amino-4,6-dinitrophenol 4,30 and then L-ascorbic
acid to effect the second reduction to 2,6-diamino-4-nitro-
B
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
oxobenzene-bridged related radicals, notably 2b,c (2, R = H,
F),21d,e and are consistent with estimates obtained from DFT
calculations (Table 1). Taken collectively, the results indicate
little dependence of the value of the spin distribution on the
nature of the basal ligand, a conclusion consistent with the fact
that the radical SOMO is nodal at the site of substitution.
Scheme 1
Table 1. Hyperfine Coupling Constants, Electrochemical
Potentials and Ion Energetics
radical
2a, R = NO2
2b, R = H
2c, R = F
a
aN, mT
0.331 (0.234)
0.032 (0.045)
−0.218
0.227
0.354 (0.277)
0.203 (0.372)
0.333 (0.277)
0.674 (0.837)
a
aR, mT
E1/2(−1/0), V
b
c
c
−0.525
−0.479
phenol 5.31 Finally, double Herz cyclocondensation32 of 5 with
sulfur monochloride in acetonitrile affords the desired
oxobenzene-bridged bisdithiazolylium framework in the form
of the black, insoluble chloride salt [2a][Cl]. Metathesis of this
crude material with silver triflate (AgOTf) in hot acetonitrile
affords a deep red solution (λmax = 505 nm) of the
corresponding triflate salt [2a][OTf], which crystallizes upon
cooling as lustrous green shards.
b
E1/2(0/+1), V
0.035
0.208
d
d
E
cell, V
0.45
0.56
0.69
e
IP, eV
4.852
4.610
3.376
3.787
3.569
1.234
0.822
1.041
4.784
3.542
3.825
3.658
1.242
0.959
e
EA, eV
3.632
e
EA′, eV
4.160
e
EA″, eV
4.045
e
U (eV)
1.220
e
U′, eV
U″, eV
0.692
Chemical reduction of [2a][OTf] to the neutral radical can
be effected using octamethylferrocene (OMFc) in acetonitrile
(MeCN) or propionitrile (EtCN) solution. As with other
oxobenzene-bridged bisdithiazolyls,21c,f there is a marked
tendency for 2a to crystallize as a solvate. Thus, when
acetonitrile is employed, black needles of the 1:1 solvate 2a·
MeCN are produced, while with propionitrile, black plates of
the 4:1 solvate, that is, 2a·1/4EtCN, are formed. The structural
identity of both materials has been established by single crystal
X-ray diffraction, but the formulations are also supported by
elemental analysis and, in the case of 2a·MeCN, by
thermogravimetric analysis (see Figure S1), which shows a
rapid weight reduction starting near 70 °C corresponding to the
loss of one MeCN molecule per radical unit. Differential
scanning calorimetry show a single phase transition at the same
temperature range.
e
0.806
1.126
a
Observed hyperfine coupling constants aN and aR (R = 14NO2, H
1
and 19F), with calculated (U)B3LYP/6-311G(d,p) values in
b
c
parentheses. Volts versus SCE. Reduction to the anion is irreversible,
Epc values are cited. Ecell estimated from difference in Epc values.
(U)B3LYP/6-311G(d,p)/PCM calculated values, with MeCN as
d
e
solvent. Values of EA, EA′ and EA″ refer respectively to formation of
the closed shell singlet, triplet and open shell singlet anion (see Figure
1); U, U′ and U″ follow suit. Data for R = H and F are from refs 21d, f,
and 27.
In contrast to the EPR data, which reflect a similar ground
state structure, the three radicals 2a,b,c display significant
differences in their electrochemical properties. While the
cationic and anionic states can be easily generated electro-
33
chemically for all three radicals, the cell potentials Ecell are
markedly dependent on the nature of the R-group. For
example, cyclic voltammetry on a solution of [2a][OTf] in
MeCN (Figure 2b) indicates two reversible waves, the +1/0
couple with E1/2 = +0.227 V vs SCE and the −1/0 couple with
E
1/2 = −0.218 V vs SCE. The resulting value for Ecell = 0.45 V is
the lowest ever observed for a radical of this type (Table 1);
those for 2b and 2c are some 100 and 200 mV higher,
respectively.21d,f Such variations are not observed in the Ecell
values of N-alkyl-bridged bisdithiazolyls 1, which are uniformly
higher and independent of the nature of the R2-group.19
The reason for the marked dependence of Ecell of 2 on the
nature of the R-group can be understood by considering the
different anionic states accessible upon reduction. To
demonstrate this point, we have carried out DFT calculations
at the (U)B3LYP/6-311G(d,p)/PCM level to determine IP,
EA, EA′ and EA″ values of 2a,b,c, and hence the respective
(solvent screened) Coulombic barriers U, U′ and U″. The
results, summarized in Table 1, indicate that the value of U,
which corresponds electrochemically to formation of the closed
shell singlet anion, or a SOMO-to-SOMO electron transfer in
the solid state, remains essentially constant across the series. By
contrast reduction to a triplet or open shell singlet anion (EA′
and EA″ respectively), corresponding to a SOMO-to-LUMO
transfer in the solid state, becomes increasingly easier along the
series R = F < H < NO2, in keeping with the expected changes
Figure 2. (a) X-Band EPR spectrum of 2a in toluene at ambient
temperature. (b) CV scan of [2a][OTf] in MeCN, with Pt electrodes,
0.1 M n-Bu4NPF6 electrolyte, scan rate 100 mV s−1.
EPR Spectroscopy, Electrochemistry and Ion Ener-
getics. The electronic structure of 2a has been probed at the
molecular level by EPR spectroscopy, cyclic voltammetry and
computational methods. The X-band EPR spectrum (g =
2.0097) in toluene (Figure 2a) displays the expected pentet fine
structure pattern arising from hyperfine coupling (aN = 0.331
mT) to the two equivalent 14N nuclei (I = 1) within the
heterocyclic framework; weaker, unresolved coupling (aN
=
0.032 mT) to the nitro group can be extracted by spectral
simulation. These values are similar to those seen in related
C
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
in the SOMO-LUMO gap Δε occasioned by the π-electron
donating/releasing power of the R-group. At the DFT level, the
triplet anion is preferred over the open shell singlet for all three
radicals, so that U′ < U″, but regardless of this ordering, the
observed trends in the values of Ecell of 2a,b,c can be viewed as
a confirmation of the involvement of the LUMO of the radical
in defining the nature and accessibility of the anionic state, and
hence the value of Ecell, a feature not possible for 1. Taken
together, the electrochemical and computational results support
the idea that, in contrast to 1, the solid state Coulombic barrier
in 2 is not only small, but tunable; the value of Ueff (the
composite of U, U′ and U″) is increased relative to R = H by π-
donor ligands (R = F) and decreased by π-acceptor ligands (R
= NO2). On this basis, it can be reasoned that 2a represents a
promising building block for a neutral radical-based metal.
Table 2. Crystal Data
2a·MeCN
2a·1/4EtCN
formula
fw
C6N3O3S4·C2H3N
331.38
4(C6N3O3S4)·C3H5N
1216.40
a, Å
3.4719(3)
30.764(4)
11.0663(13)
90.2099(7)
1182.0(2)
1.862
25.0506(2)
23.9588(2)
14.9339(2)
117.7784(5)
7930.13(14)
2.038
b, Å
c, Å
β, deg
V, Å3
ρ(calcd), g cm−3
space group
Z
Figure 3. (a) Unit cell drawing of 2a·MeCN, viewed parallel to the a-
axis. (b) Out-of-register slipped π-stacking of radicals. (c) Ribbon-like
arrays of radicals running parallel to the c-axis, and intermolecular
contacts.
P21/n
C2/c
4
8
temp, K
μ, mm−1
λ, Å
100(2)
296(2)
0.811
0.955
0.71073
0.71073
solution method
R, Rw (on F2)
direct methods
0.0360, 0.0757
direct methods
0.0430, 0.0731
Ambient Pressure Crystallography. As described above,
crystallization of 2a by in situ reduction of the triflate salt
[2a][OTf] in MeCN and EtCN, affords the solvates 2a·MeCN
and 2a·1/4EtCN respectively. The crystal structures of these
two materials have been determined by single crystal
diffraction; crystal data are provided in Table 2. Black needles
of 2a·MeCN belong to the monoclinic space group P21/n, and
consist of slipped π-stacks of evenly spaced radicals running
parallel to the a-axis (Figure 3); there is no indication of a
charge density wave (CDW) or Peierls distortion.34 The
associated tilt angle (τ = 76.4°) with respect to the stacking
direction and mean-plane separation (δ = 3.374 Å) of
neighboring radicals are consistent with nearly direct super-
position of radicals along the stacking direction. Each radical is
coordinated laterally to an acetonitrile molecule by a pair of
short (inside the van der Waals separation)35 S···N′ contacts, in
a fashion similar to that found in 2·MeCN (R = Cl, Br, I)21c,f
and 2·EtCN (R = I).21f There is also a network of close
intermolecular S···O′ and S···N′ interactions that connect
adjacent radicals along the n-glides to produce a ruffled ribbon
arrangement running parallel to the c-axis. This pattern is
reminiscent of the packing observed in 2b.21e Intramolecular
bond lengths and angles are nominal.
Figure 4. (a) Unit cell of 2a·1/4EtCN, and pinwheel arrangement of
the asymmetric unit in the ab plane. (b) Side-on views of the unit cell
of 2a·1/4EtCN, showing 4-decker repeat unit, and association of
individual radicals into [AC], [B]2 and [D]2 dimers. (c) Local views of
dimers, with intermolecular S−S distances.
The crystal structure of 2a·1/4EtCN belongs to the
monoclinic space group C2/c, and with Z = 8 there are four
independent molecules (A−D) in the asymmetric unit, which
consists of an approximately planar pinwheel arrangement of
four radicals, as shown in Figure 4a; the nitrogen atom of the
single EtCN solvent molecule is located at the center of this
pinwheel. Within each pinwheel there is a spider-web like
D
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
network of close (inside van der Waals separation)35 S···O′ and
S···N′ interactions (listed in Figure S3), which creates a slightly
ruffled layer of radicals lying parallel to the ab plane. There are
four such layers stacked along the c-direction in the unit cell
(Figure 4b), with nearest neighbor layers related by 2-fold
rotations, and second nearest layers related by c-glides. Radicals
in different layers within this 4-decker structure are linked into
three unique dimer pairs (Figure 4c) via one or more close S−S
bonds, with radicals A and C overlapping to form [AC]
heterodimers, while radicals B and D associate as [B]2 and [D]2
homodimers.
Figure 6. Plots of σ vs 1000/T for radicals 2a,b,c (R = NO2, H, F),
from measurements using 4-probe technique on pressed pellet
samples, with derived thermal activation energies Eact. For 2a·
MeCN, results from 2-probe single crystal measurements (A) are
also provided. Data for 2b,c are from refs 21d, e.
significant deterioration of the surface of the crystals, possibly
as a result of desolvation (loss of MeCN). We were, however,
able to perform 2-probe single crystal measurements, and these
results, along with those obtained using the 4-probe technique
on pressed pellets, are shown in Figure 6. By way of
comparison, the corresponding data (4-probe, pressed pellet)
for 2b,c are illustrated; thermal activation energies Eact are also
provided. Overall, the results indicate a high, albeit activated
conductivity for 2a·MeCN. Indeed its performance is the best
seen to date for a heterocyclic radical conductor, with a room
temperature conductivity σRT slightly higher and Eact slightly
lower than that of 2c (the previous best).21d The dimerized
solvate 2a·1/4EtCN also shows good charge transport behavior
for a closed shell semiconductor.
Figure 5. Plots of χ vs T and χT vs T (inset) at a field H = 1000 Oe for
(a) on 2a·MeCN and (b) 2a·1/4EtCN. A nonlinear fit (in red) to the
data for 2a·MeCN affords a Curie defect concentration of 1.7% and
χTIP = 6 × 10−4 emu mol−1.
Magnetic Susceptibility Measurements. The results of
variable temperature magnetic susceptibility (χ) measurements
on 2a·MeCN and 2a·1/4EtCN, performed over the range 2−
300 K at a field H = 1000 Oe, are presented in Figure 5 in the
form of plots of χ and χT (corrected for diamagnetic
contributions) versus T. In the case of 2a·MeCN the magnetic
profile shows no indication of a phase transition to a Peierls
distorted (CDW) or to a magnetically ordered state. Instead,
the material displays weak temperature independent para-
magnetism (TIP) over the entire temperature range, with a
small Curie tail apparent at low temperature that may arise
from defects. From a nonlinear Curie fit the concentration of
these defects is estimated to be near 2%. The derived value for
χTIP = 6 × 10−4 emu mol−1 is suggestive of Pauli paramagnet
behavior, with a magnitude similar to that found in organic
charge transfer salts,36 spiro-bisphenalenyl12 and metal-bridged
bis-dithiolate radicals37 possessing a near metallic state.
In accord with the presence of four weakly associated dimers
in its crystal structure, the magnetic profile for 2a·1/4EtCN
points to essentially diamagnetic behavior, with a weak Curie
tail at low temperature arising from defects, but otherwise a
near zero value of χ up to 200 K, and a very slight increase
thereafter (more readily seen in the χT versus T plot). The
latter feature probably reflects a gradual, thermally driven
increase in the degree of dimer dissociation.
Figure 7. Single crystal reflectivity of 2a·MeCN at ambient pressure,
measured using incident radiation polarized parallel (∥) and
perpendicular (⊥) to the a-axis (the needle direction).
While we experienced difficulties obtaining single crystal
conductivity data on 2a·MeCN, measurements of polarized
infrared reflectance spectra on single crystals proved more
successful. Figure 7 shows the reflectivity obtained from the flat
face of a needle, with incident radiation polarized parallel (∥)
and perpendicular (⊥) to the needle direction, which
corresponds to the π-stacking a-axis. With light polarized ∥
to the a-axis the response in reflectivity is Drude-like, with a
pronounced minimum in reflectivity near 1 eV, which may be
assigned to a plasma edge, followed by a rapid increase. By
contrast, with polarization ⊥ to the a-axis, reflectivity is weak
and featureless. Overall this response is consistent with the
behavior of 1D metallic charge transfer salts.38
Conductivity Measurements. We have performed 4-
probe, variable temperature conductivity measurements on
pressed pellet samples of both 2a·MeCN and 2a·1/4EtCN.
Given the relatively large size and needle-like morphology of
crystals of 2a·MeCN, we also attempted measurements on
single crystals. However, not only were the needles extremely
fragile, but attachment of wires with silver paint caused
High Pressure Structural and Conductivity Measure-
ments. Given the impressive ambient pressure transport
property data on 2a·MeCN, that is, the Pauli paramagnet-like
E
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
temperature dependence of its magnetic susceptibility, and its
high conductivity coupled with a small activation energy, we
wished to explore the effect of applied pressure on its
performance. At what pressure might a metallic state be
observed? As a first step in this direction, however, we needed
to establish the effect of pressure on the crystal structure itself.
Given the proclivity of the material to lose solvent by heating to
just 70 °C, we were concerned that pressure might also lead to
a loss of solvent, and hence deterioration in crystallinity and
conductivity.
High-pressure (HP) crystallographic studies were therefore
performed on 2a·MeCN, using synchrotron radiation and
diamond anvil cell (DAC) techniques. Powder diffraction data
were collected at room temperature as a function of increasing
pressure up to 8 GPa. The retention of resolution and the
smooth and steady evolution of the positions of the diffraction
peaks with increasing pressure indicated that neither sample
degradation (solvent loss) nor phase change was occurring over
this pressure range. A series of data sets so obtained were
indexed, and the structures solved and refined in DASH using a
rigid body molecular model of the radical taken from the single
crystal structure solution. Initial attempts to include the solvent
molecule in the model were eventually abandoned, as its
presence did not affect the outcome of the refinement. During
the final Rietveld refinements (in GSAS) only the unit cell
parameters were optimized; a sampling of the crystal data at
selected pressures is provided in Table S1. Variations in the
unit cell dimensions as a function of pressure are shown in
Figure 8a. All three axes contract, the most significant response
being parallel to the stacking direction (the a-axis), as found
previously for 2b and 2c,29 an observation which reflects the
relative ease of compression of the layered π-stacks.
rapid rise in conductivity up to ∼5 GPa, then a more gradual
increase thereafter, with σRT reaching 101 S cm−1 at 8 GPa.
Although this limiting value lies well below the Mott−Ioffe−
Regel limit for a metallic state,39 it is similar in magnitude to
that observed upon compression of related radicals, and may be
ascribed to the formation of a highly correlated or “bad metal”
state. Upon decompression, the high pressure limiting value of
σRT is maintained down to near 1 GPa, whereupon it rapidly
drops to near 10−2 S cm−1, that is, close to the value obtained
from the ambient pressure 4-probe measurements. While this
delay in recovery may stem from clamping of the sample, it may
also reflect the elimination of grain boundaries during
compression, so that the “bad metal” state is in fact preserved
down to near 1 GPa. The results of separate measurements of
the thermal activation energy Eact, performed over the
temperature range −50 °C to +20 °C (in order to avoid
solvent loss), were consistent with the conductivity measure-
ments. Thus, as illustrated in the inset to Figure 8b, Eact steadily
decreases with increasing pressure, and is essentially eliminated
near 6 GPa, heralding closure of the charge gap ΔC and
formation of bad metal state.
Figure 9. (a) Frontier Kohn−Sham π-MOs of 2a and 2b, and energy
level sketch (not to scale) illustrating the orbital energy changes
occasioned by the incorporation of the NO2 group. (b) Dispersion of
CO energy bands arising from a single π-stacked array of 2a.
Band Structure Calculations. The relatively high
conductivity and diamagnetic behavior of 2a·1/4EtCN militate
in favor of a small band gap semiconductor description, but
with four molecules in the asymmetric unit, its unit cell is too
large and the resulting solid state electronic structure too
complex to allow a detailed theoretical examination. By
contrast, the 1D π-stacked architecture of 2a·MeCN is more
amenable to both qualitative and quantitative analysis. To begin
such an exercise, it is useful to identify the relevant frontier
molecular orbitals of the building block 2a, which are easily
derived from those of the prototypal radical 2b (Figure 9a).
While attachment of a nitro group does not, to first order,
perturb the SOMO (ϕ0, energy ε0), the LUMO in 2a is split
into a new LUMO (ϕ1, energy ε1), which is lowered in energy
relative to 2b and based largely on the heterocycle, and a higher
lying LUMO+1 level (ϕ2, energy ε2), which is more localized
on the nitro group. The idealized band structure arising from a
perfectly superimposed 1D π-stacked array of 2a can then be
easily anticipated. The three MOs ϕ0−2 will generate three
largely unhybridized crystal orbitals (COs), whose individual
bandwidths will reflect the relative magnitude of the respective
Figure 8. (a) Contraction in unit cell dimensions of 2a·MeCN over
the pressure range 0−8 GPa. (b) Pressure dependence of the room
temperature conductivity σRT of 2a·MeCN during compression (blue)
and decompression (green). The inset shows the pressure dependence
of the thermal activation energy Eact, measured over the range −50 °C
to +20 °C.
Given the well-behaved structural response of 2a·MeCN to
physical pressure, we carried out high pressure conductivity
measurements using multianvil press techniques. Data collected
at room temperature in both compression and decompression
modes over the pressure range 0−8 GPa are shown in Figure
8b. We attribute the initially low value of σRT observed during
compression to grain boundary effects arising from incomplete
compaction of the sample. With increasing pressure there is a
F
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
π-stack hopping integrals t0−2. Because ϕ0 and ϕ1 are both
heterocycle-centered, the values of t0,1 should be comparable,
and both are expected to be numerically larger than t2, which
arises from interactions between orbitals that are primarily
ligand-based.
SUMMARY AND CONCLUSIONS
■
Historically, achieving high conductivity in purely organic
radical-based molecular materials has been impeded by the
energetic imbalance between the high potential energy cost (U)
of site-to-site transfer of an unpaired electron and the limited
kinetic stabilization (W) afforded by charge delocalization into
a half-filled energy band. To break out of the Mott insulating
state that this condition imposes, a variety of approaches have
been pursued. The use of spin delocalization to lower U, as
Haddon originally proposed,2 and the incorporation of heavy
heteroatoms (chemical pressure) to increase W,13,14 have both
been extensively explored. Performance can also be improved
by relaxing the classical one-orbital, one-electron model, as in
mixed-valence spiroconjugated PLYs9,10 and doped radical ion
salts,18 in which the effective Coulombic barrier is reduced as
additional channels for charge transport become available.
In the case of oxobenzene-bridged bisdithiazolyls 2 improved
conductivity arises from a multiorbital effect, that is, the
presence of a low-lying LUMO (Figure 1) which destabilizes
the Mott state by generating a lower effective Coulombic
barrier Ueff.27 In some cases, such as 2c, the 2D electronic
structure allows for mixing (hybridization) of the SOMO and
the LUMO, which leads to an enhanced kinetic stabilization Ek
of the metallic state occasioned by a lowering the Fermi level
To quantify this description we have performed single point
DFT band structure calculations with the Quantum Espresso
package, using atomic coordinates taken from crystal structures
determined at ambient and elevated pressure; to simplify the
analysis the solvent molecules were not included. The results of
the 0 GPa calculation, based on a nonmagnetic state, are shown
in Figure 10a in the form of a CO band dispersion diagram over
several reciprocal space directions within the first Brillouin
zone. The pattern of bands near the Fermi surface follows
directly from frontier orbital and CO ordering described above
(Figure 9), with three bands each comprised of four COs
arising from the SOMO (ϕ0), LUMO (ϕ1) and LUMO+1
(ϕ2). It is immediately evident that the electronic structure is
very 1D, as expected from the crystal structure and the optical
response, with strong band dispersion of the ϕ0 and ϕ1 bands
(∼1 eV) from Γ → X, loosely speaking the π-stacking direction,
and virtually none along Γ → Y, Γ → Z vectors, that is,
corresponding to interactions transverse to the stacking axis.
The nonintersecting appearance of these two bands is in
marked contrast to the band structure of 2c, where the 2D
packing of radicals gives rise to strong mixing (hybridization) of
relative to that of the Mott state.28 29 Together or separately,
a,
these two effects contribute to reducing the charge gap ΔC =
Ueff − Ek. In this respect these materials provide an interesting
parallel with neutral, nominally radical (S = 1/2) Au(III)
bisdithiolate complexes, where a high-lying HOMO may play a
similar role.37,40
the COs.28 Given the position of the Fermi level it is
a
nonetheless clear that both ϕ0 and ϕ1 bands are both partially
occupied, giving rise to a less than 1/2-filled band description,
although the involvement of the latter is significantly less. As
expected the ϕ2 band is relatively narrow, and separated from
the lower-lying ϕ0 and ϕ1 bands. The effect of pressure on the
electronic structure of 2a is illustrated in Figure 10b, which
shows the band structure at 4.4 GPa. It is immediately apparent
that the 1D nature of the material is preserved under pressure;
band dispersion from Γ → X increases slightly with
compression along the stacking axis, but there remains little
or no dispersion along Γ → Y or Γ → Z.
In addition to structural modifications, which can reduce the
charge gap ΔC by increasing electronic dimensionality and
hence Ek, the performance of multiorbital radicals 2 can, in
principle, be improved through a lowering of Ueff. In this work
we have demonstrated that significant changes in Ecell, the
electrochemical signature of Ueff, can be induced by substituent
effects, that is, by attachment of a strongly electron withdrawing
nitro-group (R = NO2). However, while the ion energetics of
2a suggest that it should serve as an excellent building block for
a radical-based metal, our attempts to incorporate it into a
crystalline lattice have yielded the two solvates 2a·MeCN and
2a·1/4EtCN, neither of which represents, from a structural
perspective, an ideal candidate for a radical-based metal. In the
1/4 EtCN solvate the radicals are dimerized, and while they are
not dimerized in the MeCN solvate, the almost directly
superimposed, 1D π-stack packing pattern leads to a highly 1D
electronic structure.
The transport properties of 2a·MeCN are thus characteristic
of the highly correlated “inter-regnum” region found between
Mott insulating and metallic states, in which there is a
separation of the spin and charge behavior of the unpaired
electrons.41 While the Pauli-like magnetic properties of this
material are consistent with a metallic state, pressed pellet and
single crystal measurements indicate thermally activated
conductivity, with a small charge gap 0.05−0.10 eV which
can be closed by the application of 6 GPa pressure. Polarized
single crystal reflectance measurements confirm strong
electronic anisotropy, a result supported by DFT band
structure calculations, which support the highly 1D electronic
structure description.
Figure 10. DFT computed band structure of 2a·MeCN at (a) 0 GPa
and (b) 4.4 GPa, showing dispersion of bands based on unhybridized
COs arising from the SOMO (ϕ0) in blue, the LUMO (ϕ1) in green
and LUMO+1 (ϕ2) in black. The Fermi level is shown with a dashed
red line.
In summary, it is perhaps ironic that the structure of fluoro-
substituted radical 2c combines a somewhat larger Ecell with a
G
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
mmol) in 400 mL of 1.0 M NaOH was heated at 70 °C for 5 h. The
deep red mixture was hot-filtered on a large Buchner funnel and
cooled to 5 °C on an ice−water bath. Glacial acetic acid was added
dropwise to the rapidly stirred, cold mixture, until pH ∼ 5 was
obtained. After 2 h a light brown solid was collected by filtration and
air-dried. Recrystallization from hot H2O afforded dark orange needles.
Yield: 2.46 g (14.5 mmol, 58% yield); mp 168−169 °C. ESI MS (1:1
MeOH/H2O + 0.1% formic acid) m/z = 170.08 (M+H+). This
material was dried at 100 °C in vacuo. Anal. Calcd for C6H7N3O3: C,
42.61; H, 4.17; N, 24.84. Found: C, 43.00; H, 4.37; N, 24.66. 1H NMR
(DMSO-d6) δ: 6.883 (s, C−H), 5.890 (broad, N−H). IR (cm−1):
3489 (m), 3471 (m), 3250 (br, s), 3121 (w), 3105 (w), 3077 (w),
1783 (w), 1635 (m), 1615 (m), 1595 (m), 1554 (s), 1516 (s), 1444
(m), 1401 (m), 1338 (vs), 1304 (s), 1237 (s), 1143 (m), 1112 (m),
1066 (m), 995 (m), 935 (m), 894 (w), 880 (w), 856 (vw), 817 (w),
806 (w), 771 (w), 736 (m), 712 (m), 676 (s), 590 (vw), 491 (w).
Preparation of [2a][Cl]. A solution of sulfur monochloride (4.0
mL, 6.8 g, 50.1 mmol) in 25 mL MeCN was added dropwise to a
slurry of 2,6-diamino-4-nitrophenol (1.70 g, 10.0 mmol) in 200 mL
MeCN and the mixture was heated under gentle reflux for 6 h. The
black precipitate of crude [2a][Cl] was filtered off, washed with
MeCN, hot DCE, CS2 and DCM, then dried in vacuo, yield 2.29 g
(7.03 mmol, 70%); mp >250 °C. IR (cm−1): 1710 (s), 1626 (br, s),
1503 (vs), 1343 (m), 1291 (sh, m), 1277 (s), 1099 (vw), 1074 (w),
1017 (vw), 949 (w), 863 (w), 852 (w), 825 (w), 762 (m), 746 (w),
632 (vw), 613 (vw), 602 (vw), 568 (vw), 493 (m), 465 (w).
highly developed 2D-electronic structure, in essence the reverse
combination of potential and kinetic energy terms found in 2a·
MeCN. The challenge for the future will be to design materials
which enjoy simultaneous optimization of both terms, that is, to
find the right building block and the right crystal structure.
EXPERIMENTAL SECTION
■
General Methods and Procedures. The reagents picric acid 3,
sulfur monochloride, silver triflate (trifluoromethanesulfonate),
ammonium sulfide solution (40−48 wt %), L-ascorbic acid and
octamethylferrocene (OMFc) were obtained commercially. All were
used as received save for OMFc, which was sublimed in vacuo and
recrystallized from acetonitrile prior to use. The solvents acetonitrile
(MeCN), propionitrile (EtCN), dichloroethane (DCE), dichloro-
methane (DCM), carbon disulfide (CS2), ethyl acetate (EtOAc),
glacial acetic acid (HOAc) were of at least reagent grade. MeCN was
dried by distillation from P2O5 and CaH2 before use. EtCN was
purified by prewashing with aqueous HCl, then dried over K2CO3 and
molecule sieves, followed by distillation from CaH2.42 All reactions
were performed under an atmosphere of dry nitrogen. Melting points
are uncorrected. Infrared spectra (Nujol mull, KBr optics) were
recorded on a Nicolet Avatar FTIR spectrometer at 2 cm−1 resolution,
and UV−visible spectra were collected on samples dissolved in MeCN
using a PerkinElmer Lambda 35 UV−vis spectrophotometer. 1H NMR
spectra were run on a Bruker Avance 300 MHz NMR spectrometer,
and low resolution Electro-spray Ionization (ESI) mass spectra were
recorded on a Micromass Q-TOR Ultima Global LC/MS/MS system.
Elemental analyses were performed by MHW Laboratories, Phoenix,
AZ, 85018. Thermogravimetric analysis (TGA) and differential
scanning calorimetry (DSC) on 2a·MeCN was performed in a
platinum crucible using a TA Instruments SDT Q600 under nitrogen
gas with a heating (and cooling) rate of 1 °C min−1.
Preparation of [2a][OTf]. Silver triflate (4.10 g, 16.0 mmol) was
added to a slurry of crude [2a][Cl] (4.55 g, 14.0 mmol) in 200 mL
MeCN, to afford a deep red solution, which was heated at gentle reflux
for 90 m, then filtered to remove gray-black precipitate of AgCl. The
deep red filtrate was concentrated to 45 mL and crude [2a][OTf], a
dark green-black solid, was collected by filtration. Recrystallization
from MeCN afforded dark green shards (1.98 g, 4.50 mmol, 32%); mp
>250 °C. Anal. Calcd for C7F3N3O6S5: C, 19.13, N, 9.56. Found: C,
19.30; H, 0.1; N, 9.41. IR (cm−1): 1692 (vs), 1515 (vs), 1402 (m),
1290 (s), 1246 (br, vs), 1224 (m), 1181 (m), 1157 (m), 611 (w), 574
Preparation of Ammonium 2-amino-4,6-dinitrophenolate.30
Ammonium sulfide solution (50 mL) was added slowly over 20 min to
a yellow solution of picric acid 3 (36.0 g, 0.157 mol) in 600 mL of 20%
aq. NH3 at 60 °C, while never allowing the temperature to exceed 70
°C, to give a deep red mixture. The mixture was heated for 3 h, and
the resulting gray solid filtered off and washed with 300 mL of hot
distilled H2O. The filtrate was concentrated to a third of the volume in
vacuo at 60 °C, then cooled on an ice-bath for 2 h, and the resulting
dark purple needles collected by filtration. The crude product was
boiled into 500 mL distilled H2O, the mixture hot filtered and the
filtrate concentrated in vacuo to 300 mL. The filtrate was allowed to
cool to room temperature, whereupon 50 mL 28% aq. NH3 was added
and the mixture allowed to stand overnight. The dark purple-black
needles of the ammonium salt of 4 were filtered off and dried in air.
Yield 28.1 g, (0.130 mol, 83%). 1H NMR (DMSO-d6) δ: 8.153 (d, C−
(w), 520 (w), 502 (vw), 488 (w), 470 (w). UV−vis (MeCN), λmax
505 nm (log ε = 5.4).
=
Preparation of 2a·MeCN. Method 1: Bulk Material for
Conductivity and Magnetic Measurements. A solution of
[2a][OTf] (0.270 g, 0.614 mmol) in 100 mL of degassed MeCN (4
freeze−pump−thaw cycles) was combined with a solution of OMFc
(0.240 g, 0.805 mmol) in 85 mL of similarly degassed MeCN to yield a
dark blue solution with a dark green-black microcrystalline precipitate.
After 30 min of stirring at room temperature, the microcrystalline
product [2a][MeCN] was filtered off, washed with MeCN (4 × 20
mL) and dried in vacuo (0.145 g, 0.499 mmol, 81% yield); mp >250
°C. Anal. Calcd for C6N3O3S4·(C2H3N): C, 29.00; H, 0.91; N, 16.91.
Found: C, 28.80; H, 1.19; N, 16.78. IR (cm−1): 1595 (br, vs), 1536
(br, s), 1490 (s), 1293 (br, vs), 1234 (s), 1129 (m), 1059 (m), 974
(w), 867 (m), 860 (sh, m), 812 (m), 758 (m), 738 (m), 635 (vw), 595
(w), 557 (w), 489 (m), 465 (w), 443 (w), 422 (w). Complete removal
of MeCN solvent was effected by heating the solvate in vacuo at 90 °C
for 30 m. Anal. Calcd for C6N3O3S4: C, 24.82; N, 14.47. Found: C,
25.00; H, trace; 14.11.
Method 2: Slow Diffusion for Single Crystals. A solution of
[2a][OTf] (44 mg, 0.100 mmol) in 12 mL of degassed MeCN (4
freeze−pump−thaw cycles) was allowed to diffuse slowly into a
similarly degassed solution of OMFc (32 mg, 0.107 mmol) in 15 mL
MeCN over a 16 h period, affording [2a][MeCN] as lustrous green-
black needles.
Preparation of 2a·1/4EtCN. A solution of [2a][OTf] (30 mg,
0.068 mmol) in 15 mL of degassed EtCN (4 freeze−pump−thaw
cycles) was allowed to diffuse slowly into a similarly degassed solution
of OMFc (28 mg, 0.094 mmol) in 15 mL EtCN over a 16 h period,
affording 2a·1/4EtCN as lustrous black plates suitable for both X-ray
work and bulk property measurements. Anal. Calcd for
C8H3N4O3S4·1/4(C3H5N): C, 26.66; H, 0.41; N, 14.97. Found: C,
26.59; H, 0.79; N, 15.00. IR (cm−1): 1579 (br, s), 1491 (m), 1292 (s),
1224 (s), 1112 (w), 1058 (m), 860 (s), 808 (m), 723 (s), 559 (m).
+
H, J4 = 2.7 Hz, 1H), 7.012 (s, broad; NH4 , 4H), 6.955 (d, C−H, J4 =
2.7 Hz, 1H), 5.087 (s, broad; N−H, 2H). IR (cm−1): 3407 (m), 3313
(w), 3277 (br, m), 3158 (br, m), 1614 (sh, m), 1589 (m), 1537 (s),
1311 (s), 1262 (br, vs), 1189 (w), 1123 (w), 1069 (m), 997 (vw), 936
(w), 888 (w), 856 (w), 828 (w), 734 (m), 707 (w), 558 (w), 486 (m).
Preparation of 2-Amino-4,6-dinitrophenol 4.31 Glacial acetic
acid (20 mL) was added dropwise to a stirred solution of ammonium
2-amino-4,6-dinitrophenolate (18.9 g, 87.4 mmol) in 250 mL warm
H2O (∼60 °C). The mixture was cooled on an ice−water bath and the
resulting orange precipitate of 4 was collected by filtration and air-
dried. Yield: 15.7 g (78.8 mmol, 79%). The product was recrystallized
from EtOAc as orange needles, mp 169−171 °C. 1H NMR (CDCl3) δ:
11.178 (s, O−H, 1H), 8.392 (s, C−H, 1H), 7.730 (s, C−H, 1H),
4.478 (s, broad; N−H, 2H). IR (cm−1): 3485 (m), 3470 (m), 3250
(br, s), 3117 (w), 3104 (w), 3072 (w), 1783 (w), 1615 (s), 1595 (s),
1557 (vs), 1507 (s), 1401 (m), 1338 (vs), 1299 (vs), 1233 (vs), 1141
(s), 1110 (s), 1064 (s), 994 (s), 934 (s), 895 (m), 880 (w), 855 (w),
816 (m), 804 (m), 771 (w), 736 (m), 725 (w), 711 (m), 675 (s), 569
(vw), 491 (w).
Preparation of 2,6-Diamino-4-nitrophenol 5. A mixture of 2-
amino-4,6-dinitrophenol 4 (5.00 g, 25.1 mmol), L-ascorbic acid (22.2
g, 126.0 mmol), and a catalytic amount of FeSO4·7H2O (0.80 g, 2.9
H
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Cyclic Voltammetry. Cyclic voltammetry was performed using a
PINE Bipotentiostat, Model AFCClBP1, with scan rates of 50−250
mV s−1 on a solution of [2a][OTf] in anhydrous MeCN degassed with
argon and containing 0.1 M tetra-n-butyl-ammonium hexafluorophos-
phate. Potentials were scanned with respect to the quasi-reference
electrode in a single compartment cell fitted with Pt electrodes and
referenced to the Fc/Fc+ couple of ferrocene at 0.38 V vs SCE.43
EPR Spectroscopy. The X-Band EPR spectrum of 2a was
recorded at ambient temperature on a Bruker EMX-200 spectrometer
using a sample of 2a·MeCN dissolved in degassed toluene. Hyperfine
coupling constants were obtained by spectral simulation using
Simfonia and WinSim.44
Single Crystal Crystallography. For 2a·MeCN a crystal was
mounted on a MiTeGen MicroMount using Fomblin oil, and data
were collected at 100 K on a Bruker D8 Venture diffractometer with
Mo Kα (λ = 0.71073 Å) radiation collected using Rigaku CrystalClear
software45 and processed with Bruker APEX2 software46 and
SADABS.47 For 2a·1/4EtCN a crystal was mounted on a glass fiber
and data were collected at 296 K using omega and phi scans with a
Bruker kappa APEX II CCD detector and Mo Kα (λ = 0.71073 Å)
radiation. Data reduction and structure solution were performed using
APEX2 software. During refinement of 2a·MeCN the presence of
channels of residual e-density (possibly disordered MeCN) running
parallel to the a-axis and located at (x, 1/2, 0) and (x, 0, 1/2) was
observed. Analysis of the occupancy of these channels in PLATON
suggested the presence of 24 electrons per unit cell. These were
excluded from the final refinement using SQUEEZE.48 An alternative
refinement based on a model in which the channels were 1/2-occupied
by two disordered carbon atoms afforded R = 0.0510 and Rw = 0.1268,
that is, somewhat higher values than those provided in Table 1 for the
SQUEEZE model, which is consistent with the results of elemental
and thermogravimetric analysis.
Powder Crystallography. High pressure diffraction experiments
on 2a·MeCN were carried out on beamline HMXA at the Canadian
Light Source, using synchrotron radiation (λ = 0.509176 Å) and a
powdered sample mounted in a diamond anvil cell (DAC), with
helium as the pressure transmitting medium. (At room temperature,
helium stays in a liquid state up to 14 GPa, hence providing
hydrostatic pressure conditions.) A series of data sets was collected
over the range 2θ = 1−20° at room temperature and as a function of
increasing pressure from 0 to 10.0 GPa. Space group determinations,
indexing, and refinement of cell parameters were performed using
DASH 3.1.49 The structures were solved starting from molecular
models derived from the ambient pressure crystal structures. During
the initial refinement in DASH a rigid-body constraint was maintained.
Final Rietveld50 refinement, using fixed atomic positions and isotropic
thermal parameters with an assigned value of 0.025 was performed in
GSAS.51 Atomic positions obtained from DASH were not further
refined in GSAS, as a result of which standard deviations for atomic
coordinates are not reported in the CIF files.
for the 18/11 and 14/8 (octahedral-edge-length (mm)/truncated-
edge-length (mm)) cell configurations adopted in these experiments
were determined from prior calibrations of the applied hydraulic load
against pressures of structure transformations in standards at room
temperature (Bi I ↔ II at 2.55 GPa, Bi III ↔ V at 7.7 GPa, Sn I ↔ II at
9.4 GPa, and Pb I ↔ II at 13.4 GPa). A powder sample was densely
packed in a boron nitride (σBN = 10−11 S cm−1) cup with Pt disk
electrodes in direct contact with the samples at both ends. Four wire
AC (Solartron 1260 Impedance Analyzer) resistance measurements
were made at a frequency of 1 kHz. A series of resistance
measurements was performed at pressures up to 8 GPa at
temperatures over the range 223−298 K. Low temperatures were
achieved in the multianvil press by selectively and conductively
removing heat from the sample volume through six of the eight WC
cubes in direct contact with the octahedral pressure cell. Cooling fins
made of Cu were sandwiched between, and in thermal contact with,
neighboring anvil faces and were each connected to a dedicated Cu
heat exchanger chamber through which liquid nitrogen flowed.54 The
contiguous cylinder-shaped sample was extracted from the recovered
pressure cell and the sample geometry was measured to convert
resistance to conductivity.
Optical Measurements. Polarized reflectance spectra of a single
crystal of 2a·MeCN were measured using a Continuum Thermo-
Nicolet FTIR microscope equipped with an MCT detector and
integrated with a Nexus-670 FTIR Nicolet spectrometer.
Molecular Electronic Structure Calculations. All DFT
calculations on the radical and ionic states of 2a,b,c were performed
with the Gaussian 09W suite of programs,55 using the (U)B3LYP
hybrid functional and polarized, split-valence basis sets with triple-ζ (6-
311G(d,p)) functions. Full geometry optimization was invoked for the
calculation of the total SCF energies of each radical, its cation and the
various possible states of the anion, that is, closed shell singlet (CSS),
triplet (TS) and broken symmetry singlet (BSS). The energy of the
open shell singlet (OSS) anion (and hence EA″ and U″) was
calculated from the energies of the TS and OSS states, and their
respective ⟨S2⟩ expectation values, using the expression EOSS = ETS
−
2[ETS − EBSS]/[⟨S2⟩TS − ⟨S2⟩BSS].56 The Polarized Continuum Model
(PCM) with acetonitrile as solvent was invoked to simulate screening
effects in solution and the solid state.
Band Electronic Structure Calculations. Band structure
calculations on 2a·MeCN were carried out with the QUANTUM
ESPRESSO57 package using Perdew−Burke−Ernzerhof functionals58
and ultrasoft pseudopotentials with a plane-wave cutoff of 25 Ry and a
250 Ry integration mesh. Coordinates were taken from the single
crystal structures and GSAS refined high pressure powder data. The
SCF calculations employed a 4 × 4 × 4 Monkhorst−Pack k-point
mesh.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b11779.
■
S
Magnetic Susceptibility Measurements. DC magnetic suscept-
ibility measurements on 2a·MeCN and 2a·1/4EtCN were performed
at a field of 1000 Oe over the temperature range 2−300 K on a
Quantum Design MPMS SQUID magnetometer. Diamagnetic
corrections were made using Pascal’s constants.52
Ambient Pressure Conductivity Measurements. Four-probe
conductivity measurements on cold pressed pellet (1 × 1 × 5 mm)
samples of 2a·MeCN and 2a·1/4EtCN, and two-probe measurements
on single crystals of 2a·MeCN were performed over the range 200−
300 K using home-built equipment (for convenience, the term σRT
refers to the conductivity measured at 298 K). Silver paint (Leitsilber
200) was used to apply the electrical contacts.
Multianvil Press Conductivity Measurements. High pressure
conductivity experiments on 2a·MeCN were performed in a 3000-ton
multianvil press using a Cr2O3-doped MgO octahedron as the pressure
transmitting medium.53 The pressure was generated by three electric
oil pumps, transmitted through a split-cylinder module to six steel
wedges, then to eight tungsten carbide (WC) cubes with 32 mm edge
length, and finally through the eight truncated corners of these cubes
to the octahedral pressure medium. The force-pressure relationships
Details of powder crystallographic data collection and
structure refinement, TGA/DSC and EPR measure-
ments, and DFT calculations (PDF)
Single crystal diffraction data for 2a·MeCN (CIF)
Single crystal diffraction data for 2a·1/4EtCN (CIF)
PXRD data for 2a·MeCN at 2.0 GPa (CIF)
PXRD data for 2a·MeCN at 4.4 GPa (CIF)
PXRD data for 2a·MeCN at 6.1 GPa (CIF)
AUTHOR INFORMATION
Corresponding Author
*oakley@uwaterloo.ca
■
ORCID
Richard T. Oakley: 0000-0002-7185-2580
I
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(g) Beer, L.; Reed, R. W.; Robertson, C. M.; Oakley, R. T.; Tham, F.
S.; Haddon, R. C. Org. Lett. 2008, 10, 3121.
(12) Pal, S. K.; Itkis, M. E.; Tham, F. S.; Reed, R. W.; Oakley, R. T.;
Notes
The authors declare no competing financial interest.
Haddon, R. C. Science 2005, 309, 281.
ACKNOWLEDGMENTS
(13) (a) Cordes, A. W.; Haddon, R. C.; Oakley, R. T. In The
Chemistry of Inorganic Ring Systems; Steudel, R., Ed.; Elsevier:
Amsterdam, 1992; pp 295−321. (b) Oakley, R. T. Can. J. Chem.
1993, 71, 1775.
■
We thank the Natural Sciences and Engineering Research
Council of Canada for financial support, the Government of
Canada for a Tier I Canada Research Chair to J.S.T. and the
Canada Foundation for Innovation for funding a 3000 ton
multianvil press. This work is dedicated to the memory of
Professor Robert Haddon, a true friend and colleague.
(14) (a) Rawson, J. M.; Alberola, A.; Whalley, A. J. Mater. Chem.
2006, 16, 2560. (b) Hicks, R. G. In Stable Radicals: Fundamentals and
Applied Aspects of Odd-Electron Compounds; Hicks, R. G., Ed.; John
Wiley & Sons, Ltd.: Wiltshire, 2010; pp 317−380. (c) Ratera, I.;
́
Veciana, J. Chem. Soc. Rev. 2012, 41, 303. (d) Boere, R. T.; Roemmele,
T. L. Comp. Inorg. Chem. II 2013, 1, 375. (e) Rawson, J. M.; Hayward,
J. J. Handbook of Chalcogen Chemistry: New Perspectives in Sulfur,
Selenium and Tellurium; Devillanova, F., Du Mont, W.-W., Eds.; Royal
Society of Chemistry: Cambridge, 2013; Vol 2; pp 69−98. (f) Haddon,
R. C. ChemPhysChem 2012, 13, 3581. (g) Saito, G.; Yoshida, Y. Bull.
Chem. Soc. Jpn. 2007, 80, 1. (h) Saito, G.; Yoshida, Y. Top. Curr. Chem.
2011, 312, 67.
REFERENCES
■
(1) McCoy, N. H.; Moore, W. C. J. Am. Chem. Soc. 1911, 33, 273.
(2) (a) Haddon, R. C. Nature 1975, 256, 394. (b) Haddon, R. C.
Aust. J. Chem. 1975, 28, 2333. (c) Haddon, R. C. Aust. J. Chem. 1975,
28, 2343.
(3) Hubbard, J. Proc. R. Soc. London, Ser. A 1963, A276, 238.
(4) This condition assumes a uniform distribution of orbitals in the
energy band. A more rigorous treatment of a 1D system affords W >
(π/4) U. See: Whangbo, M. H. J. Chem. Phys. 1979, 70, 4763.
(5) (a) Mott, N. F. Proc. Phys. Soc., London, Sect. A 1949, 62, 416.
(b) Mott, N. F. Metal-Insulator Transitions; Taylor and Francis:
London, 1990.
(15) (a) Bendikov, M.; Wudl, F.; Perepichka, D. F. Chem. Rev. 2004,
104, 4891. (b) Cortizo-Lacalle, D.; Skabara, P. J.; Westgate, T. D.
Handbook of Chalcogen Chemistry: New Perspectives in Sulfur, Selenium
and Tellurium; Devillanova, F., Du Mont, W.-W., Eds.; Royal Society
of Chemistry: Cambridge, 2013; Vol 2; pp 99−126.
(16) (a) Rawson, J. M.; Banister, A. J.; Lavender, I. Adv. Heterocycl.
Chem. 1995, 62, 137. (b) Haynes, D. A. CrystEngComm 2011, 13,
4793. (c) Melen, R. L.; Less, R. J.; Pask, C. M.; Rawson, J. M. Inorg.
Chem. 2016, 55, 11747.
(17) (a) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Oakley, R. T.;
Palstra, T. T. M.; Schneemeyer, L. F.; Waszczak, J. V. J. Am. Chem. Soc.
1992, 114, 1729. (b) Cordes, A. W.; Haddon, R. C.; Oakley, R. T.;
Schneemeyer, L. F.; Waszczak, J. V.; Young, K. M.; Zimmerman, N. M.
J. Am. Chem. Soc. 1991, 113, 582. (c) Cordes, A. W.; Haddon, R. C.;
Hicks, R. G.; Oakley, R.T.; Palstra, T. T. M.; Schneemeyer, L. F.;
Waszczak, J. V. J. Am. Chem. Soc. 1992, 114, 5000.
(18) (a) Bryan, C. D.; Cordes, A. W.; Haddon, R. C.; Glarum, S. H.;
Hicks, R. G.; Kennepohl, D. K.; MacKinnon, C. D.; Oakley, R. T.;
Palstra, T. T. M.; Perel, A. S.; Schneemeyer, L. F.; Scott, S. R.;
Waszczak, J. V. J. Am. Chem. Soc. 1994, 116, 1205. (b) Bryan, C. D.;
Cordes, A. W.; Fleming, R. M.; George, N. A.; Glarum, S. H.; Haddon,
R. C.; MacKinnon, C. D.; Oakley, R. T.; Palstra, T. T. M.; Perel, A. S. J.
Am. Chem. Soc. 1995, 117, 6880. (c) Bryan, C. D.; Cordes, A. W.;
Goddard, J. D.; Haddon, R. C.; Hicks, R. G.; MacKinnon, C. D.;
Mawhinney, R. C.; Oakley, R. T.; Palstra, T. T. M.; Perel, A. S. J. Am.
Chem. Soc. 1996, 118, 330.
(6) (a) Morita, Y.; Suzuki, S.; Sato, K.; Takui, T. Nat. Chem. 2011, 3,
197. (b) Kubo, T. Chem. Rec. 2015, 15, 218.
(7) (a) Haddon, R. C.; Wudl, F.; Kaplan, M. L.; Marshall, J. H.; Cais,
R. E.; Bramwell, F. B. J. Am. Chem. Soc. 1978, 100, 7629. (b) Haddon,
R. C.; Chichester, S. V.; Stein, S. M.; Marshall, J. H.; Mujsce, A. M. J.
Org. Chem. 1987, 52, 711. (c) Beer, L.; Mandal, S. K.; Reed, R. W.;
Oakley, R. T.; Tham, F. S.; Donnadieu, B.; Haddon, R. C. Cryst.
Growth Des. 2007, 7, 802. (d) Goto, K.; Kubo, T.; Yamamoto, K.;
Nakasuji, K.; Sato, K.; Shiomi, D.; Takui, T.; Kubota, M.; Kobayashi,
T.; Yakusi, K.; Ouyang, J. J. Am. Chem. Soc. 1999, 121, 1619.
(e) Koutentis, P. A.; Chen, Y.; Cao, Y.; Best, T. P.; Itkis, M. E.; Beer,
L.; Oakley, R. T.; Cordes, A. W.; Brock, C. P.; Haddon, R. C. J. Am.
Chem. Soc. 2001, 123, 3864.
(8) (a) Kubo, T.; Katada, Y.; Shimizu, A.; Hirao, Y.; Sato, K.; Takui,
T.; Uruichi, M.; Yakushi, K.; Haddon, R. C. J. Am. Chem. Soc. 2011,
133, 14240. (b) Anamimoghadam, O.; Symes, M. D.; Long, D. L.;
Sproules, S.; Cronin, L.; Bucher, G. J. Am. Chem. Soc. 2015, 137,
14944. (c) Huang, J.; Kertesz, M. J. Am. Chem. Soc. 2007, 129, 1634.
(9) (a) Chi, X.; Itkis, M. E.; Patrick, B. O.; Barclay, T. M.; Reed, R.
W.; Oakley, R. T.; Cordes, A. W.; Haddon, R. C. J. Am. Chem. Soc.
1999, 121, 10395. (b) Huang, J.; Kertesz, M. J. Am. Chem. Soc. 2003,
125, 13334. (c) Mandal, S. K.; Samanta, S.; Itkis, M. E.; Jensen, D. W.;
Reed, R. W.; Oakley, R. T.; Tham, F. S.; Donnadieu, B.; Haddon, R. C.
J. Am. Chem. Soc. 2006, 128, 1982. (d) Pal, S. K.; Itkis, M. E.; Tham, F.
S.; Reed, R. W.; Oakley, R. T.; Haddon, R. C. J. Am. Chem. Soc. 2008,
130, 3942.
(10) (a) Haddon, R. C.; Sarkar, A.; Pal, S. K.; Chi, X.; Itkis, M. E.;
Tham, F. S. J. Am. Chem. Soc. 2008, 130, 13683. (b) Sarkar, A.; Pal, S.
K.; Itkis, M. E.; Liao, P.; Tham, F. S.; Donnadieu, B.; Haddon, R. C.
Chem. Mater. 2009, 21, 2226. (c) Bag, P.; Itkis, M. E.; Pal, S. K.;
Donnadieu, B.; Tham, F. S.; Park, H.; Schlueter, J. A.; Siegrist, T.;
Haddon, R. C. J. Am. Chem. Soc. 2010, 132, 2684. (d) Bohlin, J.;
Hansson, A.; Stafstrom, S. Phys. Rev. B: Condens. Matter Mater. Phys.
2006, 74, 155111. (e) Bag, P.; Pal, S. K.; Itkis, M. E.; Sarkar, S.; Tham,
F. S.; Donnadieu, B.; Haddon, R. C. J. Am. Chem. Soc. 2013, 135,
12936.
(11) (a) Uchida, K.; Mou, Z.; Kertesz, M.; Kubo, T. J. Am. Chem. Soc.
2016, 138, 4665. (c) Nishida, S.; Kawai, J.; Moriguchi, M.; Ohba, T.;
Haneda, N.; Fukui, K.; Fuyuhiro, A.; Shiomi, D.; Sato, K.; Takui, T.;
Nakasuji, K.; Morita, Y. Chem. - Eur. J. 2013, 19, 11904. (d) Cui, Z. H.;
Gupta, A.; Lischka, H.; Kertesz, M. Phys. Chem. Chem. Phys. 2015, 17,
23963. (e) Kolb, B.; Kertesz, M.; Thonhauser, T. J. Phys. Chem. A
2013, 117, 3642. (f) Small, D.; Zaitsev, V.; Jung, Y.; Rosokha, S. V.;
Head-Gordon, M.; Kochi, J. K. J. Am. Chem. Soc. 2004, 126, 13850.
(19) Cordes, A. W.; Haddon, R. C.; Oakley, R. T. Phosphorus, Sulfur
Silicon Relat. Elem. 2004, 179, 673.
(20) (a) Beer, L.; Brusso, J. L.; Cordes, A. W.; Haddon, R. C.; Itkis,
M. E.; Kirschbaum, K.; MacGregor, D. S.; Oakley, R. T.; Pinkerton, A.
A.; Reed, R. W. J. Am. Chem. Soc. 2002, 124, 9498. (b) Beer, L.;
Britten, J. F.; Brusso, J. L.; Cordes, A. W.; Haddon, R. C.; Itkis, M. E.;
MacGregor, D. S.; Oakley, R. T.; Reed, R. W.; Robertson, C. M. J. Am.
Chem. Soc. 2003, 125, 14394. (c) Leitch, A. A.; Reed, R. W.;
Robertson, C. M.; Britten, J. F.; Yu, X.; Secco, R. A.; Oakley, R. T. J.
Am. Chem. Soc. 2007, 129, 7903.
(21) (a) Yu, X.; Mailman, A.; Dube, P. A.; Assoud, A.; Oakley, R. T.
Chem. Commun. 2011, 47, 4655. (b) Yu, X.; Mailman, A.; Lekin, K.;
Assoud, A.; Dube, P. A.; Oakley, R. T. Cryst. Growth Des. 2012, 12,
2485. (c) Yu, X.; Mailman, A.; Lekin, K.; Assoud, A.; Robertson, C.
M.; Noll, B. C.; Campana, C. F.; Howard, J. A. K.; Dube, P. A.; Oakley,
R. T. J. Am. Chem. Soc. 2012, 134, 2264. (d) Mailman, A.; Winter, S.
M.; Yu, X.; Robertson, C. M.; Yong, W.; Tse, J. S.; Secco, R. A.; Liu,
Z.; Dube, P. A.; Howard, J. A. K.; Oakley, R. T. J. Am. Chem. Soc. 2012,
134, 9886. (e) Wong, J. W. L.; Mailman, A.; Winter, S. M.; Robertson,
C. M.; Holmberg, R. J.; Murugesu, M.; Dube, P. A.; Oakley, R. T.
Chem. Commun. 2014, 50, 785. (f) Mailman, A.; Winter, S. M.; Wong,
J. W. L.; Robertson, C. M.; Assoud, A.; Dube, P. A.; Oakley, R. T. J.
Am. Chem. Soc. 2015, 137, 1044.
J
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(22) (a) Beneberu, H. Z.; Tianza, Y.-H.; Kertesz, M. Phys. Chem.
Chem. Phys. 2012, 14, 10713. (b) Preuss, K. E. Polyhedron 2014, 79, 1.
(23) (a) Brusso, J. L.; Derakhshan, S.; Itkis, M. E.; Kleinke, H.;
Haddon, R. C.; Oakley, R. T.; Reed, R. W.; Richardson, J. F.;
Robertson, C. M.; Thompson, L. K. Inorg. Chem. 2006, 45, 10958.
(b) Brusso, J. L.; Cvrkalj, K.; Leitch, A. A.; Oakley, R. T.; Reed, R. W.;
Robertson, C. M. J. Am. Chem. Soc. 2006, 128, 15080. (c) Leitch, A. A.;
Yu, X.; Winter, S. M.; Secco, R. A.; Dube, P. A.; Oakley, R. T. J. Am.
Chem. Soc. 2009, 131, 7112.
(24) Leitch, A. A.; Lekin, K.; Winter, S. M.; Downie, L. E.; Tsuruda,
H.; Tse, J. S.; Mito, M.; Desgreniers, S.; Dube, P. A.; Zhang, S.; Liu,
Q.; Jin, C.; Ohishi, Y.; Oakley, R. T. J. Am. Chem. Soc. 2011, 133, 6051.
(25) (a) Robertson, C. M.; Leitch, A. A.; Cvrkalj, K.; Reed, R. W.;
Myles, D. J. T.; Dube, P. A.; Oakley, R. T. J. Am. Chem. Soc. 2008, 130,
8414. (b) Robertson, C. M.; Leitch, A. A.; Cvrkalj, K.; Myles, D. J. T.;
Reed, R. W.; Dube, P. A.; Oakley, R. T. J. Am. Chem. Soc. 2008, 130,
14791. (c) Leitch, A. A.; Brusso, J. L.; Cvrkalj, K.; Reed, R. W.;
Robertson, C. M.; Dube, P. A.; Oakley, R. T. Chem. Commun. 2007,
3368.
Funct. Mater. 2002, 12, 693. (d) Tenn, N.; Bellec, N.; Jeannin, O.;
́
Piekara-Sady, L.; Auban-Senzier, P.; Iniguez, J.; Canadell, E.; Lorcy, D.
̃
J. Am. Chem. Soc. 2009, 131, 16961. (e) Le Pennec, R.; Jeannin, O.;
Auban-Senzier, P.; Fourmigue,
(41) Terletska, H.; Vucicevic,
Phys. Rev. Lett. 2011, 107, 026401.
́
M. New J. Chem. 2016, 40, 7113.
̌
̌
́
J.; Tanaskovic, D.; Dobrosavljevic, V.
́
́
(42) Chai, C. L. L.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 5th ed.; Butterworth-Heinemann: New York, 2003.
(43) Boere,
1994, 620, 1589. (b) Boere,
2000, 210, 369.
́
R. T.; Moock, K. H.; Parvez, M. Z. Anorg. Allg. Chem.
́
R. T.; Roemmele, T. L. Coord. Chem. Rev.
(44) WinEPR Simfonia, version 1.25; Bruker Instruments, Inc.:
Billerica, MA, 1996.
(45) CrystalClear; Rigaku Corporation: Tokyo, Japan, 2005.
(46) APEX2; Bruker AXS Inc.: Madison, WI, 2007.
(47) SADABS; Bruker AXS Inc.: Madison, WI, 2001.
(48) Spek, A. L. Acta Crystallogr., Sect. C: Struct. Chem. 2015, 71, 9.
(49) David, W. I. F.; Shankland, K.; van de Streek, J.; Pidcock, E.;
Motherwell, W. D. S.; Cole, J. C. J. Appl. Crystallogr. 2006, 39, 910.
(50) Rietveld, H. M. J. Appl. Crystallogr. 1969, 2, 65.
(51) Larson, A. C.; Von Dreele, R. B. General Structure Analysis
System (GSAS); Los Alamos National Laboratory Report LAUR 86−
748; 2000.
(52) (a) Carlin, R. L. Magnetochemistry; Springer-Verlag: New York,
1986. (b) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008, 85, 532.
(53) Secco, R. A.; Yong, W.; Officer, T. New 3000 Ton Multi-Anvil
Press at the University of Western Ontario, Abstract MR51A-02,
American Geophysical Union Meeting, May 14−17, 2013 Cancun,
Mexico.
(54) Secco, R. A.; Yong, W. Rev. Sci. Instrum. 2016, 87, 123901.
(55) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, Jr., J. A.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. W.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.
(56) (a) Noodleman, L. J. Chem. Phys. 1981, 74, 5737. (b) Noodle-
man, L.; Davidson, E. R. Chem. Phys. 1986, 109, 131. (c) Yamaguchi,
K.; Jensen, F.; Dorigo, A.; Houk, K. N. Chem. Phys. Lett. 1988, 149,
537. (d) Nagao, H.; Nishino, M.; Shigeta, Y.; Soda, T.; Kitagawa, Y.;
Onishi, T.; Yoshika, Y.; Yamaguchi, K. Coord. Chem. Rev. 2000, 198,
265.
(57) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.;
Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.;
Dal Corso, A.; Fabris, S.; Fratesi, G.; de Gironcoli, S.; Gebauer, R.;
Gerstmann, U.; Gougoussis, C.; Kokalj, A.; Lazzeri, M.; Martin-Samos,
L.; Marzari, N.; Mauri, F.; Mazzarello, R.; Paolini, S.; Pasquarello, A.;
Paulatto, L.; Sbraccia, C.; Scandolo, S.; Sclauzero, G.; Seitsonen, A. P.;
Smogunov, A.; Umari, P.; Wentzcovitch, R. M. J. Phys.: Condens.
Matter 2009, 21, 395502.
(26) Winter, S. M.; Hill, S.; Oakley, R. T. J. Am. Chem. Soc. 2015,
137, 3720.
(27) Wong, J. W. L.; Mailman, A.; Lekin, K.; Winter, S. M.; Yong, W.;
Zhao, J.; Garimella, S. V.; Tse, J. S.; Secco, R. A.; Desgreniers, S.;
Ohishi, Y.; Borondics, F.; Oakley, R. T. J. Am. Chem. Soc. 2014, 136,
1070.
(28) Winter, S. M.; Mailman, A.; Oakley, R. T.; Thirunavukkuarasu,
K.; Hill, S.; Graf, D. E.; Tozer, S. W.; Tse, J. S.; Mito, M.; Yamaguchi,
H. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 214403.
(29) Tian, D.; Winter, S. M.; Mailman, A.; Wong, J. W. L.; Yong, W.;
Yamaguchi, H.; Jia, Y.; Tse, J. S.; Desgreniers, S.; Secco, R. A.; Julian, S.
R.; Jin, C.; Mito, M.; Ohishi, Y.; Oakley, R. T. J. Am. Chem. Soc. 2015,
137, 14136.
(30) (a) Eugen, K. German Patent 3328002-A1 (1985). (b) Sheth, A.;
Doshi, N.; Sen, D. J.; Badmanaban, R.; Patel, C. N. J. Chem. Pharm.
Res. 2010, 2, 1.
(31) Tuzun, C.; Oklemer, A.; Yildiz, G.; Birer, S. Commun. Fac. Sci.
Univ. Ankara, Ser. B: Chem. Chem. Eng. 1988, 34, 159.
(32) (a) Herz, R. Chem. Z. 1922, 4, 948. (b) Warburton, W. K. Chem.
Rev. 1957, 57, 1011.
(33) The cell potential Ecell is defined here as the numerical difference
between the half-wave potentials for the oxidation and reduction
processes, that is, |Ecell = E1/2(ox) − E1/2(red)|.
(34) Peierls, R. C. Quantum Theory of Solids; Oxford University
Press: London, 1955; p 108.
(35) (a) Bondi, A. J. Phys. Chem. 1964, 68, 441. (b) Dance, I. New J.
Chem. 2003, 27, 22.
(36) (a) Williams, J. M.; Ferraro, J. R.; Thorn, R. J.; Carlson, K. D.;
Geiser, U.; Wang, H. H.; Kini, A. M.; Whangbo, M.-H. Organic
Superconductors (Including Fullerenes): Synthesis, Structure, Properties
and Theory; Prentice Hall: Englewood Cliffs, NJ, 1992. (b) Mori, T.
Electronic Properties of Organic Conductors; Springer: Japan, 2016.
(37) (a) Yzambart, G.; Bellec, N.; Ghassan, N.; Jeannin, O.; Roisnel,
́
́
T.; Fourmigue, M.; Auban-Senzier, P.; Iniguez, J.; Canadell, E.; Lorcy,
̃
D. J. Am. Chem. Soc. 2012, 134, 17138. (b) Le Gal, Y.; Roisnel, T.;
Auban-Senzier, P.; Guizouarn, T.; Lorcy, D. Inorg. Chem. 2014, 53,
8755.
́
(38) (a) Jankowski, D.; Swietlik, R.; Jeannin, O.; Assaf, A.;
Reinheimer, E. W.; Formigue, M. J. Raman Spectrosc. 2013, 44,
́
1765. (b) Ishikawa, T.; Kitayama, M.; Chono, A.; Onda, K.; Okimoto,
Y.; Koshihara, S.; Nakano, Y.; Yamochi, H.; Morikawa, T.; Shirahata,
T.; Misaki, Y. J. Phys.: Condens. Matter 2013, 24, 195501. (c) Dressel,
M. ISRN Condens. Matter Phys. 2012, 2012, 732973.
(58) (a) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996,
77, 3865. (b) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett.
1997, 78, 1396.
(39) Ioffe, A. F.; Regel, A. R. Prog. Semicond. 1960, 4, 237.
(40) (a) Schiødt, N. C.; Bjørnholm, T.; Bechgaard, K.; Neumeier, J.
J.; Allgeier, C.; Jacobsen, C. S.; Thorup, N. Phys. Rev. B: Condens.
Matter Mater. Phys. 1996, 53, 1773. (b) Belo, D.; Alves, H.; Lopes, E.
B.; Duarte, M. T.; Gama, V.; Henriques, R. T.; Almeida, M.; Per
Benítez, A.; Rovira, C.; Veciana, J. Chem. - Eur. J. 2001, 7, 511.
(c) Dautel, O. J.; Fourmigue, M.; Canadell, E.; Auban-Senzier, P. Adv.
́
ez-
́
K
DOI: 10.1021/jacs.6b11779
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX