Electrowinning of iron in aqueous alkaline solution using a rotating cathode

Article abstract of DOI:10.1149/1.3039998

Electrowinning of iron by electrodecomposition of iron oxide ore in aqueous alkaline electrolytes is an alternative method to reduce C O2 emissions in steelmaking. Laboratory experiments were carried out in a suspension of hematite (Fe2 O3) in concentrate

Full text of DOI:10.1149/1.3039998

D64
Journal of The Electrochemical Society, 156 ͑2͒ D64-D69 ͑2009͒
0
013-4651/2008/156͑2͒/D64/6/$23.00 © The Electrochemical Society
Electrowinning of Iron in Aqueous Alkaline Solution
Using a Rotating Cathode
a, ,z
b
a,
Boyan Yuan, * Ole Edvard Kongstein, and Geir Martin Haarberg **
a
Department of Materials Technology, Norwegian University of Science and Technology,
7
491 Trondheim, Norway
b
SINTEF, Department of Applied Mechanics and Corrosion, Division of Materials and Chemistry,
7034 Trondheim, Norway
Electrowinning of iron by electrodecomposition of iron oxide ore in aqueous alkaline electrolytes is an alternative method to
reduce CO emissions in steelmaking. Laboratory experiments were carried out in a suspension of hematite ͑Fe O ͒ in concen-
2
2
3
trated NaOH at 114°C. Fe O particles were reduced to metallic iron upon contact with the rotating cathode, while oxygen gas was
2
3
evolved on the nickel mesh anode. Iron deposits generally adhered well to the carbon cathode substrate. The current efficiency for
iron was consistently above 90%, hydrogen evolution on the cathode being the main reason for the inefficiency. The influence of
cathode rotation rate, cathodic current density, content of Fe O particles, and concentration of NaOH on the current efficiency and
2
3
the morphology of the deposit was studied.
2008 The Electrochemical Society. ͓DOI: 10.1149/1.3039998͔ All rights reserved.
©
Manuscript submitted June 20, 2008; revised manuscript received November 5, 2008. Published December 10, 2008.
9
-11
Recent studies on global warming and climate change have con-
with alkaline or alkali carbonates at a temperature above 100°C.
cluded that substantial reductions in greenhouse gas ͓notably carbon
In order to develop a new, efficient industrial process for producing
dioxide ͑CO ͔͒ emissions are necessary and urgent. In order to
iron and steel with significantly reduced CO emissions, we inves-
tigated the electrowinning of iron from an iron oxide ore in aqueous
alkaline solutions by using a rotating disk electrode ͑RDE͒ ͑cathode͒
2
2
achieve the decided target on the reductions, enormous technologi-
cal innovations in, especially, electric power generation and heavy
1,2
in a laboratory cell. As shown in Fig. 1e, hematite ͑Fe O ͒ solid
industries, including steelmaking, are required. As illustrated in
Fig. 1a, currently iron is produced almost exclusively by reducing
iron ores by coke in a blast furnace at a temperature of around
2
3
particles suspended in concentrated aqueous sodium hydroxide
NaOH͒ solutions were transferred to the cathode ͑a rotating disk
͑
graphite electrode͒ and reduced to iron metal by applying a constant
current; oxygen was evolved on an inert anode ͑a nickel mesh͒. In
order to evaluate the feasibility of this process to be industrialized,
we determined the current efficiency and the corresponding energy
consumption of the process for producing iron. Moreover, we stud-
ied influences of cell operation parameters on the current efficiency
and the morphology of deposits and determined the optimum cell
operation conditions under which compact iron deposits can be pro-
duced with high current efficiencies and productivities.
2
000°C. This carbothermic reduction process directly produces liq-
uid metal with a high efficiency but generates a large amount of
exhaust gases containing CO . The production of iron and steel
contributes to about 8% of the global emissions of CO . This indus-
try is expected to continue growing in the near future, especially due
to developments in China and India. Alternative ways of producing
iron are studied in a large project ͓Ultra Low Carbon Dioxide in
Steelmaking ͑ULCOS͔͒ supported by the European Commission and
the steel industry in Europe. Electrolysis to decompose iron ore
2
2
͑
Fe O ͒ using an inert oxygen-evolving anode is an attractive alter-
2 3
Experimental
native in terms of reducing CO emissions. However, the current
2
blast furnace reduction process is highly efficient so an electro-
chemical process must be developed to be competitive.
Figure 2a schematically illustrates the experimental setup and 2b
the sketch of the rotating disk graphite electrode ͑cathode͒ used in
this study. A nickel mesh ͑anode; 100 ϫ 100 mesh, 0.1 mm line
thickness, 99.9 wt % purity; 310 mm length, 50 mm width͒ was at-
tached to the wall of a Teflon container ͑100 mm diam, 160 mm
length͒ before adding chemicals. The electrolyte was typically a
Figures 1b-d illustrate the electrolytic processes that were histori-
cally employed for producing iron at an industrial scale. As shown
in Fig. 1b, iron was electrorefined from a pig or cast iron anode in an
3-5
aqueous FeSO –FeCl solution. The highly pure and uniform iron
4
2
−
1
5
0 wt % ͑25 mol kg ͒ NaOH aqueous solution containing 33 wt %
produced was used directly as boiler tubes and in magnet cores. The
disproportionation of ferrous chloride ͑FeCl ͒ by electrolysis using
Fe2O3. The density of the 50 wt % NaOH solution at 114°C was
measured to be 1.43 ϫ 10 kg m ; this data was reasonable as
compared to the literature value of 1.48882 ϫ 10 kg m
2
3
−3
an insoluble anode and a diaphragm cell produced iron in the cath-
3
−3
at
ode chamber and ferric chloride ͑FeCl ͒ in the anode chamber ͑Fig.
3
1
2
7
0°C. The molarity of the NaOH solution calculated using the
1
c͒. The regenerated FeCl was used to treat iron sulfide ores to
3
−
1
6
,7
measured density value was 18 mol L ͑M͒. The highly concen-
produce sulfur and FeCl2. The Pyror process shown in Fig. 1d was
based on the electrowinning of iron from an aqueous FeSO –H SO
trated NaOH solution was used mainly in order to suppress the H2
4
2
4
evolution reaction at the cathode. The H evolution reaction caused
solution, which was generated by treating iron sulfide ore by using
sulfuric acid. Iron was deposited at an iron sheet cathode and oxy-
2
reduced current efficiency for the iron deposition and also contrib-
8
uted to increased porosity of the product. The content of Fe O was
gen was generated at a lead anode in a diaphragm cell.
2
3
defined as the weight of Fe O divided by the total weight of Fe O ,
The above-mentioned electrolytic processes were efficient and
low cost with regard to the highly pure and uniform iron produced
for special purposes or for the treatment of iron sulfide ores. How-
ever, these processes cannot be directly employed to produce iron
2
3
2
3
NaOH, and H O. The particle size distribution and specific surface
2
area of the Fe O powder were analyzed by the laser light diffrac-
2
3
tion method ͑Mastersizer 2000, Malvern Instruments Ltd., United
9
Kingdom͒, and they were d10 = 0.2 ␮m, d ͑median diameter͒
50
from iron oxide ores. Estelle originally developed the electrowin-
2
−1
=
0.5 ␮m, and d90 = 1.5 ␮m, and 15 m g , respectively. d , d ,
ning of iron from iron oxide or hydroxide in a slime containing a
caustic alkaline solution. The iron oxide or hydroxide was pretreated
10 50
and d90 are the equivalent particle diameters where the cumulative
distribution percentages reach 10, 50, and 90%, respectively; there-
fore, 10, 50, and 90% of the Fe O particles were smaller than 0.2,
2
3
0
9
.5, and 1.5 ␮m, respectively. NaOH ͑VWR International, Ltd.,
9%, 716.3 g, tablets͒ was dissolved in distilled water ͑716.3 g͒
*
Electrochemical Society Student Member.
*
* Electrochemical Society Active Member.
z
E-mail: boyan.yuan@material.ntnu.no
gradually, and then the Fe O powder ͑Alfa Aesar Gmbh & Co KG,
2 3
Downloaded on 2014-07-01 to IP 134.117.10.200 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, 156 ͑2͒ D64-D69 ͑2009͒
D65
Figure 1. Schematic illustration of vari-
ous iron production processes: ͑a͒ carbo-
thermic reduction ͑current industrial pro-
cess͒; ͑b͒ electrorefinning of cast or pig
iron in FeSO –FeCl solution ͑Boucher
4
2
3
-5
process, commercialized͒;
winning in FeCl2 solution using iron
sulfide ore ͑Eustis process,
͑c͒ electro-
6
,7
commercialized͒; ͑d͒ electrowinning in
FeSO solution using iron sulfide ore ͑Py-
4
8
ror process, commercialized͒; and ͑e͒
electrowinning from iron oxide particles
suspended in aqueous alkaline solution
͑
this study͒.
9
9.5%, 716.3 g͒ was added to the NaOH solution. The cathode disk
The produced iron deposits were weighed. The deposits were
characterized by using an optical microscope ͑Nikon SMZ800, Ja-
pan͒ and a scanning electron microscope ͑JEOL, Ltd., JSM–5600LV,
Japan͒. The cross sections of the deposits without polishing were
prepared by splitting them manually, and those after polishing were
made by wet grinding the cross sections of the splits embedded in
epoxy using various grades of silicon carbide paper. The cross sec-
tions were observed by scanning electron microscopy ͑SEM͒. The
purity of the iron deposits at the cross section after polishing was
analyzed by using SEM energy-dispersive X-ray spectroscopy
was fabricated by embedding a graphite rod ͑8 mm diam͒ in a plas-
tic ͑polyetheretherketone͒ capsule ͑30 mm diam, 50 mm length͒. It
was mounted to the shaft of a modulated speed rotator ͑AFMSRCE,
Pine Research Instrumentation, USA͒ through the threads made in
the capsule and the graphite ͑around 5 mm͒ in order to maintain an
electrical contact between the shaft and the graphite electrode. The
reference electrode was a commercial Hg/HgO electrode ͑XR400,
Radiometer Analytical S.A., France͒ which was immersed in an
aqueous potassium hydroxide ͑KOH͒ solution ͑1 M͒ in a Teflon
container. A Teflon tube filled with an organic fiber was immersed in
the KOH solution and the electrolyte; the fiber was wetted in the
KOH solution beforehand. A mechanical Teflon stirrer was rotated at
a speed of 500 rpm in all experiments in order to keep the Fe O
particles from settling at the bottom of the container. A lid with five
inlets was equipped with the mechanical stirrer, the graphite disk
electrode, the lead to the nickel mesh, a thermometer, and the Teflon
tube and was attached to the electrolyte container. The cell was then
placed in a thermal circulator ͑Julabo, Germany͒ filled with oil. The
temperature of the electrolyte was maintained at 114°C in all ex-
periments. A constant current was supplied by an electrochemical
workstation ͑Zahner IM6, Germany͒ to perform the electrowinning
experiments, and the cell voltage was monitored using a multimeter.
The deposited iron was peeled off from the graphite substrate and
washed in distilled water by using an ultrasonic cleaner, and dried in
a fume hood or a vacuum oven.
͑
EDS͒.
Results and Discussion
Influence of cell operation parameters on the current efficiency
2
3
of iron deposition.— Influence of rotation rate on the current effi-
ciency.— The current efficiency for iron deposition was calculated
according to Faraday’s law using the weight of the obtained iron.
Figure 3a shows the current efficiency in the 50 wt % NaOH solu-
−
2
tion containing 33 wt % Fe2O3 at a current density of 1000 A m
and different rotation rates from 0 to 3000 rpm of a cathode. At a
rotation rate of 0 rpm ͑stationary disk electrode͒, in most of the
experiments, powdery deposits were produced and fell down into
the electrolyte and the current efficiency was very low, around 10%.
However, in some cases smooth deposits were obtained at current
efficiencies ranging from 86 to 94%. The experiments at 0 rpm ro-
Downloaded on 2014-07-01 to IP 134.117.10.200 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D66
Journal of The Electrochemical Society, 156 ͑2͒ D64-D69 ͑2009͒
Figure 3. Variation of current efficiency with cell operation parameters: ͑a͒
rotation rate of cathode, ͑b͒ current density, ͑c͒ content of Fe O in electro-
2
3
lyte, and ͑d͒ concentration of NaOH solution.
Figure 2. ͑a͒ Schematic illustration of experimental setup for electrowinning
of iron from iron oxide particles suspended in aqueous alkaline solution. ͑b͒
Mechanical sketch of rotating disk graphite electrode.
−
2
6
6
00 to 1000 A m , the current efficiency increased sharply, from
2 to 90%. Current efficiency above 88% was obtained at current
densities from 1000 to 4000 A m . At current densities above
000 A m , it was observed that the obtained deposits were fairly
tation rate of the cathode ͑stationary disk electrode͒ were conducted
multiple times ͑Ͼ10͒. These experiments gave current efficiencies
for iron deposition ranging from 10 to 94%. The poor reproducibil-
−2
−2
4
rough with big edges and included some oxide particles, and the
calculated current efficiency slightly decreased. The current effi-
ciency at 1500 rpm was from 5 to 10% lower than at 1000 rpm. This
was possibly due to the high negative cathode potentials ͑more
negative than −4 V vs the Hg/HgO electrode͒ caused by gas
bubbles occurring frequently in the experiments at 1500 rpm. This
suggests that the development of techniques for controlling the gas
bubbles in the electrolyte is important in order to achieve high cur-
rent efficiencies in this process. These results indicate that the high
ity is likely to be due to the random transfer of Fe O particles to the
2
3
cathode surface. The particles primarily tend to settle at the bottom
of the electrolyte when using a stationary cathode. Repeated use of
the same electrolyte was found to change the homogeneity of the
suspension, which may also affect the current efficiency. For rota-
tion rates below 1000 rpm the current efficiency was found to vary
from 60 to 95% with low reproducibility. At rotation rates from
1
200 to 2000 rpm the current efficiency was above 90% in repeated
experiments. When the rotation rates were further increased to
above 2200 rpm, it was observed that in some of the experiments air
bubbles might be dragged into the electrolyte. Together with the
hydrogen gas evolved, these gas bubbles might partly or fully cover
the cathode surface, and caused high cathode potentials ͑more nega-
tive than −4 V vs the Hg/HgO electrode͒. The current efficiency
was found to decrease at high rotation rates, and at 3000 rpm it was
−2
cathodic current density of 4000 A m was applicable in the elec-
trowinning cell. This resulted in the high productivity of the process,
estimated to be 22 tons of iron per square meter of cathode area per
year under the assumption of continuous operation at a constant
current efficiency of 92%.
Influence of Fe O content on the current efficiency.— Figure 3c
2
3
shows current efficiency of the iron deposition at a rotation rate of
76%. The RDE was positioned about 40 mm ͑length of plastic cap-
−
2
1
000 rpm, a current density of 4000 A m , and different contents
sule 50 mm͒ down into the electrolyte. The dragging of air into
electrolyte might be due to the onset of turbulent flow for rotation
rates above 2200 rpm ͑although theoretically it has a much higher
value͒. Gas bubbles adhering to the cathode contributed to the pas-
sivation of the cathode surface. The problems arising from the pas-
sivation of the cathode due to accumulation of gas bubbles have not
been examined in detail. These problems are related to the use of a
rotating cathode which will not be used in a real industrial cell. The
cathode potentials and cell voltages were from −1.13 to − 1.19 V
vs the Hg/HgO electrode and from 1.4 to 1.9 V, respectively. The
corresponding energy consumption can be estimated to be about
from 5 to 50 wt % Fe O particles in a 50 wt % NaOH solution.
2
3
The current efficiency increased linearly with content of Fe O par-
2
3
ticles in the solution. When the content of Fe O increased from
2
3
5
to 10 wt %, the current efficiency increased from 77 to 86%.
Above 30 wt % the current efficiency was above 90%, and at
0 wt % Fe O , the current efficiency was 98%. These results indi-
5
2
3
cate that the content of the Fe O particles in the solution had a
2
3
significant influence on the current efficiency.
Influence of the concentration of NaOH on the current
efficiency.— Figure 3d shows the current efficiency of the iron
−
1
3
kWh kg iron. The optimum rotation rates under the above-
mentioned conditions were determined to be from
000 to 2000 rpm.
deposition at a rotation rate of 1000 rpm and a current density of
2
4
000 A m− and different concentrations of NaOH from 17 ͑6 M͒
1
to 71 ͑26 M͒ wt % containing 33 wt % Fe O . The pH of the solu-
2
3
Influence of current density on the current efficiency.— Figure 3b
tion and the solubility of the Fe O particles in the electrolyte were
2 3
shows current efficiency of the iron deposition in a 50 wt % NaOH
varied with the concentration of the NaOH solution. The current
efficiency increased from 60 to 80 wt % when the concentration of
NaOH was increased from 17–25 wt %; high current efficiencies
were obtained at concentrations above 40 wt %.
solution containing 33 wt % Fe O3 at rotation rates of 1000 and
2
1
6
500 rpm and different cathodic current densities from
00 to 18,000 A m . When the current density was increased from
−
2
Downloaded on 2014-07-01 to IP 134.117.10.200 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, 156 ͑2͒ D64-D69 ͑2009͒
D67
Figure 4. ͑Color online͒ Optical microscopic images of surface of iron de-
posits obtained at different rotation rates and current densities: ͑a, b, c͒
bottom of deposit, ͑aЈ,bЈ,cЈ͒ top of deposit; and ͑aЉ,bЉ,cЉ͒ magnified image
of bottom of deposit.
Influence of cell operation parameters on the morphology of iron
deposits.— Morphology of iron deposits.— Compact deposits are
highly desired because porous deposits are more likely to be con-
taminated by inclusions of impurities. Additionally, porous deposits
cause difficulties in achieving narrower electrode spacing, which is
important in reducing cell voltage and therefore energy consump-
tion. Figure 4 shows typical optical microscopic images of the sur-
faces of the iron deposits obtained at different rotation rates and
current densities. The iron deposited on the RDE developed a ring-
shaped edge of around 1–2 mm diam, and spirals were formed on
the top surface of the deposits. The smooth and dense deposit was
obtained at a rotation rate of 1500 rpm and a current density of
−
2
1
000 A m . When the rotation rate was increased to 3000 rpm, the
deposit had an asymmetrical shape with partly dendritic growth and
occlusions of Fe O particles. It was possible that the reduced metal
2
3
grains were thrown away from the electrode; this might contribute to
the lower current efficiency obtained at high rotation rates above
2
200 rpm. The images of the iron deposited at 1000 and
−
2
3000 A m showed that the surfaces of the deposits obtained at the
higher current densities were rougher and more porous, and bigger,
discontinuous edges were developed.
Figure 5 shows SEM images of the iron deposit at ͑a͒ the sur-
face, ͑b͒ the cross section of the center, and ͑c͒ the cross section of
the edge. The deposited iron crystal was composed of six twin crys-
tals; the twin crystals were tetrahedron-shaped and converged at an
apex. At the center, the deposit tended to form crystals growing from
and perpendicular to the graphite surface by an oriented layer-by-
layer growth in a shape of a tetrahedron. At the edge, the deposit
tended to form smaller crystals growing in radial directions with
various angles, containing large pores between the crystals. Close to
the electrode surface, smaller grains were formed; however, bigger
crystals were formed with the oriented crystal growth.
Figure 5. SEM images of iron deposit at ͑a͒ surface, ͑b͒ cross section of
center, and ͑c͒ cross section of edge.
by-layer growth was facilitated by high rotation rates of the cathode.
The corresponding SEM images of the polished cross sections
showed that the iron deposits produced at higher rotation rates were
more compact than those produced at lower rotation rates. More
pores were observed in the deposits close to the surface of the elec-
trode.
Influence of cell operation parameters on morphology of iron
deposit.— Figures 6a-c and 6aЈ–cЈ show SEM images of the cross
sections of iron deposits obtained in a 50 wt % NaOH solution con-
Figures 7a-c and 7aЈ–cЈ show SEM images of the iron deposits
obtained in a 50 wt % NaOH solution containing 33 wt % Fe O at
2
3
−
2
taining 33 wt % Fe O at a current density of 1000 A m and dif-
a rotation rate of 1000 rpm and different cathodic current densities
from 1000 to 18,000 A m− without and after polishing, respec-
tively. The shape of the deposits changed from a columnar structure
2
3
2
ferent rotation rates from 0 to 3000 rpm of the cathode without and
after polishing, respectively. It was observed that the compact layer-
Downloaded on 2014-07-01 to IP 134.117.10.200 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D68
Journal of The Electrochemical Society, 156 ͑2͒ D64-D69 ͑2009͒
Figure 7. SEM images of cross sections of iron deposits obtained at different
current densities without and after polishing.
Figure 6. SEM images of cross sections of iron deposits obtained at different
rotation rates without and after polishing.
−
2
−2
͑
͑
1000 A m ͒, to islands ͑4000 A m ͒, dendritic and powdery
−
2
18,000 A m ͒. The porosity of the deposits was found to be
higher at increasing current density in the range from
000 to 18000 A m .
−
2
1
Figures 8a-c and 8aЈ–cЈ show SEM images of iron deposits ob-
tained in a 50 wt % NaOH solution at a rotation rate of 1000 and a
current density of 4000 A m−2 and different contents from
5
to 50 wt % of the Fe O3 particles without and after polishing,
2
respectively. When only 5 wt % of Fe O was added in the NaOH
2
3
solution, the iron deposit was powdery and contained individual
grains. At 50 wt %, compact individual crystals were formed at the
edge of the cathode. The corresponding SEM images of the polished
cross sections showed that the porosity of the deposits was less as
the Fe O content increased. The optimum content of Fe O for
2
3
2
3
producing a smooth and compact deposit was around 33 wt %. The
corresponding SEM images of the polished cross sections showed
that the porosity of the deposits decreased with the content of the
Fe O particles in the electrolyte.
2
3
Figures 9a-c and 9aЈ–cЈ show SEM images of iron deposits ob-
tained at a rotation rate of 1000 rpm and a current density of
−
2
4
000 A m , and different concentrations of the NaOH solutions
from 17 to 71 wt % containing 33 wt % Fe O3 without and after
2
polishing, respectively. At higher NaOH concentration the deposits
were found to be more compact, and the crystal size increased and
the porosity was less as the NaOH concentration increased. The H2
evolution reaction which contributed to increase the porosity of the
deposits was hindered in the highly concentrated NaOH solution.
The solubility of Fe O in NaOH solution increases with tem-
2
3
perature and NaOH concentration but remains fairly low. The re-
ported solubility data in the 18 M NaOH solution at 90 and 140°C
were around 3 and 8 ϫ 10−3 M, respectively.
13,14
The dissolved iron
−
13-15
species is believed to be Fe͑OH͒ .
The hydrogen evolution re-
action ͑2 H O + 2e = H + 2OH ͒, which occurred at a potential
4
Figure 8. SEM images of cross sections of iron deposits obtained at different
−
−
contents of Fe O without and after polishing.
2
2
2
3
Downloaded on 2014-07-01 to IP 134.117.10.200 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, 156 ͑2͒ D64-D69 ͑2009͒
D69
Table I. Analytical results of iron produced by electrowinning in
alkaline solution.
Concentration of
a
Rotation Current
rate, efficiency, Site of
␻ ͑rpm͒ ␧ ͑%͒ interest Fe
element i, C ͑wt %͒
i
Exp.
no.
O
Na Cu Cr Si
b
D
I
1000
1500
2000
2200
2500
90.7
92.1
91.8
86.8
82.1
Edge 99.2 0.8 n.d. n.d. n.d. n.d.
Center 99.2 0.8 n.d. n.d. n.d. n.d.
Edge 99.1 0.9 n.d. n.d. n.d. n.d.
Center 97.0 1.5 1.3 n.d. n.d. 0.2
Edge 98.7 0.7 n.d. n.d. 0.6 n.d.
Center 98.3 0.7 n.d. n.d. 1.0 n.d.
Edge 98.3 1.2 0.5 n.d. n.d. n.d.
Center 98.8 0.7 0.3 n.d. 0.2 n.d.
Edge 99.0 1.0 n.d. n.d. n.d. n.d.
Center 98.8 1.2 n.d. n.d. n.d. n.d.
L
N
O
a
Analyzed by SEM-EDS.
Not detected.
b
Conclusions
Electrowinning of iron by electrodeoxidation of solid Fe O in
2
3
aqueous alkaline electrolytes is an alternative to reduce CO emis-
2
sions during iron and steel making. Electrolysis experiments using a
laboratory cell to reduce solid Fe O particles at a rotating graphite
2
3
Figure 9. SEM images of cross sections of iron deposits obtained at different
concentrations of NaOH solution without and after polishing.
cathode in concentrated NaOH at 114°C were carried out. The cur-
rent efficiency for iron formation was consistently found to be in the
range 90–95%. The purity of iron was above 98%, oxygen being the
main contaminant. The deposits generally adhered well to the cath-
ode but dentritic growth was pronounced at high rotation rates.
These results suggest that industrial iron electrowinning can be car-
of around −1.1 V vs the Hg/HgO electrode, was considered to be
the main side reaction that contributed to the loss of current effi-
ciency for iron deposition.
−1
ried out at an energy consumption of about 3 kWh kg iron.
The mechanism of electrochemical reduction of iron oxides in
NaOH solutions was discussed according to thermodynamic
Acknowledgments
1
4
considerations but it remains unresolved. Electrodeposition of iron
The authors are grateful to the Ultra Low Carbon Dioxide in
Steelmaking Project ͑ULCOS͒ SP13 partners: Dr. Jean-Pierre Birat,
Dr. Herve Lavelaine, Dr. Antoine Allanore ͑Arcelormittal Maizieres
les Metz, France͒, Chris Treadgold ͑Corus, The Netherlands͒, José
Ignacio Barbero ͑Labein, Spain͒, Jorge Frade ͑University of Aveiro,
Portugal͒, and Dr. Sverre Rolseth ͑Foundation for Industrial and
Technical Research, Trondheim, Norway͒; Professor Jomar
Thonstad ͑Norwegian University of Science and Technology, Nor-
way͒, Dr. Ana Maria Martinez, and Karen S. Osen ͑SINTEF, Nor-
way͒ for their valuable input. Financial support provided by the
ULCOS ͑project id: 515960͒ program from the European Union is
greatly appreciated.
from dissolved species of Fe O in strong NaOH solution was
2
3
1
5
studied. An apparently similar process in molten salts to produce
metals by electrodeoxidation of solid metal oxides attached to the
16
cathode has been proposed. The electrochemical reduction is be-
lieved to occur in the solid phase so that oxygen-containing ions
diffuse through the partly reduced metal oxides and dissolve in the
electrolyte. The shape of reduced metal remaining at the cathode
resembles that of the starting material. In the aqueous alkaline elec-
trolyte the produced iron has a different morphology than the initial
hematite powder, which may suggest that the dissolution of Fe O
2
3
takes place before the formation of metallic iron. The kinetics of
electrochemical reactions are generally controlled by a charge-
transfer reaction, giving rise to activation overpotential. In some
cases the rate of the electrode reactions can be controlled by a dif-
fusion of the reactants or product. The current density will depend
on the convection, whereas for activation-controlled processes the
convection does not affect kinetics. In the present study the current
efficiency with respect to iron deposition is slightly dependent on
the rotation rate of the cathode. This can suggest that the rate of the
iron formation depends on the diffusion of dissolved iron-containing
complexes, but the supply of new particles to the reaction zone also
depends on the convection. More carefully designed experiments are
needed to determine the mechanism of iron deposition from Fe O
Norwegian University of Science and Technology assisted in meeting the
publication costs of this article.
References
1
. P. Friedlingstein, Nature (London), 451, 297 ͑2008͒.
2. R. Pielke, T. Wigley, and C. Green, Nature (London), 452, 531 ͑2008͒.
3. A. Boucher, U.S. Pat. 1,086,132 ͑1914͒.
4
5
6
. D. Belcher, Trans. Am. Electrochem. Soc., 45, 455 ͑1924͒.
. S. O. Cowper-Coles, U.S. Pat. 884,075 ͑1908͒.
. W. Palmaer and J. A. Brinell, Chem. Metall. Eng., 11, 197 ͑1913͒.
7. F. A. Eustis, Chem. Metall. Eng., 27, 684 ͑1922͒.
8. E. Mostad, S. Rolseth, and J. Thonstad, Hydrometallurgy, 90, 213 ͑2008͒.
9
. A. T. C. Estelle, U.S. Pat. 1,275,161 ͑1918͒.
0. S. J. Lloyd, Trans. Am. Electrochem. Soc., 55, 305 ͑1929͒.
1. J. A. M. LeDuc, R. E. Loftfield, and L. E. Vaaler, J. Electrochem. Soc., 106, 659
1959͒.
2
3
1
1
particles suspended in the NaOH solution.
͑
Purity of deposit produced by electrowinning in alkaline
solution.— Table I shows compositions of the produced iron depos-
its analyzed by SEM-EDS. The purities of the iron deposits were
above 98 wt %, and the deposits contained around 1 wt % of oxy-
gen. Inclusions of NaOH and the Fe O particles were not detected.
12. G. Akerlof and G. Kegeles, J. Am. Chem. Soc., 61, 1027 ͑1939͒.
13. G. Picard, D. Oster, and B. Tremillon, J. Chem. Res., Synop., 8, 252 ͑1980͒.
1
1
4. G. Picard, D. Oster, and B. Tremillon, J. Chem. Res., Synop., 8, 250 ͑1980͒.
5. A. Allanore, H. Lavelaine, G. Valentin, J. P. Birat, and F. Lapicquie, J. Electro-
chem. Soc., 154, E187 ͑2007͒.
16. G. Chen, D. J. Fray, and T. W. Farthing, Nature (London), 407, 361 ͑2000͒.
2
3
Downloaded on 2014-07-01 to IP 134.117.10.200 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).