Inorganica Chimica Acta 480 (2018) 91–100

Contents lists available at ScienceDirect

Inorganica Chimica Acta

journal homepage: www.elsevier.com/locate/ica

Research paper

Electrochemistry of TCNQF2 in acetonitrile in the presence of

[Cu(CH3CN)4]+: Electrocrystallisation and characterisation of CuTCNQF2
Nguyen T. Vo a,b, Lisandra L. Martin a,⇑, Alan M. Bond a,⇑
a
School of Chemistry, Monash University, Clayton, Victoria 3800, Australia
b
Danang University of Education, Danang, Viet Nam

a r t i c l e i n f o a b s t r a c t

Article history: The bulk electrochemical reduction of TCNQF2 (where TCNQF2 = 2,5-difluoro-7,7,8,8-tetracyanoquin-
Received 10 January 2018 odimethane) in acetonitrile (0.1 M Bu4NPF6) in the presence of [Cu(CH3CN)4]+ leads to the electrocrys-
Received in revised form 29 March 2018 tallisation of TCNQF1
and TCNQF2
materials, identified and proposed as CuITCNQFI
and
2 2 2
Accepted 7 April 2018
CuI2(TCNQFII

2 )(CH 3CN) 2, respectively. The existence of two forms of each solid was established by cyclic
Available online 21 April 2018
voltammetry. The low solubility of both CuITCNQFI
I II

2 and Cu2(TCNQF2 )(CH3CN)2 solids, facilitated detection
of a solid-solid transformation in the presence of [Cu(CH3CN)4]+. CuITCNQFI
2 was synthesized chemically as
Keywords:
a dark blue microcrystalline solid by reaction of TCNQF2 and CuI in CH3CN, as well as electrochemically.
Electrocrystallisation of CuITCNQF2I
and
CuI2(TCNQF2II
)(CH3CN)2
Electronic and vibrational spectroscopic methods confirmed the CuITCNQFI
2 product obtained by either

Chemical synthesis of CuITCNQF2I
method was structurally identical. Powder X-ray diffraction studies of CuITCNQFI
2 gave a closely related

Cyclic voltammetry pattern to that for the thermodynamically stable CuITCNQI
phase II (a coordination polymer) rather than
Electronic and vibrational spectroscopy the kinetically favoured CuITCNQI
phase I. Scanning electron microscopy established the dominant mor-
Morphology phology, derived from both electrocrystallized and chemically synthesised samples, were the same. The

6
conductivity of CuITCNQFI
2 as a film on FTO glass was 6.0  10 S cm
1, which lies in the semiconducting
range.
Ó 2018 Elsevier B.V. All rights reserved.

1. Introduction of new dianionic materials. A family of dihalogenated dibromo-,

Organic charge-transfer materials derived from TCNQ (TCNQ =
7,7,8,8-tetracyanoquinodimethane, Fig. 1) have been widely inves-
tigated and exhibit a range of practically important properties
[1,2]. For example, TCNQ charge-transfer complexes with transi-
tion metals ions, such as copper (I) or silver (I) are semi-conduc-
tors, and have been utilised in optical, electrical and magnetic
devices [3–7]. TCNQF4-based materials (TCNQF4 = 2,3,5,6-tetraflu-
oro-7,7,8,8-tetracyanoquinodimethane, Fig. 1) have also been stud-
ied recently [8,9]. The presence of four fluorine atoms enhances the
electron affinity of TCNQF4, which facilitates reduction to TCNQF1

4 was synthesized and shown to be a fully ionic material derived
and TCNQF2
4 and increases the stability of derived anions, espe-
cially the dianionic, two electron reduced form. This property
allows TCNQF2
4 -based materials to be generated in the air [10–
13], while those of the dianionic parent compound, TCNQ2
usually
need to be synthesized anaerobically [14–16]. The enhanced stabil-
ity of the TCNQF4 derivative therefore offers the exciting prospect
⇑ Corresponding authors.
E-mail addresses: Lisa.Martin@monash.edu (L.L. Martin), Alan.Bond@monash.
edu (A.M. Bond).
dichloro- and difluoro-TCNQ complexes also have been reported
[1,17–19]. However examples of the difluorinated, TCNQF2
(TCNQF2 = 2,5-difluoro-7,7,8,8-tetracyanoquinodimethane, Fig. 1)
derivatives are relatively rare.
TCNQF2-based materials are expected to possess intermediate
electronic properties between TCNQ and TCNQF4.[20] For example,
its electron affinity is 3.02 eV, which lies midway between 2.85 and
3.20 eV for TCNQ and TCNQF4, respectively [21,22]. TCNQF1
2 salts
of TTF+ (TTF = tetrathiafulvalene) and its derivatives have been
reported [17–19] and a TCNQF2-nucleobase, cytosine material
from the TCNQF1
2 monoanion and the cytosine cation [23]. How-
ever, studies on the interaction between reduced forms TCNQF2
and transition metals have yet to be reported. In this study, the
reductive electrochemistry of TCNQF2, in the presence of [Cu(CH3-
CN)4]+ is reported in acetonitrile. The formation of CuITCNQFI
2
based on the electrochemical investigation is described along with
the its synthesis, electronic and vibrational spectroscopy and struc-
tural characterization. In addition, the electrocrystallisation of
CuI2(TCNQFII

2 )(CH3CN)2 is also described.

https://doi.org/10.1016/j.ica.2018.04.010
0020-1693/Ó 2018 Elsevier B.V. All rights reserved.
92 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100
Fig. 1. Molecular structures of TCNQ, TCNQF2 and TCNQF4.
2. Experimental1
2.1.Chemicals
TCNQF2 (98%, TCI Tokyo), [Cu(CH3CN)4]PF6 (98%, Aldrich), ace-
tonitrile (CH3CN or alternatively MeCN; HPLC grade, Omnisolv),
isopropanol (BHD) and acetone (suprasolv, Merck KGaA) were used
as received from the manufacturer. Bu4NPF6 (Aldrich), used as the
supporting electrolyte in electrochemical studies, was recrystal-
lized twice from 96% ethanol (Merck) and then dried at 100 °C
under vacuum for 24 h prior to use. Copper iodide (CuI) was
obtained from Strem Chemicals (98%, Newburyport). Microanalysis
was carried out at the Campbell Microanalytical Laboratories,
University of Otago, New Zealand.
2.2. Electrochemistry
Voltammetric experiments were undertaken at room tempera-
ture (22 ± 1 °C) using a Bioanalytical Systems (BAS) 100 W work-
station. A standard three electrode cell configuration, comprising
a glassy carbon (GC, 1 or 3 mm diameter) working electrode, an
Ag/Ag+ (1.0 mM Ag+) reference electrode (RE) and a 1.0 mm diam-
eter platinum wire counter electrode, was employed in most
experiments. For some experiments, working electrodes were
BAS gold or platinum (Au or Pt, 1.6 mm or Au 10 lm diameter),
indium tin oxide (ITO)- or fluorine tin oxide (FTO)-coated glass
plates (0.1–0.2 cm2) with a resistance of 10 X/sq, as specified by
the manufacturer (Prazisions Glas and Optik GmbH). The data
reported in detail on GC are almost independent of electrode mate-
rial; GC, Au, Pt and FTO. Prior to each experiment, the working
electrode was polished with an aqueous 0.3 lm Al2O3 slurry using
a polishing cloth, rinsed with water followed by sonication in an
ultrasonic bath for 30 s and dried under a stream of nitrogen.
The RE was constructed from Ag wire in contact with acetonitrile
solution (0.1 M Bu4NPF6) containing 1.0 mM AgNO3 and separated
from the test solution using a salt bridge. The potential of this ref-
erence electrode was
124 mV vs the ferrocene/ferrocenium
(Fc0/1+) couple. All solutions were purged with nitrogen gas for at
least 10 min prior to each experiment and a stream of nitrogen
was maintained above the solutions during the course of the
voltammetric experiments. In bulk electrolysis experiments, a
three-compartments cell was used with a large area Pt mesh work-
ing electrode, a Ag/Ag+ (1.0 mM Ag+) reference electrode and a Pt
mesh counter electrode. In this case, each compartment was sepa-
rated by a glass frit.
2.3. Synthesis of CuITCNQFI

2 FEI Nova NanoSEM 450 FEGSEM instrumentation using an acceler-
CuITCNQFI
2 was prepared electrochemically as follows. Initially
a 5.0 mM solution of TCNQF1
2 was prepared quantitatively by the
reductive bulk electrolysis of 10 ml of 5.0 mM TCNQF2 in acetoni-
trile (0.1 M Bu4NPF6). The potential of the Pt, ITO or FTO working
electrode was held at
100 mV vs Ag/Ag+ until the current reached
1% of its initial value. A dark blue precipitate formed immediately
1
In equations, acetonitrile soluble species are designated by use of the subscript
(MeCN), whereas solids use the subscript (s).
upon addition of 0.75 ml of 100 mM [Cu(CH3CN)4]+. After stirring
for 10 min, the solid was collected by filtration and washed several
times with CH3CN. Finally, the solid was dried under vacuum over-
night before further characterisation.
Solid CuITCNQFI
2 was also chemically synthesized by a redox
reaction between TCNQF2 and CuI (in a 1:1.5 stoichiometric ratio)
in acetonitrile with stirring for 3 h. CuI acts as the reductant and
the resulting dark blue solid, formed by the reaction given in Eq.
(1), was filtered, washed with CH3CN and dried as described above,
prior to further characterisation.
3CuI þ 2TCNQF2 ! 2CuI TCNQ FI

2 þ 3CuI3 ð1Þ
2.4. Conductivity of CuTCNQF2
The conductivity was measured on a CuITCNQFI
2 film. A
Cu(metal) film was firstly sputter-coated using a Quorum Q150TS
sputter coater instrument over a 1.0 cm  1.0 cm area of FTO-
coated glass (1.0 cm  3.0 cm). The modified FTO glass was then
soaked in an acetonitrile solution containing 10 mM TCNQF2 for
12 h, resulting in the CuITCNQFI
2 film as describe in Eq. (2).
CuðmetalÞ þ TCNQF2ðMeCNÞ ! CuI TCNQFI

2ðsÞ ð2Þ
Films on the FTO glasses were rinsed briefly with acetonitrile, to
remove excess TCNQF2, followed by a further rinsing step using
copious amounts of water, before being dried under a stream of
nitrogen and stored under vacuum. For conductivity measure-
ments, two CuITCNQFI
2 -coated FTO glass pieces were stacked
together so that the two CuITCNQFI
2 films were in contact. The
FTO glasses were then clamped carefully to minimize the contact
force as shown in Scheme 1. Constant potential measurements
were performed for 60 s at potentials from 50 to 500 mV with
50 mV intervals. The resistance (R) was calculated as R = U/I
(where U (V) is the applied potential and I (A) is the measured cur-
rent). The DC conductivity, r, was calculated from the relationship
r = t/R.S where t (cm) is the thickness of the layer formed by two
CuITCNQFI
2
2 films, and S (cm ) is the cross-sectional area. A VMP3
multi-channel potentiostat from BioLogic Instruments was used
for the resistance measurements. The thickness of the CuITCNQFI
2
films was measured using a VeeCo Dektak 150 profilometer. All
of these measurements were performed in triplicate on two differ-
ent thicknesses of FTO, at the same temperature, humidity and
light conditions and corrected for the background conductivity
contribution from FTO.
2.5. Other instrumentation
UV–Vis spectra were recorded with a Varian Cary 5000 UV–Vis
NIR spectrophotometer with a 1.0 cm path length quartz cuvette. A
Varian UMA600 IR microscope and FTS7000 optics bench with 128
scans and a resolution of 8 cm
1 was used for IR spectra measure-
ments. Raman spectra were recorded on a Renishaw Invia Raman
spectrograph with an Argon ion laser with excitation at 633 nm.
After being coated with iridium, SEM images were collected with
ating voltage of 5.0 kV. The X-ray powder diffraction (XRD) pattern
was collected using an Oxford Diffraction Supernova diffractome-
FTO CuTCNQF2 FTO
+
-
FTO CuTCNQF2
Scheme 1. Experimental configuration used for conductivity measurements.
 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100 93

ter. The theoretical XRD powder pattern for CuITCNQI
phase II was
calculated from the structure obtained from single crystal X-ray
diffraction data [24].
3. Result and discussion
3.1. Cyclic voltammetry of the TCNQF0/1
reduction process in the
2
presence of [Cu(CH3CN)4]+
In acetonitrile (0.1 M Bu4NPF6), [Cu(CH3CN)4]+ can be reduced
to Cu metal and oxidized to Cu2+ (Fig. 2a) [11]. The Cu+(MeCN) to
Cu(metal) reduction component is the Cu+/0 process and occurs at

900 mV. The oxidative Cu(metal) to Cu+(MeCN) counterpart is seen as
a stripping process at
300 mV in cyclic voltammograms shown in
Fig. 2. The oxidation process, at 800 mV with its reductive counter-
part at
650 mV, represents the fully solution based Cu+/2+
(MeCN) couple.
A comparison of cyclic voltammograms for [Cu(CH3CN)4]+ and
TCNQF2 in acetonitrile (Fig. 2a and b, respectively) reveals that the
potentials for reduction and oxidation of [Cu(CH3CN)4]+ are well-
removed from the two reversible TCNQF0/1
and TCNQF1
/2
reduc-
2 2
tion steps. Therefore, the electrochemical reduction of TCNQF2 in the
presence of [Cu(CH3CN)4]+ can be investigated provided the applied
In this Section 3.1, the potential range chosen led to only the
TCNQF0/1
2 process occurring and not the TCNQF1
/2

2 one.
Cyclic voltammograms for a series of acetonitrile solutions (0.1
M Bu4NPF6) containing mixtures of TCNQF2 and [Cu(CH3CN)4]+
were recorded at a scan rate of 50 mV.s
1. For TCNQF2 (1.0 mM),
in the presence of either 5.0 mM or 10.0 mM [Cu(CH3CN)4]+, the
TCNQF0/1
2 process retained its reversible, diffusion controlled
characteristics (Fig. 3a). On increasing the concentration of TCNQF2
to 2.0 mM, still in the presence of 10.0 mM [Cu(CH3CN)4]+ and
scanning over the same potential range, the TCNQF0/1
2 process is
not significantly modified. However, a new oxidation process at
385 mV, designated as Ox1, is clearly observed after switching
the potential, and scanning in the positive direction (Fig. 3b). Pro-
cess Ox1 was enhanced when the solution contained 5.0 mM
TCNQF2 and 10.0 mM [Cu(CH3CN)4]+ (Fig. 3b). On cycling the
potential over the range relevant to Fig. 2b, the peak current ratio
ip ox/ip red for the TCNQF0/1

2 process decreased to slightly less than
unity, and the magnitude of the oxidative current gradually
decreased (Fig. S1). Simultaneously, the current magnitude for
the Ox1 process increased. These observations imply that a chem-
ical reaction is coupled to the TCNQF0/1
2 process, which consumes
TCNQF1
2 , as evidenced by the decrease of the oxidative current

potential lies between the reduction and oxidation of [Cu(CH3CN)4]+. when the potential was switched at
100 mV. The new product

Fig. 2. Cyclic voltammograms obtained in acetonitrile (0.1 M Bu4NPF6) at a scan rate of 100 mV.s
1 for (a) 10.0 mM [Cu(CH3CN)4]+ with a 1.0 mm diameter GC electrode, and
(b) 1.0 mM TCNQF2 with a 3.0 mm diameter GC electrode.

Fig. 3. Cyclic voltammograms obtained with a 1.0 mm diameter GC electrode at a scan rate of 50 mV s
1 in acetonitrile (0.1 M Bu4NPF6) over the potential range where the
TCNQF0/1

2 process occurs for (a) 1.0 mM TCNQF2 in the presence of 10.0 mM of [Cu(CH3CN)4]+, (b) designated TCNQF2 concentrations in the presence of 10.0 mM of [Cu
(CH3CN)4]+.
94 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100

is oxidized at more positive potentials than the TCNQF0/1

2 process rates. The fact that a memory effect exists on cycling the potential
(Ox1 in Fig. 3b). After completion of cyclic voltammetric experi- implies that not all CuITCNQFI

2(s) is removed in the electrocrystalli-
ments, as described above, a dark blue solid was evident on the sation-stripping sequence as also noted in quartz crystal microbal-
GC electrode surface, which was subsequently identified as CuI- ance and surface plasmon resonance studies on CuITCNQFI
4(s) [11].
TCNQFI
0/1

2 . This suggests that reduction of TCNQF2 in the presence A mechanism for the reduction of TCNQF02 to TCNQF1
2 in the
of [Cu(CH3CN)4]+ leads to electrocrystallisation of CuITCNQFI
2 on presence of [Cu(MeCN)4]+, in acetonitrile, can be written as
the electrode surface which is then partially oxidised to TCNQF2 follows:
and [Cu(CH3CN)4]+ in process Ox1.
TCNQF2ðMeCNÞ þ e

TCNQF1

2ðMeCNÞ ð3Þ
3.1.1. Effect of scan rate
þ
Fig. 4a displays cyclic voltammograms of 5.0 mM of TCNQF2 and TCNQF1
I I

2ðMeCNÞ þ CuðMeCNÞ
Cu TCNQF2ðsÞ;phase I ð4Þ
10.0 mM of [Cu(CH3CN)4]+ in CH3CN (0.1 M Bu4NPF6) at scan rates
over the range of 20–100 mVs
1. Process Ox1, detected at 385 CuI TCNQFI
I I

2ðsÞ;phase I
Cu TCNQF2ðsÞ;phase II ð5Þ
mV, with scan rates of 20 and 50 mVs
1, diminishes at higher scan
rates, 100 mVs
1. Moreover, at the slow scan rate of 20 mVs
1, a þ

CuI TCNQFI

2ðsÞ;phase I
CuðMeCNÞ þ TCNQF2ðMeCNÞ þ e
new process, labelled Ox2, becomes visible as a shoulder on the
TCNQF1
/0
2 oxidation process. Oxidation processes, Ox1 and Ox2 ðprocess Ox2Þ ð6Þ
are attributed to the oxidation CuITCNQFI
2 solid in two phases. It
is known that the coordination polymer CuITCNQI
exists in two þ

CuI TCNQFI

2ðsÞ;phase II
CuðMeCNÞ þ TCNQF2ðMeCNÞ þ e
phases, each with different structures, electrochemistry and mor-
ðprocess Ox1Þ ð7Þ
phologies [25–28]. Phase I of CuITCNQI
is the kinetically favoured
phase and phase II is thermodynamically favoured [26,27]. In the
case of electrocrystallisation of CuITCNQFI
2 , the thermodynami- 3.2. Cyclic voltammetry of the TCNQF1
/2

2 process in the presence of
cally favoured phase II formed on longer timescales (slow scan [Cu(CH3CN)4]+
rates) is oxidised via process Ox1, and the initially formed, kineti-
cally favoured phase I, is oxidised in process Ox2 and detected at The second TCNQF2 reduction step generates TCNQF2
2 and
faster scan rates. occurs at a significantly more negative potential than the
TCNQF0/1

2 process. This allows the TCNQF1
/2

2 process to be inves-
3.1.2. Phases of CuITCNQFI

2 tigated in the presence of [Cu(CH3CN)4]+ by scanning the potential
In order to further probe the origin of the process assigned to to more negative values (600 to
600 mV). In order to prevent the
oxidation of phase I of CuITCNQFI
2 (Ox2 in Fig. 4a), a higher con- electrocrystallisation of the sparingly soluble CuITCNQFI

2 material,
centration solution containing 8.0 mM of TCNQF2 and 8.0 mM of lower concentrations of TCNQF2 and [Cu(CH3CN)4]+ were
[Cu(CH3CN)4]+ was prepared. At a scan rate of 50 m Vs
1, the first employed. Thus with an acetonitrile solution containing 1.0 mM
positive potential direction scan is now dominated by the stripping TCNQF2 and 2.0 mM [Cu(CH3CN)4]+ and a slow scan rate of 20
process Ox2 at 250 mV, with process Ox1 at 385 mV being less sig- mVs
1, a simple reversible TCNQF1
/02 process was observed, as is
nificant (Fig. 4b). However, on cycling the potential, the magnitude the case in the absence of [Cu(CH3CN)4]+ when the potential
of the Ox2 peak current progressively decreases, and is not switched at
120 mV, that is, prior to the TCNQF1
/2
2 process
detected in the 3rd cycle. Simultaneously, the Ox1 peak current (Fig. 5). Furthermore, in these experiments, no CuITCNQFI
2 mate-
gradually increases in magnitude with each cycle of the potential. rial precipitated on the electrode surface, as observed with higher
At the higher scan rate of 100 mVs
1, process Ox2 is not concentrations of [Cu(CH3CN)4]+. Therefore, under these dilute [Cu
detected (Fig. S2). The scan rate interdependence of Ox1 and Ox2 (CH3CN)4]+ conditions, the second reduction process can be inves-
supports the existence of two phases of CuITCNQFI
2(s) with the pro- tigated without interference from precipitation from the initial
cess Ox2 associated with the thermodynamically favoured phase, process.
that is only detected at slow scan rates. Process Ox1 is associated The TCNQF1
/2

2 process changes dramatically for 1.0 mM
with the kinetically favoured phase, and is dominant at faster scan TCNQF2 in the presence of 2.0 mM [Cu(CH3CN)4]+ by extending

Fig. 4. Cyclic voltammograms obtained in acetonitrile (0.1 M Bu4NPF6) for: (a) 5.0 mM TCNQF2(MeCN) in the presence of 10 mM [Cu(CH3CN)4]+ at designated scan rates; (b) 8.0
mM TCNQF2(MeCN) in the presence of 8.0 mM [Cu(CH3CN)4]+ at a scan rate of 50 mV s
1 with a 1.0 mm diameter GC electrode.
 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100 95

TCNQF2ðMeCNÞ þ e

TCNQF1

2ðMeCNÞ ð8Þ

TCNQF1
2

2ðMeCNÞ þ e
TCNQF2ðMeCNÞ ð9Þ

þ
TCNQF2

2ðMeCNÞ þ 2½CuðCH3 CNÞ4 ðMeCNÞ

CuI2 ðTCNQFII

2 ÞðCH3 CNÞ2ðsÞ;phase III þ 6CH3 CN ð10Þ

CuI2 ðTCNQFII
I II

2 ÞðCH3 CNÞ2ðsÞ;phase III
Cu2 ðTCNQF2 ÞðCH3 CNÞ2ðsÞ;phase IV

ð11Þ

CuI2 ðTCNQFII

2 ÞðCH3 CNÞ2ðsÞ;phase III þ 2CH3 CN
þ


CuI TCNQFI

2ðsÞ;phase I þ ½CuðCH3 CNÞ4  þ e ð12Þ

þ

CuI TCNQFI

2ðsÞ;phase I þ 4CH3 CN
½CuðCH3 CNÞ4  þ TCNQF2ðMeCNÞ þ e

ð13Þ

Fig. 5. Cyclic voltammograms obtained using a 1.0 mm diameter GC electrode for CuI2 ðTCNQFII

2 ÞðCH3 CNÞ2ðsÞ;phase IV
TCNQF2ðMeCNÞ
the TCNQF0/1
process in acetonitrile (0.1 M Bu4NPF6) containing 1.0 mM TCNQF2
2
and 2.0 mM [Cu(CH3CN)4]+ at designated scan rates.
þ 2½CuðCH3 CNÞ4 þ þ 2CH3 CN þ 2e
ð14Þ
The scan rate dependence (Fig. 7) also illustrates the inter-rela-
tionship of the various electrocrystallized and solution soluble
the potential scan range from 600 mV to
600 mV. Using
materials. Notably, on increasing the scan rate, a decrease in the
a slow scan rate of 20 mV s
1, the reduction of TCNQF1
2(MeCN) to
Ox3 current is accompanied by the simultaneous loss of Ox1 and
TCNQF2(MeCN)2
is detected at
275 mV in the presence of 2.0 mM
at scan rates 200 mV s
1, only two (Ox4 and Ox3) major pro-
[Cu(CH3CN)4]+ (Fig. 6a), which occured at
448 mV in the absence
cesses are evident in cyclic voltammograms when the potential
of [Cu(CH3CN)4]+ (cf. Fig. 2). Thus, it is now much easier to reduce
is scanned in the positive direction. These results imply that
TCNQF1
2(MeCN). Furthermore, a sharp oxidation process (Ox3) is
CuI2(TCNQFII

2 )(CH3CN)2(s), phase III, effectively can be converted to
observed at 169 mV, on the reverse scan, which replaces the
CuITCNQFI
2 phase I in a solid state transformation reaction.
TCNQF2
/1

2(MeCN) oxidation reaction originally detected at
373 mV
(s),

(cf. Fig. 2). Again, this voltammetric behaviour supports formation

of an insoluble material accompanying the TCNQF1
/2
process. 3.3. Cyclic voltammetry of CuI2(TCNQFII

2 )(CH3CN)2(s) solid in the
2
The sharp oxidation process, seen at 169 mV (Ox3, in Fig. 6) has presence of [Cu(CH3CN)4]+
the characteristics of a ‘stripping process’, derived from the oxida-
tion of solid electrocrystallised copper(I)-TCNQF2
material. The In order to further probe the solid-solid transformation of
2
inferred material, CuI2(TCNQFII
CuI2(TCNQFII
I I

2 )(CH3CN)2 to Cu TCNQF2 cyclic voltammetry experiments
2 )(CH3CN)2 (assigned in this paper
as phase IV,2 would be formed by reaction of TCNQF2
were performed in CH3CN (0.1 M Bu4NPF6) containing a very high
2(MeCN) with
[Cu(CH3CN)4]+ on or near to the electrode surface, and then reoxi- 50.0 mM [Cu(CH3CN)4]+, but with solid CuI2(TCNQFII
2 )(CH3CN)2

dised to soluble TCNQF1
+ deposited on the electrode surface at
600 mV, prior to sweeping
2(MeCN) and [Cu(CH3CN)4] in the stripping
step (Ox3). In Fig. 6a, the shoulder present at 210 mV, following pro- the potential in the positive direction. Representative cyclic voltam-
cess Ox3, is the solution phase oxidation of TCNQF1
/0 mograms are shown in Fig. 8.
2(MeCN). The strip-
ping process observed at 348 mV corresponds to process Ox1 which When the potential was scanned from the initial value
involves oxidation of CuITCNQFI
solid to [Cu(CH3CN)4]+ and of
600 mV to 600 mV (Fig. 8a), the first scan is dominated by
2
TCNQF02(MeCN), as described in Section 3.1. The detection of Ox1 can three oxidation processes. On the cycling the potential, these
be attributed to CuI2(TCNQFII
processes were replaced by the relatively small, solution phase
2 )(CH3CN)2 (phase IV) solid being oxi-
dized to solid CuITCNQFI
TCNQF0/1
and TCNQF1
/2
diffusion controlled processes.
2 . Once again, on increasing the scan rate,
2 2

a new oxidation process (Ox4) appears at
192 mV (Fig. 6b), imply- However, when the potential is switched at 300 mV (Fig. 8b),
ing the existence of a second form of a CuI2TCNQFII
which is prior to TCNQF1
/0 oxidation, solid-solid transformation
2 based material,
2

as observed [11] for CuI2TCNQFII
of CuITCNQFI
I II

2 and Cu2(TCNQF2 )(CH3CN)2 is detected for several
4 (phase B or phase III in the termi-
nology used here, see below). In the investigation of CuI2TCNQFII
cycles of the potential, although the current magnitudes decrease
4 ,
the thermodynamically favoured material was assigned as as CuI2(TCNQFII

2 )(CH3CN)2 solid slowly dissolves. When the same

CuI2(TCNQFII
experiment as in Fig. 8b was achieved but in the absence of

4 )(CH3CN)2 (phase B) and the kinetically favoured phase
A was not unambiguously identified, but are named here as phases [Cu(CH3CN)4]+, no oxidative current could be detected because dis-
IV and phase III, respectively. Nethertheless, one should be cautious solution of CuI2(TCNQFII

2 )(CH3CN)2 solid is rapid.

in assigning these as phases, as they may well differ by solvation.

The mechanism for the reduction of TCNQF2 to TCNQF2
2 in the
presence of [Cu(CH3CN)4]+ is proposed to be as follows;
2
By analogy with similar studies with TCNQF4 and [Cu(CH3CN)4]+, this material is
assigned as the thermodynamically stable CuI2(TCNQF2II
)(CH3CN)2 (phase IV). The
kinetically favoured form is assigned a another phase of CuI2(TCNQF2II
)(CH3CN)2
(phase III), or the non-solvated CuI2TCNQF2II
. So, here describe these solids as phases
formed in the related electrochemistry the TCNQF4 in the presence of [Cu(CH3CN)4]+
(see ref. [11]).
3.4. Powder X-ray crystallography (XRD) of CuITCNQI

Unfortunately, extensive efforts to grow crystals of CuITCNQFI
2
suitable for single-crystal X-ray diffraction were unsuccessful.
However, an XRD powder pattern for the chemically synthesized
material was obtained as shown in Fig. 9. The CuITCNQFI
2 experi-
mental diffraction pattern displays broader peaks than for the cal-
culated [24] data for the coordination polymer CuITCNQI
phase II
[26]. However, their positions are in most cases almost identical to
96 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100

Fig. 6. Cyclic voltammograms obtained for 1.0 mM TCNQF2(MeCN) and 2.0 mM [Cu(CH3CN)4]+ in acetonitrile (0.1 M Bu4NPF6) with a 1.0 mm diameter GC electrode at a scan
rate of (a) 20 mV s
1 or (b) 200 mV s
1.

Fig. 7. Cyclic voltammograms obtained in acetonitrile (0.1 M Bu4NPF6) containing Fig. 9. X-ray diffraction powder pattern of chemically synthesized CuITCNQFI
2 (red)
1.0 mM TCNQF2 and 2.0 mM [Cu(CH3CN)4]+ at designated scan rates using a 1.0 mm compared with that calculated single crystal data from CuITCNQI
phase II (black)
diameter GC electrode. [8,18,24,27,28]. (For interpretation of the references to colour in this figure legend,
the reader is referred to the web version of this article.)

Fig. 8. Cyclic voltammograms obtained in acetonitrile (0.1 M Bu4NPF6) containing 50.0 mM [Cu(CH3CN)4]+ when CuI2(TCNQFII

2 )(CH3CN)2(s) is initially deposited onto a 3.0 mm
diameter GC electrode and the potential is scanned from (a)
600 to 600 mV and (b)
600 to 350 mV at a scan rate of 50 mV s
1.
 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100
those of CuITCNQI
phase II, which crystallized in a monoclinic unit
cell [26].
3.5. Spectroscopic characterization of CuITCNQFI

2
In order to obtain samples of CuITCNQFI
2 for spectroscopic
characterization, an ITO electrode was used to electrocrystallize
the solid by holding the potential for 10 min at
100 mV, which
lies between the TCNQF0/1
and TCNQF1
/2
processes. The cyclic
2 2
voltammetry was similar to that observed on GC, although the
much larger surface area contributed to larger currents, hence
larger IR drop. The surface-immobilised, dark blue crystals of
CuITCNQFI
2 , were washed with CH3CN and dried under vacuum
overnight prior to characterisation by vibrational spectroscopy.
Fig. 10 provides a comparison of the FTIR and Raman spectra of
TCNQF2 and electrocrystallized CuITCNQFI
2 . Bands that are charac-
teristic for the neutral TCNQ(F)n family have been identified
[10,29,30]. In the case of TCNQF2, the CAH stretch band is clearly
visible at 3064 cm
1 in the FTIR spectrum (Fig. 10a). Also, present
are other major IR bands for the C„N stretch at 2227 cm
1, ring
C@C stretch at 1574 cm
1, exocyclic C@C stretch at 1548 cm
1
and the C-F and ring CAC stretch at 1392 cm
1. The Raman spec-
trum for TCNQF2 (Fig. 10b) also provides characteristic bands for
the C„N stretch at 2230 cm
1, the ring C@C stretch at 1633 cm
1,
and the exocyclic C@C stretch at 1435 cm
1. The vibrational bands
observed for the neutral TCNQ(F)n family, generally shift to lower
energies for the TCNQF1
2 radical anion. Thus the IR bands associ-
ated with TCNQF1
2 in CuITCNQFI
2 are mmax/cm
1 2214, 2182
(CN), 1515 (ring C@C), 1477 (exocyclic C@C), 1339 (C-F and ring
CAC). The Raman bands are mmax/cm
1 2219 (CN), 1618 (ring
C@C) and 1401 (exocyclic C@C). In Fig. 11, the IR and Raman spec-
tra for the electrocrystallized and the chemically synthesized
material are compared and shown to be identical, within experi-
Fig. 10. Vibrational (a) FTIR and (b) Raman spectra of neutral TCNQF2 and chemically synthesized CuITCNQFI

97
mental uncertainty. Similarly good agreement was found in the
Raman spectra obtained from both the electrocrystallisation and
synthetic methods of preparation.
3.6. Elemental microanalysis of CuITCNQFI

2
The elemental analysis data of C = 46.76%, H = 0.82% and
N = 18.18% are consistent with the formulation being
CuTCNQF20.25H2O (calculated: C = 46.66%, H = 0.86%,
N = 18.54%). This suggests the chemically synthesized CuITCNQFI

2
contains a small amount of adventitious water.
3.7. Scanning electron microscopy of CuITCNQFI

2
The morphologies of the chemically synthesized and electro-
chemically crystallized (ITO) CuITCNQFI
2 samples were examined
using the SEM imaging technique (Fig. 12). Two morphologies
were observed with the electrochemically crystallized material
which was prepared from an acetonitrile (0.1 M Bu4NPF6) solution
containing 2.0 mM TCNQF2 and 10.0 mM [Cu(CH3CN)4]+ by reduc-
tive electrolysis at
100 mV (vs Ag/Ag+) for 10 min. The major
morphology consisted of block-shaped microcrystals of micron
size (Fig. 12a) and the minor morphology, clusters of rod-like
microcrystals (Fig. 12b). However, only block-shaped microcrystals
were observed for the chemically synthesized sample. The two
morphologies found by electrochemical crystallization could be
derived from the two phases of CuITCNQFI
2 identified by cyclic
voltammetry. The block-shaped morphologies of CuITCNQFI
2
(Fig. 12 a and c) are similar to the CuITCNQI
phase II formed by
extensive refluxing of CuITCNQI
phase I in acetonitrile [26,27].
These data, together with the very similar XRD powder
pattern for both materials further supports the assignment as
CuITCNQFI
2 phase II.
2 .
98 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100

Fig. 11. Comparison of vibrational (a) FTIR and (b) Raman spectra of chemically (black) synthesized and electrocrystallized (red) CuITCNQFI

2 onto an ITO electrode. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 12. Comparison of SEM images of CuITCNQFI

2 of electrocrystallized material onto an ITO electrode, showing two different locations (a) and (b) and chemically
synthesized CuITCNQFI

2 (c).

3.8. Solubility of CuITCNQFI

2 centrifugation leaving a clear solution that was diluted 1:60 with
The solubility of the chemically synthesized CuITCNQFI
2 solid
was determined in the presence and absence of 0.1 M Bu4NPF6
supporting electrolyte in acetonitrile. In the absence of electrolyte
these measurements used UV–Vis spectroscopy whereas the elec-
trochemical technique of steady state microelectrode voltammetry
was used for solutions containing electrolyte. To achieve a
saturated solution, CuITCNQFI
2 solid was dissolved in 2.0 ml of ace-
tonitrile and sonicated for 5 min. For UV–vis measurements, con-
taining no electrolyte, the excess solid was removed by
acetonitrile. The concentration of TCNQF1
2 was then determined
by UV–vis spectroscopy using the absorption band with kmax at
424 nm and reference to a calibration curve. For electrochemical
measurements, a saturated solution of CuITCNQFI
2 was prepared
in acetonitrile containing 0.1 M Bu4NPF6 and near steady state
voltammetry employing a 10 lm diameter Au microdisk electrode
used. No dilution was necessary for these measurements which
gave linear limiting currents versus concentration calibration data
for the TCNQF0/1

2 process. In this manner, the resulting concentra-
tion of TCNQF1
2 , hence the solubility of Cu TCNQF2
I I

was

Fig. 13. Current and potential versus time data obtained with a (a) CuITCNQFI
2
2 film on FTO with a cross-sectional area of 0.82 cm and a thickness of 840 nm, and (b) current
vs potential data calculated from data derived from (a).
 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100
determined to be 5.25 ± 0.13  10
5 M in the absence of electrolyte
(UV–vis spectroscopy) and 9.05 ± 0.09  10
5 M in the presence of
electrolyte (steady state voltammetry). Thus, the solubility prod-
ucts were calculated to be 2.75 ± 0.21  10
9 M2 and 8.19 ± 1.02
 10
9 M2 in the absence and presence of 0.1 M Bu4NPF6 elec-
trolyte, respectively. These compare with the slightly more soluble
CuITCNQFI
4 , which gave solubility products of 1.30 ± 0.07 
10
8 M2 in the absence and 3.81 ± 0.25  10
8 M2 in the presence
of electrolyte. In the case of CuITCNQI
values are 1.96  10
8 M2
and 4.9  10
7 M2 in acetonitrile with and without electrolyte,
respectively [25].
I
3.9. Conductivity of Cu TCNQFI

2 Electroanal. Chem. Interfacial Electrochem. 247 (1) (1988) 193–202.
The linear dependence of current on potential obtained with a
CuITCNQFI
2 film on a FTO-coated glass is shown in Fig. 13. The
DC resistivity of CuITCNQFI

2 bulk material was determined from
the slope of the plot to give an average conductivity of 6.0 
10
6 S cm
1 which lies at the low end of the semi-conducting
range. These data were corrected for the underlying conductivity
of the FTO glasses and employed several measurements with sam-
ples having an average thickness of 840 nm. The conductivity of

5
the CuITCNQFI
2 samples is slightly lower than that of 1.3  10
S cm
1 found for CuITCNQI
phase II [26].
4. Conclusion
CuITCNQFI
I II

2 and Cu2(TCNQF2 )(CH3CN)2 are electrocrystallized
in voltammetric experiments when TCNQF2 is reduced to TCNQF1
2
and TCNQF2
+
2 , respectively, in the presence of [Cu(CH3CN)4] . The
existence of two forms of both materials is revealed by variable scan
rate cyclic voltammetry.
CuITCNQFI
2 was both chemically and electrochemically
synthesized and characterised using spectroscopic and imaging
methods. The XRD powder pattern of CuITCNQFI

2 is similar to
CuITCNQFI
phase II which has a monoclinic unit cell. The conduc-
tivity determined from a film of CuITCNQFI
2 indicates that it is in
the semiconducting range. The vibrational spectroscopies of
CuITCNQFI
2 materials synthesized by either electrochemical or
chemical methods are identical implying that the same material
is generated via either method.
Acknowledgements
Financial support from the Australian Research Council
(DP120101066 and DP170103477) to L.L.M. and A.M.B. is gratefully
acknowledged. We thank Associate Professor Brendan Abrahams
and his group at the University of Melbourne for measurement
and calculated X-ray powder diffraction data. N.T.V expresses her
appreciation for the award of MGRS and MIPRS Monash University
scholarships.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.ica.2018.04.010.
References
[1] N. Lopez, H. Zhao, A.V. Prosvirin, W. Wernsdorfer, K.R. Dunbar, A homologous
heterospin series of mononuclear lanthanide/TCNQF4 organic radical
complexes, Dalton Trans. 39 (18) (2010) 4341–4352.
[2] A. Nafady, A.P. O’Mullane, A.M. Bond, Electrochemical and photochemical
routes to semiconducting transition metal-tetracyanoquinodimethane
coordination polymers, Coord. Chem. Rev. 268 (2014) 101–142.
[3] A. Nafady, A.M. Bond, A. Bilyk, A.R. Harris, A.I. Bhatt, A.P. O’Mullane, R. De
Marco, Tuning the electrocrystallization parameters of semiconducting Co
99
[TCNQ](2)-based materials to yield either single nanowires or crystalline thin
films, J. Am. Chem. Soc. 129 (8) (2007) 2369–2382.
[4] A. Nafady, A.P. O’Mullane, A.M. Bond, A.K. Neufeld, Morphology changes and
mechanistic aspects of the electrochemically-induced reversible solid-solid
transformation of microcrystalline TCNQ into Co TCNQ (2)-based materials
(TCNQ = 7,7,8,8-tetracyanoquinodimethane), Chem. Mater. 18 (18) (2006)
4375–4384.
[5] H.L. Peng, C.B. Ran, X.C. Yu, R. Zhang, Z.F. Liu, Scanning-tunneling-microscopy
based thermochemical hole burning on a new charge-transfer complex and its
potential for data storage, Adv. Mater. 17 (4) (2005) 459–464.
[6] C. Ran, H. Peng, W. Zhou, X. Yu, Z. Liu, Thermochemical hole burning on a series
of N-substituted Morpholinium 7,7,8,8-tetracyanoquinodimethane charge-
transfer complexes for data storage, J. Phys. Chem. B 109 (47) (2005) 22486–
22490.
[7] A. Yasuda, J.E. Seto, Electrochromic properties of vacuum evaporated organic
thin films: part I. Electrochromic thin films of the three primary colors, J.
[8] N.L. Haworth, J. Lu, N. Vo, T.H. Le, C.D. Thompson, A.M. Bond, L.L. Martin,
Diagnosis of the redox levels of TCNQF4 compounds using vibrational
spectroscopy, ChemPlusChem 79 (7) (2014) 962–972.
[9] T.H. Le, J. Lu, A.M. Bond, L.L. Martin, Identification of TCNQF4 redox levels
using spectroscopic and electrochemical fingerprints (TCNQF = 2,3,5,6-
tetrafluoro-7,7,8,8- tetracyanoquinodimethane), Inorg. Chim. Acta 395
(2013) 252–254.
[10] T.H. Le, A. Nafady, A.M. Bond, L.L. Martin, Electrochemically directed synthesis
of Co2+ and Ni2+ complexes with TCNQF2
4 (TCNQF4 = 2,3,5,6-Tetrafluoro-
7,7,8,8-tetracyanoquinodimethane), Eur. J. Inorg. Chem. 2012 (33) (2012)
5534–5541.
[11] T.H. Le, A. Nafady, N.T. Vo, R.W. Elliott, T.A. Hudson, R. Robson, B.F. Abrahams,
L.L. Martin, A.M. Bond, Electrochemically directed synthesis of CuI2(TCNQFII4 –)
(MeCN)2 (TCNQF4 = 2,3,5,6-Tetrafluoro-7,7,8,8-tetracyanoquinodimethane):
voltammetry, simulations, bulk electrolysis, spectroscopy, photoactivity, and
X-ray crystal structure of the CuI2(TCNQFII4 –)(EtCN)2 analogue, Inorg. Chem. 53
(6) (2014) 3230–3242.
[12] T.H. Le, A.P. O’Mullane, L.L. Martin, A.M. Bond, A new method for
electrocrystallization of AgTCNQF4 and Ag2TCNQF4 (TCNQF4 = 2,3,5,6-
tetrafluoro-7,7,8,8-tetracyanoquinodimethane) in acetonitrile, J. Solid State
Electrochem. 15 (11) (2011) 2293–2304.
[13] A. Nafady, T.H. Le, N. Vo, N.L. Haworth, A.M. Bond, L.L. Martin, Role of water in
the dynamic disproportionation of Zn-based TCNQ(F4) coordination
polymers (TCNQ = Tetracyanoquinodimethane), Inorg. Chem. 53 (4) (2014)
2268–2275.
[14] M.C. Grossel, A.J. Duke, D.B. Hibbert, I.K. Lewis, E.A. Seddon, P.N. Horton, S.C.
Weston, An investigation of the factors that influence the decomposition of
7,7‘,8,8‘-tetracyanoquinodimethane (TCNQ) and its salts to, and structural
characterization of, the a, a-Dicyano-p-toluoylcyanide Anion, Chem. Mater. 12
(8) (2000) 2319–2323.
[15] A. Lombardo, T.R. Fico, Mechanism of conversion of 1:1 anion-radical salts of
tetracyano-p-quinodimethane to their 1:2 analogs, J. Org. Chem. 44 (2) (1979)
209–212.
[16] M.R. Suchanski, R.P. Van Duyne, Resonance Raman Spectroelectrochemistry. 4.
Oxygen decay chemistry of tetracyanoquinodimethane dianion, J. Am. Chem.
Soc. 98 (1) (1976) 250–252.
[17] T.J. Emge, F.M. Wiygul, J.S. Chappell, A.N. Bloch, J.P. Ferraris, D.O. Cowan, T.J.
Kistenmacher, Crystal structures for the electron donor
dibenzotetrathiafulavalene, DBTTF, and its mixed-stack charge-transfer salts
with the electron acceptors 7,7,8,8-tetracyano-p-quinodimethane, TCNQ, and
2,5-difluoro-7,7,8,8-tetracyano-p-quinodimethane, 2,5-TCNQF2, Mol. Cryst.
Liq. Cryst. 87 (1–2) (1982) 137–161.
[18] J. Lieffrig, O. Jeannin, T. Guizouarn, P. Auban-Senzier, M. Fourmigué,
Competition between the C-H. . .N hydrogen bond and C–I. . .N Halogen Bond
in TCNQFn (n = 0, 2, 4) salts with variable charge transfer, Cryst. Growth Des.
12 (8) (2012) 4248–4257.
[19] H. Miyasaka, N. Motokawa, S. Matsunaga, M. Yamashita, K. Sugimoto, T. Mori,
N. Toyota, K.R. Dunbar, Control of charge transfer in a series of Ru2II, II/TCNQ
two-dimensional networks by tuning the electron affinity of TCNQ units: a
route to synergistic magnetic/conducting materials, J. Am. Chem. Soc. 132 (5)
(2010) 1532–1544.
[20] N. Vo, N.L. Haworth, A.M. Bond, L.L. Martin, Investigation of the redox and
acid-base properties of TCNQF and TCNQF2: electrochemistry, vibrational
spectroscopy, and substituent effects, ChemElectroChem (2018), https://doi.
org/10.1002/celc.201701387.
[21] T. Emge, M. Maxfield, D. Cowand, T.J. Kistenmacher, Mol. Cryst. Liq. Cryst. 65
(1981) 161.
[22] M.F. Wiygul, J.P. Ferraris, T.J. Emge, T.J. Kistenmacher, Crystal, molecular and
electronic structure of the electron acceptor 2,5-difluoro-7,7,8,8-Tetracyano-
p-quinodimethane, 2,5-TCNQF2, Mol. Cryst. Liq. Cryst. 78 (1) (1981) 279–293.
[23] T. Murata, G. Saito, K. Nishimura, Y. Enomoto, G. Honda, Y. Shimizu, S. Matsui,
M. Sakata, O.O. Drozdova, K. Yakushi, Complex formation between a
nucleobase and tetracyanoquinodimethane derivatives: crystal structures
and transport properties of charge-transfer solids of cytosine, Bull. Chem.
Soc. Jpn. 81 (3) (2008) 331–344.
[24] C.F. Macrae, I.J. Bruno, J.A. Chisholm, P.R. Edgington, P. McCabe, E. Pidcock, L.
Rodriguez-Monge, R. Taylor, J. van de Streek, P.A. Wood, Mercury CSD 2.0 –
new features for the visualization and investigation of crystal structures, J.
Appl. Cryst. 41 (2008) 466–470.
100 N.T. Vo et al. / Inorganica Chimica Acta 480 (2018) 91–100
[25] A.R. Harris, A.K. Neufeld, A.P. O’Mullane, A.M. Bond, R.J.S. Morrison,
Voltammetric, EQCM, spectroscopic, and microscopic studies on the
electrocrystallization of semiconducting, phase I, CuTCNQ on carbon, gold,
and platinum electrodes by a nucleation-growth process, J. Electrochem. Soc.
152 (9) (2005) C577–C583.
[26] R.A. Heintz, H.H. Zhao, O.Y. Xiang, G. Grandinetti, J. Cowen, K.R. Dunbar, New
insight into the nature of Cu(TCNQ): solution routes to two distinct
polymorphs and their relationship to crystalline films that display bistable
switching behavior, Inorg. Chem. 38 (1) (1999) 144–156.
[27] A.K. Neufeld, I. Madsen, A.M. Bond, C.F. Hogan, Phase, morphology, and particle
size changes associated with the solid-solid electrochemical interconversion of
TCNQ and semiconducting CuTCNQ (TCNQ = tetracyanoquinodimethane),
Chem. Mater. 15 (19) (2003) 3573–3585.
[28] A.P. O’Mullane, A.K. Neufeld, A.M. Bond, Distinction of the two phases of
CuTCNQ by scanning electrochemical microscopy, Anal. Chem. 77 (17) (2005)
5447–5452.
[29] R. Bozio, A. Girlando, C. Pecile, Vibrational analysis of spectra of quinonoid
molecular ions. Part 3. -Vibrational spectra and assignment of 7,7,8,8-
tetracyanoquinodimethane radical anion, J. Chem. Soc., Faraday Trans. 2:
Mol. Chem. Phys. 71 (1975) 1237–1254.
[30] T. Takenaka, Molecular vibrational spectra of tetracyanoquinodimethane and
tetracyanoquinodimethane-d4 crystals, Spectrochim. Acta Part A: Mol.
Spectrosc. 27 (9) (1971) 1735–1752.