Journal of The Electrochemical Society, 162 (10) E237-E243 (2015) E237

Electrochemical Properties and Analyses of CeCl3 in LiCl-KCl

Eutectic Salt
D. Yoonz and S. Phongikaroon
Department of Mechanical & Nuclear Engineering, Virginia Commonwealth University, Richmond,
Virginia 23284, USA

Thermodynamic and electrochemical properties of cerium in LiCl-KCl eutectic salt have been measured and studied at different
concentrations (0.5 – 4 wt%) and temperatures (698 K – 798 K) via both cyclic voltammetry (CV) and electrochemical impedance
spectroscopy (EIS) techniques as a part of developing a fundamental understanding and methodology in materials detection and
accountability for pyroprocessing technology. CV experiments were performed to determine the diffusion coefficient and apparent
standard potential of CeCl3 on the tungsten working electrode. The diffusion coefficient was calculated by using Delahay equation,
and raging from 0.48 × 10−5 to 1.01 × 10−5 cm2 s−1 . Results reveal that the calculated diffusion coefficient of CeCl3 in the salt
follows the Arrhenius temperature relationship and it is weakly affected by the changes in concentration of CeCl3 . The apparent
standard potentials were calculated from peak potentials showing linear relationship with temperature. Exchange current density
values of Ce3+ /Ce couple in the salt were obtained from EIS experiments, ranging from 0.0076 A cm−2 to 0.18 A cm−2 . The results
indicate that these values follow Arrhenius temperature dependence and increase when CeCl3 concentration increases.
© The Author(s) 2015. Published by ECS. This is an open access article distributed under the terms of the Creative Commons
Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any
medium, provided the original work is properly cited. [DOI: 10.1149/2.0401510jes] All rights reserved.

Manuscript submitted June 5, 2015; revised manuscript received July 10, 2015. Published July 24, 2015.

Pyroprocessing technology has been proposed as another promis-
ing method for the recovery and recycle of uranium and actinide
elements from the used nuclear fuel. An essential step in this tech-
nology is the electrorefining process in which uranium is selectively
recovered by using solid cathodes in chloride-based molten salt at
high temperature.1 Then, co-recovery of uranium and transuranic el-
ements can be accomplished by replacing the solid cathodes with a
liquid cadmium cathode because the reduction potentials of the ele-
ments become close when the liquid cadmium is used as a cathode
electrode.2,3 Since uranium is the major element in most nuclear fuel
cycle paths as well as pyroprocessing technology, the assessment of
accurate thermochemical data for the element in the molten salt is
extremely important.4
Many studies on the thermochemical properties of uranium have
been done in LiCl-KCl molten eutectic salt in different tempera-
ture ranges. Masset et al.5,6 investigated diffusion coefficients of ac-
tinides and lanthanides in LiCl-KCl via cyclic voltammetry (CV) and
chronopotentiometry (CP). Kuznetsov et al.7,8 studied the electro-
chemical behaviors of actinides and rare-earth metals in LiCl-KCl
salt. They performed CP and chronoamperometry (CA), and linear
sweep voltammetry to determine the diffusion coefficients. Hoover
et al. in 20149 extended the uranium concentration in LiCl-KCl molten
salt up to 10 wt% and observed the electrochemical and thermo-
dynamic behaviors of uranium using CV, CP, and anodic stripping
voltammetry. These data are valuable to a development of kinetic
models, which can be useful for understanding the main features of
actinide deposition at the electrode surface, and also for prediction
of material distribution in an electrorefiner of a safeguarding aspect.
Zhang10 developed a kinetic model for electrorefining system show-
ing that the model is capable of predicting the kinetic features and
material fluxes of nuclear materials in the electrorefiner.
However, an exchange current density of uranium, which is es-
sential to the physic-based model, has not been well measured and
understood due to challenges of the measurement. Only few studies
on the exchange current densities of uranium in LiCl-KCl have been
published.11–13 Choi et al. in 200911 performed a linear polarization
method in LiCl-KCl at 773 K and reported the exchange current den-
sity of uranium ranging from 0.3 to 0.5 A cm−2 . Later, Ghosh and
co-workers reported that the exchange current density of uranium is 8
± 2 mA cm−2 , by measurements of Tafel plot.12 Rose et al. in 201513
measured the exchange current density of uranium from Tafel plot,
ranging from 69.5 ± 9 to 220 ± 32 mA cm−2 . From these literature
z
E-mail: yoond2@vcu.edu
results, a meaningful comparison of the reported data is not possi-
ble owing to the dispersed values of the exchange current density of
uranium. This may be due to challenges of measuring reliable slope
on Tafel plot and electrode surface area where uranium deposition
occurs. Therefore, further studies must be conducted on measurement
of the exchange current densities of uranium.
These aforementioned challenges provide a motivation for this
study to investigate another element that may exhibit similar uranium
characteristic and to establish proper electrochemical techniques that
can be applied toward uranium. Cerium is one of the common elements
to be used as a surrogate material for uranium. The main reason is that
cerium has similar ionic size with uranium, and its reduction poten-
tial is closer to uranium than other lanthanide materials.14 Therefore,
cerium was selected for the purpose of developing methods to evalu-
ate the electrochemical and thermodynamic properties of uranium in
the molten salt. Several studies on the electrochemical behaviors of
cerium in the molten salt at high temperatures have been conducted.
In 1998, Iizuka15 conducted CP to determine diffusion coefficient of
CeCl3 at different temperatures. Marsden and Pesic in 201114 mea-
sured apparent standard potentials and diffusion coefficients of CeCl3
by CV. They also determined the exchange current densities of CeCl3
using the linear polarization method.
Thus, the main goal of this work is to measure and analyze thermo-
dynamic and electrochemical properties of CeCl3 in LiCl-KCl eutectic
salt at different concentrations and temperatures using two methods:
1 CV for measuring the diffusion coefficients and apparent standard
potentials and 2 electrochemical impedance spectroscopy (EIS) for
determining the exchange current densities. This study will provide
useful insight into these properties with a unique feature of EIS tech-
nique by reducing uncertainty of electrode area measurement because
very small current is applied at around an open circuit potential.
Experimental
The electrochemical experiments were performed in an argon-
atmosphere glove box, as shown in Figure 1a. The oxygen and water
concentrations were monitored and maintained below 0.5 ppm (Figure
1b) for all the experiments. Within the glove box, a Kerrlab melting
furnace (Figure 1c) was used to melt and maintain the electrolyte
at the desired temperatures. The electrochemical measurements were
performed using a VSP-300 potentiostat/galvanostat from Biologic
Science Instrument (Figure 1d) at five different temperatures (698,
723, 748, 773, and 798 K). Figure 2 shows the experimental setup
within the furnace. The LiCl-KCl-CeCl3 electrolyte was loaded in
a tapered alumina crucible (Coorstek, 99.8% Al2 O3 ). This alumina

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 E238 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015)
Figure 1. Experimental setup and instruments. (a) Innovative Technology ar-
gon atmosphere glove box, (b) Oxygen and moisture display by sensor in the
glove box, (c) Kerrlab furnace, (d) VSP-300 potentiostat/galvanostat, Biologic
Science Instrument.
crucible was placed in a secondary crucible designed to contain any
molten salt upon possible failure of the primary crucible. Once the salt
was melted, the cathode, anode, and reference electrodes were lowered
into the salt through the alumina oxide sheaths. The salt temperature
was monitored via an inserted thermocouple (see Figure 2).
Anhydrous lithium chloride (LiCl) and potassium chloride (KCl)
(99.995%) were purchased from Alfa Aesar, and then mixed to pre-
pare a LiCl-KCl (58.2:41.8 in mol%) eutectic salt. The amount of
cerium chloride (0.102, 0.42, and 0.84 g ± 0.1 mg) was weighed and
added into 20 g ± 0.17 mg of LiCl-KCl yielding 0.0011 ± 9.8 × 10−4 ,
0.0047 ± 2.4 × 10−4 , and 0.0094 ± 1.2 × 10−4 mole fraction CeCl3
(equivalent to 0.5, 2, and 4 wt% of CeCl3 in LiCl-KCl salt), respec-
tively. Silver chloride (Alfa Aesar, 99.998%) and cerium metal chips
(Sigma-Aldrich 99.9%) were used as reference and counter electrode,
respectively.
Tungsten rod (1.5 mm and 2 mm in diameter) was used as the work-
ing electrode. The length of the working electrode submerged into the
salt was measured, and the measured surface areas were ranging from
0.32 cm2 to 0.63 cm2 depending on experimental runs. Prior to using
the counter electrode, an oxide layer on the cerium chips was elim-
inated using sand paper under argon environment. The cerium chips
Furnace lid
Furnace
Alumina
crucible
Safety crucible
Cerium chips in
Anode basket
Thermocouple
Cathode
LiCl-KCl
Reference electrode
eutectic salt
(1 mm Ag wire in 5
mol% Ag/AgCl)
Figure 2. Schematic sketch of the experimental setup of all electrochemical
experiments.
Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
0.4
0.2
Current [A]
0.0
-0.2
-0.4
-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0
Potential [V]
Figure 3. Cyclic-voltammogram of pure LiCl-KCl eutectic salt at 773 K at
the scan rate of 0.1 V s−1 . Tungsten rod (2 mm in diameter) was used as
cathode electrode, and the surface area was 0.471 cm2 .
were then loaded in a molybdenum basket and lowered into the pre-
pared salt. Silver-silver chloride (Ag/AgCl) reference electrode was
prepared by contacting a 1 mm Ag wire with LiCl-KCl-5 mol% AgCl
in a 7 mm diameter Pyrex tube. At the tip of the Pyrex tube, the thick-
ness of the wall was made thin enough (less than 0.5 mm in thickness)
allowing ionic conduction between the solution and electrolyte. Prior
to each experiment, the working electrode was anodically cleaned by
stripping at a potential of −0.1 V versus the reference electrode for 3
minutes. Then, an open circuit potential (OCP) was checked to ensure
the equilibrium condition has reached in the system. It should be noted
that all salts were dried at around 523 K for 5 hours to remove possible
moisture despite having the salt in sealed glass ampoules under argon
prior the melting processes. The furnace was heated at 5 K/min to
avoid thermal shock on the alumina crucible.
Results and Discussion
Cyclic voltammetry (CV) of the LiCl-KCl-CeCl3 system.— The CV
technique was first applied to the pure LiCl-KCl system to identify that
no other reaction occurs in the range between 0 V and −2.5 V versus
Ag/AgCl reference. The voltammogram of pure LiCl-KCl (in Figure
3) shows that Li reduction starts at −2.55 V (vs. 5 mol% Ag/AgCl). No
red-ox reaction between 0 to −2.4 V was observed and residual current
in that region was less than 2 mA. Therefore, it was safe to perform the
CV experiments over that potential range without worry from other
reactions. Figure 4 shows the cyclic voltammograms of CeCl3 (0.5
wt%, 2 wt%, and 4 wt%) in LiCl-KCl at 773 K. Cerium reduction and
oxidation peaks were observed at around −2.2 and −2.09 V versus
the Ag/AgCl reference electrode, respectively. For 0.5wt% CeCl3 ,
the peak potentials stay at the same potential under different scan
rates representing the reversibility of the reaction in the range of the
scan rate. However, the peak potentials move slightly in the negative
direction according to the scan rate when the concentration of CeCl3
was increased to 4 wt%. This may be considered as a quasi-reversible
reaction. The difference between peak potential and half peak potential
can be used to calculate the number of electron transferred by the
following expression,15
RT
E p − E 2p = −0.77 [1]
nF
where Ep is the peak potential (V), Ep/2 is the half peak potential (V),
R is the universal gas constant (J mol−1 K−1 ), and T is the absolute
temperature (K), F is the Faraday constant (C mol−1 ), and n is the
number of electrons transferred. The calculated number of electron
transferred, n, was ranging from 2.5 to 3.1 agreeing with the expected
value for the reduction process of Ce3+ /Ce. The cathodic peak cur-
rents were plotted with respect to the square root of the scan rate to
 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015) E239

0.08 0.4
(a) 50 mV/s (b) 50 mV/s
100 mV/s 100 mV/s
0.06 0.3
150 mV/s 150 mV/s
200 mV/s 200 mV/s
Current [A]

Current [A]
0.04 0.2

0.02 0.1

0.00 0.0

-0.02
-0.1
-0.04
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0
Potential [V] Potential [V]

0.3 (c) 50 mV/s

100 mV/s
150mV/s
0.2
200mV/s
Current [A]

0.1

0.0

-0.1

-0.2
-2.5 -2.0 -1.5 -1.0 -0.5 0.0

Potential [V]
Figure 4. Cyclic-voltammogram of CeCl3 in LiCl-KCl at 773 K at scan rates of 0.05, 0.1, 0.15, and 0.2V s−1 : (a) 0.5 wt% CeCl3 , with a working electrode
(tungsten) area of 0.534 cm2 . (b) 2 wt% CeCl3 with a working electrode (tungsten) area of 0.6283 cm2 . (c) 4 wt% CeCl3 with a working electrode (tungsten) area
of 0.3240 cm2 .

calculate diffusion coefficient of CeCl3 in LiCl-KCl using Delahay reversible soluble/insoluble system, the cathodic peak potential can
equation which is known for soluble-insoluble process:16 be expressed as5
 
n Fv D 1/2 RT   RT
i p = 0.611 n F SC0 [2] E p = E Ce
0∗
3+ /Ce ln X CeCl3 − 0.854 [4]
RT nF nF
where ip is the peak cathodic current (A), S is the electrode area (cm2 ), where Ep is the peak potential (V) obtained from the cathodic side
C0 is the bulk concentration of CeCl3 (mol cm−3 ), v is the scan rate in this case, X is the mole fraction, and E0∗ Ce 3+ /Ce is the apparent
(V s−1 ), and D is the diffusion coefficient (cm2 s−1 ). Here, the concen- standard potential. The apparent standard potentials versus Cl− /Cl2
tration of CeCl3 (mol cm−3 ) is a function of temperature. The diffusion reference electrode were calculated by using the potential difference
coefficient of CeCl3 was determined at different concentrations, as in- between Ag/AgCl (5 mol%) and Cl− /Cl2 reference electrode from
dicated in Figure 5. Small decrease of the values could be observed
by increasing concentration from 0.5 wt% to 2 wt%; however, the dif-
fusion coefficients of CeCl3 in LiCl-KCl salt were approximately the 1.1
same between the concentration of 2 and 4 wt%. Present study shows
1.0 0.5 wt%
smaller values for the diffusion coefficients comparing with those
2 wt%
from Marsden and Pesic14 and Iizuka.17 But these values possess a
0.9 4 wt%
similar trend. The diffusivity generally follows Arrhenius temperature
D [× 10 cm s ]
2 -1

relationship, which can be expressed as

  0.8
−E a
D = D0 ex p [3] 0.7
5

RT
where, D0 is the pre-exponential factor, and Ea is then an activation 0.6
energy (kJ mol−1 ) for the diffusion. Therefore, the activation energy
can be calculated from the slope when ln(D) is plotted versus 1/T. The 0.5
values of R-squared between the fitted regression lines and experi-
0.4
mental points were all greater than 0.96 indicating a good fit to the 1.1 1.2 1.3 1.4 1.5 1.6
data sets. Table I lists the diffusion coefficients of CeCl3 and the aver-
-1
age activation energies with different temperatures at three different 1000/T [K ]
concentrations.
From the cyclic voltammogram, the apparent standard potential Figure 5. Plots of diffusion coefficient of Ce3+ versus inverse temperature at
of CeCl3 was calculated from the cathodic peak potentials. For a different CeCl3 concentrations (0.5, 2, and 4 wt%).

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 E240 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015)

Table I. Diffusion coefficients (D) measured from CV experiments and activation energy (Ea ) at different concentrations and temperatures.

CeCl3 0.5 wt % 2 wt % 4 wt %

T [K] D [×105 cm2 s−1 ] Ea [kJ mol−1 ] D [×105 cm2 s−1 ] Ea [kJ mol−1 ] D [×105 cm2 s−1 ] Ea [kJ mol−1 ]
698 0.479 0.430 0.418
723 0.545 0.547 0.544
748 0.653 30.7 0.675 31.6 0.672 33.4
773 0.751 0.690 0.700
798 1.012 0.875 0.860

-3.06 the double-layer capacitance and W is the diffusion related resistance

(Warburg).
-3.08 A frequency ranging from 50 kHz to 50 mHz was used, and the ap-
plied potential amplitude was set at 10 mV. Figure 8 shows impedance
spectra for 0.5 wt% of CeCl3 in which the potential was gradually in-
Potential [V]

-3.10
-3.12
-3.14
-3.16
700 720 740 760
Temperature [K]
Figure 6. Plot of apparent standard potentials versus temperature.
the study by Yang and Hudson.18 The calculated apparent standard
potentials are plotted in Figure 6 showing a proportional relationship
with respect to an increase in temperature. The apparent standard
potentials for the concentration of 0.5 wt% and 2 wt% agree with
each other (staying within similar ranges of values), but the apparent
standard potential for 4 wt% of CeCl3 was slightly more negative. It
is suspected that this is due to the characteristic of quasi-reversibility
in the CV curves at high concentration of CeCl3 and it might be better
to use a soluble/insoluble irreversible expression for this calculation
in the future study.
Electrochemical impedance spectroscopy (EIS).— The EIS tech-
nique was selected and performed to calculate exchange current den-
sity (i0 ) of Ce3+ /Ce couple in LiCl-KCl salt. Compared to the linear
polarization technique, EIS has an advantage that the electrode surface
area is almost maintained the same due to an extremely small current
that is being applied at OCP or in that proximity. Therefore, uncer-
tainty of electrode area can be reduced significantly in determining
i0 . For analyzing impedance spectra, a simple equivalent circuit was
proposed as shown in Figure 7 where Rs is the solution resistance,
Rct is the charge transfer resistance on the electrode surface, Cdl is
creased from the equilibrium potential (−2.169 V). In general, an
impedance should be measured at an equilibrium potential to prop-
erly calculate i0 . However, at the equilibrium potential, the impedance
swiftly increases at the high frequency and downward distortion was
0.5 wt% observed at low frequency as shown in Figure 8 because no ion transfer
2 wt% can occur between the tungsten electrode and cerium ions. Therefore,
4 wt% minimum overpotentials (η = 1–5 mV) was applied to the cell for the
cerium reduction to occur at the electrode surface. For an example,
780 800 in Figure 8, by increasing η from equilibrium potential, a transition
point can be observed at −2.172 V (open circles in Figure 8), where
also a diffusion related impedance (Warburg impedance) started to be
seen at low frequency region. This indicates that electrons transfer and
diffusion from the bulk salt to the electrode surface started to occur
at that potential. In this case, current density flows through the EIS
experiment was only less than 1.5 mA cm−2 . The electron exchange
was confirmed by OCP measured right after the EIS experiments. Af-
ter performing EIS at the potential of −2.172 V, OCP was maintained
at the equilibrium potential for 500 s while OCP was released from
the equilibrium potential when the applied potential was lower than
−2.172 V. Therefore, minimum η for Ce3+ /Ce reduction to occur were
found and Rct were measured at those voltages by fitting the Nyquist
plot to the equivalent circuit.
The measured and fitted impedance spectra of Ce3+ /Ce for the
three different concentrations of CeCl3 at the different temperatures
are indicated in Figure 9. Instead of using capacitance and Warburg
impedance, constant phase element (CPE) was introduced, which is19
1
ZC P E = [5]
T ( jω)φ
10
-Im [Ohm]

Cdl 5 -2.169 V
-2.170 V
-2.171 V
-2.172 V
-2.173 V
-2.174 V
0
0 5 10 15 20 25
Rs
Re [Ohm]
Rct W
Figure 8. Nyquist plot for 0.5 wt% of CeCl3 in LiCl-KCl at 723 K on a
Figure 7. Equivalent circuit for the electrochemical cell showing bulk solution tungsten electrode. The frequency was from 50 kHz to 50 mHz, and the
resistance, double layer capacitance, charge transfer resistance, and diffusion amplitude of applied sinus potential was 10 mV. Applied potentials were
related resistance. ranging from −2.169 to −2.174 V with an equilibrium potential of −2.169 V.

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015) E241

3.5 1.2
(a) (b) Measured (698 K) Fitted (698 K)
Measured (698 K) Fitted (698 K) Measured (723 K) Fitted (723 K)
3.0 Mesured (723 K) Fitted (723 K) 1.0 Measured (748 K) Fitted (748 K)
Measured (748 K) Fitted (748 K) Measured (773 K) Fitted (773 K)
2.5 Measured (773 K) Fitted (773 K) 0.8 Measured (798 K) Fitted (798 K)
-Im [Ohm]

-Im [Ohm]
2.0
0.6
1.5
0.4
1.0
0.2
0.5

0.0 0.0
0 1 2 3 4 5 6 0.0 0.4 0.8 1.2 1.6
Re [Ohm] Re [Ohm]
1.0
(c) Measured (698 K) Fitted (698 K)
Measured (723 K) Fitted (723 K)
0.8
Measured (748 K) Fitted (748 K)
Measured (773 K) Fitted (773 K)
Measured (798 K) Fitted (798 K)
-Im [Ohm]

0.6

0.4

0.2

0.0
0.0 0.4 0.8 1.2 1.6
Re [Ohm]
Figure 9. Measured and fitted Nyquist plots at temperatures of 698, 723, 748, 773, and 798 K: (a) 0.5 wt% CeCl3 with a working electrode (tungsten) area of
0.534 cm2 ; (b) 2 wt% CeCl3 with a working electrode (tungsten) area of 0.6283 cm2 ; and (c) 4 wt% CeCl3 with a working electrode (tungsten) area of 0.3240 cm2 .

where T is a constant in F cm−2 sφ−1 is the constant, φ is the number
constant between −1 and 1, and ω is the frequency. CPE is useful in
fitting the equivalent circuit because it can represent resistor, inductor,
capacitor, Warburg response, and combination of these impedances
by changing the value of φ. First, the measured spectra were automat-
ically fitted by using randomize and simplex method in Z-fit software
(Bio-Logic), then a manual adjustment was done by changing the val-
ues of the equivalent circuit components. As the manual curve fitting
was performed, the relative error could be minimized below a fraction
of 10−1 . The measured Rct S and η are summarized in Table II. From
the measured Rct , i0 can be readily computed by using20
RT RT
Rct = = 2 2 0 1−α
n Fi 0 n F Sk CCeCl3
where k0 is the rate constant for Ce3+ /Ce, and α is the transfer coef-
ficient of Ce3+ /Ce. From the measured charge transfer resistance, k◦
can be calculated by assuming α is 0.5 based on the observation from
CV experiments that Ce3+ /Ce reaction follows reversible behaviors
with a weak diffusion effect (at slow scan rates). Table III summaries
i0 and k◦ calculated from Rct . Figure 10 plots the exchange current
densities of Ce3+ /Ce reaction which can be characterized with con-
centrations and temperatures. The results indicate that the exchange
current densities with 0.5 wt% of CeCl3 are in between 0.0076 A cm−2
and 0.016 A cm−2 , agreeing well with repeated experimental runs. By
increasing the concentration of CeCl3 to 4 wt%, the exchange current
density appears to increase up to 0.18 A cm−2 . Marsden and Pesic14
reported the exchange current density of CeCl3 at 4 wt% concentration
using the linear polarization method. The values of i0 from this study
are slightly higher, but both studies show similar range of values for
the exchange current density of CeCl3 in LiCl-KCl salt.
Based on the given data sets, Arrhenius temperature dependence
form can be applied to further looking into temperature effects on the
[6]
exchange current density using the expressing i0 = I0 exp(-Ea /RT)
where I0 is the pre-exponential factors (often referred to as an ex-
change current density at an infinite temperature). Figure 11 shows
a plot of the logarithm of i0 against the inverse temperature. Here, a
straight line can be seen for all three different CeCl3 concentrations. Ea
and I0 were calculated from the slope of the straight lines and the inter-
cept of ln i0 , respectively. Ea values for Ce3+ /Ce were 34.5, 30.9, and

Table II. The applied overpotential (η) and measured charge transfers (Rct ) times electrode surface area (S) values at different concentrations
and temperatures.

CeCl3 0.5 wt % 2 wt % 4 wt %

T [K] η [V] Rct S [ cm2 ] η [V] Rct S [ cm2 ] η [V] Rct S [ cm2 ]
698 0.006 2.63 0.004 0.34 0.003 0.24
723 0.006 2.29 0.004 0.30 0.003 0.19
748 0.006 2.09 0.004 0.24 0.002 0.18
773 0.006 1.62 0.004 0.22 0.002 0.15
798 0.005 1.45 0.004 0.20 0.001 0.13

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 E242 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015)

Table III. Calculated exchange current density (i0 ) and rate constant (k0 ) for the charge transfer at different concentrations and temperatures.

CeCl3 0.5 wt % 2 wt % 4 wt %

T [K] i0 [A cm−2 ] k0 [×105 cm s−1 ] i0 [A cm−2 ] k0 [×105 cm s−1 ] i0 [A cm−2 ] k0 [×105 cm s−1 ]
698 0.0076 0.450 0.058 1.71 0.086 1.69
723 0.0091 0.536 0.068 1.99 0.107 2.20
748 0.0103 0.610 0.088 2.57 0.112 2.44
773 0.0137 0.817 0.098 2.88 0.146 3.03
798 0.0158 0.944 0.113 3.33 0.177 3.69

0.020
0.25 0.5wt%
0.5 wt% (repeat)
2 wt%
0.20 4 wt% 0.015
Ref. [14]
i0 [A cm-2]

0.15

0.10
0.05
0.00
1.0 1.1 1.2 1.3 1.4
1000/T [K-1]
Figure 10. Plots of exchange current densities versus inverse temperature
from various experiment sets. Experiments for 0.5 wt% of CeCl3 ( = 3.35 ×
10−5 mol cm−3 at 773 K) has a cathode surface area of 0.534 cm2 and 0.685
cm2 (repeat). Experiments for 2 wt% of CeCl3 ( = 1.38 × 10−4 mol cm−3 at
773 K) has a cathode surface area of 0.6283 cm2 . Experiments for 4 wt% of
CeCl3 ( = 2.77 × 10−4 mol cm−3 at 773 K) has a surface area of 0.3240 cm2 .
32.4 kJ mol−1 (R2 > 0.98) for 0.5 wt%, 2 wt% and 4 wt%, respectively.
These values are similar to the activation energy for U3+ /U measured
by Roes et al.,13 which is 34.5 kJ mol−1 . As expected, higher activa-
tion energy is required for the charge transfer at lower concentration
of CeCl3 . Interestingly, the activation energy values from the diffusion
coefficients shown in Table I are within similar range in comparing to
those for the charge transfer, but behave in an opposite trend. That is,
the activation energy for the diffusion increases with increasing the
concentration of CeCl3 , suggesting that it would be due to interaction
between particles at high concentration. Dimensionless quantities of
i0 /I0 are plotted versus exp(-Ea /RT) in Figure 12. Marsden and Pesic14
reported the values of i0 in broad range, so that values were being av-
eraged, calculated, and superimposed onto Figure 12 for comparison.
i0/I0
0.010
0.5 wt%
2 wt%
0.005
4 wt%
Ref. [14]
0.005 0.010 0.015 0.020
1.5 1.6 1.7
Exp(-Ea/RT)
Figure 12. Plot of i0 /I0 versus exp(-Ea /RT).
All exchange current densities from three different concentrations are
laid on a single straight line. Although the data point at 773 K in the
study by Marsden and Pesic14 is slightly off from the trend line, it
is shown here that the exchange current densities from both studies
exhibit a similar trend on temperature effect.
Analysis on practical application.— The results of this work pro-
vide the fact that cerium is a good surrogate material for uranium
since they show similar electrochemical and thermodynamic behav-
iors in LiCl-KCl eutectic salt. By comparing the properties of cerium
with those of uranium, both are very stable in the trivalent form in
LiCl-KCl salt and reduced to metal form by gaining three electrons at
certain potentials. However, the standard reduction potential of UCl3
is ranging from -2.4 to -2.6 V versus Cl2 /Cl− reference electrode,5–9
which is about 0.7 V more positive than the standard reduction poten-
tial for CeCl3 . The diffusion coefficients for UCl3 in LiCl-KCl molten
salt have been reported by many researchers,4–9,11 which are shown in
the Figure 13. Although the values for UCl3 are generally higher than

5
-1 CeCl3-0.5wt% 2wt% 4wt%
0.5 wt%
2 wt% 4 Ref. [4] Ref. [5] Ref. [6]
4 wt% Ref. [7] Ref. [8] Ref. [9]
-2
D [×10 cm s ]

Ref. [11]
-1

3
2

-3
ln i0

2
5

-4 1

-5 0
0.00125 0.00130 0.00135 0.00140 0.00145 1.20 1.25 1.30 1.35 1.40 1.45
-1 -1
1/T [K ] 1000/T [K ]
Figure 11. Plot of ln (i0 ) versus inverse temperature showing Arrhenius de- Figure 13. Plot of diffusion coefficients for UCl3 from other studies, compar-
pendency. ing with the diffusion coefficients of CeCl3 in this study.

Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
 Journal of The Electrochemical Society, 162 (10) E237-E243 (2015)
0.6
CeCl3(0.5 wt%)
CeCl3(2 wt%)
CeCl3 (4 wt%)
0.4
Ref. [11]
i0 [A cm ]
-2
Ref. [12]
Ref. [13]
0.2
0.0
1.1 1.2 1.3 1.4 1.5 1.6
-1
1000/T [K ]
Figure 14. Plot of the exchange current densities for U3+ /U reaction from
other research studies comparing with those of Ce3+ /Ce measured in this
study.
those for CeCl3 , the diffusion coefficients for both UCl3 and CeCl3
are in the same order of magnitude and can be correlated with the
temperature. The activation energies for the diffusion of UCl3 have
been reported, ranging from 24.2 to 34.4 kJ mol−1 ,5,7 which is in a
good agreement with the activation energy for CeCl3 . The similarity
between both the activation energies for the diffusion may be owing
to the similar ionic size of uranium and cerium. The exchange cur-
rent densities of uranium in the molten salt have been reported with
few different methods.11–13 As shown in Figure 14, the reported i0 for
U3+ /U reaction is scattered in the order from 10−2 to 10−1 ; thus, it is
difficult to compare each data or observe its trend. The main reason of
the deviation could be that the authors performed the different meth-
ods to measure i0 at different temperatures and concentrations, and
also be the challenges that have been previously mentioned. There-
fore, further study is necessary to determine the exchange density for
U3+ /U reaction, and observe its trend regarding to the concentration
and temperature. EIS technique in this study shows an alternative
method in determining i0 of CeCl3 , and can also be an important
tool to investigate the exchange current density of UCl3 in LiCl-KCl
eutectic salt.
Conclusions
Electrochemical properties of CeCl3 in LiCl-KCl eutectic salt at
different concentrations and temperatures have been studied by using
CV and EIS techniques. From the CV experiments, it has been shown
three-electron exchange for the reduction of Ce3+ /Ce. The diffusion
coefficients of CeCl3 in LiCl-KCl were measured and calculated to be
from 0.48 × 10−5 to 1.01 × 10−5 cm2 s−1 , which can be correlated
with temperature using Arrhenius expression. The results reveal that
the concentration of CeCl3 has a weak effect on the diffusion coeffi-
cients. Comparing with the values for UCl3 , the diffusion coefficients
of CeCl3 are slightly smaller than those of UCl3 . Apparent standard
potentials were also computed by using peak potentials from the CV
experiments, which were linearly dependent on temperature. EIS ex-
Downloaded on 2017-11-09 to IP 190.111.30.254 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).
E243
periments were performed to determine exchange current density of
Ce3+ /Ce couple in LiCl-KCl molten salt system. Minimum η was
applied for Ce reduction to happen and the charge transfer resistance
was measured to calculate the exchange current density. The exchange
current densities range from 0.0076 A cm−2 and 0.18 A cm−2 , which
can be related to temperature and concentration (see Table III). From
Arrhenius temperature dependence, the activation energy for Ce3+ /Ce
exchange was determined though EIS data, which is in the same range
obtaining through CV data sets and in similar range with the activa-
tion energy for U3+ /U.13 By plotting dimensionless quantities of the
exchange current density, the exchange current densities of Ce3+ /Ce
reaction in this work are in good agreement with those measured by
linear polarization method.14 In comparison with the exchange cur-
rent densities of U3+ /U measured by other researchers, the values for
Ce3+ /Ce are in a similar order, but a meaningful comparison is hard
to be made due to the dispersed data for uranium.
Acknowledgments
This study is being performed using funding received from the
DOE Office of Nuclear Energy’s Nuclear Energy University Pro-
grams (NEUP) in collaboration with the Ohio State University
(NEUP-13-4908).
References
1. J. J. Laidler, J. E. Battles, W. E. Miller, J. P. Ackerman, and E. L. Carls, Progress in
Nuclear Energy, 31, 131 (1997).
2. K. Uozumi, M. Iizuka, T. Kato, T. Inoue, O. Shirai, and T. Iwai, Journal of Nuclear
Materials 325, 34 (2004).
3. T. Koyama, M. Iizuka, Y. Shoji, R. Fujita, H. Tanaka, T. Konayashi, and M. Tokiwai,
Journal of Nuclear Science and Technology, 34, 384 (1997).
4. G. Kim, D. Yoon, S. Paek, S. Kim, T. Kim, and D. Ahn, Journal of Electroanalytical
Chemistry, 682, 128 (2012).
5. P. Masset, D. Bottomley, R. Konings, R. Malmbeck,A. Rodrigues, J. Serp, and
J. Glatz, Journal of The Electrochemical Society, 152(6), A1109 (2005).
6. P. Masset, R. J. M. Konings, R. Malmbeck, J. Sert, and J. Glatz, Journal of Nuclear
Materials, 344, 173 (2005).
7. S. A. Kuznetsov, H. Hayashi, K. Minato, and M. Gaune-Escard, Electrochimica Acta,
51, 2463 (2006).
8. S. A. Kuznetsov, H. Hayashi, K. Minato, and M. Gaune-Escardc, Journal of The
Electrochemical Society, 152(4), C203 (2005).
9. R. O. Hoover, M. R. Shaltry, S. Martin, K. Sridharan, and S. Phongikaroon, Journal
of Nuclear Materials, 452, 389 (2014).
10. Jinsuo Zhang, Kinetic model for electrorefining, Progress in Nuclear Energy, 70, 279
(2014).
11. I. Choi, B. E. Serrano, S. X. Li, S. Herrmann, and S. Phongikaroon, Paper 9045
presented at Proceedings of Global 2009, Paris, France, Sept. 6–11 (2009).
12. S. Ghosh, S. Vandarkuzhali, N. Gogoi, P. Venkatesh, and G. Seenivasan, Electrochim-
ica Acta, 56, 8204 (2011).
13. M. A. Rose, M. A. Williamson, and J. Willit, ECS Electrochemistry Letters, 4(1), C5
(2015).
14. K. C. Marsden and B. Pesic, Journal of The Electrochemical Society, 159(6), F111
(2011).
15. D. S. Poa, Z. Tomczuk, and R. K. Steunenberg, Journal of The Electrochemical
Society, 135, 1161 (1988).
16. P. Delahay, New Instrumental Methods in Electrochemistry: Theory, Instrumenta-
tion and Application to Analytical and Physical Chemistry, Interscience, New York
(1954).
17. M. Iizuka, Journal of Electrochemical Society, 145, 84 (1998).
18. L. Yang and R. G. Hudson, Journal of The Electrochemical Society, 106, 986 (1959).
19. A. Lasia, Electrochemical Impedance Spectroscopy and Its Applications, New York:
Springer (2014).
20. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Appli-
cations, New York: Wiley (2001).