Two-equivalent electrochemical reduction of a cyano-complex [Tl III(CN)2]+ and the novel di-nuclear compound [(CN)5PtII-TlIII]0

Article abstract of DOI:10.1016/j.electacta.2005.02.006

Extending our recent insights in two-electron transfer microscopic mechanisms for a Tl^III/Tl^I redox system [D.E. Khoshtariya, et al., Inorg. Chem. 41 (2002) 1728], the electrochemical response of glassy carbon electrode in acidified solutions of Tl^III (ClO ₄)₃ containing different concentrations of sodium cyanide has been extensively studied for the first time by use of cyclic voltammetry and the CVSIM curve simulation PC program. The complex [Tl^III(CN) ₂]⁺ has been thoroughly identified electrochemically and shown to display a single well-defined reduction wave (which has no anodic counterpart), ascribed to the two-equivalent process yielding [Tl ^I(aq)]⁺. This behavior is similar to that of [Tl ^III(aq)]³⁺ ion in the absence of sodium cyanide, disclosed in the previous work, and is compatible with the quasi-simultaneous yet sequential two-electron transfer pattern (with two reduction waves merged in one), implying the rate-determining first electron transfer step (resulting in the formation of a covalently interacting di-thallium complex as a metastable intermediate), and the fast second electron transfer step. Some preliminary studies of the two-equivalent reduction of directly metal-metal bonded stable compound [(CN)₅Pt^II-Tl^III]⁰ has been also performed displaying two reduction waves compatible with a true sequential pattern.

Full text of DOI:10.1016/j.electacta.2005.02.006

Electrochimica Acta 50 (2005) 4444–4450
Two-equivalent electrochemical reduction of a cyano-complex
[TlIII(CN)2]+ and the novel di-nuclear compound [(CN)5PtII−TlIII]0
Tina D. Dolidzea, Dimitri E. Khoshtariyaa,b,∗
,
c
c
d,∗∗
˚
¨
Martin Behm , Goran Lindbergh , Julius Glaser
a
Institute of Inorganic Chemistry and Electrochemistry, Georgian Academy of Sciences, Jikiya 7, Tbilisi 0186, Georgia
b
Institute of Molecular Biology and Biophysics, Georgian Academy of Sciences, Gotua 14, Tbilisi 0160, Georgia
c
Department of Applied Electrochemistry, The Royal Institute of Technology (KTH), S-10044 Stockholm, Sweden
d
Department of Inorganic Chemistry, The Royal Institute of Technology (KTH), S-10044 Stockholm, Sweden
Received 22 June 2004; received in revised form 29 November 2004; accepted 6 February 2005
Available online 2 March 2005
Abstract
Extending our recent insights in two-electron transfer microscopic mechanisms for a TlIII/TlI redox system [D.E. Khoshtariya, et al.,
Inorg. Chem. 41 (2002) 1728], the electrochemical response of glassy carbon electrode in acidified solutions of TlIII (ClO4)3 containing
different concentrations of sodium cyanide has been extensively studied for the first time by use of cyclic voltammetry and the CVSIM curve
simulation PC program. The complex [TlIII(CN)2]+ has been thoroughly identified electrochemically and shown to display a single well-
defined reduction wave (which has no anodic counterpart), ascribed to the two-equivalent process yielding [TlI(aq)]+. This behavior is similar
to that of [TlIII(aq)]3+ ion in the absence of sodium cyanide, disclosed in the previous work, and is compatible with the quasi-simultaneous yet
sequential two-electron transfer pattern (with two reduction waves merged in one), implying the rate-determining first electron transfer step
(resulting in the formation of a covalently interacting di-thallium complex as a metastable intermediate), and the fast second electron transfer
step. Some preliminary studies of the two-equivalent reduction of directly metal–metal bonded stable compound [(CN)5PtII–TlIII]0 has been
also performed displaying two reduction waves compatible with a true sequential pattern.
© 2005 Elsevier Ltd. All rights reserved.
Keywords: Thallium(III); Cyano-complexes; Two-electron transfer; Cyclic voltammetry; Di-nuclear complexes
1. Introduction
equivalent redox chemistry of thallium(III) is rich in various
unresolved puzzles. Both homogeneous and electrode pro-
cesses of aqueous complex [TlIII(aq)]3+ have been studied
in the past (see Refs. [8–11] and references sited therein).
Thallium(III) is known to form very strong complexes of dif-
ferent ligand compositions, in particular with CN−, while
thalliun(II) and thallium(I) do not [8,9,12,13]. The homo-
geneous systems containing cyanide ions were investigated
in the context of the electron exchange kinetics [8] and the
ligand uptake stability constants [9]. However, electrochem-
ical studies of thallium(III) cyano-complexes have not been
performed so far.
The diverse mechanisms of multi-electron transfer are of
fundamental interest due to their significant, not fully rec-
ognized, role in biological and chemical processes [1–4].
However, notwithstanding the significant progress in their
theoretical understanding [5–7], the number of even well-
characterized two-electron transfer processes is much smaller
compared with that of corresponding single-electron pro-
cesses. In particular, despite some earlier efforts, the two-
∗
∗∗
Corresponding author. Tel.: +7 995 32 386077; fax: +7 995 32 939157.
Co-corresponding author.
In our previous work [11] we accomplished the de-
tailed study of electrochemical two-equivalent reduction of
[TlIII(aq)]3+ to [TlI(aq)]+ in acidified solutions of sodium per-
E-mail addresses: dimitri.k@joker.ge (D.E. Khoshtariya),
julius@inorg.kth.se (J. Glaser).
0013-4686/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2005.02.006

T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
4445
chlorate containing different concentrations of TlIII(ClO4)3.
On the basis of experimental results, obtained by use of cyclic
voltammetry (CV) and rotating disk electrode (RDE) tech-
niques, and the subsequent data analysis by the curve sim-
ulation PC program (CVSIM) [14], it has been suggested
that this two-equivalent reduction process occurs through the
quasi-simultaneous sequential mechanism, probably via for-
mation of covalently interacting (yet unstable) di-thallium(II)
intermediate [11]. The overall reaction involves three steps,
viz.:
TlIII(ClO4)3 + 7.28 M HClO4 + TlIClO4, latter unavoidably
present as an impurity (≤1%)] was prepared according to the
procedure described earlier [21]. Ultrapure water was sup-
plied by a Millipore Milli-Q system. Synthesis of the solid
compound [(CN)5PtII−TlIII]0 also described elsewhere [20].
All other chemicals used were commercial products of high-
est purity available. Computer simulation procedures were
run using CVSIM Program [14].
3. Results and discussion
2 [TlIII(aq)]3+ + 2e− ⇒ 2[TlII(aq)]2+ (slow)
2[TlII(aq)]2+ ⇒ [TlII–TlII]4+ (fast)
[TlII–TlII]4+ + 2e− ⇒ 2[TlI(aq)]+ (fast)
(1)
(2)
(3)
3.1. Overview (including data for a [TlI(aq)]+/Tl0
couple)
Typical CV experiments (at scan rates covering the range
of 0.01–0.5 V/s) were performed for working solutions con-
taining TlIII(ClO4)3 at 2 × 10−3 to 5 × 10−2 M, TlIClO4
at ca. 2 × 10−5 to 5 × 10−4 M, without (see Ref. [11]) or
with the addition of 0.025–0.05 M NaCN, within the broad
potential range of +2.5 to −0.8 V (most experiments were
run within +1.2 to −0.5 V). Under these conditions, within
the potential range of +1.0 to −0.5 V, voltammograms ex-
hibited a single reduction wave, peaked at ca. +0.75 V (in
the absence of NaCN), or at ca. +0.05 V (in the presence of
NaCN), Fig. 1, attributable to the reduction of TlIII species
present in two different coordination forms (vide infra). No
anodic counterpart for either of these waves (within the whole
potential range extended up to +2.5 V) was observed on the
reverse scan in both cases. However, when the potential range
was extended down to −0.8 V, the reduction and oxidation
waves peaked at −0.6 and −0.4 V, respectively, could be
observed (Fig. 2), attributable to the [TlI(aq)]+/Tl0 couple
[10]. Fig. 2 simultaneously represents CV curves for both,
the [TlIII(aq)]3+/[TlI(aq)]+ and TlI(aq)]+/Tl0 couples for the
solution initially containing only ca. 1% of TlI compared to
This pattern, showing up in a single (merged) reduction wave,
results from the facts that: (1) the [TlII]2+ ion (whatever set of
redox potentials for individual steps is chosen, see Ref. [11])
is a stronger oxidant than [TlIII]3+ [9–11]; (2) according to
the solid-state structural data [15,16], there should exist at
least the 6s–6s bonding interaction between the TlII species
[11], by the analogy with two relative iso-electronic systems,
Au0–Au0 and HgI–HgI [17,18].
The aim of the present work was an essential extension
of our previous studies of aqueous thallium(III) ([Tl(aq)]3+)
electrochemistry [11]. In particular, (a) an electrochemi-
cal identification of thallium(III) cyano-complexes, and the
novel di-nuclear compound [(CN)5PtII−TlIII]0, disclosed in
earlier work mostly by the multinuclear NMR techniques (see
Refs. [13] and [19,20], respectively), and (b) detailed study
of electrochemical mechanisms for two-equivalent reduction
for these compounds in comparison with the mechanism of
aqueous thallium(III) investigated earlier [11].
2. Experimental
The electrochemical experiments were carried out on
a PAR 273 Potentiostat/Galvanostat, controlled by EG&G
Model Software. Two kinds of cells with standard three-
electrode configurations containing 50 and 5 ml of working
solution were used. Glassy carbon working electrodes (ar-
eas 0.1256 and 0.0314 cm2) were cleaned before each series
of experiments by polishing with 1.0; 0.3; and 0.05 ␮m alu-
minum granules from Buehler on the Buehler polishing cloth,
followed by sonification in purified water (vide infra). The
saturated calomel reference electrode and/or a silver/silver
chloride reference electrodes were separated from the main
cell by glass frits. All potentials quoted in this paper are ref-
ered to a normal hydrogen electrode (NHE). Measurements
were performed at room temperature (24 1) ◦C.
Fig. 1. Cyclic voltammograms for glassy carbon electrode (area:
0.1256 cm2): (a) 1 M NaClO4 + 0.273 M HClO4 + 0.01 M Tl(ClO4)3 at
scan rate v = 0.1 V s−1, and (b) in1 M NaClO4 + 0.073 M HClO4 + 0.01 M
Tl(ClO4)3 + 0.025 M NaCN at the scan rates (from top to bottom) 0.2, 0.1
and 0.05 V s−1, respectively.
Working solutions were prepared from the stock so-
lution of TlIII(ClO4)3 using aqueous solution of HClO4
or aqueous acidified solution of NaClO4 by adding of
appropriate amounts of NaCN. Stock solution [1.02 M

4446
T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
adsorbed and deposited species, respectively, and, hence are
not directly comparable with each other regarding the actual
bulk concentrations of two kinds of reactant ions under the
comparison.
3.2. Reduction of [TlIII(aq)]3+ [11]
The overall electrochemical mechanism for this two-
equivalent process has been thoroughly analyzed in the pre-
vious work [11]. The RDE technique, along with the CV
voltammetry, latter furnished by the extended CV curve sim-
ulation procedure (CVSIM Program [14]), were applied to-
wards the elucidation of the microscopic two-equivalent pat-
tern. In brief, theoretical analysis (cross-testing) included the
following procedures (this analysis turned to be equally valid
also for the branch case of the two-equivalent [TlIII(CN)2]+
reduction, see Section 3.3):
1. Application of basic general Eqs. (4)–(7) valid for irre-
versible electrochemical processes [22–24]:
Fig. 2. Simultaneous representation of CV curves for the [TlIII 3+/[TlI]+
]
(reduction wave is numbered as 1) and [TlI]+/Tl0 (reduction and oxida-
tion waves numbered as 2 and 3, respectively) couples (glassy carbon elec-
trode, area: 0.1256 cm2, scan rate: 0.5 V s−1) in solution containing 1 M
HClO4 + 0.01 M TIIII(CIO4)3 + 0.0001 M TlIClO4. Solid curve: total first
scan; thin curve 1 represents the early return path for the first wave; thin
curves 2 and 3 represent the total sixth scan. See Sections 3.1 and 3.2 for
further details.
47.7
Ep − E(p/2)
=
(mV),
(4)
(5)
αnα
ꢀ
ꢁ
1/2
RT
v1
(Ep)1 − (Ep)2 =
ln
αnαF
v2
Ip = 2.99 × 105n(αnα)1/2Ac0D1/2v1/2
(6)
(7)
TlIII. The intensity of both the reduction and oxidation waves
for the latter couple increases with the increase of TlI concen-
tration (caused either by beforehand electrochemical reduc-
tion of TlIII, or by simple adding of TlClO4 to the solution,
see Fig. 2), and exhibits the shape (an abrupt current rise and
a loop) typical for metal deposition reaction due to the nucle-
ation overpotential. At the same time, the oxidation wave is
sharp and symmetrical, indicating that reactant is deposited
on the electrode (see, e.g., Ref. [22] and fig. 6.24 therein).
This couple, which starting component, TlI, initially main-
tained at the concentration of ca. 1% (subsequently increas-
ing as a result of electrochemical reduction of TlIII, Fig. 2),
has been used in some further experiments as an internal
marker (vide infra). An observation that the magnitudes of
peak currents for the [TlI(aq)]+/Tl0 couple present as an im-
purity are comparable to the one for the reduction of the main
component (TlIII, coordinated either with water or CN−, re-
spectively, Figs. 1 and 2) can be readily explained by different
intrinsic charge-transfer mechanisms behind these CV waves
(numbered by 1–3 in Fig. 2). In particular, wave 1 represents
outer-sphere electron transfer in the course of the reduction
of freely diffusing TlIII species (vide infra), whereas waves
2 and 3 probably represent the intrinsic steps of reduction
(complete discharge of weakly adsorbed TlI that follows the
preceding ion transfer step, see Refs. [6,22]), and oxidation
of the deposited Tl0 that can be considered as strongly ad-
sorbed particle [22]. Consequently, the cases of CV waves 2
and 3, in contrast with wave 1 in which case the peak current
is given by Eq. (6) (see the next sub-section below) these peak
currents are proportional to surface concentrations of weakly
4.405Ip
k10 =
nFAc0 exp[(−αnαF/RT)(E10 − Ep)]
where Ep is the peak potential, and Ep/2 is the half peak
potential, α is the transfer coefficient, nα is the number of
electrons transferred in the rate-determining step, n is the
total number of electrochemically transferred electrons, A
istheareaofworkingelectrodeandc0 isthereactant’sbulk
concentration, visthescanrate, Ip isthepeakcurrent, E10 is
the standard potential for the rate-determining (here first)
electron transfer step, and k10 is the standard rate constant
forthesamestep, Tistheabsolutetemperature, RandFare
the gas and Faraday constants, respectively. The following
observations were essential: the plot of peak current as a
function of square root of the scan rate, is linear and passes
through the origin (Fig. 3). The peak potential is shifting
to the negative potentials with increase of the scan rate, but
a bias for the half peak potential shift remains constant.
Also, the invariance of ratios Ip/v1/2 and ꢀEp/2/ꢀlg v
over the entire range of scan rates applied indicate that
chemical (e.g. dismutatio/disproportionation) reactions
does not alter kinetic aspects of the electrode (rate-
determining) process [11,22,23]. Eqs. (4) and (5), as well
as the Tafel equation applied to the RDE data (Ref. [11]),
allowed for the direct determination of the value for a
product αnα. Furthermore, after setting of a realistic value
for D (of two possible choices), Eqs. (4)–(6) allow for the
direct and unequivocal determination of the following val-
ues: n = 2, nα = 1, α = 0.66, Table 1. This set of parameters

T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
4447
ference method) has been proven to be a powerful tool for
the analysis of CV experiments. The shape of a CV curve
reflects the intrinsic features of all the stages, including the
interface charge transfer and solution chemical reactions
(that are coupledto the charge transfer), contributingtothe
overall process. Thus, through the CV study, normally one
can deduce a great deal of information regarding the con-
stituent stages. Simulationscanbe veryhelpful, bothinthe
preliminary stages of CV studies (in deciding the general
mechanism)and in the extraction of rate and equilibrium
parameters of the chosen mechanism by comparing ex-
perimental CV curves with those successively generated
by the program. The CVSIM Program embodies all the
accumulated knowledge on the electrochemical mecha-
nisms collected up to early nineteen nineties, including
the multi-electron mechanisms, and enables one to gener-
ate CV curves for nearly any desired mechanism and any
realistic sets of intrinsic parameter [14] (see also Ref. [25]
for the related information).
Fig. 3. Dependence of peak current of cyclic voltammograms for glassy
carbon electrode on the square root of scan rate in the solutions of (᭹) 1 M
NaClO4 + 0.273 M HClO4 + 0.01 M Tl(ClO4)3, and (ꢀ) the same solution,
after addition of 0.05 M NaCN.
indicates that both electrons are transferred electrochem-
ically, and that the rate-determining step includes transfer
of one electron. In addition, a value of E10 has been roughly
determined from the visual analysis of CV curves, allow-
ing for a rough determination of k10 according to Eq. (7),
Table 1. However, a transfer of the second electron cannot
be considered as an independent process, since the Tl2+
ion is a stronger oxidant than Tl3+. This case is considered
in the literature to be characterized by a single (merged)
wave, the peak current of which is determined by two elec-
tron equivalents, and the shape of which is determined by
the parameters (α1, E10, k10) of the first electron transfer
step in the sequence [11,22,24].
Assuming the above deduced mechanism of two-electron
reduction for the aqueous [TlIII(aq)]3+ species (as well as
[TlIII(CN)2]+ species to be considered below), and using the
estimated values of α1, E10, k10, n and nα, as starting values for
the simulation program, the other constituent values, viz., α2,
k10, k20, were smoothly varied within the reasonable ranges,
aiming the reproduction of experimental CV curves at the
maximal accuracy. Further iterative steps included slight (ad-
justing) variation of the values for “known” parameters, α1,
E10 and k10 such to achieve the desirable coincidence of simu-
lated and experimental CV curves. As a result, the experimen-
tal curves were surprisingly well reproduced by the values of
α1, E10 and k10 that felt within rather narrow ranges very close
to the corresponding initial values (Fig. 4, Tables 1 and 2).
We note that the decrease of the simulated value for E10 cor-
2. ApplicationoftheCVSIMProgram. Computersimulation
by the CVSIM Program (based on the explicit finite dif-
Table 1
Kinetic parameters estimated through the cross-testing analysis of experimental data for two-equivalent reduction of [TlIII(aq)]3+ and [TlIII(CN)2]+ (see text
for details)
Reactant
Technique
α1
D (×106 cm2 s−1
3.72 0.32
3.52 0.10
3.32 0.30
)
lg(k10) (cm s−1
)
E10 (V)
[TlIII(aq)]3+*
[TlIII(aq)]3+*
[TlIII(CN)2]+***
CV**
RDE
CV**
0.66 0.04
0.65 0.01
0.33 0.03
−5.4 0.3
−5.5 0.1
−6.4 0.6
1.04
1.04
0.89
*
Data from Ref. [11].
Estimates based on Eqs. (4)–(7).
This work.
**
***
Table 2
CVSIM simulation results (for CV curves) for two-equivalent reduction of [TlIII(aq)]3+ and [TlIII(CN)2]+ (see text for details)
E10 (V)
E20 (V)
lg(k10) (cm s−1
)
lg(k20) (cm s−1
)
α1
α2
**
**
Reactant
[TlIII(aq)]3+*
[TlIII(CN)2]+
1.04 0.1
0.89 0.1
1.46***
1.46
−5.7 0.6
−6.5 0.6
−5.4 1.2
−5.4 1.2
0.65 0.05
0.35 0.05
0.5 0.2
0.5 0.2
*
Data from Ref. [11].
Rough estimates.
**
Estimated by using the equation E0 = (E10 + E20)/2 = 1.25 V [10].
***

4448
T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
Our further analysis of the two-equivalent electrochemical
pattern for the [TlIII(CN)2]+ complex ion is based on two
reasonable assumptions:
(a) The overall general mechanism deduced earlier for aque-
ous thallium(III) species, [TlIII(aq)]3+, does not change
upon the complexation with two CN− ligands (leading
to the [TlIII(CN)2]+ species acting as a main component
instead if [TlIII(aq)]3+).
(b) The [TlIII(CN)2]+ ions lose both cyano-ligands through
the first electron transfer step, because the intermediate
(TlII) and product (TlI) species are known to be inert
regarding such a complexation [8,12,13].
Hence, the overall two-equivalent pattern can be presented
as:
2 [TlIII(CN)2]+ + 2e−
Fig. 4. Experimental data of Fig. 1 both at scan rate of v = 0.1 V s−1
:
solid curves (after minor background corrections). Curve a: reduction of
[TlIII(aq)]3+; curve b: reduction of [TlIII(CN)2]+. Simulated CV curves. (1)
⇒ 2[TlII(aq)]2+ + 4CN− (slow)
(8)
Open circles: n = 2, nα = 1, E10 = 1.04 V, α1 = 0.65, k10 = 2 × 10−6 cm s−1
,
E20 = 1.46 V, α2 = 0.5, k20 = 2 × 10−6 cm s−1
; (2) filled circles: n = 2,
2 [TlII(aq)]2+ ⇒ [TlII–TlII]4+ (fast)
[TlII–TlII]4+ + 2e− ⇒ 2 [TlI(aq)]+ (fast)
(9)
nα = 1, E10 = 0.89, α1 = 0.3, k10 = 3 × 10−7 cm s−1, E20 = 1.46 V, α2 = 0.5,
k20 = 2 × 10−6 cm s−1
.
(10)
where steps (9) and (10) probably are identical to steps (2)
and (3) for the reduction of [TlIII(aq)]3+ in the absence of
NaCN.
relates with the increase of the corresponding value of k10
within these narrow ranges indicated in Table 2. The param-
eters characterizing the second electron transfer step were
fixed with much lower exactness (Table 2). This is due to
the fact that the shapes of CV waves are much less affected
by the intrinsic kinetic characteristics of the second electron
transfer step. However, the peak current is affected equally
by both the steps [26].
The peak shift observed for the cyano-complex ion
[TlIII(CN)2]+ (Fig. 1), generally may be due to the changes
of all three major parameters α1, k10 and E10 of the first
slow electron transfer step. The cross-testing analysis pro-
cedure, including steps (1) and (2) described in the previ-
ous sub-section, has been applied to the CV data ascribed
to the [TlIII(CN)2]+ species completely. The values of basic
parameters obtained from the theoretical analysis through
Eqs. (4)–(7), and through the CVSIM simulation, are pre-
sented in Tables 1 and 2, respectively. We found that the
change of ␣1 from 0.66 to 0.33 is mainly responsible for the
change of the peak shape, and can be only in part responsi-
ble for the observed peak shift. Another contribution comes
from the change of the value of E10 from 1.04 to ca. 0.89 V
(Tables 1 and 2). Note that as in the case of [TlIII(aq)]3+, in the
course of the reduction of [TlIII(CN)2]+, we could not observe
individual waves for the first and second electro-reduction
steps, but only a single well-defined reduction wave. Evi-
dently, in both cases the metastable intermediate containing
the [TlII]2+ ion(s) is a stronger oxidant than TlIII (complexed
or uncomplexed with the cyanide), E20 > E10 [9–11]. Under
this condition, if the potential separation between successive
reductions is more than 180 mV (second electron transfer
is easier, e.g. takes place at more positive potentials than
first step), the individual waves should merge into one wave,
the shape, position and kinetic parameters of which are de-
termined by the first electron transfer step in the sequence
[17,19]. The formal redox potential of this step, Eq. (11),
accompanied by the activationless formation of a thermo-
dynamically significant di-thallium(II) intermediate, Eq. (8),
3.3. Two-electron reduction of [TlIII(CN)2]+
Upon the addition of 0.025 ÷ 0.05 M NaCN to working
solutions containing TlIII(ClO4)3 at 2 × 10−3 to 5 × 10−2 M,
the above discussed two-electron reduction wave peaked at
ca. +0.75 V, is shifted to +0.05 V (Fig. 1a and b, respec-
tively). No sign of the former [TlIII(aq)]3+ reduction wave
remains on the CV scan. The shape of a “new” wave is
different from the original one. However, the peak current
again changes linearly as a function of square root of the
scan rate (Fig. 3). All these fact are indicative of TlIII com-
plexation by the CN− ions according to the results of former
NMR studies [13]. The previous detailed computer analysis
of systematic NMR data, using the PC Program “Medusa”,
allowed for the calculations of concentrations for different
kind of thallium(III) cyanocomplexes at different solution
compositions (concentrations of solution components such
as TlIII, CN−, H+). According to the previous results [13],
verified also by the new calculations, under the specified
range of CN− concentrations, almost 100% of TlIII is present
as the species: [TlIII(CN)2]+. Hence, the newly detected re-
duction wave should be ascribed to this complex ion (vide
infra).

T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
4449
can be defined as [11]:
100% content can be expected at the concentrations of
NaCN exceeding 0.15 M (at otherwise similar conditions).
The reduction wave for latter specie was found to overlap
strongly with one belonging to the reduction of Tl+ (un-
avoidably present in the solution), and hence has not been
investigated in detail due to this complication. We were not
able to disclose the reduction waves of two other species,
[TlIII(CN)]2+ and [TlIII(CN)3]0 in a whole range of avail-
able NaCN concentrations. The only observed well-defined
wave in the presence of NaCN was proven to belong to the
[TlIII(CN)2]+ complex with the peak intensity corresponding
to its molar fraction in the solution, calculated according to
[13].
ꢂ
ꢃ
[TlIII(CN)2
[Tl42+]1/2
]
+
RT
F
E1 = E10 +
ln
.
(11)
3.4. Microscopic reversibility of the two-equivalent
process
The overall mechanism of two-equivalent reduction of
[TlIII(CN)2]+ represented by Eqs. (8)–(10) implies that, anal-
ogously to the reduction process of [TlIII(aq)]3+, all three
steps including two electrode and one chemical, presumably
take place at the outer Helmholtz plane (OHP), as can be nor-
mally expected for strongly solvated redox-active ions [6,11].
According to the principle of microscopic reversibility, for
the reverse oxidation process of [TlI(aq)]+ to [TlIII(aq)]3+
(or to [TlIII(CN)2]+, vide infra) one can expect the following
common starting steps [11]:
3.6. Reduction of [(CN)5PtII−TlIII]0 (“compound I”)
We have obtained also preliminary results for the elec-
trochemical reduction of this long-lived bi-nuclear cyano-
complexes at nearly saturating concentration of 10−3 M,
0.1 M in HClO4 (no salt added). The cyclic voltammograms
display two cathodic waves at ca. 0.3 and −0.3 V, Fig. 5,
which can be ascribed to the manifestation of stepwise re-
duction of this compound (“compound I”, according to the
classification of Ref. [19,20]) by two electrons. These waves
have no anodic counterparts even if the scan is reversed
before reaching potentials where the TlI deposition begins.
This mechanism is confirmed by formation of Tl+ ions de-
tected by appearance of characteristic Tl+ + e− ⇔ Tl0 re-
duction and oxidation waves gradually increasing in inten-
sity in the course of repetitive cycling (note that no Tl+
ions were present in the starting solution), Fig. 5. We pro-
pose the following mechanism for the two step reduction of
compound I:
2 [TlI(aq)]+ − 2e− ⇒ 2 [TlII(aq)]2+ (slow)
2 [TlII(aq)]2+ ⇒ [TlII − TlII]4+ (fast)
(12)
(13)
The estimated standard potential for this reaction is ca. 1.5 V
[11,28]. In fact, no anodic wave due to this process was de-
tected within the potential range of 1.0–1.5 V, in the absence
or presence of cyanide ions (Ref. [11] and the present work).
The most probable reason for this observation is a violation of
the principle of microscopic reversibility due to the essential
asymmetry in the solvation features of reactant and product
species. According to Koper and Schmickler [6], for soft and
easily desolvable univalent ions such as [TlI(aq)]+, the more
probable redox path implies an ion transfer to the electrode
followed by a strong specific adsorption on the electrode as
a rate-determining step:
0
[(CN)5PtII − TlIII] + e− ⇒ [(CN)5PtII − TlII]•
(16)
−
[TlI(aq)]+ − δe− ⇒ Tl(ads)(1+δ) (slow)
(14)
(15)
Tl(ads)(1+δ) − (1 − δ)e− ⇒ TlII(ads)2+ (slow)
In this case, the TlII(ads)2+ ion formed as an intermediate,
is probably trapped in a local potential minimum on the
electrode surface. This would shift the standard reduction
potential to much higher value (E20 ≥ 2V) compared to that
expected for the hypothetical (microscopically reversible)
pattern, Eqs. (12) and (13). Unfortunately, studies in this
potential range are complicated by the processes of oxygen
evolution and formation of oxide films.
3.5. Other cyano-complexes of TlIII
According to the established equilibrium relationships for
the complexation between [TlIII]3+ and CN− [13], at the
concentrations of NaCN below or higher the frames quoted
above, mostly the mixtures of other thallium cyanocom-
plexes, viz.: [TlIII(CN)]2+ and [TlIII(CN)3]0 together with
the already characterized compound [TlIII(CN)2]+ can be
formed. Formation of compound [TlIII(CN)4]− at nearly
Fig. 5. Cyclic voltammograms for glassy carbon electrode (area:
0.0314 cm2) in the solution: 0.001 M [(CN)5PtII–TlIII]0 + 0.01 M HClO4;
bold curve: first cycle; gray curve: sixth cycle; scan rate v = 0.1 V s−1
.

4450
T.D. Dolidze et al. / Electrochimica Acta 50 (2005) 4444–4450
[(CN)5PtII − TlII]•− + e−
unavoidable adsorbed states, seem to be extremely promising
regarding the redox couples under the present interest.
⇒ [PtII(CN)4]2− + Tl+ + CN−
(17)
(We note that the [PtII(CN)4]2− complex is not electrochem-
ically active within the potential range of +0.8 to −0.8 V, un-
published results). In contrast to the above considered cases
of mononuclear thallium(III), in the absence or presence
of cyanide ions, the electrochemical steps now seemingly
have the reverse order of formal redox potentials (i.e. here
E10 > E20) leading to the appearance of two cathodic waves
[24,27].
Acknowledgements
The continuous support from the Swedish Natural
Sciences Research Council (NFR) and the European Com-
mission through INTAS Project No. 96-1162 are kindly
acknowledged. The authors are grateful to Dr. E. Ahlberg
for introducing them to the CVSIM program and helpful
discussions.
References
4. Conclusions
[1] R.K. Clayton, Photosynthesis. Physical Mechanisms and Chemical
Patterns, Cambridge University Press, Cambridge, 1980.
[2] D.S. Bendall (Ed.), Protein Electron Transfer, BIOS Scientific Pub-
lication, Oxford, 1996.
[3] P.E.M. Siegbahn, M.R.A. Blomberg, Chem. Rev. 100 (2000) 421.
[4] D. Uhrhammer, F.A. Schultz, J. Phys. Chem. A 106 (2002) 1630.
[5] L.D. Zusman, D.N. Beratan, J. Phys. Chem. A 101 (1997) 4136.
[6] M.T.M. Koper, W. Schmickler, J. Electroanal. Chem. 450 (1998) 83.
[7] E.G. Petrov, Ya.R. Zelinskyy, V. May, J. Phys. Chem. B 108 (2004)
13208.
[8] A.G. Lee, The Chemistry of Thallium, Elsevier, Amsterdam, 1971,
Chapter 9.
[9] J. Glaser, Adv. Inorg. Chem. (1995) 43.
[10] M.I. Bellavance, B. Miller, in: A.J. Bard (Ed.), Encyclopedia of
Electrochemistry of the Elements, Marcel Dekker, New York, 1975,
vol. IV, Chapter 4.
[11] D.E. Khoshtariya, T.D. Dolidze, L.D. Zusman, G. Lindberg, J.
Glaser, Inorg. Chem. 41 (2002) 1728.
[12] E. Penna-Franca, R.W. Dodson, J. Am. Chem. Soc. 77 (1955) 2651.
[13] I. Blixt, B. Gyo¨ri, I. Glaser, J. Am. Chem. Soc. 111 (1989) 7784.
[14] D.K. Gosser, A General Simulation Program (CVSIM), VCH, New
York, 1993.
[15] R. Dronskowski, A. Simon, Angew. Chem. Int. Ed. 28 (1989) 758.
[16] S. Henkel, K.W. Klenkhammer, W. Schwarz, Angew. Chem. Int. Ed.
33 (1994) 681.
[17] W.C. Ermber, Y.S. Lee, K.S. Pitzer, J. Chem. Phys. 70 (1979) 293.
[18] D.A. Kleier, W.R. Wadt, J. Am. Chem. Soc. 102 (1980) 6909.
[19] M. Maliarik, K. Berg, J. Glaser, M. Sandsro¨m, I. To´th, Inorg. Chem.
37 (1998) 2910.
[20] M. Maliarik, J. Glaser, I. To´th, M.W. da Silva, L. Ze´ka´ny, Eur. J.
Inorg. Chem. (1998) 565.
[21] M. Maliarik, J. Glaser, I. To´th, Inorg. Chem. 37 (1998) 5452.
[22] R. Greef, R. Peat, L.M. Peter, D. Pletcher, J. Robinson, Instrumental
Methods in Electrochemistry, Southampton Electrochemistry Group,
Horwood, New York, 1986, Chapter 6.
[23] R.S. Nicholson, I. Shain, Anal. Chem. 36 (1964) 706.
[24] A.J. Bard, L.R. Faulkner, Electrochemical Methods. Fundamentals
and Application, Wiley, New York, 1980, Chapter 6.
[25] A.W. Bott, Curr. Separat. 16 (1997) 61;
For acidified solutions of TlIII(ClO4)3, containing sodium
cyanide at appropriate concentrations, in the potential range
of +1.7 to −0.7 V, the only electro-active particles are
[TlIII(CN)2]+ and [TlIII(CN)4]−. The reduction wave for
the latter complex occurs at the potentials, where reduc-
tion of [TlI]+ (present in the working solutions as an im-
purity), also takes place. Reduction of thallium(III) (present
as [TlIII(CN)2]+) to [TlI(aq)]+, as in the case of reduction of
[TlIII(aq)]3+, goes through the quasi-simultaneous sequen-
tial mechanism, provided that the transfer of the first elec-
tron is rate-determining step. So, the overall mechanism of
two-equivalent reduction does not change upon the complex-
ation with cyanide, though the parameters do change for
the kinetically significant first (slow) step. The second step
should be identical in the presence and absence of cyanide
ions in the solution. Presumably the difference between stan-
dard potentials of first and second electron transfer steps
both in the presence and absence of NaCN is more than
180 mV (E20 > E10) and, consequently, one can observe only
a single combined (merged) reduction wave (see also Ref.
[11]).
Preliminary results for the electrochemical two-equivalent
reduction of the novel di-nuclear long-lived compound
[(CN)5PtII–TlIII]0 also indicate the two-step mechanism, but,
obviously, in contrast to the above-mentioned cases of reduc-
tion of mononuclear thallium(III), the steps have the reverse
order for standard redox potentials (E10 > E20) leading to the
appearance of two separated cathodic waves.
In closing we would like to mention that the two-electron
mechanisms proposed above, especially conclusion about the
involvement of a Tl(II) dimer as an intermediate, require fur-
ther justifications by using of additional approaches and/or
techniques. A significant extension of this work is planned
with the application of a high scan rate cyclic voltamme-
try, in the combination with a methodology of modification
of electrodes by the self-assembled monolayer films of dif-
ferent composition and thickness (for illustrative cases, see
Refs. [29,30]). The latter approach, potentially allowing for
the prevention of direct interaction of reactant ions with the
electrode surface, and, hence, for an exclusion of otherwise
A.W. Bott, Curr. Separat. 18 (1999) 9.
[26] L.I. Krishtalik, Double Layer and Electrode Kinetics (in Russian),
Nauka, Moscow, 1981, Chapter 5 (p. 198–282).
[27] D.S. Polcyn, I. Shain, Anal. Chem. 38 (1966) 370.
[28] N. Hush, Trans. Faraday Soc. 57 (1961) 557.
[29] D.E. Khoshtariya, T.D. Dolidze, L.D. Zusman, D.H. Waldeck, J.
Phys. Chem. A 105 (2001) 1818.
[30] T.D. Dolidze, D.E. Khoshtariya, D.H. Waldeck, J. Macyk, R. van
Eldik, J. Phys. Chem. B 107 (2003) 7172.