+Advanced Search
网刊加载中。。。

使用Chrome浏览器效果最佳,继续浏览,你可能不会看到最佳的展示效果,

确定继续浏览么?

复制成功,请在其他浏览器进行阅读

Formation Process of CoCrFeNi High Entropy Alloy via Electro-Deoxidization of Metal Oxides in Molten Salt  PDF

  • Yang Yu 1,2
  • Luo Xiangyu 1,2
  • Ma Tongxiang 1,2
  • Hu Liwen 1,2
  • Wen Liangying 1,2
  • Hu Meilong 1,2
1. College of Materials Science and Engineering, Chongqing University, Chongqing 400044, China; 2. Chongqing Key Laboratory of Vanadium-Titanium Metallurgy and New Materials, Chongqing University, Chongqing 400044, China

Updated:2021-09-26

  • Full Text
  • Figs & Tabs
  • References
  • Authors
  • About
CN CITE
OUTLINE

Abstract

CoCrFeNi high entropy alloys (HEAs) were prepared by electro-deoxidization of metal oxides in CaCl2 molten salt at 1173 K. The phase transformation from the metal oxides to HEA under different electrolysis durations was investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy dispersive X-ray spectroscopy (EDS). The results show that the formation process of CoCrFeNi HEA includes two stages: the rapid deoxidization stage and deep deoxidization stage. In the rapid deoxidization stage, 93.93wt% oxygen in sintered oxide pellet is removed within 1 h and the current efficiency reaches 89.95%. After electrolysis of 15 h, the oxygen content of the product is 0.26wt% and the current efficiency is 17.93%. The formation process of CoCrFeNi HEA provides guidance for establishing the electrochemical route with low cost and high efficiency.

Science Press

Generally, the traditional alloys consist of one or two main elements as the matrix and several minor alloying elements for modifying the microstructure and properties[

1]. However, the high entropy alloy (HEA), a relatively new type of metallic materials, has attracted much attention for its unique composition, microstructure, and excellent performance. According to the definition proposed by Yeh et al[2], HEAs are typically composed of five or more principal elements in equimolar or near-equimolar ratios, and each principal element concentration is between 5at% and 35at%. Because the large configurational entropy is in favor of the thermo-dynamic stability in equimolar or near-equimolar ratios, the HEAs tend to have a simple solid solution of single-phase structure, and exhibit outstanding properties, such as high strength and hardness, good structure stability, excellent corrosion and oxidation resistance.

At present, the traditional preparation methods of HEAs based on the product state in the preparation process are the liquid method, the solid method, and the gas method[

3,4]. But the disadvantages, such as high preparation temperature, high cost, and long process flow, of these preparation methods have restricted the further application of HEAs. The method of the electro-deoxidization in molten salt has demonstrated that HEAs can be synthesized directly from their mixed solid oxides[5]. Compared with the traditional methods, electro-chemical reduction process has the advantages of low temperature and one-step reduction. Currently, many resear-ches about the electro-deoxidization mechanism have been made for different pure metal oxides[6-8]. Sri et al[9] investigated the reaction process during electro-deoxidization of the solid Nb2O5 discs in molten CaCl2. The results show that the reaction pathway of Nb2O5 is CaxNbyOz→NbO2+NbO→Nb. Jackson et al[10] reported that NiTi alloy prepared via electrochemical reduction in molten CaCl2 involves a series of partial reductions. Firstly, Ni and CaTiO3 are formed rapidly from NiTiO3 precursor. Then, the intermediate compounds Ni3Ti and Ni-Ti-O appear. Finally, NiTi alloy forms. In addition, the reduction of other pure metal oxides and mixed metal oxides to the pure metals and alloys via electro-deoxidization in molten salt was also investigated. However, the research on the formation process of HEAs has rarely been reported[11].

In this research, the formation process of CoCrFeNi HEA was investigated by varying the electrolysis duration. The oxygen content and the current efficiency of the specimens after different electro-deoxidization durations were obtained, which provides the guidance for one-step and high efficiency process of HEAs preparation.

1 Experiment

The solid oxide powders (commercially available solid oxide in molar ratio of CoO:Cr2O3:Fe2O3:NiO=2:1:1:2) were mixed in the ball mill for 1 h to ensure the uniformity. In order to keep the similar electro-reduction rate for specimens with different electro-deoxidization durations, NH4HCO3 (reagent grade, Sinopharma Chemical Reagent Co., Ltd, Beijing, China) was used as fugitive agent[

12]. The homogenized mixed solid oxide powders of 2 g and NH4HCO3 of 1 g were placed in a cylindrical mold of 20 mm in diameter and die-pressed at 6 MPa (disc of 3.5 mm in thickness). Then, the die-pressed specimens were dried at 353 K for 1 h to remove the NH4HCO3. Finally, the green pellets were sintered at 1173 K for 3 h.

The anhydrous CaCl2 of 500 g (reagent grade, Sinopharma Chemical Reagent Co., Ltd, Beijing, China) was dried at 443 K for 24 h to remove moisture. The reactor was vacuumed and then flowed by argon to remove oxygen. Subsequently, the temperature of the melt in alumina crucible was increased to 1173 K under argon atmosphere to melt salt. Fig.1 presents the schematic diagram of the experiment platform, electrolysis cell, and appearance of the anode and cathode. The stainless-steel wire mesh (75 μm) and the wire (0.1 mm in diameter) were wrapped around the cathode pellet. The graphite rod (10 mm in diameter and 100 mm in height) was used as the anode. The electrolytic voltage in the experiment was 3.0 V. The electrolysis time ranged from 5 min to 15 h. After electrolysis, the products were ground into powders and washed in deionized water. Finally, the specimens were vacuum-dried at 333 K.

Fig.1 Schematic diagram of experiment platform, electrolysis cell, and appearance of anode and cathode

The theoretical decomposition potentials of each oxide into the pure metal at 1173 K were calculated by HSC Chemistry 6.0 software[

13], as listed in Table 1. The electrolytic voltage was supplied by DC power supply (DP310, MESTEK, China). The open porosity of the specimens was measured by Archimedes' method. The precursors and the products were characterized by X-ray diffraction (XRD, D/max 2500PC, Rigaku, Japan), scanning electron microscopy (SEM) with backscattered electron (BSE) mode, and energy dispersive X-ray spectroscopy (EDS, TESCAN VEGA II with Oxford INCA Energy 350). The ONH analyzer (THC600, Germany) was used to measure the oxygen content of the cathodic products.

Table 1 Possible reactions and related theoretical decomposition potentials
No.ReactionΔG1173 K/kJ·mol-1E1173 K/V
1 2CoO=2Co+O2 (g) 303.036 -0.79
2 Cr2O3=2Cr+1.5O2 (g) 827.988 -1.43
3 Fe2O3=2Fe+1.5O2 (g) 516.287 -0.89
4 2NiO=2Ni+O2 (g) 267.152 -0.69
5 CaCl2=Ca+Cl2 (g) 623.491 -3.23

2 Results and Discussion

2.1 Characterization of solid oxide precursors

Fig.2 shows XRD patterns of the mixed oxide powders and the sintered pellet, indicating that no chemical reaction happens during ball milling. The phases of the sintered pellet include CoCr2O4, NiCo2O4, NiCr2O4, and NiFe2O4, suggesting that the separated oxides can be converted to complex oxides by sintering at 1173 K. This conclusion was also proposed by Sure[

14]. In addition, because NH4HCO3 escaped when the die-pressed specimens were dried at 353 K for 1 h, the mixed oxide powder is not reduced by NH4HCO3.

SEM/EDS analysis was used to verify the homogeneity of the individual oxide particles of precursors, as shown in Fig.3. The obvious sintering phenomenon among the separated original oxides occurs and the pores are distributed in the sintered pellet uniformly, which is caused by the evaporation of NH4HCO3 fugitive agent during the sintering process[

12]. The open porosity of the specimens prepared under the same condition is 48.37vol%, which provides the same ion diffusion rate in the mixed oxide pellets under different experiment conditions. Fig.3c presents SEM-BSE image of the sintered pellet. The EDS element distributions reveal that Co is distributed uniformly, Ni is aggregated in a small area, and Fe and Cr are dispersed uniformly in the area without Ni aggregation. XRD results of the sintered pellet show that not all the original oxides are converted into complex oxides in the sintering process. There is some remaining NiO, which leads to the aggregation of Ni. The uniform distribution of Fe and Cr in the area without Ni aggregation is due to the formation of spinel of CoCr2O4 and Fe2O3.

Fig.3 SEM-BSE images (a~c) and EDS element distributions of Co (d), Cr (e), Fe (f), Ni (g), and O (h) for sintered oxide pellet

2.2 Electro-deoxidization of solid oxide precursors

The theoretical decomposition voltages of CoO, Cr2O3, Fe2O3, and NiO at 1173 K are 0.79, 1.43, 0.89, and 0.69 V, respectively. Meanwhile, the generation of CO at anode offers depolarization effect in the electrolytic cell, which promotes the decrease of potential gain[

15]. The theoretical decompo-sition voltage of CaCl2 electrolyte is 3.23 V, indicating that the experiment voltage of 3.0 V in the two-electrode system is insufficient to decompose the CaCl2 electrolyte but sufficient to electro-deoxidize the metal oxides.

Fig.4 presents the current (I)-electrolysis time (t) curve of the sintered pellet electrolyzed for 15 h. The inset images show that the product powders changed from black to gray with the increase of electrolysis time. There are 4 typical stages in the I-t plot of electro-deoxidization for 15 h in the molten CaCl2 at 1173 K. Firstly, a peak current forms at the beginning of the electrolysis. Then the current decreases sharply from 4.15 A to 3.90 A in a short time. Secondly, the current gradually decreases from 3.90 A to 2.35 A during the following 0.66 h. Thirdly, the current enters a slow decrease stage from 1.77 A to 0.47 A during 0.75~3.5 h after a current quick drop. Finally, the current decreases to a plateau stage, which is close to the background current. The peak current formed at the beginning of the electrolysis mainly benefits from the large reaction area because the cathode pellet surface is covered by stainless steel wire mesh. Different electrolysis durations (5 min, 15 min, 0.5, 1, 3, 6, and 10 h) were selected to investigate the electro-deoxidization and alloy formation process. The inset curve shows the I-t curve of electrolysis in 1 h.

Fig.4 Current (I)-electrolysis time (t) curve of sintered pellet (inset images are appearance of the products after electrolysis for different time)

2.3 Structure and morphology characterization of catho-dic products

XRD patterns were used to identify the composition of the product powders from black to gray, as shown in Fig.5. According to the changes in diffraction peaks of specimens after different electrolysis durations CoCrFeNi HEA can be found at electrolysis time of 3 h. After electrolysis of 3 h, the phases do not change with further extending the electrolysis time. XRD pattern of product after electrolysis of 5 min shows that the Ni, Co, and CaCr2O4 appear. On the one hand, the formation of Ni and Co in the first 5 min benefits from the large contact area between cathode oxide pellet and stainless-steel wire mesh due to the large initial reaction interface and good conductivity. On the other hand, the low theoretical decomposition potentials and oxide valence make NiO and CoO easy to be electro-reduced into pure metal. CaCr2O4 in the reduced specimens is generated, and calcium ions play an active role in the reduction process and are temporarily found in the cathode[

15,16]. Eq.(1) and Eq.(2) reveal the reactions as follows:

4Cr2O3+3Ca2++6e-=2Cr+3CaCr2O4 (1)
CaCr2O4+6e-=2Cr+4O2-+Ca2+ (2)

Cr diffraction peak cannot be found in XRD patterns due to the trace content of Cr. Fig.6 shows the SEM-secondary electron (SE) and SEM-BSE images of the product after electrolysis of 5 min. It can be observed that a small number of other phases (light color phase) are distributed in the specimens. SEM-SE and SEM-BSE images show that the particles precipitated from the matrix in other phases are about 3 μm in size. EDS spectrum of point 1 indicate that the precipitated particles are the alloy phase consisting of Ni, Co, and a little Cr. The EDS spectra of point 2 and 3 indicate that the matrix phase consists of complex oxides and CaCr2O4.

Fig.6 SEM-SE (a, c) and SEM-BSE (b, d) images of specimen after electrolysis for 5 min; EDS annlysis results of point 1 (e), 2 (f), and 3 (g) in Fig.6d

XRD pattern of specimen after electrolysis for 15 min shows that CoCrFeNi HEA, Fe2O3, FeO, and Fe form in this stage. The appearance of CoCrFeNi HEA with face-centered cubic (fcc) structure in specimen is ascribed to the presence of Fe. The calcium ferrite does not form in the electro-reduction process of Fe2O3. The reduction of solid Fe2O3 is carried out by a two-step reduction: Fe2O3→FeO→Fe. This result agrees with the conclusion of Gao[

17].

Hence, the metal of Fe is electrolyzed later than Ni, Co, and Cr. When Fe metal is produced in the molten CaCl2 at 1173 K, the HEA is likely to form among the metal of Ni, Co, Cr, and Fe due to the high entropy effect[

18,19]. Fig.7 shows SEM-BSE images and related EDS spectra of the specimen after electrolysis for 15 min. From the SE images in Fig.7, the dendritic structure (Fig.7a and 7d) and the precipitated particles (Fig.7g) can be observed. EDS analyses were used to identify the special phases in the product. The results of point 1 and 3 indicate that the dendritic structure is composed of Fe, and different morphologies in dendritic structure may be caused by Cr and Co. The results of point 2 and 4 indicate that the matrix phase is mainly composed of CaCr2O4 and complex oxides. The results of point 5 and 6 represent the phase of Ni and CoCrFeNi HEA, and the morphology of CoCrFeNi HEA is nodular structure, as identified by XRD results.

XRD pattern of specimen after electrolysis for 0.5 h indicates that the spinel phase in the sintered pellet disappears after electrolysis for 0.5 h. The main oxide phases in the cathode product are CaCr2O4 and Cr3O, because the theoretical decomposition potential of Cr2O3 is higher than that of other metal oxides, which leads to a slow electro-deoxidization process. Besides the main phases of CoCrFeNi HEA, Fe also exists in the cathode product because the dendritic iron does not form CoCrFeNi HEA due to the long diffusion distance with other elements. Fig.8 shows SEM-SE and SEM-BSE images and EDS spectra of the product after electro-deoxidization for 0.5 h. As shown in Fig.8b, it can be observed that the main phase has nodular structure. EDS analyses of point 2 and 4 indicate that the nodular structure phase is CoCrFeNi HEA in non-equimolar ratio. EDS analyses of point 1 and 3 exhibit the unreduced CaCr2O4 and Fe metal, respectively, which do not form HEA.

The cathode product after electrolysis for 1 h consists of CoCrFeNi HEA, metal Fe, and NiCoCr phases. The oxide phases disappear after electrolysis for 1 h in CaCl2 molten salt at 1173 K. Fig.9 shows the SEM-SE and SEM-BSE images and EDS spectra of the product after electrolysis for 1 h. Fig.9a and 9d show that the phase of dendritic structure and coupling shape exists in the nodular structure matrix phase. Fig.9b, 9c, 9e, and 9f present the phases of dendritic structure and coupling shape. EDS analyses of point 1 and 2 indicate that the phase of dendritic structure and the coupling shape is Fe-rich phase, which is similar to the phases observed in Fig.7a and 7b. This result indicates that Fe is reduced and aggregated firstly, and then diffused and alloyed with other elements during the formation process of HEA in molten salt electrolysis.

XRD pattern of the product after electrolysis for 3 h shows that the CoCrFeNi HEA is the only phase in the product. Fig.10a~10h show SEM-BSE images of the products after electrolysis for 3~15 h. All images reveal a nodular structure with the individual particles closely joined together at their interfaces, and the morphology of CoCrFeNi HEA is not affected by the processing time. Fig.10i~10l display element distributions of the CoCrFeNi HEAs reduced for different durations. The composition of different CoCrFeNi HEAs is listed in Table 2. The result indicates that Fe, Co, and Ni are uniformly distributed and only a small amount of Cr is aggregated in the certain areas. The similar element distributions were also reported by other studies[

20,21]. Dai[22] pointed out that the compounds of Cr7C3 and Cr23C3 are formed during the electro-deoxidization process, and C element is produced by the side reactions of soluble CO32- and calcium ions.

Fig.10 SEM-BSE images (a~h) and corresponding element distributions (i~l) in Fig.10e~10h of specimen after electrolysis for 3 h (a, e, i), 6 h (b, f, j), 10 h (c, g, k), and 15 h (d, h, l)

Table 2 Chemical composition of CoCrFeNi HEAs after diffe-rent electrolysis durations (at%)
Electrolysis time/hCrFeCoNi
3 26.81 25.14 25.25 22.80
6 32.34 19.61 27.45 20.60
10 27.20 30.08 23.85 18.87
15 26.54 23.98 27.54 21.94

2.4 Oxygen content, current efficiency, and energy con-sumption of HEAs

There are 4 typical stages of CoCrFeNi HEA formation in the I-t plot. In the second stage, the current gradually decreases from 3.90 A to 2.35 A as electrolysis time ranges from 5 min to 0.5 h. XRD patterns of the specimens after electrolysis for 5 min~0.5 h show that metal oxide phases always exist. However, metal oxide phases disappear in the specimen electrolyzed for 1 h after the current drops. As shown in Fig.11, it can be concluded that the second stage is the rapid deoxidization stage, in which 93.93wt% oxygen in sintered oxide pellet was removed within 1 h and the current efficiency reaches 89.95%. The decrease of oxygen content of the products after electrolysis for 1~15 h is only 3.11wt% (the oxygen content of specimens after electrolysis for 15 h is 0.26wt%), but the current efficiency decreases from 89.95% to 17.93%. It can be demonstrated that the third and fourth stages in I-t plot is the deep deoxidization stage. The main reason for the low current efficiency is the high background current caused by the side reactions of soluble CO32- and calcium ions[

23]. In general, the side reactions can be inhibited effectively and the current efficiency is improved significantly by an inert anode[24,25].

The energy consumption of CoCrFeNi HEA preparation with different oxygen contents is also plotted in Fig.11. The results show that the energy consumption of specimen after electrolysis of 1 h (3.37wt% oxygen) is 4 kW·h/kg, which is significantly lower than that treated by iron metallurgy. However, the energy consumption increases linearly as oxy-gen content decreases, and eventually reaches 14.7 kW·h/kg with 0.26wt% oxygen, which is higher than that treated by iron metallurgy. It has been reported that oxygen and carbon doping in HEAs can significantly enhance the strength and ductility[

26,27]. In addition, a mixture of the individual metal oxides and graphite was deoxidized in a melt of CaCl2 at 1173 K to synthesize nanoscale (TiNbTaZrHf)C HEA using modern additive manufacturing techniques, and the preliminary tests also indicate a possible application in supercapacitors[28]. The oxygen and carbon doping in HEAs not only solves the problems of cathode carbon pollution and deep deoxidization, but also improves the current efficiency significantly. This result verifies the feasibility of HEAs preparation at low processing temperature, which is a low-energy and high efficiency process.

3 Conclusions

1) The formation process of CoCrFeNi high entropy alloy (HEA) in molten salt during electro-deoxidization of metal oxides consists of two stages: the rapid deoxidization stage and the deep deoxidization stage. In the first stage, 93.93wt% oxygen in sintered oxide pellet is removed within 1 h and the current efficiency reaches 89.95%. In the second stage, the oxygen content of the products after electrolysis for 15 h is 0.26wt%, but the current efficiency is only 17.93%.

2) In the electro-deoxidization process, NiO and CoO are reduced into pure metal more easily than Cr2O3 and Fe2O3. Fe is reduced and aggregated firstly, and then diffused and alloyed with other elements to form HEA during electro-chemical deoxidization. Because the mechanical properties of HEAs can be enhanced by oxygen and carbon doping, the disadvantages of electrochemical reduction with graphite, such as carbon pollution of cathode, difficulty of deep deoxidization, and low current efficiency, can be conquered.

References

1

Sure J, Vishnu D S M, Schwandt C. Applied Materials Today[J], 2017, 9: 111 [Baidu Scholar

2

Yeh J W, Chen S K, Lin S J et al. Advanced Engineering Materials[J], 2004, 6(5): 299 [Baidu Scholar

3

Zhang Yong, Zuo Tingting, Tang Zhi et al. Progress in Materials Science[J], 2014, 61: 1 [Baidu Scholar

4

Zhang W R, Liaw P K, Zhang Y. Science China Materials[J], 2018, 61(1): 2 [Baidu Scholar

5

Chen G Z, Fray D J, Farthing T W. Nature[J], 2000, 407(6802): 361 [Baidu Scholar

6

Chen G Z, Gordo E, Fray D J. Metallurgical and Materials Transactions B[J], 2004, 35(2): 223 [Baidu Scholar

7

Qiu Guohong, Ma Meng, Wang Dihua et al. Journal of the Electrochemical Society[J], 2005, 152(10): 328 [Baidu Scholar

8

Abdelkader A M, Kilby K T, Cox A et al. Chemical Reviews[J], 2013, 113(5): 2863 [Baidu Scholar

9

Sri M V D, Sanil N, Shakila L. Electrochimica Acta[J], 2013, 100: 51 [Baidu Scholar

10

Jackson B, Jackson M, Dye D et al. Journal of the Electro-chemical Society[J], 2008, 155(12): 171 [Baidu Scholar

11

Yang Y, Ma T X, Hu M J et al. TMS 2020 149th Annual Meeting & Exhibition Supplemental Proceedings[C]. San Diego: Springer, 2020: 1593 [Baidu Scholar

12

Li Wei, Jin Xianbo, Huang Fulong et al. Angewandte Chemie International Edition[J], 2010, 49(18): 3203 [Baidu Scholar

13

Suzuki R O, Ono K, Teranuma K. Metallurgical and Materials Transactions B[J], 2003, 34(3): 287 [Baidu Scholar

14

Sure J, Vishnu D S M, Schwandt C. Journal of Alloys and Compounds[J], 2019, 776: 133 [Baidu Scholar

15

Schwandt C, Fray D J. Zeitschrift für Naturforschung A[J], 2007, 62(10-11): 655 [Baidu Scholar

16

Hyslop D J S, Abdelkader A M, Cox A et al. Acta Materialia[J], 2010, 58(8): 3124 [Baidu Scholar

17

Gao Haiping, Jin Xianbo, Zou Siwei et al. Electrochimica Acta[J], 2013, 107: 261 [Baidu Scholar

18

Yeh J W. Annales de Chimie, Science des Materiaux (Paris)[J], 2006, 31(6): 633 [Baidu Scholar

19

Yeh J W. JOM[J], 2013, 65(12): 1759 [Baidu Scholar

20

Maiti S, Steurer W. Acta Materialia[J], 2016, 106: 87 [Baidu Scholar

21

Wang Bin, Huang Jian, Fan Jinhang et al. Journal of the Electrochemical Society[J], 2017, 164(14): 575 [Baidu Scholar

22

Dai Lei, Lu Yuan, Wang Xianyu et al. International Journal of Refractory Metals and Hard Materials[J], 2015, 51: 153 [Baidu Scholar

23

Jiao H D, Wang M Y, Tu J et al. Journal of the Electrochemical Society[J], 2018, 165(11): 574 [Baidu Scholar

24

Jiao S Q, Fray D J. Metallurgical and Materials Transactions B[J], 2010, 41(1): 74 [Baidu Scholar

25

Hu L, Song Y, Ge J et al. Journal of the Electrochemical Society[J], 2016, 163(3): 33 [Baidu Scholar

26

Lei Zhifeng, Liu Xiongjun, Wu Yuan et al. Nature[J], 2018, 563(7732): 546 [Baidu Scholar

27

Wang Z, Baker I, Cai Z et al. Acta Materialia[J], 2016, 120: 228 [Baidu Scholar

28

Sure J, Vishnu D S M, Kim H K et al. Angewandte Chemie International Edition[J], 2020, 59(29): 11 830 [Baidu Scholar