Study on the cathodic process of co-existing LiF and KF in Na₃AlF₆–Al₂O₃ molten salt with various cryolite ratios

Abstract

Selective and efficient electrochemical analytical methods for examining the deposition of metallic aluminum are significant and necessary. Herein, the electrochemical behaviour of aluminium was studied at an inert tungsten electrode in an Na₃AlF₆–Al₂O₃–LiF–KF eutectic melt at 1253 K by means of transient electrochemical techniques. Electrochemical measurements are based on the potentiodynamic steady-state polarization curve plotting, continuous potential pulse method, constant potential method, constant current method and open-circuit chronopotentiometry. The cathodic overpotential for the deposition of metallic aluminum first decreased and then increased, while the activity of aluminum gradually decreased, as the cryolite ratio increased. Na⁺ ions discharged and precipitated when the cryolite ratio exceeded 3.0. KF played a decisive role in modifying electrode reactions, whereas the effect of LiF was not significant in the Na₃AlF₆–Al₂O₃–LiF–KF melt. The addition of KF not only influenced the current efficiency, but also affected the purity of the metallic aluminum, and the simultaneous addition of KF and LiF increased the dissolution loss of aluminum.

---

Introduction

The electrolytic mechanisms involved in producing metallic aluminum has garnered increasing attention from scholars with the gradual maturation of the aluminum electrolysis industry and the irreplaceable role of the cryolite–alumina electrolyte system.^1–4 Fundamental research has been conducted to better guide industrial applications.^5–8 Concurrently, a growing number of research methods and measurement techniques have emerged, among which electrochemical measurement methods constitute one of the primary approaches.^9–13 These methods were mainly used to study the electrode processes at the anode and cathode during aluminum electrolysis.^14–18

Tie et al.¹⁹ utilized cyclic voltammetry and potential step testing methods to investigate the role of aluminum–tungsten alloying during the electrolysis process. The study elaborated the alloying effect of metallic aluminum on the electrode substrate in a cryolite–alumina melt, concluding that the rate of aluminum loss increased with a rise in electrode surface activity. Raj, as well as Sum and Skyllas-Kazacos, employed cyclic voltammetry and chronopotentiometry to examine the current efficiency during aluminum electrolysis on a titanium diboride electrode and the dissolution loss of metallic aluminum, alongside the formation mechanism of aluminum–tungsten alloys on a tungsten wire working electrode and the reaction process of aluminum ions.^20–23 The reaction mechanism of aluminum ions was also elucidated. In subsequent studies, cyclic voltammetry, reverse current chronopotentiometry with delay time, and chronopotentiometry were utilized to investigate the effects of temperature, alumina concentration, and anode bubbles on the dissolution loss of metallic aluminum in cryolite–alumina molten salt during the electrolysis process. Tellenbach and Landolt employed electrochemical methods such as cyclic voltammetry and sequential potential pulse experiments to study current efficiency. Thonstad and Rolseth utilized a galvanostatic method, sequential potential pulse technique, and electrochemical impedance spectroscopy to investigate and discuss the formation mechanism of cathode overvoltage.^24,25 The analysis concluded that cathode overvoltage is primarily caused by diffusion overvoltage, with charge transfer overvoltage having a little effect. In subsequent research, steady-state overvoltage measurements, electrode short-circuiting, and electromotive force measurement methods were applied to study the relationship between cathode overvoltage and the cryolite ratio. Kisza and Kazmierczak investigated the effects of CaF₂ and AlF₃ additives on the reaction mechanism and kinetics of an aluminum cathode through AC impedance and galvanostatic relaxation methods using liquid aluminum as the working electrode.²⁶

While these studies have laid a solid foundation for fundamental theoretical research on aluminum electrolysis,^27–30 there has been relatively limited research on the influence of LiF and KF on the electrode process, and available electrochemical data in this area are scarce. The influence of LiF and KF on the cathode process during the electrolysis of aluminum in the Na₃AlF₆–Al₂O₃ molten salt system has been studied and discussed in detail.^31–34 However, LiF and KF do not exist alone in the actual industrial electrolyte—both forms are present. Therefore, it is necessary to discuss the influence of the coexistence of LiF and KF on the electrolytic cathode processes occurring in the Na₃AlF₆–Al₂O₃ molten salt.

In this paper, the influence of cryolite ratios on the cathodic electrolysis process in the Na₃AlF₆–Al₂O₃–LiF–KF molten salt system was studied. Electrochemical measurements such as potentiodynamic steady-state polarization curve determination, combined with continuous potential pulse, constant potential, constant current and open-circuit chronopotentiometry methods were used to study the electrochemical reaction processes in the electrolyte melt.

Experiment

Preparation of the electrolyte

Na₃AlF₆–Al₂O₃–KF–LiF electrolytes with various cryolite ratios were prepared. The Na₃AlF₆–Al₂O₃–KF–LiF electrolytes with CR values of 1.5, 2.0, 2.5, 3.0, 3.5 and 4.0 (containing 3 wt% Al₂O₃, 3 wt% KF, 3 wt% LiF) were used. Before use, anhydrous KF and LiF were dried at 250 °C under vacuum for 12 h. AlF₃ was pretreated by heating at 200 °C for 5 h and then at 500 °C for 6 h to remove most of the Al₂O₃ impurities. After pretreatment, the reagents with various cryolite ratios were homogeneously mixed in a grove box by continuous grinding with a muller. The mixture with a total mass of 120 g was then transferred to a corundum crucible fitted in a graphite crucible. The crucible was heated in a flow of argon at the working temperature.

Experimental apparatus

Electrochemical studies were conducted in a three-electrode system. The working and reference electrodes were identical tungsten wire electrodes (Φ 2 mm, dia. >99.9% pure), which were mechanically inserted into a corundum tube with an inner diameter of 6 mm. The working electrode (A = 1 cm²) was immersed to a depth of 15 mm; the working area of the working electrode was 1 cm². Prior to electrochemical measurements, the tungsten electrodes were polished with increasingly finer grades of emery paper, followed consecutively by a polishing with alumina compounds down to 0.05 μm. The corundum crucible, with an 8 mm diameter hole at the base, was fitted in a graphite crucible that functioned as the high-surface-area counter electrode; this was, in turn, fitted in the stainless steel crucible. The melt temperature was measured with a nickel–chromium–silicon thermocouple. The electrochemical cell was kept under a continuous argon flow during the experiment.

Electrochemical measurements

The working electrode was characterized by means of potentiodynamic steady-state polarization, continuous potential pulse, chronoamperometry, chronopotentiometry and open-circuit chronopotential techniques. The apparatus used for electrochemical measurements consisted of an American Princeton PARSTST2273 electrochemical workstation and a KEPCO bipolar operational power amplifier (BOOSTER 20A). The data was analyzed by using PowerSuit software and fitted using Origin8.0 software. The experimental setup is illustrated in Fig. 1.

Fig. 1 Experimental setup for electrochemical deposition: 1-auxiliary electrode; 2-working electrode; 3-reference electrode; 4-shielding gas (Ar); 5-water jacket; 6-stainless steel crucible; 7-graphite crucible; 8-corundum crucible; 9-temperature thermocouple; and 10-controlling thermocouple.

Results and discussion

Potentiodynamic steady-state polarization curves

The potentiodynamic steady-state polarization curves of the Na₃AlF₆–Al₂O₃–LiF–KF melt with 3 wt% concentrations of both LiF and KF under different cryolite ratios are shown in Fig. 2. The potential scanning range and the scanning rate were 0– −1.5 V and 5 mV s⁻¹, respectively.

Fig. 2 Potentiodynamic cathodic polarization curves of Na₃AlF₆–Al₂O₃–LiF–KF at the tungsten electrode as a function of cryolite ratio.

As can be observed from Fig. 2, the curves were divided into three feature regions. As the cryolite ratio changed, the biggest difference in the curve was observed in region B. According to the analysis, region B involves the deposition of metallic aluminum. When the cryolite ratio was between 1.5 and 3.0, the metallic aluminum began to deposit at an electrode potential of −0.7 V, and then entered the steady-state diffusion region, reaching the limit current density of metallic aluminum. Here, the electrode process was completely controlled by the diffusion process.

When the cryolite ratio was increased to between 3.5 and 4.0, the curves in region B did not reach the limit current density of metallic aluminum deposition showing a straight line with a lower slope, which changed from diffusion process control to electrochemical process control, and the conductivity of molten salt was enhanced. It can be inferred that the addition of LiF and KF increased the transfer rate of aluminum-containing complex ions when the cryolite ratio was between 1.5 and 3.0. With the increase in the cryolite ratio, the initial potential of the linear region C shifted positively, indicating that the cathode overvoltage of the sodium metal reaction decreased. The slope of the characteristic line presented by linear region C increased first and then decreased with the increase in cryolite ratio, indicating that the conductivity of the molten salt first increased and then decreased. The characteristic curves of the reduction of lithium and potassium metals were not detected on the potentiodynamic steady-state polarization curves.

The comparative analysis indicated that in the Na₃AlF₆–Al₂O₃–LiF–KF melt, KF played a decisive role in the electrode reaction, while the role of LiF was not obvious.^31,33 With the increase in the cryolite ratio, the initial deposition potential of aluminum metal moved first negatively and then positively, which proved that the cathode overvoltage decreased first and then increased with the increase in the cryolite ratio in the Na₃AlF₆–Al₂O₃–LiF–KF melt.

Continuous potential pulse curve

The continuous potential pulse curves of the Na₃AlF₆–Al₂O₃–LiF–KF melt with 3 wt% concentrations of both LiF and KF under different cryolite ratios are shown in Fig. 3. The cathodic reduction and anodic oxidation potentials were E_c = −1.1 V and E_a = −0.8– −0.6 V, respectively.

Fig. 3 Recurrent potential double pulses of Na₃AlF₆–Al₂O₃–LiF–KF at the tungsten electrode as a function of the cryolite ratio.

As seen from Fig. 3, a potential step was set at a scanning time of 5 s, and the cathodic reduction mainly occurred before the potential step at 5 s corresponding to the reduction reaction of the anions in the melt on the tungsten wire electrode. Combined with the oxidation part, it can be inferred that the reduction of the anions was mainly the reduction reaction of Al³⁺ ions. After the step at 5 s, the working electrode was converted to an anode reaction corresponding to the oxidation of the cathode product. Continuous potential pulse curve measurements are used to calculate the reaction electricity required by the cathodic reduction and the anodic oxidation by integrationto evaluate the dissolution loss of metallic aluminum in the process of aluminum electrolysis. Combined with the measurement of steady-state polarization curves discussed in the previous section, the cathode potential deposition in aluminum electrolysis was examined. The potential range selected for the anode was −0.8 to −0.6 V. In this potential range, oxidation of the tungsten electrode on the anode can be avoided, and the anode dissolution of the working electrode can be fully recorded. At the end of the anode reaction, the current density eventually approached zero, indicating that the product generated on the cathode was completely oxidized. When the electrolyte melt contained both LiF and KF, no obvious current plateau was observed during electrode oxidation.

Through the above analysis and integration calculation from Fig. 3, the values of Q_c and Q_a can be calculated (where Q_c is the cathodic reduction reaction charge, Q_a is the anode oxidation reaction charge). The specific values are listed in Table 1, and the trends of the Q_a/Q_c ratios are shown in Fig. 4. The dissolution loss data of metallic aluminum at different cryolite ratios in Na₃AlF₆–Al₂O₃–LiF–KF molten melt is shown in Table 1. The reduced cathode products are completely reacted during the whole oxidation process shown in Fig. 4. In addition, it can be concluded from the data analysis that when the cryolite ratio was between 1.5 and 2.5, the Q_a/Q_c ratio was basically unchanged. When the cryolite ratio was greater than 2.5, the Q_a/Q_c ratio decreased slightly. However, the dissolution loss of aluminum in the Na₃AlF₆–Al₂O₃–LiF–KF melt was more serious than that in Na₃AlF₆–Al₂O₃–LiF and Na₃AlF₆–Al₂O₃–KF melts based on analysis of the ratio of Q_a/Q_c.^31–34

Table 1 The dissolution loss of aluminum in Na₃AlF₆–Al₂O₃–LiF–KF molten salt as a function of cryolite ratio

Cryolite ratio	Anode oxidation potential (V)	Reduction charge (Qc)	Oxidation charge (Qa)	Q a/Qc (%)
1.5	-0.6	1.246	0.712	57.14
	−0.7	1.270	0.519	40.87
	−0.8	1.283	0.536	41.78
2.0	−0.6	1.056	0.598	56.62
	−0.7	1.064	0.430	40.41
	−0.8	0.982	0.317	32.28
2.5	−0.6	1.321	0.757	57.30
	−0.7	1.404	0.550	39.17
	−0.8	1.285	0.404	31.44
3.0	−0.6	1.710	1.000	58.48
	−0.7	1.700	0.620	36.47
	−0.8	1.696	0.475	28.01
3.5	−0.6	1.803	0.969	53.74
	−0.7	1.814	0.367	20.23
	−0.8	1.743	0.322	10.40
4.0	−0.6	1.751	0.793	45.29
	−0.7	1.800	0.194	10.78
	−0.8	1.730	0.180	18.47

---

Fig. 4 The ratio of Q_a/Q_c as a function of the cryolite ratio.

The rate at which aluminum dissolved into the electrolyte can be expressed by the formula (2.1):

In formula (2.1), N_M is the dissolution loss of metallic aluminum, and D_M is the diffusion coefficient of metallic aluminum dissolution; C_s and C_b represent the concentration of metallic aluminum on the electrode surface and into the molten salt, respectively; δ is the thickness of the diffusion layer, which determines the natural convection on the electrode surface, C_b = 0 is use in this experiment,

The loss current density (i_loss = 3FN_M) was a part of the cathode current density, representing the loss amount of metallic aluminum that was incorporated into the molten salt. The corresponding loss in electricity Q_loss can be expressed as: Q_loss = ∫i_lossdt.

In the electrolysis process, the total amount of reduction electricity required for the deposition of metallic aluminum is Q_c, as follows:

Q_c = Q_dep + Q_loss,c (2.2)

In eqn (2.2), Q_dep represents the actual amount of reduction electricity required to reduce metallic aluminum on the cathode within time t_c; , represents the electricity required for the loss of metallic aluminum during cathodic deposition. The current efficiency can be expressed by θ_c, as follows:

The oxidation energy required during the anode oxidation process is represented by Q_a. The anode oxidation of metallic aluminum was only an electrochemical reaction process at this point under a given potential, while the physical dissolution of metallic aluminum occurred throughout the entire anode oxidation process. The corresponding anode oxidation current density was i_loss,a.

The anode oxidation energy can be expressed as:

Q_a = Q_dep − Q_loss,a (2.5)

Combining eqn (2.3) and (2.5), eqn (2.6) can be obtained:

From the eqn (2.6), Q_loss,a can approximately be eliminated in eqn (2.6) for Q_loss,a ≈ i_loss,at_a, Q_a ≈ i_at_a, and eqn (2.7) can be obtained:

If i_a ≫ i_loss,a in the equation, then .

It was shown that the dissolution of aluminum metal included chemical dissolution, electrochemical dissolution and physical dissolution, all of which can affect the dissolution loss of aluminum metal. Through the above analysis, it can be concluded that chemical, electrochemical, and physical dissolution of metallic aluminum occurred simultaneously during the process of aluminum electrolysis. In this paper, the chemical and physical dissolution was accompanied by the reduction of metallic aluminum, while only physical dissolution occurred during the oxidation process.

According to the physical dissolution model of aluminum metal, the dissolution rate of metallic aluminum did not depend on a given electrode potential, and the same applies to cathode deposition and anode oxidation. Therefore, i_loss,c = i_loss,a = i_loss can be obtained, and i_loss,c may vary with the thickness of the diffusion layer under specific experimental conditions.

The molten salt contained a certain concentration of metallic aluminum in the industrial electrolyte, which included physical dissolution and the reaction between metallic aluminum and CO₂ (secondary reaction of metallic aluminum dissolution). According to eqn (2.1), the dissolution rate of aluminum metal is directly proportional to C_s − C_b. However, the concentration of aluminum metal in the molten salt was zero, which means that C_b = 0. Therefore, the dissolution rate of aluminum metal will be higher under laboratory conditions. It can also be inferred from eqn (2.1) that the increase in the natural convection rate will also increase the dissolution loss of metallic aluminum and reduce its current efficiency.

Chronoamperometry curve

The constant potential curves of the Na₃AlF₆–Al₂O₃–LiF–KF melt with 3 wt% concentrations of both LiF and KF under different cryolite ratios are shown in Fig. 5. The cathode potential E_c was −1.1 V, and the time of constant potential electrolysis was 2 h.

Fig. 5 Chronoamperometry curves for 2 h at the tungsten electrode in Na₃AlF₆–Al₂O₃–LiF–KF as a function of the cryolite ratio.

In the initial stage of electrolysis, the current density decreased significantly. When the cryolite ratio was less than 3.5 after the current attenuation. The curves tended to remain in a constant current range, and the electrode was in a stable diffusion state. However, when the molecular ratio was greater than 3.5, the current density did not level off after the attenuation, but began to linearly increase with time before finally reaching a stable state. Changes in the cryolite ratio led to different reactions on the electrode surface. When the cryolite ratio was in the range of 1.5–3.0, only the deposition reaction of metallic aluminum was observed on the electrode. However, when the cryolite ratio was increased to 3.5, other ion deposition reduction reactions occurred on the electrode surface, mainly the Na⁺ ion discharge reaction.

The joint intervention of Li⁺ and K⁺ ions in the molten melt can inhibit the discharge of Na⁺ ions and weaken the current conduction ability of Na⁺ ions; thus, the cathode overvoltage was reduced before the discharge of Na⁺ ions. Moreover, the Na⁺ ion content in the molten salt became higher with increasing cryolite ratio, and the Na⁺ ions began to discharge, resulting in a larger Na⁺ ion concentration gradient between the electrode surface and the electrolyte molten salt, which increased the cathode overvoltage. At the same time, the increase in the cryolite ratio led to an enhancement of the double-layer charging effect, a reduction in the electrochemical reaction rate and diffusion process to varying degrees, and an increase in the cathode overvoltage.

Chronopotentiometry curves

The chronopotentiometry curves of the Na₃AlF₆–Al₂O₃–LiF–KF melt with 3 wt% concentrations of both LiF and KF under different cryolite ratios are shown in Fig. 6. The cathode current density range was 100 to 350 mA cm⁻².

Fig. 6 Galvanostatic curves of Na₃AlF₆–Al₂O₃–LiF–KF at the tungsten electrode as a function of applied current density. (a) CR = 1.5, (b) CR = 2.0, (c) CR = 2.5, (d) CR = 3.0, (e) CR = 3.5, (f) CR = 4.0.

From the results of the measured chronopotentiometry curve (Fig. 6), it can be inferred that when LiF and KF coexist in the electrolyte molten melt, the initial deposition current of metallic sodium decreased with increasing cryolite ratio, as a result of the joint action of LiF and KF.

Open-circuit chronopotential curves

The chronopotentiometry curves of the Na₃AlF₆–Al₂O₃–LiF–KF melt with 3 wt% concentrations of both LiF and KF under different cryolite ratios are shown in Fig. 7. The working electrode was polarized for 600 s at a cathode potential E_c = −1.1 V.

Fig. 7 Open-circuit potential transient curves after application of a potential of E_c = −1.1 V versus the reference electrode by electrolysis for 600 s at the tungsten electrode in Na₃AlF₆–Al₂O₃–LiF–KF as a function of the cryolite ratio. (a) CR = 1.5, (b) CR = 2.0, (c) CR = 2.5, (d) CR = 3.0, (e) CR = 3.5, (f) CR = 4.0.

During the open-circuit potential scanning process, the potential decreased over time. As shown in Fig. 7, three potential plateaus (1, 2, 3) are observed, which correspond to the spontaneous physical dissolution of cathodic products on the polarized electrode.

Plateau 1 corresponds to the spontaneous dissolution of metallic aluminum, while plateau 2 corresponds to the spontaneous dissolution of the Al–W intermetallic compound. The trend of the open-circuit potential change with increasing cryolite ratio in the Na₃AlF₆–Al₂O₃–LiF–KF molten salt system was consistent with that observed in the Na₃AlF₆–Al₂O₃–LiF and Na₃AlF₆–Al₂O₃–KF molten salt systems.^31–34 The duration of plateaus 1 and 2 decreased as the cryolite ratio increased. In other words, the total duration of plateaus 1 and 2 also decreased with an increase in the cryolite ratio.

Therefore, the thickness of the cathodic products became progressively thinner, and the open-circuit potentials of plateaus 1 and 2 gradually shifted positively as the cryolite ratio increased, which indicated a gradual decrease in the activity of the metallic aluminum. Plateau 3 represented an open-circuit potential state where the substances on the electrode surface had completely spontaneously dissolved, and the tungsten wire electrode had returned to its pre-polarization state.

Electrode morphology and surface state characterization

SEM images and line scan analysis of the tungsten wire working electrode after potentiostatic electrolysis for 2 hours in the Na₃AlF₆–Al₂O₃–LiF–KF electrolyte melt with 3 wt% concentrations of both LiF and KF are presented in Fig. 8.

Fig. 8 Scanning electron micrographs and the corresponding line scanning analysis of a tungsten cathode at 1253 K after 2 h of aluminum electrolysis in Na₃AlF₆–Al₂O₃–LiF–KF molten salt at different cryolite ratios: (A) CR = 1.5, (B) CR = 2.0, (C) CR = 2.5, (D) CR = 3.0, (E) CR = 3.5, (F) CR = 4.0.

It was found that a stratification phenomenon appeared on the tungsten wire electrode (Fig. 8), and the electrode can be roughly divided into four layers. The first layer was the metal tungsten layer, the second was the aluminum–tungsten intermetallic compound, the third layer was metallic aluminum, and the fourth layer was the electrolyte attached to the electrode. Through the SEM analysis, it was observed that the thickness of the Al–W intermetallic compound remained essentially unchanged as the cryolite ratio increased. However, when the cryolite ratio exceeded 3.5, the thickness of the Al–W intermetallic compound decreased significantly, which indicated that the simultaneous addition of LiF and KF to the melt can slow the diffusion rate of Al³⁺ ions, reduce the activity of Al³⁺ ions on the tungsten wire substrate surface, and decrease the physical dissolution of metallic aluminum. Additionally, the aluminum metal layer became gradually thinner, and the density of the aluminum metal was significantly affected, becoming increasingly porous as the cryolite ratio increased. Analysis of the elemental distribution on the surface of the working electrode showed that although oxygen was present in the aluminum metal layer, no sodium was detected, which suggested that aluminum oxyfluoride complex ions accumulated on the surface of the working electrode during the electrolysis process. While the fluorine element escaped from the electrode surface, the remaining components remained on the surface. However, it cannot be conclusively determined whether oxygen existed in the form of aluminum–oxygen complex ions. The presence of oxygen may also be attributed to the oxidation of metallic aluminum on the electrode surface, or it could be due to an increase in the cryolite ratio, which affected the discharge of Al³⁺ ions. As a result, some oxygen ions may not be released in time and instead adhere to the electrode surface.

Conclusion

Electrochemical measurement techniques, such as steady-state polarization curve plots and continuous potential pulse, potentiostatic, galvanostatic, and open-circuit chronopotentiometry methods, were employed to investigate the influence of the cryolite ratio on the cathodic process in Na₃AlF₆–Al₂O₃–LiF–KF molten salt. The main research findings were as follows: the reduction peaks of metallic potassium and lithium were not observed in the steady-state polarization curve. KF played a decisive role in modifying the electrode reactions, while the effect of LiF was not significant; however, the dissolution loss of metallic aluminum was more severe when both KF and LiF were present compared to when the salts were present individually. When the cryolite ratio was less than 3.0, there was no Na⁺ ion discharge on the working electrode; discharge occurred only when the cryolite ratio was within the range of 3.5 to 4.0 in the Na₃AlF₆–Al₂O₃–LiF–KF melt. As the cryolite ratio increased, the aluminum activity gradually decreased. The addition of KF to the high cryolite ratio molten salt affected not only the current efficiency of aluminum electrolysis but also impacted the purity of the metallic aluminum.

Author contributions

Shaohu Tao: writing – original draft. Yilin Shao: conceptualization. Zhuang Hao: data curation. Ruiting Ma: software. Naixiang Feng: methodology. Pengyan Mao and Wu Zhang: writing – review & editing.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

All experimental data and results generated or analysed during this study are included in this published article. No additional external data were used.

Acknowledgements

The work was financially supported by the Shenyang Ligong University Introduce High-level Talents Research Support Program Funding (No. 1010147001032), 2022 Basic Scientific Research Project of the Education Department of Liaoning Provincial (Local Service) (No. LJKFZ20220185), and Basic Scientific Research Project of Higher Institution of Higher Education of Education Department of Liaoning Province (No. LJKMZ20220585).

References

1. P. Cui, B. Qin and G. M. Haarberg, J. Electrochem. Soc., 2019, 116, 559–563.
2. W. Zeng, J. Peng and Y. Wang, et al. , Ionics, 2019, 25, 1735–1745.
3. X. Zhang, S. Bao and Y. Luo, et al. , J. Light Met., 2022, 486–493.
4. C. Li, S. Bao and F. Zhang, et al., TMS Annual Meeting & Exhibition, 2023, pp. 731–736.
5. A. Y. Nikolaev, A. V. Suzdaltsev and Y. P. Zaikov, J. Electrochem. Soc., 2019, 166, 784–791.
6. P. Cui, A. Solheim and G. M. Haarberg, ECS Trans., 2014, 64, 351–355.
7. S. K. Padamata, A. S. Yasinskiy and P. V. Polyakov, Metall. Res. Technol., 2019, 116, 410–420.
8. A. Yasinskiy, P. Polyakov and Y. Yang, et al. , Electrochim. Acta, 2021, 366, 137436–137447.
9. P. S. Pershin, A. V. Suzdaltsev and Y. P. Zaikov, Russ. Metall., 2021, 2021, 213–218.
10. V. A. Lebedev and A. A. Shoppert, Solid State Phenom., 2018, 284, 839–844.
11. H. Ma, J. Xue and J. Li, et al. , J. Mater. Process. Technol., 2016, 4, 199–206.
12. E. Kubiňáková, V. Danielik and J. Híveš, J. Electrochem. Soc., 2018, 165, 274–278.
13. B. Gao, F. He and G. Li, et al. , J. Light Met., 2020, 12, 17–23.
14. V. N. Nekrasov, A. V. Suzdaltsev and O. V. Limanovskaya, et al. , Electrochim. Acta, 2012, 75, 296–304.
15. H. M. Kan, Z. W. Wang and X. Y. Wang, et al. , Trans. Nonferrous Met. Soc. China, 2010, 20, 158–164.
16. V. B. Silva and A. Rouboa, J. Electroanal. Chem., 2012, 761, 58–66.
17. G. M. Haarberg and P. Cui, J. Light Met., 2013, 729–752.
18. R. R. Nazmutdinov, T. T. Zinkicheva and S. Y. Vassiliev, et al. , Chem. Phys., 2013, 412, 22–29.
19. J. Tie and Z. X. Qiu, Non-ferrous Metals, 1994, vol. 3, pp. 15–17.
20. S. C. Raj and M. Skyllas-Kazacos, J. Appl. Electrochem., 1993, 23, 257–264.
21. E. Sum and M. Skyllas-Kazacos, J. Appl. Electrochem., 1989, 19, 485–494.
22. E. Sum and M. Skyllas-Kazacos, J. Appl. Electrochem., 1988, 18, 731–738.
23. E. Sum and M. Skyllas-Kazacos, Electrochim. Acta, 1989, 34, 677–690.
24. J. M. Tellenbach and D. Landolt, J. Appl. Electrochem., 1988, 18, 639–643.
25. J. Thonstad and S. Rolseth, Trans. Inst. Min. Metall., Sect. C, 2005, 114, C188–C191.
26. A. Kisza, J. Kazmierczak and J. Thonstad, et al. , J. Appl. Electrochem., 2002, 32, 305–310.
27. Y. Zhang, X. Zhai and Y. Wang, et al. , Guocheng Gongcheng Xuebao, 2011, 11, 755–760.
28. D. Shi, B. Gao and Z. Wang, et al. , J. Light Met., 2014, 4, 22–26.
29. T. Tan, J. Huang and Y. Di, et al. , J. Light Met., 2020, 4, 33–37.
30. Y. Wang, F. He and F. Zhang, et al. , J. Light Met., 2022, 2, 18–21.
31. S. Tao, J. Peng and Y. Di, et al. , Anal. Lett., 2015, 48, 371–381.
32. S. Tao, Y. Di and J. Peng, et al. , Huagong Xuebao, 2014, 65, 2211–2218.
33. S. H. Tao, Y. Z. Di and J. P. Peng, RSC Adv., 2015, 5, 23915–23921.
34. S. H. Tao, Y. Z. Di and J. P. Peng, Rare Met., 2018, 37, 40–46.

---