Electrochimica Acta 367 (2021) 137566

Contents lists available at ScienceDirect

Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta

Coulometric response of solid-contact anion-sensitive electrodes

Tingting Han, Ulriika Mattinen, Zekra Mousavi, Johan Bobacka∗
Åbo Akademi University, Johan Gadolin Process Chemistry Centre, Laboratory of Molecular Science and Engineering, Biskopsgatan 8, Åbo FI-20500, Finland

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

Article history: In this work, the coulometric response of nitrate, perchlorate, and sulfate solid-contact anion-sensitive
Received 8 September 2020 electrodes was investigated. The coulometric transduction method was originally introduced and so far
Revised 1 November 2020
mainly studied for solid-contact cation-selective sensors using poly(3,4-ethylenedioxythiophene) (PEDOT)
Accepted 24 November 2020
doped with poly(styrene sulfonate) as transducer. Here we provide additional proof-of-concept for the
Available online 27 November 2020
coulometric transduction method by focusing on the electrochemical characteristics of anion sensors. PE-
Keywords: DOT was electrodeposited in the presence of small anions, including chloride, nitrate, sulfate, and per-
Solid-contact anion-sensitive electrodes chlorate. The counterion was found to influence the yield of electroactive PEDOT, which is an important
Anion sensors parameter for the coulometric response. Anion-sensitive electrodes were prepared by coating the PEDOT
Coulometric responses solid contact with plasticized PVC-based anion-sensitive membranes by drop-casting or spin-coating. The
Impedance influence of the thickness of the PEDOT film and the anion-sensitive membrane on the coulometric re-
PEDOT
sponse was studied in detail. The solid-contact anion sensors showed fast charge transfer and ion trans-
port properties, making them suitable for coulometric sensing. Also for anion-sensitive electrodes, the
analytical signal was amplified by increasing the thickness of the PEDOT solid contact, which is fully
consistent with earlier works on cation-selective electrodes. This work shows that the coulometric trans-
duction principle is feasible and robust for various types of solid-contact ISEs.
© 2020 The Author(s). Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)

1. Introduction
Ion-selective electrodes (ISEs) are well-known electrochemical
sensors that are used in a broad range of applications, such as
clinical diagnostics, environmental analysis, and industrial pro-
cess monitoring [1–5]. Conventional ISEs with internal filling so-
lution have limitations as they have to be used in the verti-
cal position, need regular refilling, and are not easily minia-
turized. This has resulted in extensive research on solid-contact
ion-selective electrodes (SCISEs) over the last three decades [1].
Among many other materials, the conducting polymer poly(3,4-
ethylenedioxythiophene) (PEDOT) has often been used as a solid
contact (ion-to-electron transducer) in SCISEs [1,6–10]. The redox
capacitance related to the reversible oxidation of PEDOT has been
utilized to improve the potential stability [6,11,12] and to control
the standard potential [13,14] of SCISEs. The morphology and ca-
pacitance of PEDOT can, however, be influenced by the doping
counterion during electropolymerization [7,11,15,16]. For example,
PEDOT films doped with negatively charged MWCNTs were found
to exhibit higher redox capacitance than PEDOT doped with chlo-
∗ the solid contact (PEDOT) and a transient current is measured and
Corresponding author.
E-mail address: johan.bobacka@abo.fi (J. Bobacka).
ride ions [11]. PEDOT has been used less frequently as ion-to-
electron transducer in anion sensors [17–21], compared to cation-
selective SCISEs. Actually, oxidized (p-doped) conducting polymers,
such as PEDOT, should be well suited as transducers in anion-
sensitive SCISEs. Therefore, the lower number of anion-selective
SCISEs can be related mainly to the lack of ionophores that are
highly selective to anions.
ISEs are generally used in the potentiometric transduction
regime where the potential is related to the ion activity of pri-
mary ion in solution, according to the Nernst equation [1]. In ad-
dition to potentiometry, several other electrochemical approaches
have been explored for ISEs. Various electrochemical interroga-
tion methods for ISEs have been introduced and developed by
Bakker’s group [9,22,23]. Amemiya et.al [24] was able to determine
nanomolar concentrations of perchlorate in drinking water by us-
ing ion-transfer stripping voltammetry. Recently we introduced a
coulometric transduction method for SCISEs, which is the topic of
this work [25]. In this method, the potential between the SCISE
and the reference electrode is held constant and the current be-
tween the SCISE and a counter electrode is measured. A change
of the ion activity in solution gives rise to oxidation/reduction of
integrated to obtain the charge which is the analytical measured

https://doi.org/10.1016/j.electacta.2020.137566
0013-4686/© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
T. Han, U. Mattinen, Z. Mousavi et al.
signal [25]. The cumulated charge can be amplified by increasing
the thickness (capacitance) of the solid contact [26, 27]. In order
to decrease the response time, the resistance of the ion-selective
membrane (ISM) was lowered by enlarging the geometric surface
area [28] and by using spin-coated thin-layer ISMs [29]. Theoretical
modeling of the coulometric response based on chloride-sensitive
SCISEs and K+ -selective SCISEs was presented [29, 30]. Further-
more, coulometric SCISEs were investigated in detecting small pH
changes in seawater [26] and K+ in serum [26,30]. The results ob-
tained so far indicate that the coulometric transduction method
can significantly improve the sensitivity of SCISEs. The coulomet-
ric response mode was recently shown to work also with SCISEs
for the detection of the divalent cations Ca2+ and Pb2+ [31].
In this work, the general feasibility of the new coulometric
transduction method is explored for anion sensing. PEDOT doped
with chloride, nitrate, sulfate and perchlorate are studied with
cyclic voltammetry (CV). Solid-contact anion-sensitive electrodes
(anion-sensitive SCISEs) are prepared utilizing either PEDOT(Cl)
or PEDOT(NO3 ) as solid contact that are covered with drop-
cast (thick) or spin-coated (thin) PVC-based ion-selective mem-
branes. Nitrate, perchlorate and sulfate-sensitive SCISEs with vari-
able thickness of PEDOT(Cl) are investigated by chronoamperom-
etry and coulometry. Additionally, nitrate-sensitive SCISEs with
PEDOT(NO3 ) as solid contact are studied, while sulfate-sensitive
SCISEs are also characterized by electrochemical impedance spec-
troscopy (EIS).
2. Experimental part
2.1. Materials
Tridodecylmethylammonium chloride (TDMACl), 3,4-
ethylenedioxythiophene (EDOT, 97%), sodium sulfate, sodium
nitrate, potassium perchlorate, sodium chloride and potassium
chloride were purchased from Sigma-Aldrich. 2-nitrophenyl octyl
ether (o-NPOE), poly(vinyl chloride) (PVC) of high molecular
weight and tetrahydrofuran (THF, >99.5%) were purchased from
Fluka. De-ionized water (ELGA Purelab Ultra, resistivity 18.2
MΩcm) was used for preparing all the aqueous solutions.
2.2. Electrode preparation
Glassy carbon (GC) disk electrodes (SIGRADUR R
G, Germany)
with 3 mm diameter (made by inserting a GC rod in a PVC cylin-
der) were polished by using abrasive papers of different coarse-
ness, diamond pastes with particle diameters of 15 μm, 9 μm,
3 μm and 1 μm, respectively, and finally with 0.3 μm Al2 O3 paste.
Cleaning of the GC electrode surface was completed by immersing
them in 1.0 M HNO3 for 5 min, rinsing with deionized water, ultra-
sonication in ethanol for 5 min and finally ultrasonication in deion-
ized water for 5 min. The solution for polymerization of PEDOT
solid contacts (PEDOT(Cl), PEDOT(NO3 ), PEDOT(SO4 ), PEDOT(ClO4 ))
contained 0.01 M EDOT and 0.1 M of KCl, NaNO3 , Na2 SO4 and
KClO4 , respectively. The polymerization solution was kept under
stirring overnight to ensure complete dissolution of the monomer.
Before polymerization, the solution was purged with N2 gas for
10 min to remove oxygen from the solution. PEDOT was electro-
chemically deposited on the electrode surface using an Autolab
general purpose electrochemical system (PGSTAT20, FRA2, AUTO-
LAB, Eco Chemie, B.V., The Netherlands), connected to a conven-
tional three electrode cell. Each GC electrode was connected as
working electrode, and a GC rod was used as counter electrode.
A Metrohm double junction Ag/AgCl/3 M KCl//1 M KCl was used
as the reference electrode. PEDOT doped with various counterions,
i.e. chloride, nitrate, sulfate and perchlorate ions, were electrode-
posited galvanostatically on GC by applying a constant current of
2
Electrochimica Acta 367 (2021) 137566
0.014 mA (0.2 mA/cm2 ) for 71 s and 710 s corresponding to a total
polymerization charge of 1 and 10 mC, respectively. Additionally,
PEDOT(Cl) and PEDOT(NO3 ) were electrodeposited by applying the
same current for 355 s to obtain 5 mC polymerization charge.
The resulting GC/PEDOT(Cl), GC/PEDOT(NO3 ), GC/PEDOT(SO4 )
and GC/PEDOT(ClO4 ) electrodes were dried overnight in air at
room temperature. After drying, GC/PEDOT(Cl) and GC/PEDOT(NO3 )
electrodes were covered with a drop-cast thick-layer or a spin-
coated thin-layer ion-selective membrane (ISM). Three replicates
of each anion-sensitive SCISE were prepared for all the experi-
ments. The spin-coated thin-layer ISM was applied by adding 3
drops of the membrane cocktail onto the rotating GC electrode
surface with a rotation speed of 1500 rpm. The volume of each
drop was ca 15 μl but most of the volume was swept away from
the rotating electrode surface leaving only a thin-layer of ISM. The
anion-sensitive SCISEs with drop-cast thick membranes were pre-
pared by adding the corresponding cocktail in a 50 μl aliquot cov-
ering the whole electrode disk surface (GC electrode and the PVC
body). The composition of the cocktail in %(w/w) was 15% TDMACl,
51% o-NPOE and 34% PVC, dissolved in THF (dry mass = 15%).
The selectivity of this type of anion-sensitive membrane follows
the Hofmeister series [32-34]. The anion-sensitive SCISEs covered
with drop-cast thick or spin-coated thin ISMs were dried overnight
followed by an overnight conditioning in 0.1 M NaNO3 , KClO4 or
Na2 SO4 solution before the measurements. All the electrodes were
conditioned in their corresponding solutions between different ex-
periments.
2.3. Potentiometric and chronoamperometric measurements
Prior to chronoamperometric measurement, all the SCISEs were
checked using the potentiometric method to verify the proper
functioning of the electrodes. Both potentiometric and chronoam-
perometric measurements were done following the protocol de-
scribed in our previous work [28]. The nitrate and sulfate-
sensitive SCISEs were studied in NaNO3 and Na2 SO4 solutions, re-
spectively, without any background electrolyte. The perchlorate-
sensitive SCISE was studied in KClO4 solutions containing 0.1 M
NaCl as constant ionic background. The 0.1 M NaNO3 , 0.01 M
Na2 SO4 , and 0.1 M KClO4 starting solutions were sequentially
(stepwise) diluted either with deionized water (in the case of ni-
trate and sulfate-sensitive SCISEs) or with 0.1 M NaCl (in the case
of perchlorate-sensitive SCISEs) resulting in log aNO− or log
3
aClO− or log aSO2− = 0.18 decades/dilution step. A Metrohm dou-
4 4
ble junction Ag/AgCl/3 M KCl//1 M KCl was used as reference elec-
trode for nitrate and perchlorate-sensitive SCISEs and Ag/AgCl/3 M
KCl//0.1 M Na2 SO4 was used as reference electrode for sulfate-
sensitive SCISEs in all potentiometric and chronoamperometric
measurements.
2.4. Chronoamperometry using applied potential steps
For comparison purposes, the chronoamperometric response of
sulfate-sensitive SCISEs with 5 mC PEDOT(Cl) as solid contact and
spin-coated thin-layer membrane were also studied by applying
potential steps (10 mV/step) in 0.1 M Na2 SO4 using a conventional
three-electrode cell, which was connected to the Autolab instru-
ment described earlier. The applied potential for each step was
10 mV and ranged from 0.1 V to 0.21 V (oxidation of PEDOT) and
in the reverse direction from 0.21 V to 0.1 V (reduction of PEDOT).
The time interval between each potential step was 60 s.
2.5. Electrochemical impedance spectroscopy
Electrochemical impedance spectroscopy measurements were
performed for sulfate-sensitive SCISEs in 0.1 M Na2 SO4 at the open
T. Han, U. Mattinen, Z. Mousavi et al.
Fig. 1. Cyclic voltammograms of 1 mC and 10 mC of GC/PEDOT(Cl) (a, e), GC/PEDOT(SO4 ) (b, f), GC/PEDOT(NO3 ) (c, g) and GC/PEDOT(ClO4 ) (d, h) films measured in 0.1 M
NaNO3 (black), 0.1 M KClO4 (red), 0.1 M Na2 SO4 (blue), 0.1 M KCl (green) with 50 mV/s scan rate and a potential window from −0.2 V to 0.5 V. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
circuit potential (OCP) using the same Autolab (with FRA) instru-
ment described above and a conventional three electrode cell. A
Metrohm single junction Ag/AgCl/3 M KCl was used as reference
electrode and a glassy carbon rod as the counter electrode. The fre-
quency range was 100 kHz – 10 mHz and the excitation amplitude
was 10 mV (RMS) around OCP.
2.6. Cyclic voltammetry
Cyclic voltammetry (CV) for GC electrodes covered with 1
mC and 10 mC of PEDOT(Cl), PEDOT(NO3 ), PEDOT(SO4 ) and
PEDOT(ClO4 ) solid contact were carried out in 0.1 M KCl, NaNO3 ,
Na2 SO4 and KClO4 solutions. In these measurements, the Autolab
general purpose electrochemical system connected to a conven-
tional one-compartment three-electrode electrochemical cell was
used. A Metrohm single junction Ag/AgCl/3 M KCl was used as ref-
erence electrode and a glassy carbon rod as the counter electrode.
The cyclic voltammograms were recorded in the potential range of
−0.2 V to +0.5 V with a scan rate of 50 mV/s.
3. Results and discussions
In the coulometric transduction method for SCISEs, the poten-
tial between the SCISE and the reference electrode is held con-
stant and the current between the SCISE and the counter elec-
trode is measured [25-28]. A change of the ion activity in solu-
tion gives rise to oxidation/reduction of the solid contact (PEDOT)
and a transient current is measured and integrated to obtain the
charge which is the analytical signal. The working principle of the
constant potential coulometric method has been clearly and thor-
oughly described in our previous works for cation-selective SCISEs
[25–28].
For an anion-sensitive SCISE, the oxidation/reduction of the
PEDOT-based solid contact will ideally follow the reaction shown
in Eq. (1) where a nitrate-sensitive SCISE with a PEDOT(NO3 ) solid
contact is used as an example.
PEDOTn+ (NO3 − )n + xe− ↔ PEDOT(n−x)+ (NO3 − )n−x + xNO3 −
(1)
3
Electrochimica Acta 367 (2021) 137566
For example, a decrease of the nitrate activity in the solution
will induce a positive potential change at the ISM | solution inter-
face, which in turn gives a negative current transient that shifts the
equilibrium in Eq. (1) to the right. As a consequence of the poten-
tiometric response (Nernst equation) of SCISEs, the obtained cu-
mulated charge of anion-sensitive and cation-selective SCISEs has
opposite signs, but otherwise the operation principle is analogous.
3.1. Cyclic voltammetry of PEDOT
The cyclic voltammograms of GC/PEDOT(Cl), GC/PEDOT(NO3 ),
GC/PEDOT(SO4 ), and GC/PEDOT(ClO4 ) in various electrolyte solu-
tions are shown in Fig. 1. In the potential range of −0.2 V to
0.5 V, the cyclic voltammograms of PEDOT doped with various
counterions show no apparent oxidation or reduction peaks but
a capacitive-like behavior typical for PEDOT-based films in aque-
ous electrolytes [7]. In the coulometric measurement mode, the
capacitive-like current of the PEDOT solid contact is of importance
since the analytical signal is the cumulated charge deriving from
the redox reaction of the PEDOT solid contact.
The voltammetric response of 1 mC GC/PEDOT electrodes cy-
cled in various electrolyte solutions are compared in Fig. 1a-d.
The cyclic voltammograms of GC/PEDOT(Cl), GC/PEDOT(SO4 ) and
GC/PEDOT(NO3 ) were relatively independent of the supporting
electrolyte (KCl, Na2 SO4 , NaNO3 , KClO4 ) used during CV (Fig. 1a-
c). However, the cyclic voltammograms of GC/PEDOT(ClO4 ) showed
slightly higher currents and also some dependence on the support-
ing electrolyte used during CV (Fig. 1d).
As the loading of the PEDOT film was increased to 10 mC
(Fig. 1e-h), the differences in CV became more pronounced for
the studied electrolytes. As shown in Fig. 1e-h, the capacitive-like
current response of 10 mC GC/PEDOT electrodes in various elec-
trolyte solutions is clearly dependent on the anion present during
electropolymerization. The capacitive currents of 10 mC GC/PEDOT
electrodes tends to increase in the sequence: GC/PEDOT(Cl) ≈
GC/PEDOT(SO4 ) < GC/PEDOT(NO3 ) < GC/PEDOT(ClO4 ). The differ-
ences in current can be related to different amounts of PEDOT or
different electroactivity of PEDOT at a given polymerization charge,
depending on the doping anions used in the electropolymerization.
This implies that for a given polymerization charge, a higher yield
T. Han, U. Mattinen, Z. Mousavi et al.
Fig. 2. Electrochemical impedance spectra recorded for sulfate-sensitive SCISEs
with 1, 5, and 10 mC PEDOT(Cl) solid contact covered with drop-cast thick-layer ISM
(a) and spin-coated thin-layer ISM (b). The supporting electrolyte is 0.1 M Na2 SO4 .
The measurement is done at open circuit potential with 10 mV (RMS) perturbation
amplitude and 100 kHz – 10 mHz frequency range. (For interpretation of the refer-
ences to colour in this figure legend, the reader is referred to the web version of
this article.)
Table 1
Low-frequency capacitance (CL ) values for sulfate-sensitive SCISEs with
PEDOT(Cl) solid contact calculated from the impedance spectra in Fig. 2.
Polymerization charge of PEDOT(Cl) / mC ISM CL / μF
1 Spin-coated 16
1 Drop-cast 15
5 Spin-coated 114
5 Drop-cast 116
10 Spin-coated 223
10 Drop-cast 245
of electroactive PEDOT(ClO4 ) and PEDOT(NO3 ) is obtained, in com-
parison to PEDOT(Cl) and PEDOT(SO4 ).
The results in Fig. 1e-h show also that the supporting elec-
trolyte used during CV plays a role on the resulting current re-
sponse, especially for GC/PEDOT(ClO4 ) and GC/PEDOT(NO3 ). For the
electropolymerization of PEDOT, the doping counterions can re-
markably affect the structural feature, morphology and capacitance
of the resulting polymer film [7,11,15]. This phenomenon may be
related to anion size and hydrophobicity (Hofmeister series). Inter-
estingly, the voltammograms of GC/PEDOT(Cl) are relatively inde-
pendent of the supporting electrolyte used during potential cycling
(CV). Therefore, GC/PEDOT(Cl) was considered to be a suitable solid
contact when comparing various anion-sensitive SCISEs.
3.2. Electrochemical impedance spectroscopy
Electrochemical impedance spectroscopy (EIS) measurements
were performed for sulfate-sensitive SCISEs with 1, 5, and 10 mC
PEDOT(Cl) as solid contact covered with drop-cast thick-layer or
spin-coated thin-layer membranes. The EIS measurements were
performed at the OCP of the respective SCISE. As shown in Fig. 2,
the diameter of the high-frequency semicircle decreases when the
thickness of the anion-sensitive membrane (ISM) is decreased. This
shows that the high-frequency semicircle represents the bulk re-
sistance (in parallel with the geometric capacitance) of the ISM.
The thickness of the PEDOT solid contact does not influence the
impedance at high frequencies, as the resistance of the PEDOT
solid contact film is expected to be negligible compared to that
of the ISM [6].
The low frequency (almost vertical) part of the impedance
spectra (10 mHz) is related to the bulk redox capacitance of the
PEDOT(Cl) solid contact [7]. The results in Table 1 show that
the redox capacitance (CL ) values of sulfate-sensitive SCISEs with
4
Electrochimica Acta 367 (2021) 137566
Fig. 3. Low-frequency capacitance (CL ) values for the sulfate-sensitive SCISEs (taken
from Table 1) vs polymerization charge of the PEDOT(Cl) solid contact for spin-
coated and drop-cast ISM. For 1 and 5 mC PEDOT(Cl) the CL values for spin-coated
and drop-cast ISM overlap.
drop-cast thick-layer and spin-coated thin-layer ISM are roughly
the same, which indicates that the oxidation/reduction (dop-
ing/undoping) process of PEDOT is not limited by ion transport
through the ISM. As clearly shown in Fig. 3, the redox capacitance
(CL ) of PEDOT in the SCISEs increases linearly proportional to the
polymerization charge (thickness) of the PEDOT(Cl) solid contact in
a predictable manner.
3.3. Coulometric response of nitrate-sensitive SCISEs
The chronoamperometric and coulometric response recorded
for the nitrate-sensitive SCISEs with 1, 5, and 10 mC PEDOT(NO3 )
as solid contact covered with drop-cast and spin-coated ISM is
shown in Fig. 4. The starting solution is 0.1 M NaNO3 that was
stepwise diluted with water. As shown in Fig. 4, the electrode with
larger polymerization charge of PEDOT gives a larger cumulated
charge (Q) and requires longer time to reach equilibrium after each
change in the activity of the primary anion (NO3 - ) in the solu-
tion. The signal amplification is proportional to the polymerization
charge of the PEDOT-based solid contact, in analogy with cation-
selective SCISEs [27].
Nitrate-sensitive SCISEs with PEDOT(NO3 ) as solid contact and
spin-coated membrane give higher current peaks due to a lower
resistance of the thin ISM (Fig. 4c), compared to the thick drop-
cast ISM (Fig. 4a). These results are in good qualitative agreement
with the results from our previous work using K+ -SCISEs [26].
The coulometric response of nitrate-sensitive SCISEs with
PEDOT(NO3 ) solid contact and thin spin-coated ISM is very fast
and even for the 10 mC PEDOT(NO3 ) solid contact the cumulated
charge (Q) stabilizes within 2 min (Fig 4c, d). On the contrary,
the coulometric response of the nitrate-sensitive SCISE with 10 mC
PEDOT(NO3 ) solid contact and drop-cast (thick) ISM does not quite
reach equilibrium within 5 min (Fig. 4a, b)
Amplification of the cumulated charge Q with increasing thick-
ness of the conducting polymer is also confirmed by nitrate-
sensitive SCISEs with PEDOT(Cl) solid contact and spin-coated thin-
layer ISM, as shown in Fig. 5. The cumulated charge Q is linearly
proportional to log aNO− in the studied range of −1 to −2.45.
3
As it was shown in Fig. 1, electrolyte solutions used in the
electropolymerization of PEDOT-based solid contact films and in
the cyclic voltammetry measurements affect the current responses.
In Fig. 6, a comparison of the cumulated charge Q vs time of
nitrate-sensitive SCISEs with 1 mC PEDOT(Cl) or 1 mC PEDOT(NO3 )
solid contact and drop-cast membranes upon stepwise dilutions is
shown. The two replicates of electrodes with PEDOT(Cl) solid con-
T. Han, U. Mattinen, Z. Mousavi et al. Electrochimica Acta 367 (2021) 137566

Fig. 4. Chronoamperograms for the first three dilutions (a, c) and the corresponding chronocoulograms (b, d) of nitrate-sensitive SCISEs with 1, 5, and 10 mC PEDOT(NO3 )
covered with drop-cast thick-layer and spin-coated thin-layer ISM. The starting solution is 0.1 M NaNO3 diluted with water to have log aNO−3 = 0.18 decades/dilution step.
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. Chronocoulograms (a) and the cumulated charge Q vs log aNO−3 (b) of nitrate-sensitive SCISEs with 1, 5, and 10 mC PEDOT(Cl) covered with spin-coated thin-layer
ISM. The Q values in Fig. 5b correspond to the saturated signals in Fig. 5a. (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)

tact show good electrode-to-electrode reproducibility in the coulo-
metric measurement mode.
Despite the equivalent charge of the 1 mC PEDOT solid con-
tacts, the cumulated charge Q vs time curve of the electrode with
PEDOT(NO3 ) solid contact is significantly larger than that of the
SCISEs with PEDOT(Cl). A larger cumulated charge Q corresponds
to a larger redox capacitance of the PEDOT solid contact. However,
the difference between the coulometric response of the SCISEs
with 1 mC PEDOT(NO3 ) and 1 mC PEDOT(Cl) as solid contacts
(Fig. 6) is even larger than expected from the CVs recorded in
aqueous solutions (Fig. 1a, c). This suggests that the capacitance of
PEDOT depends also on the contacting medium, i.e. aqueous elec-
trolyte vs ISM.

5
T. Han, U. Mattinen, Z. Mousavi et al.
Fig. 6. Chronocoulograms for nitrate-sensitive SCISEs based on 1 mC PEDOT(Cl) and
PEDOT(NO3 ) solid contact covered with drop-cast ISM during sequential dilution
steps. The starting solution (0.1 M NaNO3 ) is diluted with water (log aNO−3 = 0.18
decades/step). (For interpretation of the references to colour in this figure legend,
the reader is referred to the web version of this article.)
3.4. Coulometric response of sulfate-sensitive SCISEs
In Fig. 7, the chronoamperometric and coulometric responses of
sulfate-sensitive SCISEs with spin-coated thin-layer and drop-cast
thick-layer membranes are compared. The polymerization charge
of thePEDOT(Cl) solid contact for these electrodes was 1 and 5
mC. Again, the sulfate-sensitive SCISE with spin-coated thin-layer
ISM results in higher current peaks and faster responses (shorter
equilibration time) than the drop-cast ISM for both thicknesses of
the solid contact. These obtained results are in good agreement
with the results described above (Fig. 2, Fig. 4) and results from
our earlier works [26,29]. The chronocoulograms obtained by inte-
gration of the current response is shown as dashed lines in Fig. 7.
For sulfate-sensitive SCISEs with spin-coated membranes and 1 mC
PEDOT(Cl) as solid contact, the chronocoulograms stabilize within
less than 1 min after each dilution step (Fig. 7a).
The chronoamperograms recorded for sulfate-sensitive SCISEs
with 1, 5, and 10 mC PEDOT(Cl) as solid contact covered with spin-
coated thin-layer membrane are shown in Fig. 8. The starting so-
lution (0.01 M Na2 SO4 ) is stepwise diluted with water. It can be
seen from Fig. 8a that the current peaks after each dilution are
rather reproducible. Also here, thicker PEDOT(Cl) solid contacts re-
quire longer equilibration time after each dilution step (Fig. 8a).
The chronocoulograms and the cumulated charge Q vs log SO24− are
presented in Fig. 8b. The cumulated charge Q increases with in-
creasing polymerization charge (thickness) of PEDOT(Cl) solid con-
Fig. 7. Chronoamperograms for 1st and 2nd dilution steps (solid line) and the corresponding chronocoulograms (dash line) for sulfate-sensitive SCISEs with 1 mC (a) and 5
mC (b) PEDOT(Cl) solid contact covered with drop-cast thick-layer and spin-coated thin-layer ISM. The starting solution is 0.01 M Na2 SO4 diluted with water. (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
6
Electrochimica Acta 367 (2021) 137566
tact and is linearly proportional to log aSO2− in the studied range
4
of −2 to −2.9.
When comparing the coulometric response of the sulfate-
sensitive SCISEs (Fig. 8) and the nitrate-sensitive SCISEs (Fig. 5),
it can be seen that the cumulated charge Q of the SCISEs that
are sensitive to the divalent anion (sulfate) is half of the cumu-
lated charge Q for the monovalent anion (nitrate), as expected from
Nernst equation.
Fig. 9 shows the chronoamperograms recorded for the sulfate-
sensitive SCISEs, based on 5 mC PEDOT(Cl) solid contact with spin-
coated ISM, when 10 mV potential steps are applied in both di-
rections. The current response of the SCISEs upon oxidation with
an increasing potential step (No. 1) and reduction with a decreas-
ing potential step (No. 2) are shown in Fig. 9a, while the corre-
sponding charge is shown in Fig. 9b. The 10 mV potential step
results in ca 1 μC cumulated charge Q as shown in Fig. 9b. This
is approximately twice the ca 0.6 μC charge observed for one di-
lution step (log aSO2− = 0.18 decades/step) that corresponds to
4
ca 5 mV change at the ISM | sample interface. The change of the
applied potential by the potentiostat and by varying the phase
boundary potential between solution and ISM by changing the ac-
tivity of the primary ion are thus in good agreement. Furthermore,
the results shown in Fig. 9 prove the reversibility of the oxida-
tion/reduction reaction of the PEDOT solid contact. When consid-
ering that the results shown in Fig. 9 correspond to two individual
SCISEs (No.1 and No. 2), it can be concluded that the electrode-
to-electrode reproducibility is good. The differences observed be-
tween No. 1 and No. 2 in Fig 9 include any possible irreproducibil-
ity in the electrode preparation, including electropolymerization of
PEDOT(Cl) and spin-coating of the ISM.
3.5. Coulometric response of perchlorate-sensitive SCISEs
A comparison of the chronocoulograms of perchlorate, ni-
trate, and sulfate-sensitive SCISEs with 10 mC PEDOT(Cl) solid
contact covered with spin-coated thin-layer ISM during sequen-
tial dilution steps is shown in Fig. 10. The starting solutions
are 0.01 M Na2 SO4 for sulfate-sensitive SCISE and 0.1 M NaNO3
for nitrate-sensitive SCISE that were diluted with water. In the
case of perchlorate-sensitive SCISEs, the starting solution is 0.1 M
KClO4 + 0.1 M NaCl as background electrolyte that was diluted
with 0.1 M NaCl. The coulometric response of perchlorate-sensitive
SCISEs was not influenced by the presence of the interfering ion
(Cl− ) in the solution. The perchlorate-SCISE was also observed
to give a Nernstian potentiometric response to perchlorate, in
presence of 0.1 M NaCl, within the concentration range 10−1 to
T. Han, U. Mattinen, Z. Mousavi et al.
Fig. 8. Chronoamperograms (a) and chronocoulograms (b) for sulfate-sensitive SCISEs with 1, 5, and 10 mC PEDOT(Cl) solid contact covered with spin-coated thin-layer ISM.
Cumulated charge Q vs log aSO2− is shown in the inset of Fig. 8b. The starting solution 0.01 M Na2 SO4 was diluted with water. (For interpretation of the references to colour
4
in this figure legend, the reader is referred to the web version of this article.)
Fig. 9. Chronoamperograms (a) and chronocoulograms (b) for sulfate-sensitive SCISE with 5 mC PEDOT(Cl) solid contact covered with spin-coated thin-layer ISM, measured
in 0.1 M Na2 SO4 when applying 10 mV potential steps using the potentiostat. The time interval between potential steps is 60 s. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
Fig. 10. Chronocoulograms for perchlorate, nitrate and sulfate-sensitive SCISEs with
10 mC PEDOT(Cl) solid contact covered with spin-coated ISM during sequential
dilution steps. The starting solutions for sulfate and nitrate-sensitive SCISEs are
0.01 M Na2 SO4 and 0.1 M NaNO3 , respectively (diluted with water). The starting so-
lution for the perchlorate-sensitive SCISE is 0.1 M KClO4 + 0.1 M NaCl background
electrolyte diluted with 0.1 M NaCl. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
10−3.4 . This can be explained by the fact that the selectivity of
the TDMACl ion-exchanger-based ISM towards different anions cor-
relates with their relative lipophilicity, following the Hofmeister
7
Electrochimica Acta 367 (2021) 137566
− −
series: ClO4 > SCN− > I− > NO− − −
3 > Br > Cl ≈ HCO3 > H2 PO4 >
2−
2−
SO4 [32-34]. The selectivity coefficient (log KPot ClO4,Cl
) for the
perchlorate-sensitive SCISE used here is expected to be in the
range −3 to −4 [4].
The results presented in Fig. 10 for nitrate, perchlorate, and
sulfate-sensitive SCISEs are logical and follow the Nernst equa-
tion as the coulometric response for the monovalent perchlorate-
sensitive SCISE is roughly similar to nitrate-sensitive SCISE and ap-
proximately double of that of the divalent sulfate-sensitive SCISE.
Conclusions
In this work, the electrochemical characteristics of solid-contact
anion-sensitive electrodes based on PEDOT as solid contact and
a plasticized PVC-based anion sensitive membrane were investi-
gated. Results obtained by cyclic voltammetry of GC/PEDOT elec-
trodes showed that the yield of electroactive PEDOT depended on
the supporting electrolyte used during electrodeposition as well
as the electrolyte used during the subsequent potential cycling of
the GC/PEDOT electrodes. The different yield of PEDOT for a given
polymerization charge was reflected in the coulometric response
of the solid-contact anion-sensitive electrodes where PEDOT was
used as the solid contact (GC/PEDOT/ISM). The coulometric re-
sponse of GC/PEDOT/ISM was shown to be faster for spin-coated
(thin) ISM than for drop-cast (thick) ISM. Furthermore, the am-
T. Han, U. Mattinen, Z. Mousavi et al.
plification of the coulometric signal was proportional to the poly-
merization charge (thickness) of PEDOT-based solid contact and
inversely proportional to the charge of the anion. When apply-
ing a potential step to the GC/PEDOT/ISM electrode, the cumulated
charge was reversible and of similar magnitude to the charge ob-
tained for an activity change in solution causing a similar change
of potential. Results obtained by electrochemical impedance spec-
troscopy (EIS) for GC/PEDOT/ISM clearly showed that the high-
frequency semicircle originated from the ISM and was independent
of the PEDOT thickness. The thickness of the ISM also did not influ-
ence the low-frequency capacitive part of the impedance spectra,
which was related to the redox capacitance of the PEDOT film. The
low-frequency capacitance value increased linearly with increas-
ing thickness of PEDOT solid contact. The coulometric results were
thus in line with experimental results obtained by EIS and followed
the theoretical predictions of the Nernst equation. Altogether, the
results presented in this work showed that the coulometric trans-
duction method is appropriate and applicable for anion-sensitive
solid-contact electrodes.
Credit author statement
Tingting Han: Methodology, Investigation, Data analysis, Writ-
ing - Original Draft. Ulriika Mattinen: Data analysis, Writing - Re-
view & Editing, Supervision. Zekra Mousavi: Data analysis, Writing
- Review & Editing, Supervision. Johan Bobacka: Conceptualiza-
tion, Methodology, Writing - Review & Editing, Supervision, Project
administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgements
Financial support by the Academy of Finland (Project No.
317829) is gratefully acknowledged. This work is part of the activ-
ities of the Johan Gadolin Process Chemistry Centre (PCC) at Åbo
Akademi University.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.electacta.2020.137566.
References
[1] J. Bobacka, A. Ivaska, A. Lewenstam, Potentiometric ion sensors, Chem. Rev. 108
(2008) 329–351, https://doi.org/10.1021/cr068100w.
[2] A. Lewenstam, Routines and challenges in clinical application of electrochem-
ical ion-sensors, Electroanalysis 26 (2014) 1171–1181, https://doi.org/10.1002/
elan.20140 0 061.
[3] E. Bakker, P. Bühlmann, E. Pretsch, Carrier-based ion-selective electrodes and
bulk optodes. 1. General characteristics, Chem. Rev. 97 (1997) 3083–3132,
https://doi.org/10.1021/cr940394a.
[4] P. Bühlmann, E. Pretsch, E. Bakker, Carrier-based ion-selective electrodes and
bulk optodes. 2. Ionophores for potentiometric and optical sensors, Chem. Rev.
29 (1998) 1593–1688, https://doi.org/10.1021/cr970113+.
[5] C. Zuliani, D. Diamond, Opportunities and challenges of using ion-selective
electrodes in environmental monitoring and wearable sensors, Electrochim.
Acta 84 (2012) 29–34, https://doi.org/10.1016/j.electacta.2012.04.147.
[6] J. Bobacka, Potential stability of all-solid-state ion-selective electrodes using
conducting polymers as ion-to-electron transducers, Anal. Chem. 71 (1999)
4932–4937, https://doi.org/10.1021/ac990497z.
[7] J. Bobacka, A. Lewenstam, A. Ivaska, Electrochemical impedance spectroscopy
of oxidized poly(3,4-ethylenedioxythiophene) film electrodes in aqueous
solutions, J. Electroanal. Chem. 489 (20 0 0) 17–27, https://doi.org/10.1016/
S0 022-0728(0 0)0 0206-0.
8
Electrochimica Acta 367 (2021) 137566
[8] J. Bobacka, A. Ivaska, A. Lewenstam, Potentiometric ion sensors based on con-
ducting polymers, Electroanalysis 15 (2003) 366–374, https://doi.org/10.1002/
elan.20 0390 042.
[9] G.A. Crespo, M. Cuartero, E. Bakker, Thin layer ionophore-based membrane for
multianalyte ion activity detection, Anal. Chem. 87 (2015) 7729–7737, https:
//doi.org/10.1021/acs.analchem.5b01459.
[10] E. Bakker, Electroanalysis with membrane electrodes and liquid-liquid inter-
faces, Anal. Chem. 88 (2016) 395–413, https://doi.org/10.1021/acs.analchem.
5b04034.
[11] Z. Mousavi, J. Bobacka, A. Lewenstam, A. Ivaska, Poly(3,4-
ethylenedioxythiophene) (PEDOT) doped with carbon nanotubes as
ion-to-electron transducer in polymer membrane-based potassium
ion-selective electrodes, J. Electroanal. Chem. 633 (2009) 246–252,
https://doi.org/10.1016/j.jelechem.20 09.06.0 05.
[12] J. Bobacka, Conducting polymer-based solid-state ion-selective electrodes, Elec-
troanalysis 18 (2006) 7–18, https://doi.org/10.10 02/elan.20 0503384.
[13] U. Vanamo, J. Bobacka, Electrochemical control of the standard potential of
solid-contact ion-selective electrodes having a conducting polymer as ion-to-
electron transducer, Electrochim. Acta 122 (2014) 316–321, https://doi.org/10.
1016/j.electacta.2013.10.134.
[14] U. Vanamo, J. Bobacka, Instrument-free control of the standard potential of po-
tentiometric solid-contact ion-selective electrodes by short-circuiting with a
conventional reference electrode, Anal. Chem. 86 (2014) 10540–10545, https:
//doi.org/10.1021/ac501464s.
[15] X. Wang, P. Sjöberg-Eerola, J.E. Eriksson, J. Bobacka, M. Bergelin, The effect
of counter ions and substrate material on the growth and morphology of
poly(3,4-ethylenedioxythiophene) films: towards the application of enzyme
electrode construction in biofuel cells, Synth. Met. 160 (2010) 1373–1381,
https://doi.org/10.1016/j.synthmet.2010.01.033.
[16] S.A. Spanninga, D.C. Martin, Z. Chen, X-ray photoelectron spectroscopy study
of counterion incorporation in poly(3,4-ethylenedioxythiophene) (PEDOT) 2:
polyanion effect, toluenesulfonate, and small anions, J. Phys. Chem. C. 114
(2010) 14992–14997, https://doi.org/10.1021/jp104591d.
[17] P. Sjöberg-Eerola, J. Bobacka, A. Lewenstam, A. Ivaska, All-solid-state chlo-
ride sensors based on electronically conducting, semiconducting and insulat-
ing polymer membranes, Sens. Actuat. B Chem. 127 (2007) 545–553, https:
//doi.org/10.1016/j.snb.20 07.05.0 04.
[18] T.A. Bendikov, T.C. Harmon, Long-lived solid state perchlorate ion selective
sensor based on doped poly(3,4-ethylenedioxythiophene) (PEDOT) films, Anal.
Chim. Acta. 551 (2005) 30–36, https://doi.org/10.1016/j.aca.2005.07.004.
[19] Z. Jarolímová, G.A. Crespo, X. Xie, M. Ghahraman Afshar, M. Pawlak, E. Bakker,
Chronopotentiometric carbonate detection with all-solid-state ionophore-
based electrodes, Anal. Chem. 86 (2014) 6307–6314, https://doi.org/10.1021/
ac5004163.
[20] K. Martin, S.A. Kadam, U. Mattinen, J. Bobacka, I. Leito, Solid-contact acetate-
selective electrode based on a 1,3-bis(carbazolyl)urea-ionophore, Electroanaly-
sis 31 (2019) 1061–1066, https://doi.org/10.10 02/elan.20180 0790.
[21] A. Michalska, A. Galuszkiewicz, M. Ogonowska, M. Ocypa, K. Maksymiuk,
PEDOT films: multifunctional membranes for electrochemical ion sens-
ing, J. Solid State Electrochem. 8 (2004) 381–389, https://doi.org/10.1007/
s10 0 08-0 03-0459-8.
[22] M. Cuartero, G.A. Crespo, E. Bakker, Ionophore-based voltammetric ion activity
sensing with thin layer membranes, Anal. Chem. 88 (2016) 1654–1660, https:
//doi.org/10.1021/acs.analchem.5b03611.
[23] S. Jansod, M. Cuartero, T. Cherubini, E. Bakker, Colorimetric readout for poten-
tiometric sensors with closed bipolar electrodes, Anal. Chem. 90 (2018) 6376–
6379, https://doi.org/10.1021/acs.analchem.8b01585.
[24] Y. Kim, S. Amemiya, Stripping analysis of nanomolar perchlorate in drink-
ing water with a voltammetric ion-selective electrode based on thin-layer
liquid membrane, Anal. Chem. 80 (2008) 6056–6065, https://doi.org/10.1021/
ac8008687.
[25] E. Hupa, U. Vanamo, J. Bobacka, Novel ion-to-electron transduction principle
for solid-contact ISEs, Electroanalysis 27 (2015) 591–594, https://doi.org/10.
10 02/elan.20140 0596.
[26] T. Han, U. Mattinen, J. Bobacka, Improving the sensitivity of solid-contact
ion-selective electrodes by using coulometric signal transduction, ACS Sens. 4
(2019) 900–906, https://doi.org/10.1021/acssensors.8b01649.
[27] U. Vanamo, E. Hupa, V. Yrjänä, J. Bobacka, New signal readout principle for
solid-contact ion-selective electrodes, Anal. Chem. 88 (2016) 4369–4374, https:
//doi.org/10.1021/acs.analchem.5b04800.
[28] T. Han, U. Vanamo, J. Bobacka, Influence of electrode geometry on the re-
sponse of solid-contact ion-selective electrodes when utilizing a new coulo-
metric signal readout method, ChemElectroChem 3 (2016) 2071–2077, https:
//doi.org/10.10 02/celc.20160 0575.
[29] Z. Jarolímová, T. Han, U. Mattinen, J. Bobacka, E. Bakker, Capacitive model for
coulometric readout of ion-selective electrodes, Anal. Chem. 90 (2018) 8700–
8707, https://doi.org/10.1021/acs.analchem.8b02145.
[30] Y.O. Kondratyeva, E.G. Tolstopjatova, D.O. Kirsanov, K.N. Mikhelson, Chronoam-
perometric and coulometric analysis with ionophore-based ion-selective elec-
trodes: a modified theory and the potassium ion assay in serum samples, Sens.
Actuat. B Chem. 310 (2020) 127894, https://doi.org/10.1016/j.snb.2020.127894.
[31] T. Han, Z. Mousavi, U. Mattinen, J. Bobacka, Coulometric response char-
acteristics of solid contact ion-selective electrodes for divalent cations,
J. Solid State Electrochem. 24 (2020) 2975–2983, https://doi.org/10.1007/
s10 0 08- 020- 04718- 8.
T. Han, U. Mattinen, Z. Mousavi et al. Electrochimica Acta 367 (2021) 137566

[32] W.E. Morf, The Principles of Ion-Selective Electrodes and of Membrane Trans- [34] K.N. Mikhelson, Ion-Selective Electrodes, Springer, Heidelberg, 2013 https://doi.
port, Akademiai Kiado, Budapest, 1981. org/10.1007/978- 3- 642- 36886- 8_2.
[33] A.L. Grekovich, K.N. Mikhelson, An anomalous behavior of anion-exchange
membranes with low concentration of quaternary ammonium sites: an ap-
parent selectivity to bicarbonate and phosphate, and its true nature, Electro-
analysis 14 (2002) 1391–1396, https://doi.org/10.1002/1521-4109(200211)14:
19/201391::AID-ELAN13913.0.CO;2-S.