Electrochemical studies of heterogeneous reduction of tetracyanoquinodimethane in poly(ethylene oxide) electrolytes using ac impedance and cyclic voltammetry at an ultramicroelectrode

Article abstract of DOI:10.1039/b004235h

Electrochemical studies of the TCNQ⁰/TCNQ^- couple have been carried out using ac impedance spectroscopy and cyclic voltammetry at platinum ultramicroelectrodes (UME). Liquid poly(ethylene oxide) (PEO) CH₃-O-(CH₂-CH₂-O4)-CH₃ has been used as the solvent with different concentrations of the TCNQ⁰/TCNQ^- couple and LiClO₄ as the supporting electrolyte. On the basis of the ac impedance results at the UME it has been found that the double layer capacitance and standard heterogeneous rate constant are independent of the presence of electroactive species and supporting electrolyte concentrations, indicating that adsorption of electroactive species onto the electrode is not significant. The standard heterogeneous rate constant ks was found to be 0.109 ± 0.005 cm s^-1. A similar value of k(s) = 0.094 cm s^-1 was obtained for the second reduction step TCNQ^-/TCNQ^2-. Diffusion coefficients of both TCNQ and TCNQ^- are equal, D = 1.0 x 10^-6 cm² s^-1 for a 0.5 M LiClO₄ solution. Higher diffusion coefficients are obtained in less concentrated supporting electrolyte. Comparison is made between these results and those reported previously for PEO-400 HO-(CH₃CH₂O)₈-OH. The different end groups significantly influence the viscosity and hence k(s).

Full text of DOI:10.1039/b004235h

View Article Online / Journal Homepage / Table of Contents for this issue
Electrochemical studies of heterogeneous reduction of
tetracyanoquinodimethane in poly(ethylene oxide) electrolytes using
ac impedance and cyclic voltammetry at an ultramicroelectrode
P. ¢os,a G. Zabinska,b A. Kisza,b L. Christie,c A. Mountd and P. G. Brucec
a L aboratory of Electroanalytical Chemistry, School of Medicine, 53-128 W roc¡aw, ul. Szewska
38, Poland
b Faculty of Chemistry, University of W roc¡aw, W roc¡aw, ul. Joliot-Curie 14, Poland
c School of Chemistry, University of St. Andrews, St. Andrews, UK
d Department of Chemistry, University of Edinburgh, Edinburgh, UK
Received 26th May 2000, Accepted 4th October 2000
First published as an Advance Article on the web 9th No¿ember 2000
Electrochemical studies of the TCNQ0/TCNQ~ couple have been carried out using ac impedance
spectroscopy and cyclic voltammetry at platinum ultramicroelectrodes (UME). Liquid poly(ethylene oxide)
(PEO) CH ÈOÈ(CH ÈCH ÈO) ÈCH has been used as the solvent with di†erent concentrations of the
3
2
2
4
3
TCNQ0/TCNQ~ couple and LiClO as the supporting electrolyte. On the basis of the ac impedance results at
4
the UME it has been found that the double layer capacitance and standard heterogeneous rate constant are
independent of the presence of electroactive species and supporting electrolyte concentrations, indicating that
adsorption of electroactive species onto the electrode is not signiÐcant. The standard heterogeneous rate
constant k was found to be 0.109 ^ 0.005 cm s~1. A similar value of k \ 0.094 cm s~1 was obtained for the
s
s
second reduction step TCNQ~/TCNQ2~. Di†usion coefficients of both TCNQ and TCNQ~ are equal,
D \ 1.0 ] 10~6 cm2 s~1 for a 0.5 M LiClO solution. Higher di†usion coefficients are obtained in less
concentrated supporting electrolyte. Comparison is made between these results and those reported previously
4
for PEO-400 HOÈ(CH CH O) ÈOH.1 The di†erent end groups signiÐcantly inÑuence the viscosity and hence
3
2
8
k .
s
viscous solvent as PEO. It should be noted that this couple
was used in the past to investigate the inÑuence of solvent
relaxation dynamics on the homogeneous charge transfer
process in a non-aqueous solvent.14h18
Introduction
Interest in polymer electrolytes is growing rapidly because of
their potential applications in batteries, fuel cells, sensors, etc.
As a result, understanding their fundamental electrochemistry
is becoming of considerable importance.1h13 Central to this is
a better understanding of the nature of redox reactions in
these solid solvents.
Polymer electrolytes may be prepared with the same chemi-
cal composition but very di†erent molar mass, ranging from
low molecular weight liquids through to solids with molar
masses of several million. The dynamics of the solvent varies
signiÐcantly with molar mass and as a result polymer electro-
lytes o†er an unprecedented medium for the study of the inÑu-
ence of solvent dynamics on the kinetics of electron transfer.
This is a problem of wide scientiÐc importanceÈsuch pro-
cesses lie at the core of many chemical, physical and biological
phenomena. Due to the importance of the inÑuence of solvent
molar mass on the standard rate constant, such studies have
been also reported previously.1h6,8h13 The understanding,
however, of this phenomena is far from complete.
The purpose of the present paper is to report on the basic
electrochemistry of the TCNQ0/TCNQ~ couple in the low
molecular weight PEO CH OÈ(CH CH O) ÈCH (tetramer).
3
2
2
4
3
We expect that a credible redox system for the study of the
electron transfer in viscous solvents can be established as a
result of our study. PEO is normally terminated by OH
groups. In the case of a low molecular weight liquid this
becomes a signiÐcant proportion of the solvent and can
greatly a†ect the solvent dynamics, coordination of the salt
and other properties. CH -termination ensures that the low
3
molecular weight ethers are chemically similar to high molec-
ular weight PEO. A previous study of TCNQ in low molecu-
lar weight PEO has employed an OH terminated solvent.1
Murray and co-workers have studied the kinetics of electronic
self-exchange (homogeneous e~ transfer) between TCNQ0 and
TCNQ~.12,13
In the present paper we have carried out investigations of
heterogeneous charge transfer process at solid Pt ultramicro-
electrodes (UME) in tetramer as a function of the TCNQ0/
TCNQ~ and supporting electrolyte concentrations.
Steady-state voltammetry and ac impedance methods have
been applied. Ac impedance measurements at the UME
proved to be a very e†ective method for studying the mecha-
nism and kinetics of electrochemical processes in aqueous,
non-aqueous and polymer electrolytes.4,19h21 It should be
stressed that the results presented in this work were obtained
using an argon Ðlled Mbraun glove box and distilled PEO.
Poly(ethylene oxide) (PEO)È(CH ÈCH ÈO) È has proved to
2
2
n
be a particularly e†ective solvent for polymer electrolytes. Dis-
solution of salts in this solvent is driven by the coordination
of the cations by the ether oxygens. As a result solubility is
high for salts based on monoatomic cations but lower for
polyatomic cations. A suitable redox system to study electron
transfer should be the anionic couple TCNQ0/TCNQ~ in
which both oxidised and reduced forms are soluble in PEO.
Usage of this couple avoids the complications of metal ions
and solubility in testing electron transfer models in such a
DOI: 10.1039/b004235h
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454
This journal is ( The Owner Societies 2000
5449

View Article Online
These procedures have been employed to minimise the inÑu-
ence of the impurities on the charge transfer kinetics studies. It
is well known (and unfortunately not sufficiently appreciated)
that the presence of impurities such as water and oxygen can
signiÐcantly inÑuence the kinetic results of solid electrodes in
non-aqueous solvents.
mental results. We have shown,4,19h21 that this equivalent
circuit does represent the experimental data, but the analysis
of such data may be greatly facilitated by using an equivalent
circuit which is appropriate to a hemispherical electrode (Fig.
1). In this case the di†usional impedance of the microdisc is
represented by a parallel combination of a Warburg imped-
ance, Z , and resistance, R , which is related to non-linear
w
nl
(hemispherical) di†usion. The appropriate equations describ-
Experimental
ing both the Z and the R can be found in the liter-
w
nl
ature.4,19h22 At high frequencies the ac impedance response of
the microdisc is dominated by the charge transfer resistance,
R , and double layer capacitance, C , and, as a consequence,
Chemicals
Tetraethyleneglycol
dimethyl
ether
ct
dl
a semicircle may be observed in the complex plane. At these
frequencies, the di†usion is generally linear and semi-inÐnite,
[CH ÈOÈ(CH ÈCH ÈO) ÈCH ], abbreviated here to tetra-
3
2
2
4
3
mer (Aldrich 99%), was puriÐed by vacuum distillation at
10~3 mbar and 147 ¡C using a Fischer HMS 500C distillation
apparatus with 90 theoretical plates. 7,7,8,8-Tetra-
cyanoquinodimethane (TCNQ, Aldrich 98%) was puriÐed by
sublimation under vacuum at 258 ¡C.
corresponding to a Warburg response, Z . At low frequencies,
w
and in contrast to the situation with electrodes of normal
dimensions, the di†usion becomes spherical (non-linear). Of
course the relative values of the electrochemical parameters
inÑuence the overall shape of the impedance. When R is
Lithium tetracyanoquinodimethane (LiTCNQ) was pre-
pared by the reaction of puriÐed TCNQ and lithium iodide in
high purity acetonitrile. The purple crystals obtained were Ðl-
tered o† and washed with acetonitrile until they became
bright green in colour. The solid was then washed with a large
volume of ether.
ct
large, the high frequency semicircle dominates since the charge
transfer kinetics largely control the electrochemical response.
On the other hand, when R is small, the di†usional imped-
ct
ance dominates and the low frequency di†usional quarter-
circle becomes prominent. However, even in this case, the
electrochemical parameter k (standard heterogeneous charge
Lithium perchlorate (LiClO , Aldrich 99]%) was dried by
s
4
transfer constant) makes a signiÐcant contribution to the
heating the salt at 160 ¡C for 48 h under vacuum.
value of the impedance in at least part of the frequency range.
The shape of the di†usional part of the Nyquist graph
depends on the non-dimensional frequency: a2u/D, where a is
the ultramicroelectrode radius, u angular frequency and D is
the di†usion coefficient of the electroactive species. It can be
shown that for high values of UME radius and low values of
the di†usion coefficient the inÑuence of non-linear di†usion is
not dominant and the steady-state impedance cannot be
achieved for frequencies as low as several mHz.
The experimentally recorded impedance spectra are always
inÑuenced by unwanted elements such as outer inductance
(which is negligible) and electrolyte resistance. Their values
were removed prior to the evaluation of the charge transfer
resistance and di†usion impedance. As seen in Fig. 2, where an
example of a Ðt obtained using BoukampÏs23 program is pre-
sented as a Bode plot, the Ðt is excellent in the whole fre-
quency range from 1 Hz to 100 kHz. It should be added that
in the present work the inÑuence of the outer inductance can
be neglected and consequently it is not presented in Fig. 1.
After removal from the recorded impedance spectra of the
contribution of the unwanted outer elements their remaining
part was worked out in terms of a simple model consisting of
a single charge transfer step with di†usion described by equa-
tions derived for the hemispherical approximation of the
UME by Los and Bruce19h21 who showed that the approx-
imation is valid over the whole frequency range used in this
All samples were transferred to an argon Ðlled Mbraun
glove box for storage. Typically the water and oxygen levels
inside the box were less than 0.05 ppm.
Electrodes and cells
Electrochemical measurements were performed using a stan-
dard two-electrode cell. The working electrode was a platinum
microdisc with a diameter of either 10 or 20 lm (Rudawski
Ventures, Poland). Before each measurement the working elec-
trode was polished using 0.05 lm alumina powder and rinsed
using tetramer. The counter/reference electrode consisted of a
coiled platinum wire immersed in an equimolar tetramer solu-
tion of 5 mmol dm~3 of both LiTCNQ and TCNQ with 0.5
M LiClO . The body of this electrode consisted of a glass
4
tube closed at one end by a Vicor glass frit attached with heat
shrinking TeÑon. The electrochemical cell was placed inside a
Faraday cage, which in turn was situated inside the glove box.
Apparatus and procedures
Experiments were carried out on equimolar solutions of
TCNQ and LiTCNQ in the concentration range 0.1È
10 ] 10~3 mol dm~3, with the concentration of the support-
ing electrolyte (LiClO ) being either 0.1 or 0.5 mol dm~3.
4
Linear sweep voltammetry, in the range ]0.3 V to [0.3 V,
was carried out using an Autolab PGSTAT-10 (Eco Chemie)
with a low current (ECD) module.
For ac impedance spectroscopy measurements the working
electrode was connected to the input of a EG&G PAR pre-
ampliÐer 181, the output of which was connected directly to
the input of a Solatron 1255 Frequency Response Analyser.
The generator output of the 1255 was connected to both the
voltage input and the counter electrode. Coaxial cables were
used to connect the EG&G preampliÐer to the 1255 and to
the cell, the sheaths of the coaxial cables were grounded. Mea-
surements were under the control of an IBM PC. An ac signal
of between 5 and 10 mV rms was employed and data was
collected in the frequency range 0.01 Hz to 250 kHz.
Analysis of ac impedance results
Fig. 1 Equivalent circuit used for analysis of the experimental ac
impedance results: R Èthe electrolyte resistance, R Èthe charge
The theoretical analysis of the di†usion process to a microdisc
has been given by Fleischmann and Pons22 who proposed an
equivalent circuit, which can be used to analyse the experi-
s
ct
transfer resistance, C Èthe double layer capacitance, Z ÈWarburg
dl
w
impedance, R Èthe resistance due to the non-linear di†usion.
nl
5450
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454

View Article Online
Fig. 4 Experimental CV curves obtained in 5 mM TCNQ/LiTCNQ
in tetramer with 0.5 M LiClO solution at platinum UMEs of 20 and
10 lm radius; sweep rate 0.05 V s~1.
Fig. 2 The ac impedance data obtained in 10 mM TCNQ/LiTCNQ
4
in tetramer solution with 0.5 M LiClO at a platinum UME of 20 lm
4
radius using BoukampÏs program.23 (L) experimental results, (])
NLLSFit.
coefficients of the oxidised and reduced forms are approx-
imately equal.
A value of 1.08 ^ 0.02 ] 10~6 cm2 s~1 (the average of 12
cyclic voltammetry measurements) was calculated for the
work. An example of such a NLLSFit obtained by a special
FORTRAN program24 is presented in Fig. 3.
TCNQ0/TCNQ~ di†usion coefficients in 0.5 M LiClO ,
4
based on the equation for steady-state current (I ) at an
Results and discussion
svs
ultramicrodisc electrode, I \ 4nFDCa, where a is the radius
svs
Fig. 4 presents voltammetry curves obtained at a sweep rate of
50 mV s~1 using 10 and 20 lm Pt UMEs and 5 mM TCNQ/
TCNQ~.
Two reduction waves are observed with approximately the
same heights for both electrodes. They are associated with the
following two one-electron reduction steps:12,13
of the UME, C the concentration of TCNQ0 and TCNQ~
and D the di†usion coefficient. The geometric radius of the
electrode was used to calculate the di†usion coefficients. As
might be expected, the steady-state response is independent of
sweep rate and the steady-state current is proportional to the
radius of UME. The di†usion coefficient is around 30%
higher in 0.1 M LiClO electrolyte (Fig. 5), which may be
explained by the lower viscosity of the more dilute electro-
lytes.
TCNQ0 ] 1e~ \ TCNQ~, E¡@(1) \ 0 V
(1)
(2)
4
TCNQ~ ] 1e~ \ TCNQ2~, E¡@(2) \ [0.55 V
The kinetics of the Ðrst reduction step has been studied by
ac impedance measurements. Fig. 6 presents ac impedance
responses with and without electroactive species.
With only supporting electrolyte present in the solution a
straight, almost vertical line is observed which can be
described by an equivalent circuit consisting of the solution
resistance R in series with the double layer capacitance C .
The separation of the potentials for the two reduction steps,
E¡@(1) [ E¡@(2) \ 0.55 V, is the same as that reported pre-
viously for TCNQ in non-aqueous and solid polymer electro-
lytes. A steady-state response is obtained at the 10 lm
electrode whereas planar di†usion contributes to the response
observed at the 20 lm UME. An almost ideal steady-state
response is observed for the TCNQ0/TCNQ~ couple within
the range 0.1È10 mM. The steady-state current is proportional
to the TCNQ0/TCNQ~ concentration and the approximately
equal magnitudes of the two waves indicate that the di†usion
s
dl
This description can be improved by replacing C with a con-
dl
stant phase element (CPE).25 The impedance of CPE is equal
to 1/[A( ju)Õ] where j \ J[1, A is a constant, / an exponent
\1 and u is the angular frequency of the ac signal. The devi-
ation from an ideal capacitor arises from the roughness of the
Fig. 3 The ac impedance data obtained in 10 mM TCNQ/LiTCNQ
in tetramer with 0.5 M LiClO solution at a platinum UME of 20 lm
Fig. 5 Experimental CV curves obtained in 5 mM TCNQ/LiTCNQ
4
radius. McDonaldÏs CNLS procedure24 was used: (L) experimental
in tetramer with 0.1 and 0.5 M LiClO solutions at a platinum UME
4
results, (]) NLLSFit.
of 10 lm radius, sweep rate 0.01 V s~1.
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454
5451

View Article Online
Fig. 7 The dc potential dependence of the ac impedance results
obtained in 10 mM TCNQ/LiTCNQ solution in tetramer with 0.5 M
LiClO at a platinum UME of 10 lm radius.
4
only two adjustable parameters are used:
C
and k .
dl
s
However, it is recommended to verify ac impedance measure-
ments using adjustable parameters (in this case D and a)
which can be determined by other methods. Table 1 presents
the actual exemplary values of these parameters obtained by
the Ðt procedures. It should be noticed that relatively low
values (lower than 5%) of the standard deviations of the
parameters were obtained. This proves the very good quality
of the Ðt and directly conÐrms the appropriate choice of the
equivalent circuit.
Fig. 6 Experimental ac impedance results obtained at (a) Pt UME of
10 lm radius; (b) Pt UME of 20 lm radius; in tetramer with 1 mM
Double layer capacitance values ranging from 20 to 40 lF
cm~2 have been obtained. Given that the measurements in the
absence of the redox couple showed that the surface is not
TCNQ/LiTCNQ with 0.5 M LiClO and in tetramer with 0.5 M
4
LiClO .
4
perfectly smooth, some variations in C extracted from the ac
dl
data obtained in the presence of the redox couple is to be
expected. The error reÑects di†erent C values (di†erent
dl
roughnesses) for di†erent electrodes, but the error on any one
electrode surface. The expression for the CPE can be related
to the fractal dimension of the surface.25 In the system studied
here, the near vertical nature of the line, corresponding to an
exponent of 0.97, implies a relatively smooth surface. Because
of this we shall treat the interface as a capacitor, the value of
which varies from 15 to 40 lF cm~2. With the electroactive
couple present in the solution the ac impedance response is
typical for the ultramicrodisc, as seen in Fig. 6. The di†usional
electrode is much smaller. It is clear from examination of the
values given in Table 2 that di†usion coefficients for the elec-
troactive species in 0.1 M LiClO solutions are higher than in
4
0.5 M LiClO electrolytes. This is consistent with the results
4
described above for the steady-state measurements.
It should be noted that the values for D were obtained, in
the case of the ac data, using the Ðtted area of the electrode
whereas the geometrical area was used in the analysis of the
CV data. This may account for the small discrepancy in the
absolute values of D measured by the two methods.
impedance, which consists of parallel combination of Z and
w
R , is represented by a quarter-circle on the complex plane.
nl
Fig. 6 shows that the uncompensated solution resistance R is
small in comparison to the faradaic impedance. Additionally,
s
in all cases R and R are in the same range (D0.1È1.0 ) cm2),
s
ct
showing high accuracy in the ac impedance evaluation of the
standard heterogeneous charge transfer constant, k .
s
According to the theory of ac impedance a minimum value
of the charge transfer resistance and di†usional impedance
should be obtained at the half-wave potential (E¡@(1) \ 0.00
V), assuming a simple charge transfer mechanism. Fig. 7
shows the dc potential dependence of the ac impedance
response.
A
minimum impedance is obtained at dc potentials
approaching 0.00 V. The experimental curves have been
analysed according to a simple mechanism consisting of a
single charge transfer step with di†usion. Fig. 8 shows that the
model described by equations derived by Los and Bruce19
provides a good Ðt to the data.
The Ðt has been carried out using four adjustable param-
eters: double layer capacitance C , heterogeneous charge
dl
transfer constant k , di†usion coefficient D and radius a of the
s
Fig. 8 The ac impedance data obtained in 10 mM TCNQ/LiTCNQ
UME. Very similar numeric results for the Ðt can be obtained
solution in tetramer with 0.5 M LiClO at a platinum UME of 20 lm
4
when using the voltammetric di†usion coefficients and geo-
metric radius of the electrodes as Ðxed parameters. In this case
radius. The equivalent circuit derived by Los and Bruce19 was used
(explanation in text): (L) experimental results, (]) NLLSFit.
5452
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454

View Article Online
Table 1 Parameters obtained by the Ðt procedurea
LiClO /M
TCNQ0/TCNQ~/mM
Dc pot./mV
C /lF cm~2
D/1.0 ] 106 cm2 s~1
r/lm
k /cm s~1
4
dl
s
0.1
0.5
0.5
0.5
0.5
5
5
10
10
10
0
0
0
32.6
27.6
42.1
35.0
24.8
1.07
0.66
0.57
0.50
0.52
11.6
11.5
12.2
12.4
11.5
0.103
0.138
0.122
0.139
0.102
[20
]20
a All values presented are actual exemplary results of the Ðt to the experimental ac impedance results. The quality of the Ðt is described by the
standard deviations of these parameters, which do not exceed 5%.
Table 2 Average values and the averagesÏ standard deviations of the parameters determined by the Ðt to the experimental results
Conc. of LiClO /M
k /cm s~1
C /lF cm~2
D/1.0 ] 106 cm2 s~1
4
s
dl
0.1
0.5
0.098 ^ 0.022
0.112 ^ 0.027
29.6 ^ 8.6
0.98 ^ 0.08a
0.75 ^ 0.05b
20.9 ^ 10.8
a Seven measurements at dc potentials ranging from [10 to ]10 mV. b Nineteen measurements at dc potentials ranging from [40 to ]40 mV.
Table 2 shows that double-layer capacitance and standard
heterogeneous rate constant are independent of the support-
ing electrolyte concentration within experimental error.
The electrochemical reduction of TCNQ is a relatively fast
reaction, consistent with observations of the homogeneous
self-exchange rate constants for that couple (around 1.0 ] 109
M~1 s~1 in a non-aqueous solvent of low viscosity such as
acetonitrile12,13). The values of the double-layer capacitance
are slightly higher than for pure supporting electrolyte but
they show no evidence of adsorption of the reactants.
some 20-fold higher (0.889 P) than that of tetramer. The inÑu-
ence of q is particularly strong when comparing these two
L
solvents because of the similarity of the dielectric constant and
hence the Pekar factors. These correlations ignore the speciÐc
chemical di†erences between the chain ends in the two sol-
vents. The OH groups are likely to bind preferentially to the
Li` ions and it is known that electron transfer rates can be
di†erent in solvents containing OH groups compared with
those that do not.6,8,12
Using the same analysis the following kinetic parameters
Conclusions
have been obtained in 0.5 M LiClO for the second reduction
4
step described by reaction (2): k \ 0.094 cm s~1,
TCNQ0/TCNQ~ appears to be a suitable couple for probing
solvent relaxation in PEO, as it appears to show no com-
plications due to complexation or adsorption at the electrode
s
D \ 1.30 ] 10~6 cm2 s~1 and C \ 10 lF cm~2. An advan-
dl
tage of using ac impedance measurements is that the faradaic
and non-faradaic parameters may be obtained within the
same experiment. The importance of Ðtting both faradaic and
in 0.1 to 0.5 M LiClO in a PEO electrolyte system. This
allows the inÑuence of solvent dynamics on the heterogeneous
charge transfer process to be probed. We have used ac imped-
4
non-faradaic and Ðnding that C is the same in the presence
dl
and absence of reactant is to gain conÐdence that the Ðtting
ance at an UME to determine k and D for the redox couple
s
procedure produces reliable results. The invariance of the
double-layer capacitance and the standard rate constant with
concentration demonstrate that there appears to be no com-
plications in the interpretation of the redox reaction (such as
ion pairing). This was especially important in our studies since
a double-layer e†ect has been suggested by Zhou et al.1 These
results are in a good agreement with the Ðndings of the Pyati
and Murray6 concerning cobalt tris(bipyridine) oxidation in
liquid polymer electrolytes of di†erent molecular weight.
Although the ac impedance data have been analysed suc-
cessfully in terms of a simple electrode reaction, we cannot
rule out the possibility that the redox active species may not
be TCNQ or TCNQ~ (as it is written in reactions (1) and (2)).
In particular ion pairing between TCNQ~ and Li` is possible
in such a low dielectric constant solvent (e \ 9). Fast sweep
cyclic voltammetry is currently being employed to investigate
further the possibility of ion pairing.
in the tetramer. Values for these parameters in 0.5 M LiClO
4
in PEO are k \ 0.11 ^ 0.05 cm s~1 and D \ 0.75 ] 10~6 cm
s
s~1. They are signiÐcantly higher than found previously for
PEG-400 where k \ 0.000 37 cm s~1 and D \ 2.5 ] 10~8 cm
s
s~1.1 These di†erences can be explained in terms of the di†er-
ent end groups (OH in the case of PEG-400) and by the great
di†erences in viscosities of the solvents used (tetramer: 0.0339
P and PEG-400: 0.889 P). Contrary to the PEG-400 study1
we Ðnd no dependence of k on the supporting electrolyte con-
s
centration. The kinetics of the second reduction step TCNQ~/
TCNQ2~ have also been studied using ac impedance at an
UME and the following values of the kinetic parameters
obtained: k \ 0.098 cm s~1 and D \ 1.3 ] 10~6 cm2 s~1.
s
Acknowledgements
This work has been supported by the Polish ScientiÐc Com-
mittee (KBN), grant No. 0344/T09/98/15.
It is interesting to compare our results with those of Zhou
et al.1 who used OH terminated PEO of molecular weight 400
which, like tetramer, is a liquid at room temperature. They
obtained the following kinetic data for the TCNQ/TCNQ-
couple in 0.1 LIClO : k \ 0.000 37 cm s~1 and
References
1
H. Zhou, L. Ding, N. Gu and S. Dong, Electrochim. Acta, 1998,
4
s
D \ 2.5 ] 10~8 cm s~1. Their standard rate constant is some
300-fold and the di†usion coefficient 40-fold lower than in
tetramer. The electron transfer rate is dependent on the longi-
43, 3353.
2
3
B. Scrosati, Nature, 1995, 373, 557.
P. G. Bruce and C. A. Vincent, J. Chem. Soc., Faraday T rans.,
1993, 89, 3187.
tudinal relaxation time of the solvent, q , which is in turn
4
5
L. Christie, P. Los and P. G. Bruce, Electrochim. Acta, 1995, 40,
L
related to the solvent viscosity.6,26h29 The di†erences in the
2159.
kinetic parameters between the two solvents are consistent
with the observation that the solvent viscosity of PEO-400 is
D. Baril, PhD Thesis, Institut Nationale Polytechnique de Gre-
noble, France, 1993.
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454
5453

View Article Online
6
7
R. Pyati and R. W. Murray, J. Am. Chem. Soc., 1996, 118, 1743.
M. Salomon, M. Xu, E. M. Eyring and S. Petrucci, J. Phys.
Chem., 1994, 98, 8234.
20 P. G. Bruce, A. Lisowska-Oleksiak, P. Los and C. A. Vincent, J.
Electroanal. Chem., 1994, 367, 279.
21 P. Los, P. G. Bruce, D. Stewart and C. A. Vincent, Recent
Advances in the Chemistry and Physics of Fullerenes and Related
Materials, ed. K. M. Kadish and R. S. Ruo†, The Electrochemi-
cal Society, Pennington, USA, 1997, pp. 117È124.
22 M. Fleischmann and S. Pons, J. Electroanal. Chem., 1988, 250,
277.
8
9
T. T. Wooster, M. L. Longmire, M. Watanabe and R. W.
Murray, J. Phys. Chem., 1991, 95, 5315.
H. Zhou, N. Gu and S. Dong, J. Solid State Electrochem., 1999, 3,
82.
10 H. Zhou and S. Dong, Electrochim. Acta, 1997, 42, 1801.
11 H. Zhou and S. Dong, J. Electroanal. Chem., 1997, 425, 55.
12 T. T. Wooster, M. Watanabe and R. W. Murray, J. Phys. Chem.,
1992, 96, 5886.
23 B. A. Boukamp, Equivalent Circuits, version 4.51, University of
Twente, Netherlands, 1993.
24 J. R. MacDonald, J. Schoonman and A. P. Lehner, J. Electroanal.
Chem., 1982, 131, 77.
13 M. Watanabe, T. T. Wooster and R. W. Murray, J. Phys. Chem.,
1991, 95, 4573.
25 P. Los, A. Lasia, H. Menard and L. Brossard, J. Electroanal.
Chem., 1993, 360, 101.
14 G. Grampp, W. Harrer and W. Jaenicke, J. Chem. Soc., Faraday
T rans., 1987, 83, 161.
26 A. Kapturkiewicz and B. Behr, J. Electroanal. Chem., 1984, 179,
15 W. Harrer, G. Grampp and W. Jaenicke, J. Electroanal. Chem.,
1986, 209, 223.
179.
27 M. J. Weaver, Chem. Rev., 1992, 92, 463.
28 W. R. Fawcett and C. A. Foss, J. Electroanal. Chem., 1989, 270,
103.
16 M. A. Komarynsky and A. C. Wahl, J. Phys. Chem., 1975, 79,
695.
17 N. Haran, Z. Luz and M. Shporer, J. Am. Chem. Soc., 1974, 96,
29 A. J. Bard, H. D. Abruna, C. E. Chidsey, L. R. Faulkner, S. W.
Feldberg, K. Itaya, M. Majda, O. Melroy, R. W. Murray, M. D.
Porter, M. P. Soriaga and H. S. White, J. Phys. Chem., 1993, 97,
7147.
4788.
18 W. Harrer, G. Gramp and W. Jaenicke, Chem. Phys. L ett., 1984,
112, 263.
19 P. Los and P. G. Bruce, Pol. J. Chem., 1997, 71, 1151.
5454
Phys. Chem. Chem. Phys., 2000, 2, 5449È5454