Copper extraction from copper ore by electro-reduction in molten CaCl2-NaCl

Article abstract of DOI:10.1016/j.electacta.2009.03.015

Sintered solid porous pellets of copper sulfide (Cu₂S) and Cu₂S/FeS were electrolysed at a cell voltage of 2.2-2.8 V to elemental Cu, S and Cu, Fe, S, respectively in molten CaCl₂-NaCl at 800 °C under the protection of arg

Full text of DOI:10.1016/j.electacta.2009.03.015

Electrochimica Acta 54 (2009) 4397–4402
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Copper extraction from copper ore by electro-reduction in molten CaCl –NaCl
2
a,b,∗
c
a
Xinlei Ge
, Xidong Wang , Seshadri Seetharaman
a
Department of Materials Science and Engineering, Royal Institute of Technology, SE 100 44, Stockholm, Sweden
Department of Physical Chemistry, University of Science and Technology, Beijing, 100083, China
College of Engineering, Peking University, Beijing, 100871, China
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Sintered solid porous pellets of copper sulfide (Cu S) and Cu S/FeS were electrolysed at a cell voltage of
2
2
Received 24 October 2008
Received in revised form 7 March 2009
Accepted 9 March 2009
◦
2.2–2.8 V to elemental Cu, S and Cu, Fe, S, respectively in molten CaCl2–NaCl at 800 C under the protection
of argon gas. The process parameters for optimal electrolysis are presented. The electrolysis products are
characterized by microscopic techniques and XRD. The product characteristics are linked to the process
parameters. The direct electrolysis of the sulfide to copper with the emission of elemental sulphur offers
an attractive green process route for the treatment of copper ore.
Available online 20 March 2009
Keywords:
Crown Copyright © 2009 Published by Elsevier Ltd. All rights reserved.
Copper extraction
Electro-reduction
Molten salts
Green process
SO2 emission
1. Introduction
have been paid on the direct reduction of pure or mixed oxides
[
5–20]. However, very little work has so far been carried out in the
Over 90% of the world’s primary copper on the earth’s crust
is present in the form of sulfide minerals, such as chalcocite
electrolytic reduction of sulfides. Li et al. [21] had investigated the
electrolysis of MoS . Chen and Fray [22] have examined the removal
2
(
Cu S), chalcopyrite (CuFeS ) and bornite (Cu5FeS ), which are
of S in liquid copper. The present work aims to demonstrate a poten-
tial of the electro-reduction route towards copper extraction from
sulfide material aiming at capturing sulphur in the elemental form.
2
2
4
always treated by pyrometallurgical processes [1]. Copper also
occurs in the form of oxidic ores (carbonates, oxides, silicates,
sulphates) to a much lesser extent. The latter can be treated by heap-
leaching together with solvent extraction and electro-winning, or
by bacteria-aided bioleaching. For sulfide ores, after concentration
by froth flotation, the extraction of copper is carried out with the
following three steps: (a) roasting, (b) matte smelting (in blast,
reverberatory, electric or flash furnaces) and (c) converting to blister
copper. These three steps can also be combined into a continuous
process [1]. However, emission of SO , despite the cost-intensive
modern techniques of SO -capture, is a difficult problem consider-
ing the increased environmental restrictions.
Thus, it is essential to find a method to capture most of the sul-
phur in a stable form, especially as elemental sulphur. In the past
few years, a generic process [2–4] for producing metal and alloy
by direct electrochemical reduction of corresponding metal com-
pound or mixtures has attracted worldwide attention, because of
its lower energy consumption and simple operation. Many efforts
2. Experimental
2.1. Thermodynamic principles
Even though the electro-reduction principles have been pre-
sented in a number of earlier publications [5–20], a brief outline is
given here in order to orient the readers to the context of this paper.
The oxide or sulfide compact, wrapped by a metal gauze is used as
a cathode in an electrolytic process using molten CaCl /CaCl –NaCl
2
2
2
2
◦
electrolyte in the temperature range of 700–1000 C. Graphite is
normally used as the anode. The oxide or sulfide gets reduced to
the metallic state.
During the above process with respect to oxide materials, the
use of graphite anodes leads inevitably to the carbon getting oxi-
dized, leading to the emission of CO and CO . This, in turn, has
2
negative environmental consequences. In addition, the consump-
tion of graphite during the process has an adverse economic impact.
On the other hand, use of inert materials would add a cost fac-
tor and has not been favoured in industrial practice. In the present
case, sulphur would be released at the anode leading to the reaction
between elemental sulphur and carbon forming carbon sulfides. It
∗
Corresponding author at: Department of Materials Science and Engineering,
Royal Institute of Technology, Brinellvagen 23, SE 100 44, Stockholm, Sweden.
Tel.: +46 08790 8326.
E-mail address: xinleige@yahoo.com.cn (X. Ge).
0
013-4686/$ – see front matter. Crown Copyright © 2009 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.03.015

4
398
X. Ge et al. / Electrochimica Acta 54 (2009) 4397–4402
is imperative that these reactions are carefully controlled, since CS2
is very harmful to fauna.
an ultrasonic treatment. For Cu S/FeS pellet, ethanol was used for
2
washing and subsequent ultrasonic treatment. The electrolysis pel-
◦
In this respect, the following reactions are considered:
let was then dried at 60–110 C quickly and stored in a desiccator.
The electrolysis Cu S/FeS pellets were subjected to analysis quickly
2
◦
C + S = CS ,
ꢀG (1073 K) = −18.23 kJ,
(1)
(2)
2
2
to reduce possible oxidation of iron. The samples were analyzed
by X-Ray Diffraction technique (Siemens D5000), and Scanning
Electron Microscope (SEM) (JSM-840, JEOL) equipped with Energy-
Dispersive X-ray Spectrometry analysis (EDS link, Oxford).
◦
C + 0.5S = CS,
ꢁG (1073 K) = 118.59 kJ,
2
The values of standard Gibbs free energy in this work were
all calculated by Factsage Software [23–24], and the theoretical
decomposition voltages of Cu S, Cu O and FeS shown in the follow-
ing sections were estimated from their corresponding Gibbs energy
values.
It can be seen that the formation of CS is not thermodynamically
favoured. On the other hand, CS formation is favourable, but Gibbs
energy change for reaction (1) is only slightly negative as compared
to the Gibbs energy changes accompanying the formation of CO2
2
2
3. Results and discussion
Fig. 1a shows the XRD patterns of reduced Cu S pellet. It is seen
2
that the XRD peaks show the presence of elemental Cu as well as
CaS. It is also seen that the CaS content decreases with the increase
of electrolysis time. The similarity in the XRD patterns for the elec-
trolysis corresponding to 10, 12 and 20 h indicates that pure copper
could be obtained after 10–12 h electrolysis at 2.2–2.8 V.
During every experimental run, a yellow powder deposit could
always be observed along the tube, gas path and in the condenser
without any liquid condensate indicating that the formation of CS2
was insignificant. The yellowish powder in the condenser was con-
firmed to be elemental sulphur after filtration in distilled water
2
◦
and CO at the same temperature, viz. ꢀG (CO , 1073 K) = −396.09 kJ
2
◦
and ꢀG (CO, 1073 K) = −206.38 kJ, respectively. Theoretically, the
main anode product should be elemental sulphur along with some
CS provided the oxygen partial pressure prevailing is very low. This
will be in the vapour form at the experimental temperatures and is
likely to condense along the tube or can be collected in a condenser.
2
It is to be noted that CS would be condensed in the liquid form (b.p.
2
(Fig. 1b). Since sulphur can condense in the low temperature zone
◦
4
6.5 C of CS ), leading to near-zero sulphur emission.
2
of the tube and gas channels, it is difficult to collect all the sulphur
in the present laboratory investigations in order to quantify this
conclusion.
2
.2. Materials and procedures
CaCl2 and NaCl powders (Anhydrous, E. Merck, Darmstadt, Ger-
◦
many, >99%) were kept in an oven at 110 C before use. The salts were
weighed and mixed evenly in the molar ratio of 1:1. The salt mix-
ture was filled in an alumina crucible (40 mm OD, 36 mm ID, 60 mm
height, supplied by KERANOVA AB, Sweden), which was dried in
◦
an oven at 110 C before use. Cu S powder (Alfa Aesar, 99.5%), was
2
pressed into pellets with a diameter 16 mm and weight of ∼3 g at a
pressure around 6 MPa, the pellets were sintered in argon gas (ICP
◦
5
.0, AGA gas AB, Sweden) at 400 C for 2 h. Similar procedure was
used for mixed powder (1:2 mole ratio) of Cu S and FeS (Alfa Aesar,
2
tech.) to form the Cu S/FeS pellet. A thin copper wire, ∼0.2 mm
2
diameter, was used to wrap the pellet. All pellets were stored in a
desiccator, before attaching to an iron rod, 5 mm diameter, as the
cathode. Graphite rod (4 mm diameter, LORRAINE PARIS) was used
as anode. The DC power supply (hp Hewlett 6632A) was employed.
Argon gas was used as the protective gas.
The crucible was filled with CaCl –NaCl salt mixture and placed
2
in an alumina reaction tube. The reaction tube was sealed with sil-
icone rubber stoppers. A thermocouple was positioned inside the
reaction tube in contact with the bottom of the alumina crucible.
The entire assembly was heated by a programmed heating pat-
tern (EUROTHERM temperature controller) under the protection of
◦
argon gas. First, the reactor was slowly heated to 300 C at a heating
◦
rate 2 C/min, held for 2 h to further remove any moisture trace left
◦
in the system. The furnace was then heated at a rate of 6 C/min to
the targeted temperature. The salts were kept at this temperature
for 0.5–1 h in order to ensure homogeneity of the molten salts. The
electrodes were then inserted into the salt bath. Since the theoreti-
cal decomposition voltages of Cu O and FeS are 0.469 V and 0.467 V,
2
respectively, while the value for Cu S is 0.529 V. Thus, we chosen
2
0
.5 V in order to electrolyze possible Cu O and FeS without elec-
2
trolyzing Cu S. Of course, the pre-electrolysis parameters should
2
be improved after more electrochemical measurements have been
done. After it, the electrolysis process was carried out at a pre-
determined cell voltage. The current change during electrolysis was
recorded.
After the electrolysis, the electrodes were held above the molten
salts and cooled to room temperature under argon gas protection.
The cathode was washed with distilled water and subjected to
Fig. 1. X-ray diffraction patterns for electro-reduction of Cu2S pellets at 2.8 V in
molten CaCl2–NaCl at 800 C: (a) cathode products under different electrolysis time,
and (b) anode product collected in the condenser after washing in distilled water.
◦

X. Ge et al. / Electrochimica Acta 54 (2009) 4397–4402
4399
For the electrolysis of Cu S, if S2−/S2 can be regarded as the ref-
2
erence potential, the electrode reactions of Cu S at 1073 K can be
2
represented as
anode : S2 − 2e = 1/2S (g),
−
E(S |S ) = 0 V
2−
(3)
(4)
2
2
+
+
cathode : Cu + e = Cu,
E(Cu |Cu) = −0.529 V
Thus, when the electrolysis starts, sulphur in Cu S cathode gets
2
ionized and is transported to the anode, which may cause the pre-
cipitation of CaS. In the research about the reduction of Cr O , Chen
2
3
et al. [25] proposed two possible mechanisms and concluded that
the calcium chromites was more likely to be formed by a chemi-
cal precipitation step. Similarly, in case of Cu S, the calcium sulfide
2
precipitation can be represented as follows:
2−
Cu S + 2e = 2Cu + S
2
(
at the interface of electrolyte-metal-sulfide)
(5)
(6)
Fig. 2. Cross-section of the reduced Cu2S pellet after 6 h electrolysis at 2.8 V in
molten CaCl2–NaCl at 800 C.
◦
S2− + Ca2+ = CaS(in the electrolyte)
This would explain the presence of CaS in the cathode product
in the early stages of electrolysis.
which, in turn would reduce Cu S to Cu according to the reaction:
2
◦
However, under a relative high voltage of 2.8 V (namely,
Ca + Cu S = CaS + 2Cu
ꢀG (1073 K) = −331.396 kJ
(8)
2
E(vsS2 |S ) = −2.8 V), the decomposition of CaS, namely, deposition
−
2
Reaction (8) would enable the re-formation of CaS. Thus, in this
case, CaS can be considered to function as a catalyst in the cathodic
reaction.
of calcium also can happen:
at 1073 K : Ca2 + 2e = Ca,
+
E(Ca2+|Ca) = −2.24 V
(7)
Fig. 3. Scanning electron microscopy of reduced Cu2S pellets: (a) the surface of pellet before electrolysis, (b) Cu powder after 20 h electrolysis at 2.2 V in molten CaCl2–NaCl,
c) three structures found on the outside layer, and (d) the tube structure.
(

4
400
X. Ge et al. / Electrochimica Acta 54 (2009) 4397–4402
If the CaS formed had a low solubility in the salt melt, it could
In the present work, an attempt was also made to explain the for-
mation of the needle-like structure close to the surface as shown
in Fig. 3c and d. For the needle structure, further investigations
show that these are actually tubes in micrometer size, the formation
of which can also be attributed to the electrochemical deposition.
It is known from earlier work [26–28] that Cu nanotube can be
synthesized by the direct electrochemical deposition on a porous
membrane with very careful treatment to obtain the ideal orien-
tation. Nanotubes without very good orientation even can grow
from a copper coil without any template and additives [29]. Under
the present experimental conditions, the number of pores on the
outer layer of the pellet is significant as can be seen in Fig. 3a. With
the removal of sulphur during the electrolysis, the porosity would
increase. The likelihood of copper getting deposited along the pores
to form a tube would be high, in analogy with a porous mem-
brane. Consequently, micrometer size pores close to the surface of
the pellet would lead to the formation of the tube structure in the
micrometer size (Fig. 3d). However, preliminary experiments sug-
gest that these two structures, which are confined to the outer layer
can, to a smaller extent, be controlled by a stringent control of the
chemistry of the salt mixture and the purity of the gas atmosphere
as further sulphur depletion from the surface can occur leading to
higher porosity.
be present in the form of Ca2+ and S . As electrolysis proceeds
2−
further, CaS will dissociate into Ca2 and S . The S ions would
be transported to the anode during the electrolysis and would be
oxidized at the anode to form elemental sulphur according to the
anodic reaction (3).
+
2−
2−
During the present series of experiments, the graphite anode
almost remained intact after every experimental run, confirming
the negligible formation of the anodic product, CS2.
EDS analysis of the different locations in a partially reduced pel-
let as shown in Fig. 2, indicates that the reduction pathway proceeds
from outside of the pellet to the inside. Content of copper decreases
towards the interior of the pellet, while Ca and S contents increase. It
should be pointed out that the sample is very clean and the amount
of salt, such as Na, Cl, was insignificant, even though the porous
structure is probably to be filled with molten salts during elec-
trolysis. It is admitted that the high solubility of CaCl2 and NaCl
in distilled water may contribute to some extent. However, it also
indicates that wetting could be poor between the reduced porous
copper and salt melts. This phenomenon means that the separation
of copper and molten salts is quite easy.
Fig. 3 shows the SEM images of Cu S pellets before and after
2
electrolysis. Fig. 3a shows the surface image of sintered pellet before
electrolysis. It is seen the surface of the sintered Cu S pellet consists
The current efficiency and energy consumption can be estimated
from the current–time plots during electrolysis in combination
with corresponding sulphur contents in the electrolysis products
2
of particles with sizes varying in a wide range. After 20 h electroly-
sis, the pellet presents a clear appearance of copper (inset of Fig. 3b).
The particles are of ∼10 ␮m, and well sintered, showing a more
complicated nodular structure. This can probably be attributed to
by chemical analysis. In the case of pure Cu S, shown in Fig. 4a,
2
◦
the relatively low melting point of copper (1083 C) and long elec-
trolysis time of 20 h. It should be noted that Fig. 3b is taken from the
inner section of the reduced pellet. This shows mainly the existence
of nodular structure. Fig. 3c shows the structures corresponding
to the outer part of the pellet. Here, it can be seen that, besides
the nodular structure, two other interesting structures, dendritic
as well as needle crystals, can be found. Fig. 3d presents a clear
view of the needle structure. The typical dendritic structure shown
in Fig. 3c is more likely to be formed by electrodeposition from the
salt, not from the reduction of Cu S.
2
In the present experiments, no special steps were taken to purify
the salt mixtures such as purification by chlorine gas. Further, high
purity argon gas was used in the electrolysis experiments as sup-
plied without taking extra precautions for the removal of traces of
oxygen, moisture and carbon dioxide. This would mean that the
following possible side reactions might occur:
◦
Cu S + 1.5O = Cu O + SO ,
ꢀG (1073 K) = −272.09 kJ
◦
(9)
(10)
(11)
2
2
2
2
Cu S + O = 2Cu + SO ,
ꢁG (1073 K) = −181.55 kJ
2
2
2
◦
2
Cu + 0.5O = Cu O,
ꢁG (1073 K) = −90.55 kJ
2
2
Even assuming that the partial pressure of SO2 as 1 atm, the
equilibrium oxygen pressures of the above reactions are around
−9
1
.5 × 10 atm, which means the above equations are highly pos-
sible.
Since CaCl2 is very sensitive to the moisture, HCl gas may be
generated by the reaction:
CaCl (salt) + 2H O (gas) = Ca(OH) (solid) + 2HCl (salt)
(12)
2
2
2
The generated HCl can react with Cu O according to the reaction:
2
◦
Cu O + 2HCl = 2CuCl + H O,
ꢀG (1073 K) = −76.89 kJ (13)
2
2
CuCl has
a
low decomposition voltage of Ed(1073 K,
CuCl) = 0.939 V, Cu can be deposited to form the dendritic crystals
on the surface. Another possibility is that Cu S may dissolve in the
2
salt melt to form Cu and S , and then Cu+ gets deposited on the
+
2−
Fig. 4. Typical current–time plots for pellet reduction in molten CaCl2–NaCl at
800 C. Pellets are sintered at 400 C in argon: (a) Cu2S pellet, and (b) Cu2S/FeS pellet.
◦
◦
surface. This would explain the observation in Fig. 3c.

X. Ge et al. / Electrochimica Acta 54 (2009) 4397–4402
4401
the current shows a sudden drop within the first 20 min. After a
short delay, the decrease in current continues and reaches a steady
value after 1–2 h lasting until the termination of electrolysis. The
current efficiency estimated from Fig. 4a is 32.8%, corresponding to
let started disintegrating during the process, leading inevitably to
material loss into the bulk electrolyte. Preliminary experiments
with chalcopyrite show that prior heating of CuFeS2 in argon
was essential, during which part of sulphur may be lost. The
matte remaining could then be pressed to form the porous pel-
let and subjected to electro-reduction. A selective leaching step by
hydrometallurgy in order to separate Cu and Fe is being currently
developed and the results are expected to be reported shortly. In
3
.60 kWh/kg Cu. However, the current efficiency is not very high
and actually it is also a major obstacle in reduction of many oxides.
The current efficiency depends on a number of factors such as pel-
let thickness and porosity, etc. Under similar conditions, thicker
and denser pellets showed a lower current efficiency. Strict control
of electrolysis procedures and conditions are also important. More
importantly, since reduction of Ca2 to a low valence state, e.g. Ca(0)
and Ca(I) is highly likely to occur, this will not only lower the current
efficiency as a side-reaction, but also increase the electronic con-
ductivity of the molten salt, which can bring about a relative high
background current [5]. Further, the removal of the last amount of
the present work, Cu S and FeS mixed powder (1:2 mole ratio)
2
was employed to avoid this problem. A similar change of current
+
was observed in the case of compacts of a mixture of Cu S and
2
FeS pellets as can be seen in Fig. 4b. The current efficiency for the
electro-reduction of Cu S/FeS compacts was estimated from Fig. 4b
2
to be 23.76%, with an energy consumption of 6.23 kWh/kg (CuFe).
XRD patterns of the reduced Cu S/FeS pellets are presented in
2
2
−
S
is a slow process in comparison with the reduction of Cu2S
Fig. 5a. One can see that iron is reduced first, probably due to a lower
phase. Thus, current efficiency is significantly influenced by the
background current. When electrolyzed at 2.8 V, the current effi-
ciency is estimated to be 74.08% for 2 h electrolysis, 46.61% for 4 h
and 16.56% for 20 h. Probably, in order to improve the current effi-
ciency, the low porosity (∼28%) in the present experiment can be
increased by employing a lower pressure (below 6 MPa) in pressing
the pellets or having additives, which would evaporate leaving cav-
ities. However, great effort is still necessary to reduce or eliminate
the background current.
decomposition voltage of E (1073 K, FeS) = 0.467 V. Even in this case,
d
CaS was found in the XRD patterns indicating the formation of the
same as an intermediate product. The amount of CaS was found
to be decreasing with the increase of electrolysis time and after
20 h electrolysis, it was totally absent. The product consisted only
of copper and iron. EDS analysis shows that the electrolysis product,
in fact, was a two-phase mixture of copper and iron, which would
be in accordance with Cu–Fe phase diagram. Fig. 5b shows a Cu-
rich particle surrounded by the CaS particles in a partially reduced
pellet.
The results obtained so far show a promising green process
route for treatment of copper ore by the electro-reduction method.
However, further work is needed to be carried out in order to
understand the reduction mechanism of pellet, formation of the
dendritic- and needle-structure, collection of anode sulphur, opti-
mization of process parameters such as cell voltage, sintering
temperature and pellet porosity to improve the current efficiency,
etc.
Attempts were also made in the present investigation to study
the electro-reduction of chalcopyrite. But the chalcopyrite pel-
4. Conclusions
In the present work, a potential process for copper extraction
from copper ores in molten CaCl –NaCl melts is put forward by
2
an electro-reduction method without SO2 emission. The sintered
solid porous Cu S or Cu S/FeS pellets can be electrolyzed to ele-
2
2
◦
mental Cu, S and Cu, Fe, S under a cell voltage 2.2–2.8 V at 800 C
by the evidence of XRD patterns. CaS can be formed as the inter-
mediate product in the early stage of electrolysis. SEM/EDS results
prove that the reduction proceeds from outside to inside of the pel-
lets. The particles are well sintered, and a small amount of deposits
with dendritic- and needle-like structures were found at the out-
side layer of the reduced pellets. The formation of these structures
is attributed to the electrodeposition of copper due to the side-
reactions. However, the electro-reduction process is less sensitive
to factors that can cause side reactions and thus, is suitable for
industrial applications. The current efficiency estimated for Cu2S
pellets is 32.8%, corresponding to 3.60 kWh/kg Cu; and 23.76%,
with an energy consumption of 6.23 kWh/kg (CuFe) for reduction of
Cu S/FeS compacts. All the results presented here show a promising
2
green process route for treatment of copper ore by this electro-
reduction method.
Acknowledgements
The Swedish Foundation for Strategic Environmental Research
(
(
Mistra, project No. 88032) and Natural Science Foundation of China
No. 50425415) are acknowledged for the financial support of this
Fig. 5. Treatment of Cu2S/FeS: (a) XRD patterns of reduced Cu2S/FeS pellet at 2.8 V
in molten CaCl2–NaCl at 800 C, and (b) SEM image of a Cu-rich particle in a partially
reduced Cu2S/FeS pellet.
work. The first author is grateful to the content of the analysis for
anode product from Li et al. [21].
◦

4
402
X. Ge et al. / Electrochimica Acta 54 (2009) 4397–4402
[16] Y. Sakamura, M. Kurata, T. Inoue, J. Electrochem. Soc. 153 (2006) D31.
References
[17] A.M. Abdelkader, A. Daher, R.A. Abdelkareem, E. El-Kashif, Metall. Mater. Trans.
B 38B (2007) 35.
[
[
[
[
[
1] A.K. Biswas, W.G. Davenport, Extractive Metallurgy of Copper, Pergamon, 2002.
2] G.Z. Chen, D.J. Fray, T.W. Farthing, Nature 407 (2000) 361.
3] D.J. Fray, G.Z. Chen, Mater. Sci. Technol. 20 (2004) 295.
4] D.H. Wang, X.B. Jin, G.Z. Chen, Annu. Rep. Prog. Chem. C 104 (2008) 189.
5] M. Ma, D.H. Wang, W.G. Wang, X.H. Hu, X.B. Jin, G.Z. Chen, J. Alloys Compd. 420
(
6] K. Pritish, W.E. James, Electrochim. Commun. 8 (2006) 1397.
7] C. Schwandt, D.J. Fray, Electrochim. Acta 51 (2005) 66.
8] K. Dring, R. Dashwood, D. Inman, J. Electrochem. Soc. 152 (2005) E104.
9] K. Dring, R. Dashwood, D. Inman, J. Electrochem. Soc. 152 (2005) D184.
[
18] K. Dring, R. Bhagat, M. Jackson, R. Dashwood, D. Inman, J. Alloys Compd. 419
(
2006) 103.
[
19] R. Bhagat, M. Jackson, D. Inman, R. Dashhood, J. Electrochem. Soc. 155 (2008)
E63.
[20] M. Kurata, T. Inoue, J. Serp, M. Ougier, J.P. Glatz, J. Nucl. Mater. 328 (2004) 97.
2006) 37.
[
21] G.M. Li, D.H. Wang, X.B. Jin, G.Z. Chen, Electrochem. Commun. 9 (2007) 1951.
[
[
[
[
[
[
[
22] G.Z. Chen, D.J. Fray, J. Appl. Electrochem. 31 (2001) 155.
23] FactSage 5.5. CRCT-ThermFact and GTT-Technologies, 2006.
24] C.W. Bale, P. Chartrand, S.A. Degterov, G. Eriksson, K. Hack, R. Ben Mahfoud, J.
Melan c¸ on, A.D. Pelton, S. Petersen, Calphad 26 (2002) 189.
[10] X.B. Jin, P. Gao, D.H. Wang, X.H. Hu, G.Z. Chen, Angew. Chem. Int. Ed. 43 (2004)
[
[
25] G.Z. Chen, E. Gordo, D.J. Fray, Metall. Mater. Trans. B 35B (2004) 223.
26] Y.H. Wang, C.H. Ye, X.S. Fang, L.D. Zhang, Chem. Lett. 33 (2004) 166.
7
33.
[11] D.T.L. Alexander, C. Schwandt, D.J. Fray, Acta Mater. 54 (2006) 2933.
[
27] D.C. Yang, G.W. Meng, S.Y. Zhang, Y.F. Hao, X.H. An, Q. Wei, Y. Min, L.D. Zhang,
Chem. Commun. (2007) 1733.
[
[
12] T. Nohira, K. Yasuda, Y. Ito, Nat. Mater. 2 (2003) 397.
13] X.Y. Yan, D.J. Fray, Adv. Funct. Mater. 15 (2005) 1757.
[
[
28] D.M. Davis, E.J. Podlaha, Electrochem. Solid State Lett. 8 (2005) D1.
[
[
14] K. Yasuda, T. Nohira, Y. Ito, J. Phys. Chem. Solids 66 (2005) 443.
15] K. Yasuda, T. Nohira, R. Hagiwara, Y.H. Ogata, Electrochim. Acta 53 (2007) 106.
29] X.F. Wu, H. Bai, J.X. Zhang, F.E. Chen, G.Q. Shi, J. Phys. Chem. B 109 (2005) 22836.