Journal of Electroanalytical Chemistry 522 (2002) 189 198

www.elsevier.com/locate/jelechem

Reduction of iodine at the organic liquid  aqueous

solution  graphite electrode three-phase arrangement
Valentin Mirceski 1, Fritz Scholz *
Ernst-Moritz-Arndt-Uni6ersitat Greifswald, Institut fur Chemie und Biochemie, Soldmannstrae 23, D-17489 Greifswald, Germany

Abstract
The electrochemistry of the iodine/iodide redox couple was studied at the three-phase arrangement organic liquid  aqueous
solution  graphite electrode. A droplet of an organic solvent, e.g. nitrobenzene, containing dissolved iodine was attached to a
graphite electrode and immersed in an aqueous electrolyte solution. The reaction mechanism has been studied in the presence of

NO
and ClO
3 , Cl , Br , F
4 ions in the aqueous supporting electrolyte, and for nitrobenzene, tetrachloromethane, and paraffin
oil as organic solvents for iodine. The reduction of iodine in the nitrobenzene droplet is either accompanied by a transfer of the
iodide ions formed from nitrobenzene to water: (i) I2(nb) + 2e ? 2I
(aq), or, in the case that the aqueous electrolyte is an alkali

chloride or bromide, by a transfer of chloride or bromide ions from nitrobenzene to water: (ii) I2(nb) + 2X
? 2I
(nb) + 2e
(nb) +

2X(aq) (X =Cl , Br , possibly in the form of I2X ). The second reaction is possible because with chloride and bromide ions
a rather fast reactive partition of the aqueous electrolyte into nitrobenzene occurs, due to the formation of I2X
(nb) ions in
nitrobenzene. 2002 Published by Elsevier Science B.V.
Keywords: Voltammetry; Iodine; Three-phase junction; Liquid interface; Ion transfer

1. Introduction
The interface between two immiscible electrolyte solutions (ITIES) has been studied extensively, especially
with the help of electrochemical techniques [1 4]. The
interest in these studies is fuelled by the importance of
liquid  liquid interfaces for separation membranes, ion
selective electrodes, solvent extraction, biomimetic studies of membrane function, etc. [5 13]. Among numerous phenomena occurring at the ITIES, the transfer of
ions across the interface is one of the most important
processes.
Voltammetric techniques appear to be a simple and
powerful tool for studying the thermodynamics and
kinetics of ion transfer processes. Over the last few
decades, the experimental approach for studying the
ion transfer was advanced significantly and many modifications of the voltammetric experiments have been
* Corresponding author. Tel.: + 49-3834-86-4450; fax: +49-383486-4451.
E-mail address: fscholz@uni-greifswald.de (F. Scholz).
1
On leave from: Institute of Chemistry, Faculty of Natural Sciences and Mathematics, Sv. Kiril i Metodij University, P.O. Box
162, 1000 Skopje, Macedonia.

proposed. Nevertheless, the general experimental arrangement for investigating the ion transfer across the
liquid interface remained the same. Most frequently,
supporting electrolytes are added to both solvents and
the system is studied utilising a four-electrode potentiostat [13]. Due to some a priori limitations of such
an experimental arrangement, which are related mainly
to the narrow potential window, the transfer of many
organic and inorganic ions is so far not accessible.
Osborne et al. [14] suggested that in the complete
absence of the electrolyte in the aqueous phase and a
micromolar concentration of electrolyte in the organic
phase, the potential window of the ITIES method is
considerably widened using a single microhole interface. Quinn et al. [12] published a theoretical treatment
of this approach. Shi and Anson [15,16] developed a
voltammetric method to study the ion transfer by using
electrodes that are covered with a film of an immiscible
organic liquid.
Recently, a novel experimental approach for exploring the ion transfer between two immiscible solutions
was developed [17 19]. According to the method employed, a droplet of a water-immiscible organic solvent
containing an electrochemically active compound is at-

0022-0728/02/$ - see front matter 2002 Published by Elsevier Science B.V.

PII: S 0 0 2 2 - 0 7 2 8 ( 0 2 ) 0 0 7 2 0 - 9

190

V. Mirceski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

tached to the graphite electrode and immersed into an

aqueous electrolyte solution. These authors assumed
that the electrochemical reaction of the electroactive
compound can proceed only along the three-phase
junction line graphite  water  organic phase, since the
organic phase does not contain a deliberately added
supporting electrolyte. With this simple experimental
set-up, the transfer of ions across the liquid  liquid
interface can be studied utilising a standard three-electrode potentiostat, and, moreover, the potential window of the experiment is extended considerably, as the
organic phase does not contain any dissolved
electrolyte.
The proposed method is interesting because simultaneous electron and ion transfers are realised in the
three-phase arrangement via two interfaces, i.e. the
electron transfer across the graphite  organic liquid interface and the ion transfer across the water  organic
liquid interface.
Recently, Marken et al. [20 23] and Schro der et al.
[24] studied the voltammetric response of droplets of
organic electroactive liquids attached to the surface of
an electrode. They also observed that the ion transfer
accompanies the redox transformation of the micro
droplets. However, in these experiments, the liquid
interface is itself electroactive, and its properties change
severely in the course of the electrochemical reaction.
Furthermore, so far it was not possible to determine the
Gibbs energy of ion transfer for these liquids.
In this paper, we describe the redox reaction of
iodine at the three-phase junction water  organic
phase  graphite, where the organic phase is nitrobenzene, tetrachloromethane, or paraffin oil. The results
obtained widen the understanding of the fundamentals
of these experiments, extending our basic knowledge of
electrochemical reactions occurring at three-phase arrangements, where an electron transfer between the
electron conductor and the organic liquid phase is
coupled with an ion transfer between the two liquid
phases. The most important news from this study is
that a fast partition of the electrolyte that is initially
present only in the aqueous phase occurs between the
two liquid phases. This allows the electrochemical reaction to proceed as an electron transfer across the electron conductor  organic liquid interface and a
simultaneous ion transfer across the organic liquid  aqueous solution interface. Thus the electrochemical reaction is no longer confined to the three-phase
junction line, but it can occur simultaneously at the two
interfaces. It is still reasonable to assume that the
reaction starts at the three-phase junction line, however,
due to the presence of partitioned electrolyte in the
organic liquid droplet, it may quickly advance through
the droplet.

2. Experimental
A 0.1 mol l 1 solution of iodine was prepared by
dissolving iodine in nitrobenzene, tetrachloromethane,
and paraffin oil. A drop of these solutions, with a
volume of 2 ml was attached to the surface of a paraffin
impregnated graphite electrode (PIGE) with the help of
an Eppendorf-type pipette. The approximate radius of
the hemispherical droplet was 0.15 cm. During squarewave voltammetric experiments, neither a significant
distortion nor a change of the colour of the droplet has
been observed. The dimension of the droplets influences
only the measured currents, whereas both the position
of the response (peak potential) and shape (half-peak
width) are not affected.
Square-wave voltammograms were recorded using
the electrochemical measuring system AUTOLAB
(Eco-Chemie, Utrecht, Netherlands). A Ag  AgCl 
saturated NaCl solution reference electrode (E=0.200
V vs. SHE) was used and a platinum wire served as the
auxiliary electrode. All salts used for preparation of the
supporting electrolytes were used as purchased. All
chemicals used were of analytical grade. The paraffin
oil used (Merck Index 11, 7139) was of IR spectroscopy grade. The preparation of PIGEs is described
elsewhere [25]. The water used was Millipore Q. All
experiments were carried out at a room temperature of
about 20 C.

3. Results and discussion

In an aqueous medium iodine can be reversibly reduced to iodide in a simple two-electron redox reaction
[26] characterised by the standard potential E I0 (s)/2I =
2
0.535 V vs. SHE and E I0 (aq)/2I = 0.621 V vs. SHE for
2
the solid and dissolved iodine, respectively.
The SW voltammetric response of a nitrobenzene
droplet containing dissolved iodine, attached to the
graphite electrode and immersed in 1 mol l 1 KNO3
aqueous solution, consists of a single well-defined SW
peak with a peak potential of 0.467 V vs.
Ag  AgCl  NaCl sat. and a half-peak width of DEp/2 =
112 mV (Fig. 1). The forward and backward components of the SW response are well-developed
bell-shaped curves characterised by almost equal
heights and a peak potential separation of 15 mV.
Over the frequency interval from 20 to 300 Hz, the
position and the shape of the response remained virtually unaltered, with a peak potential variation within
9 6 mV and half-peak width variation within 93 mV.
Within the same frequency interval, the peak current is
proportional to the square root of the SW frequency,
associated with a correlation coefficient of the linear
regression line of R 2 = 0.985. The slight scattering of
the experimental points of this plot arises from the

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

Fig. 1. Forward (1), backward (2), and net SW (3) components of the
square-wave voltammetric response of a nitrobenzene droplet with a
volume of 2 ml containing 0.1 mol l 1 iodine attached to a PIGE
which is immersed in 1 mol l 1 KNO3 solution. The experimental
conditions were: SW frequency f=50 Hz, SW amplitude Esw =50
mV, and scan increment dE=0.15 mV.

uncertainty of the dimensions of the nitrobenzene

droplet attached to the graphite electrode, which cannot be controlled more precisely. It is also important to
emphasise that the ratio of the peak currents of the
forward (reduction) and backward (oxidation) components of the SW response remained virtually constant
by varying the signal frequency. This indicates that the
electrode reaction is not complicated by a follow-up
chemical reaction, as can be expected for the reduction
of iodine [26].
The square-wave amplitude markedly affects the
height and shape of the SW voltammetric response of
iodine. Increasing the amplitude from 10 to 80 mV
caused the peak current to increase linearly. The halfpeak width DEp/2 increases from 126 to 146 mV, for the
amplitude of Esw = 10 and 120 mV, respectively. At the
same time, the peak potential scatters slightly around
the average value of 0.480 V vs. Ag  AgCl by 9 7 mV.
For the sake of brevity, the data of the foregoing
analyses are not presented in the form of diagrams.

191

On the basis of the properties of the voltammetric

response observed by varying the SW excitation signal
parameters, it can be concluded generally that the
electrode reaction of iodine in the three-phase arrangement exhibits properties of a reversible electrochemical
process. Moreover, the voltammetric behaviour is typical for planar diffusion mass transport, which appears
to be surprising, when the reaction is assumed to proceed only along the three-phase junction line. It is also
worth noting that, within the frequency interval explored, ohmic resistance, which would manifest mainly
as a deviation from peak shape, does not affect the
reaction. Before providing a reasonable explanation for
the diffusion regime, an attempt will be made to elucidate the possible electrode mechanism for reduction of
iodine at the three-phase boundary.
To elucidate the electrode mechanism of iodine reduction, the electroneutrality of the organic phase has
to be taken into account. Thus, it is reasonable to
expect that the reduction of iodine to iodide is followed
either by the transfer of electrochemically formed iodide ions from nitrobenzene into the aqueous phase, or
by entering of cations, initially present in the aqueous
phase into the nitrobenzene droplet. The postulated
scenario occurring in the three-phase system is depicted
schematically in Fig. 2, and ascribed to the following
electrode reactions:

I2(nb) + 2e ? 2I(aq)

I2(nb) + 2Cat+
? 2I(nb)
+ 2Cat+
(aq) + 2e
(nb)

(I)
(II)

Whereas the reaction (I) can occur at the three-phase

boundary, the overall process (II) can be split into two
reactions
occurring
simultaneously
at
the
graphite  nitrobenzene interface (IIa) and the nitrobenzene  water interface (IIb):

I2(nb) + 2e ? 2I(nb)
+
(aq)

2Cat

? 2Cat

+
(nb)

(IIa)
(IIb)

In order to clarify whether the transfer of cations

accompanies the redox reaction from aqueous phase to
organic droplet, the reduction of iodine was studied in
a variety of nitrate solutions containing different

Fig. 2. Electrode assembly with a droplet of nitrobenzene containing dissolved iodine attached to a PIGE, which is immersed in an aqueous
electrolyte solution. Possible electrode reactions are formulated (see text).

192

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

Fig. 3. Net SW voltammetric responses of nitrobenzene droplet containing 0.1 mol l 1 iodine attached to a PIGE, which is immersed in a 1 mol
l 1 aqueous solution of NaNO3 (1), LiNO3 (2), KNO3 (3), and Mg(NO3)2 (4). Other experimental conditions were the same as in Fig. 1.

cations (Fig. 3). As can be seen, the particular type of

cations present in the aqueous phase exhibits no influence on the position and shape of the voltammetric
response, suggesting that the entering of cations does
not preserve the electroneutrality of the organic phase.
Therefore, under these experimental conditions, the reduction of iodine at a three-phase boundary most probably proceeds according to the redox mechanism (I).
By an analogy of features of the electrode reduction
of iodine in an aqueous medium [26] it is reasonable to
suppose that the mechanism (I) could be complicated
additionally by a follow-up chemical reaction in which
I
3 is formed according to the reaction:

I2(nb) +I(nb)
? I3 (nb)

(III)

However, both products I and I

3 are electroactive
and thus both can be reoxidised back to the same
reactant I2 at almost an identical potential. If this is not
true, the backward component of the SW response
should consist of two separated or overlapped peaks.
The experimental results do not support this presumption. Therefore, the overall reaction (I) does not proceed as an EC mechanism. It should be also noted that
I
could be further reduced to 3I. In an aqueous
3
medium, the standard potential of the redox couple

I
is about 80 mV more negative than that of the
3 /3I

I2/2I couple. Therefore, it can be assumed that the

overall reduction of iodine can proceed as an ECE%
mechanism, where E represents the reduction of iodine
to I, C is the chemical step in which I
3 is created, and
E% represents its further electrochemical reduction to
3I. It is quite clear that in the present experimental
arrangement, there are no indications for such a behaviour; however this possibility cannot be excluded on
the basis of our experiments. Generally speaking, the
equilibrium (III) plays a significant role in the overall

mechanism of iodine reduction only in the presence of

concentrated iodide solutions [26].
At this point, it is also meaningful to consider the
possibility of iodine reduction at the graphite  water
interface because of the partition of iodine between the
organic and aqueous phase. In the classical experiments
with a saturated solution of I2 in the aqueous phase, the
peak currents of I2 reduction were about 23 orders of
magnitude smaller than in the droplet experiments.
Therefore, the possibility that the response of I2 in the
aqueous phase contributes considerably to the results
can be excluded.
A thermodynamic treatment of reaction (I) yields the
following Nernst equation:
E= E c0%+

RT
a
ln 2I2(nb)
2F
a I(aq)

(1)

where

0
E c0%= E I0 /2I + Dnb
aq I
2

(2)

The standard potential of the transfer of ions Xz at the

0
water  nitrobenzene interface Dnb
aq Xz is related to the
standard Gibbs energy of transfer across the liquid

0
0
nb
interface Dnb
aq G Xz through the relation: Daq Xz =

0
Dnb
G
aq Xz
(z is 1 for monovalent anions and +1 for
zF

0
1
[27],
monovalent cations). As Dnb
aq G I = 18 kJ mol
the standard potential of iodide transfer at the wa
0
ter  nitrobenzene interface is Dnb
aq I = 0.186 mV. Since

0
E I /2I = 0.335 V vs. Ag  AgCl, the position of the SW
2
voltammetric response of iodine in various nitrate solutions (Ep = 0.470 V vs. Ag  AgCl) is close to the expected formal potential of reaction (I) (E c0%= 0.521 V
vs. Ag  AgCl) suggesting that the redox reaction (I)
describes correctly the reduction pathway of iodine at
the three-phase boundary.

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

193

where the formal potential is

0
E c0%= E I0 /2I + Dnb
aq cat+ +
2

RT
lnacat+(aq).
F

(4)

The standard potential difference of a particular monovalent cation is related to the standard Gibbs energy of
transfer through the relation:

0
Dnb
aq cat+ =

Fig. 4. Net SW voltammetric responses of nitrobenzene droplet

containing 0.1 mol l 1 iodine attached to a PIGE, which is immersed
in 1 mol l 1 solution of NaCl (1), NH4Cl (2), CsCl (3), and KCl (4).
Other experimental conditions were the same as in Fig. 1.

Fig. 5. The dependence of the peak potential of the second SW peak,

observed in chloride aqueous solutions, on the concentration of CsCl
and NaCl, used as supporting electrolytes. Other experimental conditions were the same as in Fig. 1.

A similar thermodynamic treatment of the postulated

reaction mechanism (II) leads to the following form of
the Nernst equation:
E =E c0%+

RT
a
ln 2 I2(nb)
2
2F
a I(nb)a cat
+(nb)

(3)

0
Dnb
aq G cat+
F

since the Gibbs energies of transfer of all cations

present in the supporting electrolytes are positive values

0
[27], it follows that Dnb
aq cat+ B 0. Therefore, E c %(II)B

0
E c %(I), where E c %(I) and E c %(II) are the formal potentials of the redox mechanisms (I) and (II), respectively.
In other words, the preservation of the electroneutrality
of the organic phase by entering of cations requires a
higher Gibbs energy than the process in which iodide
ions leave the organic phase. Obviously, the redox
mechanism (II) cannot occur at all, since it is always
preceded by the process (I), which occurs at more
positive potentials. This, at least, holds true for the
alkali metal cations.
The most interesting results have been observed in
the presence of chloride ions in the aqueous phase. Fig.
4 represents the SW voltammetric response of iodine
recorded in a series of electrolyte solutions containing
chloride ions. Very unexpectedly, the presence of chloride gives rise to a new well-defined SW peak at more
positive potentials. The position of this peak is not
affected by the nature of the particular cations, however it depends strongly on the concentration of chloride ions in the aqueous phase. As can be seen from
Fig. 5, the relationship Ep vs. log(cCl) is characterised
by a negative slope of 43 and 53 mV using NaCl
and CsCl as supporting electrolytes, respectively.
It seems that the appearance of the SW peak at more
positive potentials is related to an electrode reaction
pathway in which the electroneutrality of the organic
phase is achieved by the transfer of chloride ions from
nitrobenzene into the aqueous solution (see Fig. 6).
Considering the fact that no supporting electrolyte was

Fig. 6. Electrode assembly with a droplet of nitrobenzene containing dissolved iodine attached to the graphite electrode, which is immersed in an
aqueous electrolyte solution containing chloride ions.

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

194

Fig. 7. Correlation between theoretically calculated and experimentally measured peak potential separations for different concentrations
of NaCl and CsCl supporting electrolytes. Other experimental conditions were the same as in Fig. 1.

added to the organic phase, the observed electrode

mechanism is rather unexpected. The postulated electrode reaction may be described by the following
scheme:

+ 2e ? 2I(nb)
+2Cl(aq)
I2(nb) +2Cl(nb)

(IV)

The following form of the Nernst equation describes

the electrode reaction:
E= E c0%+

RT aI2(nb)a 2Cl(nb)
ln
2F
a I2(nb)

(5)

with a formal potential of

E %=E

0
c

0
I2/2I

+D
nb
aq

0
Cl

RT

ln aCl-(aq)
F

(6)

According to the latest equation, the formal potential

of the redox reaction (IV) should depend linearly on the
logarithm of the activity of chloride ions in the aqueous
phase, with a negative slope of 57 mV (at 291 K).
The experimental results presented in Fig. 5 are in close
agreement with the theoretically predicted value.
Furthermore, as both the standard Gibbs energy of
transfer and the standard potential of transfer of chloride ions at the water  nitrobenzene interface are posi
0
1
tive
values
(Dnb
and
aq G Cl =35 kJ mol
0
nb
Daq Cl =0.362 V), it is clear why this process occurs
at more positive potentials than the transfer of iodide
ions (reaction mechanism I).
If both SW peaks observed in the chloride solutions
are correctly assigned to the electrode mechanisms (I)
and (IV), then the potential difference between them
should correspond to the difference of the formal potentials of the reactions (IV) and (I) defined by Eqs. (6)
and (2), respectively:

0
0
nb
DEp =Dnb
aq Cl Daq I

RT
ln aCl-(aq)
F

(7)

The peak potential separations have been measured

for different concentrations of chloride ions in both
NaCl and CsCl solutions and compared to the theoretical values calculated with Eq. (7). A simple correlation
of the measured and expected values reveals fairly good
agreement between these values (see Fig. 7). All these
findings strongly support the correctness of the postulated electrode reaction (IV).
A possible explanation for the presence of chloride
ions in the nitrobenzene droplet and their involvement
in the electrode reaction (IV) is a rather fast partition of
the supporting electrolyte ions between the aqueous
and organic phases, once the electrode with the nitrobenzene droplet is introduced into the chloride containing aqueous electrolyte solution:

Cat+
(aq) + Cl(aq) ? Cat(nb) + Cl(nb)

(V)

Governed by the large gradient of electrochemical

potentials acting at the liquid  liquid interface, a certain
amount of the supporting electrolyte ions enters the
organic phase. However, the amount of the chloride
ions in the organic phase due to partition of the supporting electrolyte is insufficient to give rise to such a
high and well-developed second SW peak. Some additional mechanism has to occur, increasing considerably
the chloride concentration in nitrobenzene. Most probably, a significant role is played by the reaction between
iodine and chloride ions to form I2Cl in the organic
phase, since the formation of interhalide compounds is
a well-known property of iodine [28]:

Cl(nb)
+ I2(nb) ? I2Cl(nb)
(VI)
Thus, a kind of reacti6e partition occurs at the
water  nitrobenzene interface increasing considerably
the amount of chloride in the organic phase. Once
present in the organic phase, at least close to the
three-phase boundary, the partitioned ions can participate in the electrochemical reaction.
The amount of ions transferred initially across the
liquid  liquid interface is determined by the Gibbs energy of transfer of both anions and cations of the
partitioned salt. Thus, ions with a high Gibbs energy of
transfer cannot give rise to SW peaks at more positive
potentials, since they cannot be partitioned between the
aqueous and the organic phase. This can be seen by a
comparative analysis carried out in the presence of
fluoride, bromide and chloride ions (Fig. 8). In a solution containing fluoride ions, the peak at more positive
potentials was not observed, indicating clearly that the
redox process associated with the transfer of fluoride
ions from the organic into aqueous phase does not
occur. Regarding the high energy of transfer of fluoride

0
1
ions Dnb
[27], it is clear that the
aq G F = 70 kJ mol
amount of partitioned fluoride ions is much too small
to allow participation in the electrode mechanism. The
influence of different cations of the partitioned halide
salts will be described in the following publication.

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

In the presence of bromide ions, an electrode reaction

with participation of bromide ions occurs only as a
shoulder on the positive side of the main SW peak
(curve 2 in Fig. 8). The formal potential of the electrode
reaction follows as:

0
E c0%=E I0 /2I + Dnb
aq Br
2

RT
ln aBr-(aq)
F

(8)

Since bromide ions are characterised by a lower

transfer energy than chloride ions, the following rela
0
0
nb
tion holds: Dnb
aq Br BDaq Cl. Hence, the process involving bromide ions is situated at more negative
potentials than that involving chloride ions (Eqs. (6)
and (8)).
The extent of partition can also be varied by the
choice of the organic solvent. Fig. 9 shows the voltammetric responses when iodine was dissolved in a droplet

Fig. 8. SW voltammetric responses of nitrobenzene droplets containing 0.1 mol l 1 iodine attached to a PIGE, which is immersed in 1
mol l 1 aqueous solution of NaF (1), NaBr (2), and NaCl (3). Other
experimental conditions are the same as in Fig. 1.

of paraffin, tetrachloromethane, and nitrobenzene. In

all three cases the aqueous electrolyte contained 1 mol
l 1 NaCl. Using paraffin as the organic solvent, only
one single SW peak occurs and this peak can be
explained by the reaction mechanism in which the
electrochemically formed iodide ions leave the organic
phase (reaction I). In this case, the partition of chloride
ions between the organic and aqueous phases does not
take place to a measurable extent and the process at
more positive potentials cannot be observed. Although
the Gibbs energy of the transfer of chloride from water
into paraffin is not known, it is reasonable to suppose
that this value will be very large, especially when taking
into account the non-polarity of paraffin.
Carrying out an experiment with tetrachloromethane,
one observes a voltammogram that is similar to that
with nitrobenzene (curves 2 and 3 in Fig. 9). However,
the peak ascribed to the transfer of chloride ions from
the organic phase into water appears at more positive
potentials than the corresponding one in the case of
nitrobenzene. This is also an expected result, since the
transfer of chloride to tetrachloromethane, as a non-polar solvent, will require a rather high Gibbs energy.
If the Gibbs energy of transfer of the supporting
electrolyte anions is close to that of iodide, then an
electrode reaction analogous to (IV) will not produce a
separate SW peak at more positive potentials, since the
formal potentials of the mechanisms (I) and (IV) will
not differ sufficiently.
In some cases, when the transfer energy of the
present anions in the aqueous phase is low enough to
allow a considerable amount of ions to enter into the
organic phase, the reaction mechanism analogous to
(IV) can become predominant and the single reaction
pathway for the reduction of iodine at the three-phase
boundary occurs. Such behaviour was observed in the
presence of perchlorate ions, which are characterised by

0
a Gibbs energy of transfer of Dnb
aq G ClO =
4
1
10 kJ mol . If the concentration of perchloric acid in
the aqueous phase is lower than 0.05 mol l 1, then the
reduction of iodine mainly occurs according to the
mechanism (I). However, at higher concentrations of
perchloric acid in the aqueous phase, the amount of
partitioned perchlorate ions in the organic droplet increases so much, that the electrode reaction switches
from mechanism (I) to mechanism analogous to (IV).
The latter is associated with the following formal
potential:

0
E c0%= E I0 /2I + Dnb
aq ClO
2

Fig. 9. SW voltammetric response of a paraffin (1), a tetrachloromethane (2), and a nitrobenzene (3) droplet containing 0.1 mol
l 1 dissolved iodine, attached to a PIGE and immersed in 1 mol l 1
solution of NaCl. Other experimental conditions were the same as in
Fig. 1.

195

RT
ln aClO(aq)
4
F

(9)

Due to the influence of the last term on the right-hand

side of Eq. (9), the voltammetric response is strongly
shifted in the negative direction (see Fig. 10). Within
the concentration range from 0.05 to 2 mol l 1, the
peak potential depends linearly on the logarithm of the

196

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

tion to the assumption that the reaction takes place

only at a single reaction line. For these reasons, the
theoretical model of Myland and Oldham based on a
cylindrical diffusion model as closest approximation to
the single reaction line, did not match the experimental
results [29].
For these reasons, in Appendix A a theory for modelling reaction (I), based on planar diffusion mass
transport is presented. A brief correlation between the
experimental and theoretical results is carried out in
order to support the presented thesis derived from the
experimental results in this paper.

Fig. 10. SW voltammetric response of a nitrobenzene droplet containing 0.1 mol dm 3 iodine attached to a PIGE, which is immersed in
a 1 mol l 1 (1), 3 mol l 1 (2), 5 mol l 1 (3), and 6 mol l 1 (4)
aqueous solution of HClO4. Other experimental conditions were same
as in Fig. 1.

concentration of perchloric acid in aqueous phase with

a slope of
DEp
= 64 mV
Dlog(cHClO4)
which is in good agreement with Eq. (9).

4. Conclusions
The above described results extend the basic knowledge of the nature of electrode reactions proceeding in
a three-phase arrangement, when an electron transfer is
simultaneously accompanied by an ion transfer across a
liquid  liquid interface. Obviously, in many cases, a fast
partitioning of the supporting electrolyte ions between
the organic and aqueous phases occurs and plays an
important role in the electrode reaction. This partitioning considerably increases the number of possible electrode mechanisms. The presence of an appreciable
amount of ions in the organic phase, at least in the
region where the electrode reaction starts, appears to be
the most crucial consequence arising from the partition
phenomenon. Consequently, the early hypothesis that
the reaction occurs exclusively at the three-phase junction line has to be modified at least by considering a
more extended reaction zone that is located around the
line where all three phases, organic solvent, aqueous
solution, and electrode meet. The reaction zone is large
enough to allow the reaction to be controlled by planar
diffusion mass-transport without an appreciable IR
drop. This fact explains the large currents observed in
this study, as well as in the previously reported experiments carried out with ferrocene and decamethylferrocene [17 19]. The current commonly measured with
this experimental arrangement was in strong contradic-

5. List of symbols and abbreviations

c*I2(nb)
cI2(nb)
cI
(aq)

D
E
F
f
I
n
R
S
T
t
x
rs

concentration of iodine in the bulk of

nitrobenzene
concentration of iodine in nitrobenzene
concentration of iodide in the aqueous
solution
diffusion coefficient
potential of the working electrode
Faraday constant
SW frequency
current
number of electrons
gas constant
electrode surface area
temperature
time
coordinate normal to the electrode surface
auxiliary constant in units of mol1 cm3

Acknowledgements
V. Mirc eski thanks A.V. Humboldt-Stiftung for provision of a Humboldt fellowship and F. Scholz acknowledges
kind
support
by
Deutsche
Forschungemeinschaft (DFG) and Fonds der Chemischen Industrie (FCI). The authors wish to thank Professor K.B. Oldham for several suggestions and
discussion.

Appendix A
It is assumed that the mass transport of all species
participating in the electrode reaction

I2(nb) + 2e ? 2I(aq)

(AI)

occurs by planar diffusion. As a first approximation it

is supposed that the diffusion coefficients of iodine and
iodide are equal, thus the following mathematical
model represents reaction (I):

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

#2cI2(nb)
#cI2(nb)
=D
#t
#x 2

(A1)

#cI(aq)
#2cI(aq)

=D
#t
#x 2

(A2)

t =0, x ] 0:

cI2(nb) =c*I2(nb), cI =0

(a)

t \0, x 8 :

cI2(nb) c*I2(nb), cI 0

(b)

x =0:

(aq)

(cI2(nb)
(x

I
(cI2(nb)
=D
nFS
(x

 

(aq)

(cI
1
(aq)
= D
(x
2
x=0

(c)
x=0

(d)

 
x=0

I
1
(cI
(aq)
= D
nFS
2
(x

(e)

x=0

Under equilibrium conditions at the three-phase junction line, the following condition holds (activity coefficients neglected):
cI2(nb) =rsc 2I exp()

(A3)

(aq)

0
where =(2F)/(RT)(E E I0 2I Dnb
aq I) is the di2
mensionless potential. The meanings of all symbols are
listed in Section 5. The solution expressed in the form
of an integral equation reads:

c*I2(nb)

&

&

I(~)

d~

nFS
D
y(t ~)
2
I(~)
d~

exp()
(A4)
nFS
D
y(t ~)
The numerical solution of the last equation was

= rs

Fig. 11. The forward (cf), backward (cb), and net SW (cnet)
components of the theoretical square-wave voltammetric response of
the reaction (I). The conditions of the simulations were: SW amplitude Esw =50 mV, scan increment dE= 1.5 mV, and concentration of
iodine c*I2(nb) = 0.1 mol cm 3. The dimensionless current is defined
as: c =I(nFSc*I2(nb)) 1(Df ) 1/2.

197

obtained by the numerical method of Nicholson and

Olmstead [30].
The theoretical SW voltammograms are functions of
the frequency and amplitude of the excitation signal, as
well as of the bulk concentration of iodine, since reaction (I) is of second order with respect to the product of
the electrode reaction.
The theoretical SW voltammetric response of reaction (I), simulated under identical instrumental parameters as those used in the experimental study, is depicted
in Fig. 11. Obviously, the shape of the theoretical
response is in accordance with the experimental SW
voltammograms (compare Figs. 1 and 11). The separation between the forward and backward components of
the theoretical SW response is DEp = 20 mV, whereas
the half-width is DEp/2 = 110 mV. The properties of the
theoretical response are close to the corresponding values of the experimental voltammogram recorded in 1
mol l 1 KNO3, which are: DEp = 15 mV and DEp/2 =
113 mV (see Fig. 1).
The analysis of the system by varying the SW frequency provides the main criterion for characterising a
reversible electrode mechanism controlled by a planar
diffusion mass transport. The peak current of the theoretical voltammograms is a linear function of the
square root of the SW frequency, while both the peak
potentials and half-peak widths are independent of
frequency. Obviously, the experimental (see Section 3)
and the theoretical behaviour are in excellent
agreement.
The signal amplitude affects mainly the height and
the shape of the theoretical voltammograms. The height
of the theoretical response increases linearly over the
interval from Esw = 10 to Esw = 40 mV, while the peak
potentials remain virtually unaltered. Above this range,
the peak current is unaffected by the signal amplitude.
At the same time, the half peak width increases from 70
to 150 mV, for amplitudes of 10 and 90 mV,
respectively.
In the experimental study, the results collected from
the analysis by varying the signal amplitude deviate
slightly from the theoretically predicted data. For instance, the peak current increases linearly over a wide
amplitude interval from 10 to 80 mV. For an amplitude
of 10 mV, the half-width of the experimental peak is
126 mV, which is larger than the theoretical value of 70
mV, while for an amplitude of 90 mV, the experimental
and theoretical values are virtually the same. Most
probably, the influence of the reaction mechanism (III)
in the presence of nitrate ions, on the main electrode
process (I), leads to a slight deviation between the
experimental and theoretical data.
In general, the theoretical model is in good qualitative agreement with the experimental results, supporting
the conclusion derived from the experimental study of
the reduction of iodine at the three-phase boundary.

198

V. Mirc eski, F. Scholz / Journal of Electroanalytical Chemistry 522 (2002) 189198

References
[1] H.H.J. Girault, D.J. Schiffrin, in: A.J. Bard (Ed.), Electroanalytical Chemistry, vol. 15, Marcel Dekker, New York, 1989, p. 1.
[2] J. Koryta, P. Vanysek, in: H. Gerischer, C.W. Tobias (Eds.),
Advances in Electrochemistry and Electrochemical Engineering,
vol. 12, Wiley, New York, 1981, p. 113.
[3] P. Vanysek, Modern Techniques in Electroanalysis, Wiley, New
York, 1996, p. 337.
[4] A.G. Volkov (Ed.), Liquid Interfaces in Chemical, Biological,
and Pharmaceutical Applications, Marcel Dekker, 2001, p. 1.
[5] M. Suzuki, J. Electroanal. Chem. 372 (1994) 39.
[6] M.A. Rahman, H. Doe, J. Electroanal. Chem. 424 (1997) 159.
[7] G. Geblewicz, A.K. Kontturi, K. Kontturi, D.J. Schiffrin, J.
Electroanal. Chem. 217 (1987) 261.
[8] B. Hundhammer, T. Solomon, T. Zerihum, M. Abegaz, A.
Bekele, K. Graichen, J. Electroanal. Chem. 371 (1994) 1.
[9] Z. Ding, R.G. Wellington, P.F. Brevet, H.H.J. Girault, J. Electroanal. Chem. 420 (1997) 35.
[10] Z. Samec, J. Langmaier, A. Trojanek, J. Electroanal. Chem. 426
(1997) 37.
[11] K. Kontturi, J.A. Manzanares, L. Murtomaki, D.J. Schiffrin, J.
Phys. Chem. B 101 (1997) 10801.
[12] B. Quinn, R. Lahtinen, L. Murtomaki, J. Electroanal. Chem.
460 (1999) 149.
[13] S. Wilke, H. Wang, J. Electroanal. Chem. 475 (1999) 9.
[14] M.C. Osborne, Y. Shao, C.M. Pereira, H.H. Girault, J. Electroanal. Chem. 364 (1994) 155.
[15] C. Shi, F.C. Anson, J. Phys. Chem. B 102 (1998) 9850.
[16] C. Shi, F.C. Anson, J. Phys. Chem. B 103 (1999) 6283.

[17] F. Scholz, S& . Komorsky-Lovric , M. Lovric , Electrochem. Commun. 2 (2000) 112.

[18] S& . Komorsky-Lovric , M. Lovric , F. Scholz, J. Electroanal.
Chem. 508 (2001) 129.
[19] S& . Komorsky-Lovric , M. Lovric , F. Scholz, Collect. Czech.
Chem. Commun. 66 (2001) 434.
[20] F. Marken, R.D. Webster, S.D. Bull, S.G. Davies, J. Electroanal. Chem. 437 (1997) 209.
[21] F. Marken, R.G. Compton, Ch.H. Goeting, J.S. Foord, S.D.
Bull, S.G. Davies, Electroanalysis 10 (1998) 821.
[22] F. Marken, A.N. Blythe, J.D. Wadhawan, R.G. Compton, S.D.
Bull, R.T. Aplin, S.G. Davies, J. Solid State Electrochem. 5
(2001) 17 22.
[23] J.D. Wadhawan, R.G. Compton, F. Marken, S.D. Bull, S.G.
Davies, J. Solid State Electrochem. 5 (2001) 348.
[24] U. Schro der, R.G. Compton, F. Marken, S.D. Bull, S.G. Davies,
S. Gilmour, J. Phys. Chem. B 105 (2001) 1344.
[25] F. Scholz, B. Meyer, in: A.J. Bard, I. Rubinstein (Eds.), Electroanalytical Chemistry. A Series of Advances, vol. 20, Marcel
Dekker, New York, 1998, p. 1.
[26] A.J. Bard, R. Parsons, J. Jordan, Standard Potentials in
Aqueous Solutions, Marcel Dekker, New York, 1985, p. 84.
[27] Y. Marcus, Ion Properties, Marcel Dekker, New York, 1997, p.
216.
[28] J.E. Huheey, Anorganische Chemie, Prinzipien von Struktur und
Reaktivita t, Walter de Gruyter, Berlin, New York, 1988, p. 853.
[29] J.C. Myland, K.B. Oldham, Electrochem. Commun. 2 (2000)
541.
[30] R.S. Nicholson, M.L. Olmstead, in: J.S. Mattson, H.B. Mark,
H.C. MacDonald (Eds.), Electrochemistry: Calculations, Simulation and Instrumentation, vol. 2, Marcel Dekker, New York,
1972, p. 120.