Materials Transactions, Vol. 46, No. 2 (2005) pp. 309 to 316 #2005 The Japan Institute of Metals
Nanocrystalline Manganese-Molybdenum-Tungsten Oxide Anodes for Oxygen Evolution in Acidic Seawater Electrolysis Ahmed A. El-Moneim1 , Naokazu Kumagai2 , Katsuhiko Asami3 , Koji Hashimoto1; * 1
Tohoku Institute of Technology, Sendai 982-8588, Japan Daiki Engineering Co Ltd, Kashiwa, Chiba 227-8515, Japan 3 Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan 2
In an attempt to tailor new oxygen evolution anodes for acidic seawater electrolysis, manganese-molybdenum-tungsten oxide anodes were prepared by anodic deposition on IrO2 /Ti substrate using 0.2 kmol m3 MnSO4 -(0.00–0.006) kmol m3 Na2 MoO4 -(0.0–0.03) kmol m3 Na2 WO4 electrolytes at pH 0:1þ1:0 and 363 K at 600 Am2 . The deposits consist of a nanocrystalline single -MnO2 type phase in the form of triple Mn1xy Mox Wy O2þxþy oxide. Anodic deposition in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -0.006 kmol m3 W6þ electrolyte at pH 0.0 resulted in the formation of an electrode with 100% and 99.8% oxygen evolution efficiencies before and after electrolysis for 691.2 ks in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 . Thus, Mn1xy Mox Wy O2þxþy electrodes were promising oxygen evolving anodes for acidic seawater electrolysis. It has been concluded that additions of both tungsten and molybdenum beneficially bring about an increase in the real electro-catalytic activity along with the formation of deposits with optimum thickness and good adherence to the IrO2 /Ti substrate. (Received October 22, 2004; Accepted December 27, 2004) Keywords: nanocrystalline, oxygen evolution efficiency, anode potential
1.
Introduction
For prevention of global warming due to CO2 emissions and for supply of abundant energy we are proposing global carbon dioxide recycling.1–6) Key materials necessary for this process are the anode and cathode for seawater electrolysis and the catalyst for CO2 conversion into CH4 . We have been tailoring these key materials. Under normal conditions of seawater electrolysis, mass transfer limitations and reaction kinetics combine to form chlorine and its oxy compounds as the main anodic reaction products.7,8) However, toxic chlorine releases for huge amounts of fuel formation must be avoided. Therefore, the preparation of new anode materials producing only oxygen without chlorine in seawater electrolysis is one of the most important but difficult subjects for the success of global carbon dioxide recycling. Among various anode materials examined so far only unique Mn1x Mox O2þx and Mn1x Wx O2þx anodes prepared by anodic deposition preferentially evolved oxygen at very high efficiency and were stable for long-term electrolysis of alkaline and neutral seawater using undivided electrolytic cell.9–13) In this regard, it has been found that the addition of molybdenum was more effective than tungsten in improving the oxygen evolution efficiency at the initial stage, but tungsten-containing anodes were more active for oxygen evolution reaction during electrolysis.13–15) Therefore, it seems interesting to examine the effect of simultaneous additions of molybdenum and tungsten on the oxygen evolution efficiency and durability of manganesebased anodes during seawater electrolysis. On the other hand, our recent results on membrane electrolysis (using two compartment laboratory cell) of alkaline and neutral 0.5 kmol m3 NaCl solution at 1000 Am2 have shown the pH decrease of the anode compartment to about 1–2 due to preferential migration of sodium ions, in *Corresponding
author, E-mail:
[email protected]
addition to hydrogen ions through membrane. In such acidic anolytes, the equilibrium potential for oxygen evolution became higher but that for chlorine evolution is independent of pH, and hence chlorine evolution is practically favored. In other words, oxygen evolution reaction became more difficult on acidic solution than in alkaline and neutral solutions. Therefore, there is a necessity to examine and optimize the behavior of manganese-based oxide anodes during acidic seawater electrolysis. In the present work, an attempt was undertaken to develop new electrodeposited manganese-molybdenum-tungsten oxide anodes with high activity and durability for oxygen evolution during acidic seawater electrolysis. Particular attention was given to the effect of deposition conditions, particularly bath composition and pH, on the anode performance. 2.
Experimental
2.1 Electrodes Punched titanium substrate mesh of 100 50 1 mm in dimension was polished in 0.5 kmol m3 HF solution for 0.3 ks and then subjected for surface roughening by etching in 11.5 kmol m3 H2 SO4 solution at 353 K until hydrogen evolution was ceased as discussed in details elsewhere.16,17) The etched substrate was coated with IrO2 as the intermediate layer using precursor of 0.1 kmol m3 chloroiridic acid in butanol solution. This layer is necessary to avoid the formation of insulating titanium oxide between electrocatalytically active manganese oxides and Ti substrate during seawater electrolysis at high current densities for a long time. The precursor mixture was applied using brush to both faces of Ti substrate; the solvent was evaporated at 353 K and the residue was calcinated in air at 723 K for 600 s in a preheated oven. This procedure was repeated three times so as to form about 10 gm2 IrO2 layer on the Ti substrate. The specimen was finally calcinated at 723 K for 3.6 ks. Before anodic
A. A. El-Moneim, N. Kumagai, K. Asami and K. Hashimoto
deposition, IrO2 /Ti electrodes of 6 mm 7 mm 1 mm in dimension and nominal surface area of about 1.68 cm2 were cleaned and activated by anodic polarization at 1000 Am2 for 0.6 ks in 10 kmol m3 NaOH and 1 kmol m3 H2 SO4 solutions, respectively. 2.2 Anodic deposition The basic electrolytes used for electro-deposition process were composed of 0.2 kmol m3 MnSO4 5H2 O, 0.001– 0.006 kmol m3 Na2 MoO4 2H2 O and 0.0–0.03 kmol m3 Na2 WO4 2H2 O. The pH of electrolytes was initially adjusted to 0:1þ1:0 by adding 18 kmol m3 H2 SO4 . A cell with separation of anode and cathode compartments was used for anodic deposition. An anode compartment was an alumina cylindrical diaphragm and a cathode was a pair of 316 stainless steel sheets set on the outside of the diaphragm in the cell. The electrodeposition was carried out for 3.6 ks at 600 Am2 and 363 K. The electrolyte was stirred at 450 rpm. The current efficiency of deposition was estimated from mass, composition and cationic valances of deposits. 2.3 Electrode characterization The composition and morphology of anodically deposited oxide electrodes thus prepared were characterized by electron probe micro analyzer (EPMA) Shimaduz EPMAC1. The structure of deposits was identified by X-ray diffraction of -2 mode (Rigaku RINT 2000/PC) using Cu-K at glancing angle of 10 . This angle is effective for the identification of thin deposits. The grain size of deposits was estimated from full width at half maximum of the most intense diffraction line using Sherrer’s equation.18) X-ray photoelectron spectroscopy (XPS) by means of a Shimaduz ESCA 850 photo electron spectrometer with Mg K (hv ¼ 1253:6 eV) excitation was employed for surface analysis of anodically deposited oxides. The binding energies of the photoelectrons were calibrated by a method described elsewhere.19,20) 2.4
Electrochemical characterization and electrode performance The performance of the electrode thus prepared was examined at 1000 Am2 in 0.5 kmol m3 NaCl solution of pH 2 stirred at 250 rpm. The oxygen evolution efficiency was estimated by the difference between the total charge passed during electrolysis, that is, 300 coulombs, and the charge of available chlorine formed during electrolysis. The term of available chlorine includes free chlorine and hypochlorate. The charge of available chlorine was measured by iodimetric titration, where liberated iodine by the addition of KI was titrated with 0.01 kmol m3 Na2 S2 O3 . To avoid reduction of chlorine and hypochlorate formed at the anode, the cathode used for the measurement of the oxygen evolution efficiency was a thin platinum wire of 0.2 mm in diameter. In order to determine the amount of available chlorine with a high accuracy, the ratio of the volume of the solution to the nominal anode surface area should be larger than 0.3 m3 m2 . The ratio used in the present investigation was about 1.8 m3 m2 . Polarization curves were measured in the galvanostatic
mode, from low to high current densities, with about ten measurements within each decade. Each measurement lasted for 0.06 ks. Correction for IR drop was made with a currentinterruption technique, where potential relaxation transients were sampled on the oscilloscope and the later mathematically processed to evaluate the IR-free potentials. During last seconds of polarization the current was repeatedly interrupted for 1:2 104 ks. All potentials were measured and reported relative to Ag/AgCl reference electrode with saturated KCl at room temperature. 3.
Results
3.1
Performance of binary oxide anodes in acidic seawater Figure 1 shows the change in the oxygen evolution efficiency of binary manganese-molybdenum and manganese-tungsten oxides deposited, respectively in electrolytes of pH 0.0 and 0.4 with the time of electrolysis in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 . It is clearly seen that the binary anodes deposited at pH 0.0 show higher activities for oxygen evolution and better durability than those deposited at pH 0.4. Regardless of the pH of deposition electrolyte, the initial oxygen evolution efficiency of the manganese-molybdenum oxide anodes is 100%, whereas that of manganese-tungsten oxide does not attain 100%. On the other hand, manganese-tungsten oxides exhibit better durability for oxygen evolution. This confirms the fact that the addition of molybdenum is effective in improving the initial oxygen evolution efficiency, but tungsten containing anode is more active for oxygen evolution reaction with longterm electrolysis.13–15)
3.2 Physicochemical properties of ternary oxide anodes XPS analysis of the deposits revealed that the peak binding
100
Oxygen Evolution Efficiency (%)
310
pH 0.0 96
pH 0.4 92
Mn-Mo oxide Mn-W oxide
88
pH 0.4 84
Electrolysis in 0.5 kmol m -2 NaCl, pH 2 at 1000 Am
0
200
400
-3
600
Time of Electrolysis, t / Ks Fig. 1 Change in the oxygen evolution efficiency of binary manganesemolybdenum and manganese-tungsten anodes deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ or 0.003 kmol m3 W6þ electrolytes at pH 0.0 and 0.4 with time of electrolysis.
Nanocrystalline Manganese-Molybdenum-Tungsten Oxide Anodes for Oxygen Evolution in Acidic Seawater Electrolysis
(a)
(b)
Mn1-x-yMoxWyO2+x+y -3
2+
-3
-3
deposited in 0.2 kmol m Mn - 0.003 kmol m Mo 6+ -2 -3 -x kmol m W at pH 0.0 and 600 Am
-3
2+
-3
6+
-x kmol m W at pH 0.0 and 600 Am
-3
0.01 kmol m
-3
0.006 kmol m
-3
0.003 kmol m
-3
0.001 kmol m
6+
0.00 kmol m W
40°
50°
60°
70°
80°
2 θ (Cu Kα)
γ -MnO
2
Ti
0.03 kmol m
-3
-3
0.01 kmol m
0.006 kmol m
-3
-3
0.003 kmol m
-3
6+
0.00 kmol m W
IrO 2/Ti
30°
6+
-2
IrO2
Intensity (arb. unit)
γ−MnO 2 IrO2 Ti
Intensity (arb. unit)
Mn1-x-yMoxWyO2+x+y
deposited in 0.2 kmol m Mn -0.006 kmol m Mo
6+
-3
311
IrO2/Ti
30°
40°
50°
60°
70°
80°
2 θ (Cu Kα)
Fig. 2 XRD patterns of Mn1xy Mox Wy O2þxþy electrodes deposited in (a) 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ and (b) 0.2 kmol m3 Mn2þ -0.006 kmol m3 Mo6þ -x kmol m3 W6þ electrolytes of pH 0.0. XRD pattern of IrO2 /Ti electrode is also presented for comparison.
energies of Mn 2p3=2 , Mo 3d5=2 and W 4f7=2 peaks were within the range of 641.2–641.95, 234.8–235.5 and 36.9– 37.8 eV, respectively. Furthermore, the multiplet splitting value for Mn 3s signal was about 4.6–5.0 eV.21,22) From these values, cations in the deposits were assigned to Mn4þ , Mo6þ and W6þ states, indicating the co-deposition of Mo6þ and W6þ during the oxidative deposition process of manganese. Therefore, hereafter, binary and ternary deposits will be referred as Mn1x Mox O2þx and Mn1xy Mox Wy O2þxþy . Figures 2(a) and (b) show X-ray diffraction (XRD) patterns for Mn1xy Mox Wy O2þxþy electrodes deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ and 0.2 kmol m3 Mn2þ -0.006 kmol m3 Mo6þ -x kmol m3 W6þ electrolytes of pH 0.0 at 600 Am2 , respectively. The XRD pattern of IrO2 /Ti substrate is also shown for comparison. All deposits consist of reflections corresponding to -MnO2 oxide. The reflections of Ti substrate are additionally detected for the electrodes deposited in 0.006 kmol m3 Mo6þ solutions (Fig. 2(b)) and 0.003 kmol m3 Mo6þ solutions with 0.0–0.001 kmol m3 W6þ (Fig. 2(a)). This indicates the insufficient coverage of the underlying IrO2 /Ti substrate with deposits. In other words, in 0.006 kmol m3 Mo6þ and 0.003 kmol m3 Mo6þ solutions with lower W6þ concentrations, the deposit may not cover completely the underlying IrO2 surface and may result in some chlorine evolution. It is also noteworthy that no extra reflections of separate oxides of molybdenum and tungsten are found, indicating that the deposits are single phase oxides. Figures 3(a–c) show the change in the cationic fractions of tungsten, molybdenum and manganese in electrodes deposited in 0.2 kmol m3 Mn2þ -(0.003 and 0.006) kmol m3 Mo6þ -x kmol m3 W6þ solutions of pH 0.0 as a function of W6þ concentration. As seen in Fig. 3(a), tungsten content in the deposits is very small and almost independent of Mo6þ
concentration in the deposition electrolytes. Meanwhile, increasing W6þ concentration leads to a decrease in the cationic fraction of molybdenum and an increase in the cationic fraction of manganese in the deposits as shown in Figs. 3(b) and (c), respectively. Furthermore, a higher content of manganese in the deposits is realized when deposition carried out in electrolyte containing 0.003 kmol m3 Mo6þ . Table 1 summarizes the change in composition, mean grain size, thickness and current efficiency of deposition with pH, and W6þ and Mo6þ concentrations in the deposition electrolyte. As can be seen, a decrease in pH and an increase in Mo6þ concentration in the deposition electrolyte decrease grain size of particles, current efficiency of manganese deposition and thickness of the deposits. On the contrary, increasing W6þ concentration leads to the formation a coarse-grained and thicker oxide. Figure 4 displays SEM images of the electrodes anodically deposited at different pH and W6þ concentrations. Increasing W6þ concentration and pH of deposition electrolytes lead to a decrease in the surface roughness of the deposited oxides. 3.3 Performance of ternary oxide anodes Figures 5(a) and (b) depict the change in the oxygen evolution efficiency at 1000 Am2 with the time of electrolysis for deposits prepared in 0.2 kmol m3 Mn2þ 0.003 kmol m3 Mo6þ -x kmol m3 W6þ and 0.2 kmol m3 Mn2þ -0.006 kmol m3 Mo6þ -x kmol m3 W6þ electrolytes at pH 0.0, respectively. All electrodes show the 100% initial oxygen evolution efficiency. However, the oxygen evolution efficiency gradually decreases during electrolysis. Comparison of these figures clearly exhibit that the electrodes prepared in the deposition electrolytes containing 0.003 kmol m3 Mo6þ have better durability than those prepared in
312
A. A. El-Moneim, N. Kumagai, K. Asami and K. Hashimoto 0.020
6+
(a)
-3
6+
-3
6+
0.003 kmol m Mo
Cationic Fraction of W
0.016
0.006 kmol m Mo
0.012
0.008
Mn1-x-y MoxWyO2+x+y -3
0.004
2+
deposited in 0.2 kmol m Mn - 0.003 -3 6+ -3 6+ and 0.006 kmol m Mo - x kmol m W -2 at pH 0.0, 363 K and 600 Am
0.000
(b)
Cationic Fraction of Mo
6+
0.12
0.08
-3
6+
-3
6+
0.006 kmol m Mo
0.04
0.003 kmol m Mo
0.00
(c) 0.96
-3
6+
-3
6+
Cationic Fraction of Mn
4+
0.003 kmol m Mo
0.92
0.006 kmol m Mo
0.88
0.000
0.005
0.010
0.015
0.020
6+
0.025
0.030
[W ] in Deposition Electrolyte, C / kmol m
-3
Fig. 3 Cationic fractions of tungsten (a), molybdenum (b) and manganese (c) in oxides anodically deposited for 3.6 ks in 0.2 kmol m3 Mn2þ -(0.003 and 0.006) kmol m3 Mo6þ -x kmol m3 W6þ electrolytes of pH 0.0 at 600 Am2 and 363 K.
the 0.006 kmol m3 Mo6þ solutions. Compared with binary Mn1x Mox O2þx oxide anodes, tungsten addition up to 0.006 kmol m3 W6þ to the 0.003 kmol m3 Mo6þ electrolyte and 0.01 kmol m3 W6þ to 0.006 kmol m3 Mo6þ electrolyte enhances the activity of the electrode for oxygen
evolution. Further tungsten addition leads to a shorter life of the electrode. Figure 6 summarizes the oxygen evolution efficiency after electrolysis at 1000 Am2 for 691.2 ks as a function of tungsten concentration in 0.2 kmol m3 Mn2þ -0.001–0.006 kmol m3 Mo6þ -x kmol m3 W6þ solutions of pH 0.0. In general, the oxygen evolution efficiency-[W6þ ] plots show maxima at 98.2, 99.8 and 99.45% oxygen evolution efficiencies at 0.003, 0.006 and 0.01 kmol m3 W6þ in 0.001, 0.003 and 0.006 kmol m3 Mo6þ solutions, respectively. Similarly, the oxygen evolution efficiency-[W6þ ] plots in Fig. 7 show a maximum at different W6þ concentrations in anodic deposition solutions of different pH. In general, all electrodes prepared in deposition electrolytes of pH 0.0 exhibit the best oxygen evolution efficiencies. In this manner, the improvement of the electro-catalytic activity by tungsten addition is not due to geometric factor but to synergistic effect of cations in the deposited oxides. Since the activity and durability are related to the electrode performance, the mass-loss of deposits after electrolysis at 1000 Am2 for 691.2 ks was examined by gravimetric measurement. The results are presented in Fig. 8. Surface observation after electrolysis revealed that the oxide layers were prone to detach from substrate when mass loss was detected. It can, therefore, be said that the partial exposure of IrO2 substrate to NaCl solution, which evolves actively chlorine, lowered the oxygen evolution efficiency. The mass loss of the electrodes increases with increasing pH and W6þ concentration in the deposition electrolytes. Since the increase in pH and W6þ concentration of the deposition electrolyte resulted in thickening of deposits as shown in Table 1, the mass loss seems due to the growing stress in the deposits leading to detachment of the oxide. By contrast, Mo6þ addition gives rises to the formation of thin and adherent oxide layers, particularly those deposited in electrolytes with 0.003 or 0.006 kmol m3 Mo6þ at pH 0.0. The addition of 0.006 kmol m3 Mo6þ , however, resulted in lower deposition efficiency to form very thin oxide layers with insufficient coverage of underlying IrO2 . It is, therefore, concluded that anodic deposition in 0.2 kmol m3 Mn2þ 0.003 kmol m3 Mo6þ -x kmol m3 W6þ solutions of pH 0.0 guaranteed stable electrodes with optimum thicknesses and excellent activity for oxygen evolution reaction. Interestingly, EPMA results, revealed no significant difference in the composition of these electrodes after electrolysis for
Table 1 Variation of deposit composition, mean grain size, thickness and efficiency of anodic deposition with pH of deposition, molybdenum and tungsten concentrations in deposition electrolytes. [Mo6þ ], kmol m3
0.003
[W6þ ], kmol m3
pH of deposition electrolyte
Composition, atomic fraction
Mean Grain size, nm
Thickness mm
Current efficiency of deposition (%)
0.006
0:1
Mn0:91 Mo0:065 W0:016 O2:08
7.50
13.5
5.30 6.70
0.006
0.0
Mn0:92 Mo0:068 W0:0079 O2:08
8.50
16.0
0.006
0.4
Mn0:93 Mo0:069 W0:0045 O2:07
9.10
20.7
0.006
1.0
Mn0:98 Mo0:062 W0:005 O2:019
9.70
91.0
41.0
33.8
14.6
0.003
0.030
0.0
Mn0:95 Mo0:03 W0:019 O2:046
0.006
0.006
0.0
Mn0:90 Mo0:09 W0:008 O2:098
12.0 8.00
9.40
9.88
3.60
Nanocrystalline Manganese-Molybdenum-Tungsten Oxide Anodes for Oxygen Evolution in Acidic Seawater Electrolysis
(a)
(b)
(c)
500 µm
(d)
500 µm
0.001 kmol m-3 W6+, pH 0.0
0.006 kmol m-3 W6+, pH 0.0
313
500 µm
500 µm
0.006 kmol m-3 W6+, pH 0.4
0.006 kmol m-3 W6+, pH 1.0
Fig. 4 SEM micrographs for electrodes deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -0.006 kmol m3 W6þ electrolytes of pH 0.0 (b), 0.4 (c) and 1.0 (c). SEM micrograph (a) for electrode deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -0.001 kmol m3 W6þ electrolyte of pH 0.0 is also presented for comparison.
-3
100
6+
0.006 kmol m W
-3
0.003 kmol m
0.01 0.001
99
(a)
0.0 kmol m
-3
98
Mn1-x-yMoxWyO2+x+y -3
97
-3
0.03 kmol m W
6+
2+
deposited in 0.2 kmol m Mn - 0.003 kmol m Mo - x kmol m W
-3
6+
-3
6+
363 K and 600 Am
-2
at pH 0.0,
-3
Electrolysis in 0.5 kmol m NaCl, pH 2.0 at 1000 A.m-2
96 0
200
400
600
800
Time of Electrolysis, t / Ks
Oxygen Evolution Efficiency (%)
Oxygen Evolution Efficiency (%)
100
-3
6+
0.01 kmol m W
(b) 0.006
99
0.003
98
Mn1-x-yMoxWyO2+x+y -3
-3
2+
deposited in 0.2 kmol m Mn - 0.006 -3
6+
-3
0.001 kmol m
6+
kmol m Mo - x kmol m W at pH 0.0,
97
-2
363 K, 600 A.m
-3
0.03 kmol m -3
Electrolysis in 0.5 kmol m NaCl, -2 pH 2.0 at 1000 A.m -3 6+ 0.0 kmol m W
96 0
200
400
600
800
Time of Electrolysis, t / ks
Fig. 5 Change in oxygen evolution efficiency of oxides anodically deposited in (a) 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ and (b) 0.2 kmol m3 Mn2þ -0.006 kmol m3 Mo6þ -x kmol m3 W6þ electrolytes at pH 0.0 and 600 Am2 with time of electrolysis in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 .
691.2 ks, indicating no preferential dissolution of oxide components. 3.4 Kinetics and electrode potential To investigate the influence of electrode composition on the kinetics, polarization curves were recorded. Figure 9 presents IR-corrected polarization curves of Mn0:908 Mo0:0907 O2:09 and Mn0:92 Mo0:068 W0:0079 O2:15 electrodes in 0.5 kmol m3 NaCl solution of pH 2. The IR-corrected polarization curve for deposited MnO2 electrode is given also for comparison. Polarization curves generally present two linear segments in both low and high overpotential regions. The increase in the Tafel slope in the high overpotential region is attributed to the change in kinetics due to change in the rate determining step.23,24) The MnO2 and Mn0:908 Mo0:0907 O2:09 electrodes exhibit similar Tafel slopes in both low and high overpotential domains, and the Mn0:908 Mo0:0907 O2:09 electrode shows lower anodic activity. The oxygen evolution efficiency of Mn1x Mox O2þx electrodes was generally much higher than that of MnO2 electrode, because more than 8% of the charge for the MnO2 electrode
was used for chlorine evolution.11–13) Thus, the chlorine evolution is responsible for the lower over potential of the MnO2 . On the other hand, Mn0:92 Mo0:068 W0:0079 O2:08 electrode interestingly shows the polarization curve with the lower Tafel slope and higher anodic activity than those for MnO2 and Mn0:908 Mo0:0907 O2:09 oxide electrodes. Figure 10 presents IR-corrected polarization curves of ternary oxides deposited at different pH. A decrease in pH leads to a decrease in the Tafel slope with a consequent increase in the anodic activity. This pH dependence is possibly related to the increase in tungsten content in the oxides as presented in Table 1. Figure 11 shows Tafel slope (b1 and b2 at low and high overpotential domains, respectively) of electrodes prepared in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ solutions of pH 0.0 as a function W6þ concentration. In this figure, the slopes are composition dependent. A minimum in Tafel slope is obtained at [W6þ ] ¼ 0:006 kmol m3 . Meanwhile, the slope higher than 120 mV/decade is observed for all electrodes, although the Tafel slope for oxygen evolution reported are 15–120 mV/decade depending
314
A. A. El-Moneim, N. Kumagai, K. Asami and K. Hashimoto 100
-3
0.010
6+
0.003 kmol m Mo
Oxygen Evolution Efficiency (%)
-3
99
-3
6+
-3
kmol m Mo - x kmol m W 363 K, 600 Am
-3
6+
at pH 0.0,
-2
6+
0.006 kmol m Mo
98
97 -3
6+
0.001 kmol m Mo
96 -3
Electrolysis in 0.5 kmol m NaCl -2 pH 2 at 1000 Am for 691.2 ks
95 0.00
0.01
Mn1-x-yMoXWyO2+x+y
-2
2+
deposited in 0.2 kmol m Mn - (0.001-0.006)
Mass Loss After Electrolysis at 1000 Am for 691.2 ks, gm
Mn1-x-yMoxWyO2+x+y
-3
0.008
-3
6+
-3
6+
0.003 kmol m Mo , pH 0.4
0.004
-3
6+
0.001 kmol m Mo , pH 0.0
0.002 -3
6+
0.003 kmol m Mo , pH 0.0 -3
6+
0.006 kmol m Mo , pH 0.0
0.01
0.02
0.03
6+
[W ] in Deposition Electrolyte, C / kmol m -3
Fig. 6 Change in oxygen evolution efficiency of electrodes after electrolysis in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 for 691.2 ks with tungsten concentration in deposition electrolytes of 0.2 kmol m3 Mn2þ (0.001–0.006) kmol m3 Mo6þ -x M W6þ at pH 0.0.
-3
Fig. 8 Change in mass-loss of electrodes after electrolysis in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 for 691.2 ks with tungsten concentration in deposition electrolytes of 0.2 kmol m3 Mn2þ -(0.001– 0.006) kmol m3 Mo6þ -x kmol m3 W6þ at different pH.
-3
0.006 kmol m W
100
6+
0.006
0.03
[ W ] in Deposition Electrolyte, C / Kmol m
-3
0.003 kmol m Mo , pH 1.0
0.000 0.00
0.02
2+
0.2 kmol m Mn - (0.001-0.006) kmol m 6+ -3 6+ -2 Mo - x kmol m W at 363 K and 600 Am
γ MnO2
6+
pH 0.0 1000 -2
pH 0.4
Current Density, I / Am
OXygen Evolution Efficieny (%)
98
96
pH 1.0
94
92
Mn1-x-yMoxWyO2+x+y 90
-3
-3
0.0 kmol m W
100
Mn 1-x-yMoxWyO2+x+y -3
2+
deposited in 0.2 kmol m Mn - 0.003 -3
6+
-3
kmol m Mo - x kmol m W
2+
deposited in 0.2 kmol m Mn - 0.003 -3 6+ -3 6+ kmol m Mo - 0.006kmol m W at -2 363 K and 600 Am
6+
363 K and 600 Am
10
6+
at pH 0.0,
-2
-3
0.5 kmol m NaCl at pH 2
88 -3
1.0
Electrolysis in 0.5 kmol m NaCl -2 pH2 at 1000 Am 86 0.00 6+
0.01
0.02
1.2
1.4
1.6
1.8
2.0
Potential, V (Ag/AgCl)
0.03 -3
[W ] in Deposition Electrolyte, C / kmol m
Fig. 7 Change in oxygen evolution efficiency of electrodes after electrolysis in 0.5 kmol m3 NaCl solution of pH 2 at 1000 Am2 for 691.2 ks with tungsten concentration in deposition electrolytes of 0.2 kmol m3 Mn2þ 0.003 kmol m3 Mo6þ -x kmol m3 W6þ at different pH.
upon the mechanism.23,24) For competitive oxygen evolution with chlorine evolution, the lower overpotential is better for oxygen evolution. Figure 12 shows the dependence of oxygen evolution overpotential (after ohmic drop correction) at 200 and 1000 Am2 as a function of tungsten concentration in the deposition electrolyte. The W6þ addition lowers the anodic potentials of electrodes and the best electrode for oxygen evolution reaction in both current density domains investigated is for
Fig. 9 IR-corrected polarization curves in 0.5 kmol m3 NaCl solution of pH 2 for electrodes deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ electrolytes of pH 0.0. Polarization curves of MnO2 electrode is also given for comparison.
that anodically deposited in the 0.006 kmol m3 W6þ electrolytes. It can, therefore, be said that the optimal addition of W6þ decreases the overpotential for oxygen evolution and hence elongates the life of electrode, in addition to saving energy. Figure 13 shows the variation of the anode potential (not IR corrected) at 1000 Am2 with time of electrolysis. The potential of the electrodes initially decreases and becomes steady after electrolysis for about 1.8 ks. Within this initial period electrolyte tends to penetrate inwardly through pores and cracks in the electrode surface, leading to an increase in the active sites for gas evolution. This means
Nanocrystalline Manganese-Molybdenum-Tungsten Oxide Anodes for Oxygen Evolution in Acidic Seawater Electrolysis
315
0.6 pH 0.0
Mn1-x-yMoxWyO2+x+y
pH -0.1
-3
Oxygen overpotential, V(Ag/AgCl)
-2
Current Density, I / Am
100
Mn 1-x-yMoxWyO2+x+y -3
2+
deposited in 0.2 kmol m Mn -0.003 -3
6+
-3
kmol m Mo - 0.006 kmol m W 363 K and 600 Am
10
6+
at ,
-2
-3
0.5 kmol m NaCl, pH 2
1.0
1.2
1.4
1.6
1.8
2+
deposited in 0.2 kmol m Mn - 0.003
pH 1.0
1000
0.5
-3
6+
-3
6+
kmol m Mo - x kmol m W at pH 0.0, -2
363 K and 600 Am
0.4
-2
1000 Am
0.3
0.2
-2
200 Am
-3
0.5 kmol m NaCl, pH 2
2.0
0.1 0.00
Potential, V (Ag/AgCl)
0.01
0.02
0.03
0.04
6+
Fig. 10 IR-corrected polarization curves in 0.5 kmol m3 NaCl solution of pH 2 for electrodes deposited in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -0.006 kmol m3 W6þ electrolytes of different pH.
-3
[W ] in Deposition Electrolyte, C / kmol m
Fig. 12 Dependence of the anode overpotential (IR corrected) at 200 and 1000 Am2 on tungsten concentration in deposition electrolytes of 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ at pH 0.0.
0.6
Mn1-x-yMoxWyO2+x+y -3
2.25
Mn1-x-y MoxWyO2+x+y
2+
deposited in 0.2 kmol m Mn - 0.003
0.5
-3
6+
-3
kmol m Mo - x kmol m W
6+
at pH 0.0,
-3
-3
Electrode Potential (Ag/AgCl)
b , V/decade
363 K and 600 Am
b2
0.4
0.3
b1 0.2 -3
0.5 kmol m NaCl, pH 2
6+
-3
kmol m Mo -x kmol m W
0.00 6+
0.01
0.02
0.03
2.15
-3
at pH 0.0,
-2
-2
Electrolysis at 1000 Am
-3
0.00 kmol m W
2.10
6+
-3
0.01 kmol m
2.05 0.001 kmol m
2.00
-3
-3
0.003 kmol m
-3
0.006 kmol m
0.04
[W ] in Deposition Electrolyte, C / kmol m
6+
363 K and 600 Am
1.95
0.1
2+
deposited in 0.2 kmol m Mn - 0.003
2.20
-2
0
2
4
6
8
10
12
Time of Electrolysis, t / ks Fig. 11 Dependence of the Tafel slope for the low, b1 , and high, b2 , overpotential domains on tungsten concentration in deposition electrolytes of 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -x kmol m3 W6þ at pH 0.0.
that the initial oxygen evolution efficiency of electrodes is due to oxidation of easily bounded water at the outer most surface of electrodes. 4.
Discussion
New Mn-Mo-W oxide anodes for oxygen evolution in acidic seawater electrolysis are prepared by anodic deposition in electrolytes containing 0.2 kmol m3 Mn2þ -(0.001– 006) kmol m3 Mo6þ -(0.0–0.03) kmol m3 W6þ of pH 0:1þ1:0 and 363 K at 600 Am2 for 3.6 ks. The deposits consist of a nanocrystalline single -MnO2 type phase in the
Fig. 13 shows the variation in the anode potential (not IR-corrected) at 1000 Am2 with time of electrolysis.
form of triple Mn1xy Mox Wy O2þxþy oxide. Deposition in electrolytes with intermediate concentration of W6þ and Mo6þ at pH 0.0 guaranteed stable electrodes with excellent oxygen evolution efficiency. It is clear that the addition of molybdenum and tungsten prevents chlorine evolution and enhances the oxygen evolution efficiency. On the other hand, the addition of tungsten into electrodeposited Mn-Mo oxide decreases the overpotential for electrolysis with enhancement of oxygen evolution. Thus, tungsten exerts its beneficial effect partly by lowering the over potential for oxygen evolution. Oxidative dissolution of MnO2 electrode as permanganate
316
A. A. El-Moneim, N. Kumagai, K. Asami and K. Hashimoto
is usually observed during seawater electrolysis.9–15) However, both binary and ternary oxide electrodes revealed no oxidative dissolution even after electrolysis at potentials far higher than equilibrium potential of the Mn4þ /Mn7þ redox couple. The formation of triple oxide Mn1xy Mox Wy O2þxþy seems responsible for enhancing both stability and oxygen evolution efficiency of electrodes. Although oxidative dissolution can be avoided by the formation of triple Mn1xy Mox Wy O2þxþy oxide, the oxygen evolution efficiency gradually decreases during electrolysis. This can be attributed to partial detachment of oxide layer from substrate with electrolysis. During electrolysis, the electrolyte will penetrate through pores and cracks into inside of the electrode surface. Because of increase in the effective surface area for oxygen evolution the apparent electrode potential decreases during electrolysis for initial 1.8 ks. Oxygen formation on the inner surfaces may responsible for the detachment of the oxide from the IrO2 substrate. During anodic deposition, oxygen evolution occurred simultaneously and became more vigorous with the decrease in the pH, mostly due to an increase in the molybdenum and/ or tungsten content of deposits. Under such vigorous oxygen evolution reaction oxides with poor adhesion to the substrate are likely to be detached during the deposition process and the oxides with better adhesion are able to stay on the electrode surface. Consequently, the oxides highly adherent to the substrate can be obtained at lower pH. The presence of adequate concentrations of Mo6þ and W6þ is a prerequisite to enhance not only the oxygen evolution but also adhesion of oxide layer to the substrate in seawater electrolysis, in addition to enhancement of oxygen evolution during anodic deposition. From these facts, it is clear that the adhesion and oxygen evolution efficiency of the deposited oxides are significantly improved by depositing at low efficiencies under vigorous oxygen evolution. 5.
Conclusions
New Mn1xy Mox Wy O2þxþy , oxide anodes for oxygen evolution in acidic seawater electrolysis were prepared by anodic deposition in electrolytes containing 0.2 kmol m3 Mn2þ -(0.001–006) kmol m3 Mo6þ -(0.0–0.03) kmol m3 W6þ at pH 0:1þ1:0 and 363 K at 600 Am2 for 3.6 ks. The following conclusions are drawn; (1) Mn1xy Mox Wy O2þxþy electrodes showed a 100% initial oxygen evolution efficiency, but the oxygen evolution efficiency gradually decreased during electrolysis at 1000 Am2 in 0.5 kmol m3 NaCl solution of pH 2. (2) Electrode prepared in 0.2 kmol m3 Mn2þ -0.003 kmol m3 Mo6þ -0.006 W6þ electrolyte of pH 0.0 and
363 K showed the best durability, that is 99.8% after electrolysis for 691.2 ks. (3) Tungsten addition was effective in lowering the overpotential for oxygen evolution reaction. This seemed effective for elongation of the life of electrode, in addition to saving energy. REFERENCES 1) K. Hashimoto: Mater. Sci. Eng. A179/A180 (1994) 27–30. 2) K. Hashimoto: Trans. Mater. Res. Soc. Jpn. 18A (1994) 35–40. 3) K. Hashimoto, E. Akiyama, H. Habazaki, A. Kawashima, K. Shimamura, M. Mori and N. Kumagai: Zairo-to-Kankyo (Corrosion Eng.) 45 (1996) 614–620. 4) K. Hashimoto, E. Akiyama, H. Habazaki, A. Kawashima, K. Shimamura, M. Mori and N. Kumagai: Sci. Rep. Res. Inst. Tohoku Univ. A43 (1997) 153–160. 5) K. Hashimoto, Y. Yamasaki, S. Meguro, S. Sasaki, H. Katagiri, K. Izumiya, N. Kumagai, H. Habazaki, E. Akiyama and K. Asami: Corros. Sci. 44 (2002) 371–383. 6) K. Hashimoto, Y. Yamasaki, K. Fujimora, T. Matsui, K. Izumiya, M. Mori, A. A. El-Moneim, E. Akiyama, H. Habazaki, N. Kumagai, A. Kawashima and K. Asami: Mater. Sci. Eng. A267 (1999) 200–206. 7) Ann Cornell, Bo Hakansson and Go¨ran Lindbergh: Electrochim. Acta 48 (2003) 473–481. 8) J. E. Bennett: Int. J. Hydrogen Energy 5 (1980) 401–408. 9) K. Izumiya, E. Akiyama, H. Habazaki, N. Kumagai, A. Kawashima, K. Asami and K. Hashimoto: Mater. Tran., JIM 39 (1998) 308–313. 10) K. Izumiya, E. Akiyama, H. Habazaki, N. Kumagai, A. Kawashima, K. Asami and K. Hashimoto: Electrochim. Acta 43 (1998) 3303–3312. 11) K. Fujimora, K. Izumiya, A. Kawashima, H. Habazaki, E. Akiyama, N. Kumagai, K. Asami and K. Hashimoto: J. Appl. Electrochem. 29 (1999) 765–771. 12) K. Fujimora, T. Matsui, K. Izumiya, N. Kumagai, E. Akiyama, H. Habazaki, A. Kawashima, K. Asami and K. Hashimoto: Mater. Sci. Eng. A267 (1999) 254–255. 13) K. Fujimora, T. Matsui, H. Habazaki, A. Kawashima, N. Kumagai and K. Hashimoto: Electrochim. Acta 45 (2000) 2297–2303. 14) H. Habazaki, T. Matsui, A. Kawashima, N. Kumagai, K. Asami and K. Hashimoto: Scr. Mater. 44 (2001) 1659–1662. 15) T. Matsui, H. Habazaki, A. Kawashima, N. Kumagai, K. Asami and K. Hashimoto: J. Appl. Electrochem. 32 (2002) 993–1000. 16) N. A. Abdel Ghany, N. Kumagai, S. Meguro, K. Asami and K. Hashimoto: Electrochim. Acta 48 (2002) 993–1000. 17) N. A. Abdel Ghany, N. Kumagai, S. Meguro, K. Asami and K. Hashimoto: Mater. Trans. 44 (2003) 2114–2123. 18) P. Scherrer: Gottingen Nachr. 2 (1918) 98–102. 19) K. Asami: J. Electron Spectrosc. 9 (1976) 469–478. 20) K. Asami and K. Hashimoto: Corros. Sci. 17 (1997) 559–570. 21) A. A. El-Moneim, B.-P. Zhang, E. Akiyama, H. Habazaki, A. Kawashima, K. Asami and K. Hashimoto: Corros. Sci. 39 (1997) 305–319. 22) A. A. El-Moneim, E. Akiyama, H. Habazaki, A. Kawashima, K. Asami and K. Hashimoto: Corros. Sci. 39 (1997) 1965–1979. 23) L. A. De Faria, J. F. C Boodts and S. Trasati: J. Appl. Electrochem. 26 (1996) 1195. 24) S. Ardizzone, G. Fregonara and S. Trasitti: Electrochim. Acta 35 (1990) 263.
