直接电脱氧制备铈镍储氢合金

发布于:2021-11-27 21:12:26

Journal of Alloys and Compounds 468 (2009) 379–385

Direct electrolytic preparation of cerium/nickel hydrogen storage alloy powder in molten salt
Bingjian Zhao a , Ling Wang a,? , Lei Dai a , Guanghua Cui a , Huizhu Zhou a , R.V. Kumar b
College of Chemical Engineering and Biotechnology, Hebei Polytechnic University, Tangshan 063009, PR China b Department of Materials Science and Metallurgy, University of Cambridge, Cambridge CB2 3QZ, UK Received 13 September 2007; received in revised form 2 January 2008; accepted 3 January 2008 Available online 8 May 2008
a

Abstract The preparation of CeNi5 by direct electrochemical reduction of mixed powders of cerium oxide (CeO2 ) and nickel oxide (NiO) (atomic Ce/Ni ratio, 1:5) in a molten mixture of CaCl2 and NaCl (mass CaCl2 /NaCl ratio, 7:3) at 820 ? C has been studied. The in?uence of process parameters, such as sintering temperature, cell voltage, and temperature of the molten salt, on the electrolysis process are reported. The current–time plots are shown at constant voltage electrolysis under different conditions. The composition and morphology of the products were analyzed by X-ray diffraction (XRD) and scanning electron microscopy (SEM), respectively. The results show that pure CeNi5 can be prepared by direct electrochemical reduction of mixed CeO2 /NiO pellets sintered at 850–1250 ? C in CaCl2 /NaCl melt at a voltage of 2.5 V or higher for 11 h. The product from the electrolysis was in the form of porous pellets and could be readily ground into CeNi5 alloy powder. At a voltage of 2.5 V or higher, the reduction of the mixed oxide starts from NiO to Ni, followed by that of CeO2 on the surface of the newly formed Ni to form CeNi5 alloy. The reduction rate increases with increasing cell voltage and temperature of molten salt. ? 2008 Elsevier B.V. All rights reserved.
Keywords: Electrochemical reduction; Molten salt; CeNi5 ; Powder

1. Introduction Nickel/metal hydride (Ni/MH) rechargeable batteries have been a popular option for power source. They have an excellent electrochemical capacity, a long life cycle and good over-charge/discharge characteristics, and are environmentally friendlier than many other competing power sources [1,2]. RENi5 (RE denotes the rare earth element) type of alloys have been widely adopted for the electrode materials of Ni/MH batteries because of their good overall properties, especially the reasonably high hydrogen storage capacity, the favorable electrochemical activation as well as improved charge and discharge kinetic properties [3–7]. However, due to the expensive manufacturing technologies, these materials are still far from being economical for large-scale applications. At present, RENi5

?

Corresponding author. Tel.: +86 315 2597148; fax: +86 315 2592170. E-mail address: tswling@126.com (L. Wang).

alloys are produced by separate extraction and re?ning the individual metals, followed by melting, alloying and casting under vacuum. The cast ingot needs to go through annealing under vacuum for several hours to eliminate compositional segregation. Afterwards, the ingot is made into a powder suitable for making the negative electrode in a battery [8]. Such multi-step processes are high in energy consumption but low in production ef?ciency, contributing to the relatively high costs of the commercial Ni/MH batteries that are mostly used in small electronic devices. A more cost-effective and novel production method will have a major impact on increased usage of the Ni/MH batteries. Since Chen et al. proposed to prepare metal Ti from TiO2 powder by molten salt electrolysis [9], the pure metals, alloys or intermetallic compounds have successfully been obtained by electrochemical reduction of pure or mixed oxide powders in molten salts, which promises a more economic technology for the metallurgical industry [10–14]. This electro-reduction method is lower in energy consumption and simpler in operation than many existing industrial technologies. With particular rele-

0925-8388/$ – see front matter ? 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2008.01.074

380

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385 perature was increased to and kept at the pre-set temperature. Electrolysis of the sandwich oxide cathode with one or two pellets proceeded at a constant voltage of 1.6–3.1 V for 1–11 h. The experimental data was collected by a PC computer aid system. After electrolysis, the sample was lifted from the molten salt and cooled naturally in a stream of argon before removal from the steel reactor, and then was immediately washed in distilled water. The composition and morphology of the products were analyzed by XRD and SEM, respectively. The experimental device is shown in Fig. 1.

vance to manufacturing alloys and intermetallic compounds, the compositions of the elements in the alloy or intermetallic compound can be controlled at high precision [15–18]. The reduction potential for direct formation of the alloy or the intermetallic compound is more positive than that for the formation of pure rare earth metal [15,16]. But the new process requires that (1) the metal oxides to be reduced are thermodynamically less stable than the oxide of the metallic element of the molten salt used and (2) electrodeposition of the metallic component of the molten salt should be avoided [19–21]. In this report, the electrolytic reduction of cerium oxide (CeO2 ) and nickel oxide (NiO) mixture into hydrogen storage alloy CeNi5 was investigated. The products from electrolysis under different conditions were analyzed by using X-ray diffraction (XRD) and scanning electron microscopy (SEM). The mechanism of the reduction reaction is also discussed.
2. Experimental 2.1. Materials and chemicals
Commercially available powders of CeO2 and NiO were used as received. Anhydrous CaCl2 , NaCl and high purity graphite (99.9%) were used for the electrolysis. Argon gas (99.99%) was used as the protective gas during high temperature experiments. All the reagents used in this paper were analytical grade reagents.

2.3. Cyclic voltammetry
Foils of molybdenum (width: 10 mm, thickness: 0.5 mm, length 20 mm, purity: 99.9%) were used as the substrate for making the metallic cavity electrode (MCE). One circular hole (1.0 mm diameter) was drilled through the foil by a mechanical drill. The oxide powders, such as CeO2 powder, NiO powder and mixed CeO2 and NiO powders (atomic Ce/Ni ratio, 1:5) were manually ?lled into the MCE cavity by repeatedly ?nger-pressing and used as the working electrode on which cyclic voltammograms (CVs) were recorded in molten mixture of CaCl2 and NaCl (mass CaCl2 /NaCl ratio, 7:3) at 820 ? C. High purity graphite (99.9%) was used as the counter electrode. The Kanthal wire was used as the pseudo-reference electrode whose stability was satisfactory on the time-scale of cyclic voltammetry. All electrochemical experiments were operated under the protection of an argon ?ow. CVs were then recorded on ZAHNER IM6e electrochemical station.

3. Results and discussion 3.1. Constant voltage electrolysis From the Gibbs free energy of formation, it can be calculated that the decomposition voltages of NaCl and CaCl2 at 820 ? C are 3.24 and 3.32 V, respectively. To avoid any decomposition of the molten salt components, the cell voltage during electrochemical reduction of the mixed oxide pellets should be controlled at less than 3.2 V. When samples sintered at 850 ? C were electrolyzed at 1.6, 2.5 and 3.1 V for 11 h, the current–time curves in the course of electrolysis are shown in Fig. 2. XRD and SEM analysis results of corresponding products are displayed in Figs. 3 and 4. Generally, the current increased with increasing applied voltage. At an applied potential of 1.6 V, the current showed a slow decline over about 450 min to a relatively stable level which then remained constant until the end of electrolysis (see Fig. 2A).

2.2. Two-electrode constant voltage electrolysis
CeO2 and NiO powders were weighed according to the molar ratio of 1:5 and then PVB (1 wt%) and acetone were added to the mixture and was thoroughly milled in a ball-milling container for 3 h. After drying, the mixture was pressed (30 MPa) into pellets of 10 mm in diameter and 3 mm in thickness. The pellets were calcined in air at 400 ? C for 4 h to remove the PVB and then were sintered at 850, 1050 and 1250 ? C for 5 h, respectively. The sintered pellets were characterised by XRD (XRERT/PRO, CuK) and SEM (KYKY/2800), respectively. An alumina crucible was ?lled with about 800 g of a dehydrated mixture of CaCl2 and NaCl (mass CaCl2 /NaCl ratio, 7:3) and placed at the bottom of a stainless steel tube reactor in a vertical furnace. A thermocouple was placed between the alumina crucible and the inner wall of the stainless steel reactor. Subsequently, argon was introduced into the reactor continuously and the tem-

Fig. 1. Schematic diagram of electro-deoxidation apparatus.

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385

381

Fig. 2. Current–time plots of electrolysis of the pellets (sintering at 850 ? C in air for 5 h) at different constant voltages in a eutectic CaCl2 /NaCl melt at 820 ? C for 11 h: (A) 1.6 V, (B) 2.5 V, and (C) 3.1 V.

Fig. 3. XRD spectra of products from electrolysis of the pellets (sintering at 850 ? C in air for 5 h) at different constant voltages in a eutectic CaCl2 –NaCl melt at 820 ? C for 11 h: (A) 1.6 V, (B) 2.5 V, and (C) 3.1 V.

However, as indicated by the small current value, the reduction reaction is relatively slow. The pellet after electrolysis showed relatively high strength and no shrinkage was observed. The XRD spectrum of the product exhibited that Ni, CeO2 , NiO and CeOCl phases exist, but no CeNi5 alloy had formed (see Fig. 3A), which suggests the reduction of only NiO NiO + 2e = Ni + O2? (1)

The existence of CeOCl may be as a result of partial reduction reaction of CeO2 CeO2 + CaCl2 + e = CeOCl + CaO + Cl? (2)

At 2.5 V, as seen in Fig. 3B, the electrolysis current goes through two stages. The current ?rstly showed a sharp decline over about 100 min to the ?rst stable plateau, and then decayed

Fig. 4. SEM images of products from electrolysis of the pellets (sintering at 850 ? C in air for 5 h) at different constant voltages in a eutectic CaCl2 /NaCl melt at 820 ? C for 11 h: (A) 1.6 V, (B) 2.5 V, and (C) 3.1 V.

382

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385

to the second stable level over the next 300 min. Such behavior is similar to previous ?ndings in potentiostatic electrolysis of transition-metal oxides [15,17]. The change trend of the current on the I–t plots during electrolysis is generally in agreement with the three-phase interlines (3PIs) mechanism. It is suggested that the reduction of the oxide pellets takes place at the metal|oxide|electrolyte 3PIs. 3PIs expand with time from the initial metal wire-oxide contact points on the pellet’s surface to the interior of the pellet. During 3PIs expansion, the amount of 3PIs ?rstly increases and then decreases, and ?nally reaches a stable value. So, the current follows the same trend. It is assumed that because of mass transfer dif?culties, the current gradually decreases [22,23]. The XRD pattern in Fig. 3B clearly exhibited the formation of pure CeNi5 phase. The electrolyzed pellet had noticeably expanded and could be ground easily into powder with pestle and mortar, which is ideal for achieving a suitable surface-to-volume ratio for electrode material in the battery. When 3.1 V was applied, the I–t plot (see Fig. 2C) exhibited features similar to those in Fig. 2B, but with higher current values suggesting a higher reduction rate. XRD patterns of the product (see Fig. 3C) also showed that pure CeNi5 phase was obtained. In practice, a higher cell voltage implies higher energy consumption but also a higher reaction rate which is often preferred to increase production. Fig. 4 showed SEM images of the products after electrolysis at three different cell voltages. The SEM images of the product electrolyzed at 2.5 and 3.1 V, as seen in Fig. 4B and C, were different from that electrolyzed at 1.6 V in Fig. 4A. These two images showed nodular metal particles. The particle sizes seem to grow with increasing electrolysis voltage, as is expected. Fig. 4A showed that sample particles electrolyzed at 1.6 V are ?ower petal-like agglomerated grains. As indicated by Qiu et al. [15], the formation of CeNi5 alloy from electrolysis of NiO or CeO2 in the mixed oxides may occur by two mechanisms: (1) NiO or CeO2 in mixed oxides is reduced to its metal Ni or Ce, respectively and then Ni and Ce form alloy; (2) the NiO with lower stability in the mixed oxide is ?rst reduced to Ni and then the other oxide CeO2 is reduced and alloys on the surface of newly formed Ni. It may be considered that alloy formation, by the latter method will possibly have lower activation energy. In order to further understand the reduction process and formation mechanism for the CeNi5 alloy, the following experiments were carried out. Firstly, electrolysis of the pure CeO2 pellet at a voltage of 3.2 V or lower was carried out. The results show that no metallic Ce formation took place. CeO2 is only partially reduced to CeOCl. Thus it can be seen that the reaction CeO2 + 4e = Ce + 2O2? (3)

Fig. 5. XRD spectra of the products from electrolysis of the pellets (sintering at 850 ? C in air for 5 h) at 3.1 V at different times in a eutectic CaCl2 /NaCl melt at 820 ? C (A) and current–time plots (B).

is dif?cult kinetically. Secondly, at low voltage (1.6 V), electrolysis of CeO2 /NiO mixture pellet was carried out. The results showed that only NiO in mixed oxide pellet is partially reduced to Ni even after the prolonged electrolysis time of 11 h. So, electrolysis was carried out at the higher voltage (3.1 V). Fig. 5 shows the XRD spectra of the products that were electrolyzed at 3.1 V for 1, 2, 4 and 11 h. After the sample is electrolyzed at a voltage of 3.1 V for 1–2 h, as seen in Fig. 5A and B, the product consists

of NiO, CeO2 , CeOCl and Ni phases but CeNi5 phase formation has still not taken place. After the sample is electrolyzed for 4 h, the product mainly contains CeNi5 , although NiO, CeO2 , CeOCl, Ni still exist in the product. These results verify that NiO reduction takes place ?rst and then CeO2 is reduced on the newly formed Ni surface to form the alloy. After electrolysis for 11 h, pure CeNi5 alloy can be obtained, which shows that in the course of electrolysis the mixture of NiO and CeO2 are completely deoxidized and form CeNi5 alloy. Fig. 6 showed the CVs of pure CeO2 powder, pure NiO powder and mixed CeO2 and NiO powders recorded in molten mixture of CaCl2 and NaCl (mass CaCl2 /NaCl ratio, 7:3) at 820 ? C by using MCE. In the absence of oxide, the CV of the bare MCE exhibited no reaction until ?1.0 V when the current started to increase, forming a reduction peak c2 at ?1.3 V which can be attributed to the reduction of NaCaCl3 [11]. The reduction current increases continuously to a very large value c3 until the potential scan is reversed, which is apparently due to the simultaneous reduction of both sodium and calcium cations. Upon reversing the potential scan, the CV of the bare MCE showed the re-oxidation peak a2 at ?1.1 V, which is obviously due to the

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385

383

Table 1 Porosity and density of the pellets after sintering at three different temperatures Sintering temperature (? C) 850 1050 1250 Porosity (%) 35.68 34.38 19.19 Apparent density (g/cm3 ) 4.3653 4.5419 5.4759

Fig. 6. CVs of the MCE in the absence and presence of CeO2 powder, NiO powder and mixed CeO2 and NiO powders in molten CaCl2 /NaCl at 820 ? C. Scan rate: 10 mV/s.

rent plateau appears between ?0.45 and ?1.0 V, which is likely caused by the reduction of CeO2 on the newly formed Ni metal particles and then the formation of the CeNi5 compound. Upon reversing the potential scan, the CV of the mixed oxides shows the re-oxidation peak a1 at ?0.95 V which is attributed to Ce oxidization in CeNi5 . This re-oxidation peak, which is absent on the CV of the bare MCE, pure CeO2 and the pure NiO, is strong evidence that the reduction of CeO2 has indeed occurred in the presence of Ni. Associated with the XRD spectra for different electrolysis time, the reaction at the cathode can be expressed as: NiO + 2e = Ni + O2? CeO2 + CaCl2 + e = CeOCl + CaO + Cl? 5Ni + CeOCl + 3e = CeNi5 + O2? + Cl? 3.2. The in?uence of sintering temperature The particle size, porosity and stability of the pellets are important for directly reducing a solid oxide powder in molten salt by the electrochemical method [12]. In order to study the (4) (5) (6)

re-oxidation of the deposited Ca or Na metal. After this peak, the current decreases to the background level at about ?0.95 V. The voltammetric feature of the CeO2 powder is similar to the one of the bare MCE, which means that no CeO2 reduction reaction occurs because of the high thermodynamic stability of CeO2 . The CV of the pure NiO powder exhibited a reduction peak c1 at ?0.45 V which can be attributed to the reduction of NiO to Ni [19]. The voltammetric features of the mixed CeO2 and NiO powders at 0 to ?0.45 V are similar to those observed on the CV of pure NiO powder. Interestingly, after the reduction peak, the current does not reach the background level. Instead, a cur-

Fig. 7. SEM images of the CeO2 /NiO pellets sintered at different temperatures before (A) and after sintering at 850 ? C (B), 1050 ? C (C), and 1250 ? C (D) for 5 h.

384

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385

Fig. 8. XRD pattern of the mixed CeO2 /NiO pellets sintered at two different temperatures for 5 h.

in?uence of sintering temperature, the mixed oxide pellets were sintered at 850, 1050, 1250 ? C for 5 h in air. As seen in Table 1, densi?cation by sintering reduces considerably the porosity and increases the particle size of a porous solid and the extent of the

densi?cation increases rapidly with temperature. Fig. 7 shows the microstructures of the cross section of the porous pellets sintered at 850, 1050, 1250 ? C, prior to electrolysis. It is obvious that porosity and surface area vary with the sintering temperature. It can be seen in Fig. 7 that raising the sintering temperatures involves both the growth of contact area among the particles and pellets strength, and the shrinkage of the solid pellets. Decreases of porosity and surface area lead to the decrease of electrolysis rate. At a sintering temperature of 1250 ? C, considerable increase in the grain size can be clearly observed in Fig. 7D. The XRD spectra of mixed oxide sintered at 850 and 1250? C in air for 5 h (Fig. 8) con?rmed the pellets after sintering still consist of the original oxides of CeO2 and NiO phases. No new mixed oxides phases form, which means the pellets are stable during sintering. The current–time plots of electrolysis of the pellets sintered at different temperatures are shown in Fig. 9A. The rate of the current decay is not constant in each current–time curve in Fig. 9A. With increase in sintering temperature, the current become lower. It is understood that the reduction involves oxygen ion diffusion within the oxide grain. A larger grain size means a longer distance for oxygen transportation and hence the slower reduction rate. The XRD spectra of the samples sin-

Fig. 9. Current–time plots of electrolysis of the pellets sintered at three different temperatures for 5 h (3.1 V, 820 ? C, 11 h, CaCl2 /NaCl melt) (A); XRD pattern of the corresponding products (B).

Fig. 10. Current–time plots of electrolysis of the pellets (sintered at 850 ? C in air for 5 h) at 3.1 V in a different temperature eutectic CaCl2 /NaCl melt for 11 h (A); XRD pattern of the corresponding products (B).

B. Zhao et al. / Journal of Alloys and Compounds 468 (2009) 379–385

385

tered at 850 and 1050 ? C after electrolysis (see Fig. 9B) show that pure CeNi5 phase is formed. As to the pellets sintered at 1250 ? C, the XRD pattern of the product after electrolysis display that the product not only includes CeNi5 , but also CeO2 and Ni phases. So, in order to realize deoxidizing CeO2 and NiO, a longer electrolysis time will be needed. 3.3. The in?uence of the molten salt temperature The electrolysis of the pellet is in?uenced by the temperature of molten salt. I–t curves at two different molten salt temperatures are shown in Fig. 10A. At 750 ? C, the electrolysis current was very small, which means a quite slow reduction reaction at the cathode. Even after 11 h of electrolysis, the pellet was not completely deoxidized. XRD pattern of the product also exhibited that besides CeNi5 , Ni, CeO2 and CeOCl phases exist (see Fig. 10B). The lower the temperature of molten salt, the slower will be the value of oxygen diffusion. During electrolysis at 820 ? C I–t plot shows much higher current at any given time, achieving a higher reaction rate for cathodic reduction. After 11 h of electrolysis, pure CeNi5 alloy was obtained. 4. Conclusions The hydrogen storage alloy CeNi5 can be prepared by direct electrochemical reduction of sintered pellets of mixed CeO2 /NiO powders in a CaCl2 /NaCl melt at a voltage of 2.5 V or higher at 820 ? C. The product from the electrolysis was in the form of porous pellets and could be manually ground into CeNi5 alloy powder. The reduction process study shows that at a voltage of 1.6 V or lower, only NiO is reduced to Ni. At a voltage of 2.5 V or 3.1 V, the reduction reaction proceeds through two steps to CeNi5 , that is, the reduction on the mixed oxide start from NiO to Ni, followed by that of CeO2 on the surface of newly formed Ni to form the CeNi5 alloy. Electrolysis of sintered pellets can lead to the preparation of CeNi5 at voltages as low as 2.5 V which is much smaller than the decomposition voltage of CaCl2 and NaCl. The preliminary results reported here shows

that molten salt electrolysis is a potential new method for commercial production of CeNi5 . Further research is being carried out for obtaining a better understanding of the fundamental principle, for optimizing the operating conditions so that the activity and hydrogen storage property of product powder is enhanced in a Ni/MH battery. Acknowledgement The authors are grateful to NNSF (50572024) for ?nancial support. References
[1] M. Geng, D.O. Northwood, Int. J. Hydrogen Energy 28 (2003) 633–636. [2] M.L. Soria, J. Chacon, J.C. Hernandez, J. Power Sources 102 (2001) 97–104. [3] C. Wang, M. Marrero-Rivera, D.A. Sera?ni, Int. J. Hydrogen Energy 31 (2006) 603–611. [4] J. Dischinger, H.J. Schaller, J. Alloys Compd. 312 (2000) 201–210. [5] L. Rong, J. Wu, X. Wang, J. Alloys Compd. 311 (2000) 40–45. [6] S.N. Klyamkin, N.S. Zakharkina, A.A. Tsikhotskaya, J. Alloys Compd. 398 (2005) 145–151. [7] X. Wang, C. Wang, C. Chen, J. Alloys Compd. 420 (2006) 107–110. [8] Y.X. Huang, H. Ye, H. Zhang, J. Alloys Compd. 330 (2002) 831–834. [9] G.Z. Chen, D.J. Fray, T.W. Farthing, Nature 407 (2000) 361–364. [10] X.B. Jin, P. Gao, D.H. Wang, Angew. Chem. Int. Ed. 43 (2004) 733–736. [11] E. Gordo, G.Z. Chen, D.J. Fray, Electrochim. Acta 49 (2004) 2195–2208. [12] K. Jiang, X. Hu, M. Ma, Angew. Chem. Int. Ed. 45 (2006) 428–432. [13] D. Wang, G. Qiu, G.Z. Chen, Angew. Chem. Int. Ed. 45 (2006) 2384–2388. [14] G.Z. Chen, D.J. Fray, Trans. Inst. Min. Metall. Sect. C—Miner. Process Extr. Metall. 115 (2006) 49–54. [15] G. Qiu, D. Wang, G.Z. Chen, Electrochim. Acta 51 (2006) 5785–5793. [16] X.Y. Yan, D.J. Fray, J. Electrochem. Soc. 152 (2005) D12–D21. [17] G. Qiu, D. Wang, G.Z. Chen, J. Electroanal. Chem. 589 (2006) 139–147. [18] B.A. Glowacki, D.J. Fray, G. Chen, Physica C 387 (2003) 242–246. [19] G. Qiu, M. Ma, D. Wang, J. Electrochem. Soc. 152 (2005) E328–E336. [20] A.J. Muir Wood, R.C. Copcutt, G.Z. Chen, Adv. Eng. Mater. 5 (2003) 650–653. [21] G.Z. Chen, D.J. Fray, J. Electrochem. Soc. 149 (2002) E455–E467. [22] Y. Deng, D.H. Wang, G.Z. Chen, J. Phys. Chem. B 109 (2005) 14043– 14051. [23] G.Z. Chen, E. Gordo, D.J. Fray, Metall. Mater. Trans. B 35 (2004) 223–233.


相关推荐

最新更新

猜你喜欢