Electrodeposition of Aluminum Coatings from AlCl3-NaCl-KCl Molten Salts with TMACl and NaI Additives

The Al coatings achieved via electrodeposition on a Cu electrode from AlCl3-NaCl-KCl (80–10–10 wt.%) molten salts electrolyte with Tetramethylammonium Chloride (TMACl) and Sodium Iodide (NaI) additives is reported. The effect of the two additives on electrodeposition were investigated by cyclic voltammetry (CV), chronopotentiometry (CP), scanning electron microscopy (SEM) and X-ray diffraction (XRD). Results reveal that compact and smooth Al coatings are obtained at 150 °C by the electrodeposition process from the electrolyte with 1% TMACl and 10% NaI. The Al coatings exhibit great corrosion resistance close to that of pure Al plate, with a corrosion current of 3.625 μA. The average particle size is approximately 2 ± 1 μm and the average thickness of the Al layer is approximately 7 ± 2 μm. The nucleation/growth process exhibits irrelevance with TMACl or NaI during the electrodeposition of Al. TMACl cannot affect and improve the electrodeposition effectively. However, the addition of TMACl and NaI can intensify the cathodic polarization, producing an inhibition of Al deposition, and contribute to form uniform Al deposits. This can increase the conductivity and facilitate in refining the size of Al particles, contributing to forming a continuous, dense and uniform layer of Al coating, which can be used as effective additives in molten salts electrolyte.


Introduction
High temperature electrolysis in molten salts is an important industrial process across many important areas. In industry, it has been widely used for metals extraction, materials processing and metallic thin film deposition. Over the years, the Hall-Héroult process has been used for the smelting of pure aluminum (Al) at industrial scale [1], the most widely used non-ferrous metal in the world. Al oxide (alumina, Al 2 O 3 ) dissolved in molten cryolite (Na 3 AlF 6 ) and consumable carbon are used as anodes in the process. The electrolysis in molten salts is not energy efficient, consuming huge amount of energy and carbon materials (11.5-13.5 kWh electricity and 0.4-0.5 kg carbon anodes for producing 1 kg Al [2]). Al production is highly energy-intensive, more than 3% of the world's entire electrical supply has been used to extraction of Al every year [3]. The primary Al industry induces As one important type of halides, ammonium halides have seldom been studied as additives in molten salts electrolytes. In industry, ammonium salts are widely used as effective cationic surfactants. It is supposed that the combination of two different types of additives, alkali halides and ammonium halides, could contribute to improving the coating quality and prolonging the electrolyte service life. In fact, in order to understand the mechanism comprehensively and achieve better Al coatings, electrochemical process and nucleation/growth mechanisms are of a great benefit. Since single additive is effective, it is reasonable to predict that the addition of multiple additives could improve the quality of Al coatings, as the sum is always greater than the parts. However, the effects of multiple additives have not been investigated yet.
In the present study, we have introduced the ammonium halides as the additives in molten salts electrolytes for the low temperature (150 • C) electrodeposition of Al. We are focusing on two specific halides-TMACl (Tetramethylammonium Chloride) was adopted to reduce the volatility of the electrolyte and NaI (Sodium Iodide) was adopted to improve the coating quality. The effects of TMACl and NaI on the deposition mechanisms, the morphology and microstructures of Al from AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salt electrolyte were studied systematically. Cyclic voltammetry (CV) and chronoamperometry (CP) have been employed to investigate the electrochemical process and the deposition mechanisms of Al. Electrodeposits were characterized using scanning electron microscopy (SEM) and X-ray diffraction (XRD). The corrosion resistance of Al coatings was tested through potentiodynamic polarization (PD) and electrochemical impedance spectroscopy (EIS).

Electrolytes Preparation
Sodium chloride (NaCl, AR, 99.9% purity, Sinopharm, Beijing, China) and potassium chloride (KCl, AR, 99.9% purity, Sinopharm) were initially dried in a vacuum oven for 72 h at 300 • C. Anhydrous Al chloride (AlCl 3 , AR, 99.8% pure, Sinopharm), NaI (AR, 99.5%, Macklin, Shanghai, China) and TMACl (AR, 99%, Macklin) were used as received. AlCl 3 , NaCl and KCl (80-10-10 wt.%) were mixed and melted at 150 • C in an electrolytic cell on a heating device with temperature control system. The mixture was purged with high-purity argon (Ar) gas. All of the electrodeposition and electrochemical experiments were performed in the glove box (Mikrouna Upure 1220/750/900, Shanghai, China). In order to investigate the effect of NaI and TMACl on the electrodeposition of Al, NaI, TMACl or their mixture were added into the electrolytes and the electrolyte continuously stirred for 4 h to yield a homogeneous yellowish liquid.

Electrolyte Volatility Tests
The electrolyte volatility tests were performed after the electrolyte had been well prepared in the glove box. The temperature-controlled heater maintained the electrolytic system at 150 • C. When we opened the lid of the electrolytic cell, the AlCl 3 started to spill out of the electrolytic cell. Then, we weighed the mass of the electrolytic cell system every 30 min, where the mass loss over the fixed period of time could be calculated as the volatile mass of AlCl 3. The electrolyte volatility tests lasted for 4 h.

Electrochemical Tests
An electrochemical workstation (CHI655D, Chenhua, Shanghai, China) was adopted in all electrochemical experiments. A typical three-electrode cell was used in the experiments, a glassy carbon (GC, 0.07 cm 2 ) was adopted using as a working electrode and an Al plate (99.99%, 350 mm 2 ) was used as a counter electrode. A pure Al bar (99.99%, diameter Φ = 0.5 mm) was adopted as the reference electrode, and it was fixed in a small glassy tube which was full of the AlCl 3 -NaCl-KCl (weight ratio 8:1:1) molten salts. The Al plate and the pure Al bar were burnished with abrasive paper, cleaned in an ultrasonic bath for 3 min, rinsed with distilled water, then dried in a vacuum oven before all the experiments and characterizations. Cyclic voltammogram (CV) tests were carried out with a scan rate of 50 mV/s.

Electrodeposition
In the electrodeposition process, the Al bar described above was adopted as the reference electrode. The Al plate described in the last section was used as anode. A pure Cu foil (the area exposed to the solution is 5 mm × 5 mm, 99.9%) was adopted as the cathode. The Cu foil, Al plate and Al bar were all cleaned with distilled water and then dried in a vacuum oven. The electrodeposition of Al was performed at 150 • C under a constant current mode. Current density was kept at 50 mA/cm 2 , and the electrodeposition process lasted for 10 min. The samples were cleaned after the electrodeposition using an ultrasound bath for 3 min, rinsed with distilled water and then dried in a vacuum oven.

Characterization of the Al Coatings
A field emission scanning electron microscope (FE-SEM, SIRION200, FEI, Shanghai, China) was used to observe the surface and cross-sectional morphologies of the Al coatings. The samples were examined by X-ray diffraction (XRD, D/MAX2000V, Rigaku, Tokyo, Japan) to identify the constituents and the crystal structure by the 2θ/θ scanning mode. The data was collected in a 2θ = 30 • −90 • range with a scanning speed of 5 • min −1 with 0.01 • (2θ) step size.
Electrochemical behavior was studied in the 3.5 wt.% NaCl solution at room temperature by potentiodynamic polarization to evaluate the corrosion resistance. A graphite electrode and a saturated calomel electrode (SCE) were used as the counter electrode and the reference electrode, respectively. All the specimens were held at open circuit for 30 min to reach the steady value prior to potentiodynamic polarization. Scans were obtained from 100 mV below open circuit potential (OCP) to and scanned upwards at a rate of 0.5 mV/s. A PARSTAT 2273 advanced electrochemical system from Princeton Applied Research (New York City, State of New York, USA) was adopted to perform the electrochemical impedance spectroscopy (EIS) tests. The EIS tests conducted at OCP with a sinusoidal signal amplitude of 10 mV. The frequency ranged from 10 5 Hz down to 1 Hz.

Electrolyte Volatility
To study the effect of TMACl on the volatility of AlCl 3 -NaCl-KCl molten salts electrolyte, the volatilization experiments were performed and the results are shown in Figure 1. It can be seen that the AlCl 3 -NaCl-KCl molten salts electrolyte volatilized continuously at 150 • C. Figure 1a shows the mass of volatile AlCl 3 could be as much as 2.180 g after a 4 h exposure. However, when TMACl was added into the electrolyte, the volatility of the electrolyte declined dramatically. As 1 wt.% TMACl, 5 wt.% TMACl, 10 wt.% TMACl was brought into the electrolyte, the mass of volatile AlCl 3 was 0.716, 0.818 and 0.765 g, respectively. The addition of TMACl could effectively inhibit the volatilization of AlCl 3 , and more TMACl in the electrolyte could not further reduce the volatility.
Different contents of NaI were added to test whether TMACl could inhibit the volatilization of AlCl 3 in the electrolyte with NaI. As shown in Figure 1b-d, the mass of volatile AlCl 3 effectively declined after the addition of TMACl, proving that the effect of inhibiting the volatilization of TMACl could not be impaired by NaI. Different content of TMACl had the similar effect on the volatilization of AlCl 3 . A reasonable explanation of the effect of TMACl is that the TMA + ionized from TMACl formed complex ions with Al 2 Cl 7 − , preventing Al 2 Cl 7 − from forming AlCl 3 .

Voltammetric Behavior
The cyclic voltammograms (CV) of Al electrodeposition on the glassy carbon electrode in the AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salts electrolyte at 150 • C has been studied, as shown in Equation (1), only one electrodeposition reaction occurred according to the composition of the electrolyte in which the main type of ions is Al 2 Cl 7 − : The crossing current loop and the oxidation/reduction peaks demonstrate that the electrodeposition of Al in AlCl 3 -NaCl-KCl molten salts is irreversible. Different contents of NaI were added to test whether TMACl could inhibit the volatilization of AlCl3 in the electrolyte with NaI. As shown in Figure 1b-d, the mass of volatile AlCl3 effectively declined after the addition of TMACl, proving that the effect of inhibiting the volatilization of TMACl could not be impaired by NaI. Different content of TMACl had the similar effect on the volatilization of AlCl3. A reasonable explanation of the effect of TMACl is that the TMA + ionized from TMACl formed complex ions with Al2Cl7 − , preventing Al2Cl7 − from forming AlCl3.

Voltammetric Behavior
The cyclic voltammograms (CV) of Al electrodeposition on the glassy carbon electrode in the AlCl3-NaCl-KCl (80-10-10 wt.%) molten salts electrolyte at 150 °C has been studied, as shown in Equation (1), only one electrodeposition reaction occurred according to the composition of the electrolyte in which the main type of ions is Al2Cl7 − : The crossing current loop and the oxidation/reduction peaks demonstrate that the electrodeposition of Al in AlCl3-NaCl-KCl molten salts is irreversible. The CV test was performed in AlCl3-NaCl-KCl (80-10-10 wt.%) molten salts electrolyte with various amount of the two additives on the glassy carbon electrode in order to study the effect of NaI and TMACl on the voltammetric behavior of Al. The CV curves in AlCl3-NaCl-KCl molten salts electrolyte without additive, with 1% TMACl, 5% TMACl, 10% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI are presented in Figure 2. The CV test was performed in AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salts electrolyte with various amount of the two additives on the glassy carbon electrode in order to study the effect of NaI and TMACl on the voltammetric behavior of Al. The CV curves in AlCl 3 -NaCl-KCl molten salts electrolyte without additive, with 1% TMACl, 5% TMACl, 10% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI are presented in Figure 2.
Every cyclic voltammogram was recorded on a fresh glassy carbon electrode surface. It shows that the cathode deposition current density increased with the addition of TMACl, indicating that TMACl cannot give rise to an inhibition of Al deposition. However, as both NaI and TMACl were added into the electrolyte, the deposition current density decreased dramatically. The related experiments [31] proved that the addition of NaI decreased the cathodic current density and gave rise to an inhibition of Al deposition. A possible reason is that the introduction of two additives leads to plenty conductive ions in the molten salts, and the TMACl does not impair the effect of NaI. As shown in the inset of Figure 2, the reduction potential of cathodic was stable at approximately 0.01 V with or without the increase of TMACl, which indicates the addition of TMACl has no prominent effect on changing the reduction potential of cathodic. While, the reduction potential of cathodic shifted obviously in a negative direction after the NaI was added into the electrolyte. The results showed that the addition of both TMACl and NaI intensified the cathodic polarization, while the addition of TMACl could not, implying the polarization effect might be caused by NaI. Generally, appropriate polarization is necessary for electrodeposition as high overpotential is needed for the nucleation. The addition of TMACl and NaI intensifies the polarization and contributes significantly to the formation of better Al coatings.
Every cyclic voltammogram was recorded on a fresh glassy carbon electrode surface. It shows that the cathode deposition current density increased with the addition of TMACl, indicating that TMACl cannot give rise to an inhibition of Al deposition. However, as both NaI and TMACl were added into the electrolyte, the deposition current density decreased dramatically. The related experiments [31] proved that the addition of NaI decreased the cathodic current density and gave rise to an inhibition of Al deposition. A possible reason is that the introduction of two additives leads to plenty conductive ions in the molten salts, and the TMACl does not impair the effect of NaI. As shown in the inset of Figure 2, the reduction potential of cathodic was stable at approximately 0.01 V with or without the increase of TMACl, which indicates the addition of TMACl has no prominent effect on changing the reduction potential of cathodic. While, the reduction potential of cathodic shifted obviously in a negative direction after the NaI was added into the electrolyte. The results showed that the addition of both TMACl and NaI intensified the cathodic polarization, while the addition of TMACl could not, implying the polarization effect might be caused by NaI. Generally, appropriate polarization is necessary for electrodeposition as high overpotential is needed for the nucleation. The addition of TMACl and NaI intensifies the polarization and contributes significantly to the formation of better Al coatings.
Additionally, the addition of the additives made the anodic current density peak shift to the positive direction in different extent, the potential difference (∆EP) [33] between oxidation and reduction became increasingly large. This means that the two additives could aggravate the irreversibility in the electrodeposition system [34]. The physical properties of the electrolyte do not change with the addition of two additives, the SEM, EDXS and XRD characterizations proved that the two additives could not be introduced in Al deposit as the impurities.

Chronoamperometric Investigations
Chronoamperometric experiments were conducted on a glassy carbon electrode on purpose of further studying the effect of TMACl and NaI on Al nucleation and growth mechanism in AlCl3-NaCl-KCl molten salts electrolyte. The experiments started soon after the short induction times, and the potential started at the value where no reduction of Al(III) took place, and then the potential Figure 2. The cyclic voltammograms of the AlCl 3 -NaCl-KCl molten salts electrolyte without additive, 1% TMACl, 5% TMACl, 10% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI, at 150 • C.
Additionally, the addition of the additives made the anodic current density peak shift to the positive direction in different extent, the potential difference (∆E P ) [33] between oxidation and reduction became increasingly large. This means that the two additives could aggravate the irreversibility in the electrodeposition system [34]. The physical properties of the electrolyte do not change with the addition of two additives, the SEM, EDXS and XRD characterizations proved that the two additives could not be introduced in Al deposit as the impurities.

Chronoamperometric Investigations
Chronoamperometric experiments were conducted on a glassy carbon electrode on purpose of further studying the effect of TMACl and NaI on Al nucleation and growth mechanism in AlCl 3 -NaCl-KCl molten salts electrolyte. The experiments started soon after the short induction times, and the potential started at the value where no reduction of Al(III) took place, and then the potential increased gradually, to a more negative value which was much more negative than the nucleation and growth of Al particles occurred. The experiments were performed in the AlCl 3 -NaCl-KCl molten salts electrolyte with TMACl and NaI at different content densities, and four typical representative current-time transients are plotted in Figure 3.
Irrespective of the different content of additives, the shapes of the peaks in the potential step current density transients are closely related to nucleation/growth process. An obvious augment owing to the formation and growth of aluminum nuclei in each current curve can be easily observed. Additionally, the process of formation and growth made Al ions collect rapidly. As the overlap [35] activates, the increasing current density receives the maximum, i m , and the time the peak appears is the time, t m . After that, all of the current density transients tend to decrease because of the consumption and diffusion of Al 2 Cl 7 − . With the applied potential increases, the i m becomes higher, while the t m becomes shorter at the same time. It indicates that the time of overlapping decreased as the nucleation Materials 2020, 13, 5506 7 of 16 density rises. In addition, the current density transients are not going to converge to the same value, the stable values of current density increase as the potential applied increases. This proves that the electrodeposition of Al in AlCl 3 -NaCl-KCl molten salts electrolyte is the electrochemical reaction controlled. The reactions determine the main ions in the AlCl 3 -NaCl-KCl molten salts electrolyte are Al 2 Cl 7 − ions, and abundant Al 2 Cl 7 − ions count for the electrochemical reaction control. Diffusion control is usually observed in most ionic liquids due to the lack of reductive ions near the cathode electrode [36].
Materials 2020, 13, x FOR PEER REVIEW 7 of 16 increased gradually, to a more negative value which was much more negative than the nucleation and growth of Al particles occurred. The experiments were performed in the AlCl3-NaCl-KCl molten salts electrolyte with TMACl and NaI at different content densities, and four typical representative current-time transients are plotted in Figure 3. Irrespective of the different content of additives, the shapes of the peaks in the potential step current density transients are closely related to nucleation/growth process. An obvious augment owing to the formation and growth of aluminum nuclei in each current curve can be easily observed. Additionally, the process of formation and growth made Al ions collect rapidly. As the overlap [35] activates, the increasing current density receives the maximum, im, and the time the peak appears is the time, tm. After that, all of the current density transients tend to decrease because of the consumption and diffusion of Al2Cl7 − . With the applied potential increases, the im becomes higher, while the tm becomes shorter at the same time. It indicates that the time of overlapping decreased as the nucleation density rises. In addition, the current density transients are not going to converge to the same value, the stable values of current density increase as the potential applied increases. This proves that the electrodeposition of Al in AlCl3-NaCl-KCl molten salts electrolyte is the electrochemical reaction controlled. The reactions determine the main ions in the AlCl3-NaCl-KCl molten salts electrolyte are Al2Cl7 − ions, and abundant Al2Cl7 − ions count for the electrochemical reaction control. Diffusion control is usually observed in most ionic liquids due to the lack of reductive ions near the cathode electrode [36].
As the different content of additives were taken into consideration, it was found that the addition of TMACl and NaI made the potential at which the reduction of Al 3+ started shift to negative direction. On the contrary, the addition of TMACl did not work effectively. The result illustrated that the NaI could increase the overpotential of electrodeposition and promotes the cathodic polarization, but TMACl could not, which agreed well with those of the above voltammograms. I − ions ionized from NaI might act as some surfactant and the adsorption of I − on the surface of the electrode benefits nucleation and inhibits reduction of Al 3+ [34]. It also revealed that additives produced an increase in response current and this effect increases with the concentration of the additive, whether TMACl or NaI. It is probably because the addition of the two additives increases the number of charged particles As the different content of additives were taken into consideration, it was found that the addition of TMACl and NaI made the potential at which the reduction of Al 3+ started shift to negative direction. On the contrary, the addition of TMACl did not work effectively. The result illustrated that the NaI could increase the overpotential of electrodeposition and promotes the cathodic polarization, but TMACl could not, which agreed well with those of the above voltammograms. I − ions ionized from NaI might act as some surfactant and the adsorption of I − on the surface of the electrode benefits nucleation and inhibits reduction of Al 3+ [34]. It also revealed that additives produced an increase in response current and this effect increases with the concentration of the additive, whether TMACl or NaI. It is probably because the addition of the two additives increases the number of charged particles and activity of the electrolyte, so that the electrical conductivity increases and the response current rises. I − ions would impede active sites on the surface of cathode, increasing the number of growing nuclei. The I − ions inhibit the growth of nuclei, which lead to an energetic homogenization of the Al coatings. While, the Cl − ions and ammonium ions ionized from TMACl could not, which implies TMACl is not a suitable additive in the AlCl 3 -NaCl-KCl molten salts electrolyte. The combined addition of TMACl or NaI would help improve the Al coatings.
During the electrodeposition of metals, the three-dimensional nucleation and hemispherical growth of the initial nuclei usually occur at the same time. A total of two different models are used to analyze three-dimensional nucleation/growth qualitatively, namely 'instantaneous' or 'progressive' [37]. The instantaneous model depicts the status that all of the nucleation sites are activated simultaneously.
In contrast, the progressive model depicts the status that the nucleation sites are gradually activated in the whole process of chronoamperometric experiment. The comparison of the experimental transients and the theoretical transients is adopted for the identification of nucleation models. The theoretical transients for instantaneous and progressive nucleation equations are listed as Equations (2) and (3), respectively: where i represents the current density, t represents the time and i m represents the maximum of the current density at t m time. To determine effect of TMACl and NaI on the nucleation/growth mechanism, the data obtained from the chronoamperometric tests were normalized and compared with the theoretical transients. Figure 4 depicts the experimental and theoretical plots of (i/i m ) 2 vs (t/t m ) with different composition of additives.
coatings. While, the Cl − ions and ammonium ions ionized from TMACl could not, which implies TMACl is not a suitable additive in the AlCl3-NaCl-KCl molten salts electrolyte. The combined addition of TMACl or NaI would help improve the Al coatings.
During the electrodeposition of metals, the three-dimensional nucleation and hemispherical growth of the initial nuclei usually occur at the same time. A total of two different models are used to analyze three-dimensional nucleation/growth qualitatively, namely 'instantaneous' or 'progressive' [37]. The instantaneous model depicts the status that all of the nucleation sites are activated simultaneously. In contrast, the progressive model depicts the status that the nucleation sites are gradually activated in the whole process of chronoamperometric experiment. The comparison of the experimental transients and the theoretical transients is adopted for the identification of nucleation models. The theoretical transients for instantaneous and progressive nucleation equations are listed as Equations (2) and (3), respectively: where i represents the current density, t represents the time and im represents the maximum of the current density at tm time. To determine effect of TMACl and NaI on the nucleation/growth mechanism, the data obtained from the chronoamperometric tests were normalized and compared with the theoretical transients. Figure 4 depicts the experimental and theoretical plots of (i/im) 2 vs (t/tm) with different composition of additives. It is obvious that the experimental curves are mostly in accordance with the instantaneous nucleation model, which suggests that the process of Al electrodeposition could be referred to instantaneous nucleation. The addition of two additives does not affect the mechanism of Al nucleation and growth, indicating that the process of nucleation/growth exhibits irrelevance with TMACl or NaI.

Morphology of Al Coatings
Al deposited samples obtained from AlCl 3 -NaCl-KCl molten salts electrolyte with varied concentration of TMACl and NaI at 150 • C were examined by SEM. The typical SEM images of the surface of Al coatings are shown in Figure 5.
It is obvious that the experimental curves are mostly in accordance with the instantaneous nucleation model, which suggests that the process of Al electrodeposition could be referred to instantaneous nucleation. The addition of two additives does not affect the mechanism of Al nucleation and growth, indicating that the process of nucleation/growth exhibits irrelevance with TMACl or NaI.

Morphology of Al Coatings
Al deposited samples obtained from AlCl3-NaCl-KCl molten salts electrolyte with varied concentration of TMACl and NaI at 150 °C were examined by SEM. The typical SEM images of the surface of Al coatings are shown in Figure 5. As no additive was introduced, shown in Figure 5a, the surface of the Al coating was still rough, and the films were nonuniform-the average particle size was about 7 ± 3 μm. When 1 wt.% TMACl was added into the electrolyte, as shown in Figure 5b, the grains became more uniform, the average particle size became smaller, the Al coating became dense. Extra experiments proved that the addition of more TMACl at 5%, 10%, and even 15%, had a similar effect on the Al coatings, but the morphology and quality of the Al deposit did not change much. This means that more TMACl could not magnify the effect which serves as an effective additive in molten salts electrolyte or improve the quality of Al coatings effectively. However, when TMACl and NaI were both introduced, whether 1% TMACl and 5% NaI or 1% TMACl and 10% NaI (Figure 5c,d), the Al particles became smaller and homogeneous, the surface of Al coating became flat and consecutive, and the size of Al particles appeared uniform. The Al particles grew neatly, and a compact and uniform electrodeposition Al coating was obtained, where the average particle size was approximately 2 ± 1 μm. The dendritic crystal and spongy are prevented effectively, indicating that NaI is helpful to refine the particle size. On account of the experiment results of cyclic voltammetry in Section 3.2, it is reasonable to speculate As no additive was introduced, shown in Figure 5a, the surface of the Al coating was still rough, and the films were nonuniform-the average particle size was about 7 ± 3 µm. When 1 wt.% TMACl was added into the electrolyte, as shown in Figure 5b, the grains became more uniform, the average particle size became smaller, the Al coating became dense. Extra experiments proved that the addition of more TMACl at 5%, 10%, and even 15%, had a similar effect on the Al coatings, but the morphology and quality of the Al deposit did not change much. This means that more TMACl could not magnify the effect which serves as an effective additive in molten salts electrolyte or improve the quality of Al coatings effectively. However, when TMACl and NaI were both introduced, whether 1% TMACl and 5% NaI or 1% TMACl and 10% NaI (Figure 5c,d), the Al particles became smaller and homogeneous, the surface of Al coating became flat and consecutive, and the size of Al particles appeared uniform. The Al particles grew neatly, and a compact and uniform electrodeposition Al coating was obtained, where the average particle size was approximately 2 ± 1 µm. The dendritic crystal and spongy are prevented effectively, indicating that NaI is helpful to refine the particle size. On account of the experiment results of cyclic voltammetry in Section 3.2, it is reasonable to speculate that the the addition of TMACl and NaI, especially the addition of NaI, could cause electrochemical polarization. A similar trend was also found in the electrodeposition of Al with ionic liquid electrolytes containing similar ammonium halides as additives [38]. While, there were a few tiny pores on the surface of the Al coating. The gap between Al particles became narrower after TMACl and NaI were added, and finally appeared in the form of voids on the coating surface.
In order to determine the elementary composition of the Al deposits, the EDXS tests were performed. The typical EDXS spectrums of the Al electrodeposits from AlCl 3 -NaCl-KCl molten salts electrolyte with additives are presented in Figure 6. The strong peak of aluminum can be clearly observed, indicating the main composition of the coating is aluminum. The tiny peak of K was also detected since there could probably be a little KCl residue from the electrolyte on the surface of Al coatings. No peak of oxygen or other element was detected after the addition of TMACl, or both TMACl and NaI (Figure 6a,b). The EDXS results proved that the electrodeposited Al coatings consisted of pure Al and were clear of the TMACl or NaI, confirming that the electrodeposition Al coatings obtained with additives were of high quality. A layer of high-purity Al coating can be obtained from AlCl 3 -NaCl-KCl molten salts electrolyte with TMACl and NaI added as additive. polarization. A similar trend was also found in the electrodeposition of Al with ionic liquid electrolytes containing similar ammonium halides as additives [38]. While, there were a few tiny pores on the surface of the Al coating. The gap between Al particles became narrower after TMACl and NaI were added, and finally appeared in the form of voids on the coating surface.
In order to determine the elementary composition of the Al deposits, the EDXS tests were performed. The typical EDXS spectrums of the Al electrodeposits from AlCl3-NaCl-KCl molten salts electrolyte with additives are presented in Figure 6. The strong peak of aluminum can be clearly observed, indicating the main composition of the coating is aluminum. The tiny peak of K was also detected since there could probably be a little KCl residue from the electrolyte on the surface of Al coatings. No peak of oxygen or other element was detected after the addition of TMACl, or both TMACl and NaI (Figure 6a,b). The EDXS results proved that the electrodeposited Al coatings consisted of pure Al and were clear of the TMACl or NaI, confirming that the electrodeposition Al coatings obtained with additives were of high quality. A layer of high-purity Al coating can be obtained from AlCl3-NaCl-KCl molten salts electrolyte with TMACl and NaI added as additive. The cross-sectional morphology of the Al coating obtained from AlCl3-NaCl-KCl molten salts with no additives, 1% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI are presented in Figure  7. It shows in Figure 7a that the thickness of the electrodeposition Al coating obtained from the electrolyte without additive varies from 2.5 to 10 μm, and the average thickness of Al electrodeposits obtained was 6 ± 4 μm. The nonuniformity of the thickness means the Al coatings are not dense and compact. In contrast, after 1% TMACl + 10% NaI added into the electrolyte, the average thickness of the achieved Al coating is approximately 7 ± 2 μm. There is no gap between Al coatings and the Cu substrate indicating the electrodeposited Al coating and the Cu substrate adhere tightly. The Faraday's law was used to calculate the cathodic current efficiency based on the mass of the electrodeposited Al, as 1% TMACl and 10% NaI were added into the electrolyte, Al coatings with a current efficiency of 99% was obtained. The cross-sectional morphology of the Al coating obtained from AlCl 3 -NaCl-KCl molten salts with no additives, 1% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI are presented in Figure 7. It shows in Figure 7a that the thickness of the electrodeposition Al coating obtained from the electrolyte without additive varies from 2.5 to 10 µm, and the average thickness of Al electrodeposits obtained was 6 ± 4 µm. The nonuniformity of the thickness means the Al coatings are not dense and compact. In contrast, after 1% TMACl + 10% NaI added into the electrolyte, the average thickness of the achieved Al coating is approximately 7 ± 2 µm. There is no gap between Al coatings and the Cu substrate indicating the electrodeposited Al coating and the Cu substrate adhere tightly. The Faraday's law was used to calculate the cathodic current efficiency based on the mass of the electrodeposited Al, as 1% TMACl and 10% NaI were added into the electrolyte, Al coatings with a current efficiency of 99% was obtained. To further characterize the microstructure of the Al coating, XRD analysis was performed and Figure 8 shows the XRD patterns of the electrodeposited Al coatings. All the strong diffraction peaks in the XRD patterns can be referred to the Al coatings and the Cu substrate. In total, four obvious diffraction peaks, (111), (200), (220), and (311), can be found from the obtained aluminum deposits. It means that the Al coatings are constituted of fcc phase of Al metal. In order to study the effect of the To further characterize the microstructure of the Al coating, XRD analysis was performed and Figure 8 shows the XRD patterns of the electrodeposited Al coatings. All the strong diffraction peaks in the XRD patterns can be referred to the Al coatings and the Cu substrate. In total, four obvious diffraction peaks, (111), (200), (220), and (311), can be found from the obtained aluminum deposits. It means that the Al coatings are constituted of fcc phase of Al metal. In order to study the effect of the two additives on the orientation and grain size of Al coatings, the texture coefficient (TC) and the grain size were adopted and calculated via XRD reflections [39], and the results are listed in Table 1. As the additive was added, the grain size of the Al decreased significantly, which is in accordance with the variation tendency of particle size. The addition of TAMCl and NaI could contribute to the formation of a preferred crystallographic orientation along (220) plane and refine the grain size effectively. Figure 7. The cross-sectional morphology of the Al coating deposited from the AlCl3-NaCl-KCl molten salts electrolyte with different additives, (a) no additives, (b) 1% TMACl, (c) 1% TMACl + 5% NaI and (d) 1% TMACl + 10% NaI.
To further characterize the microstructure of the Al coating, XRD analysis was performed and Figure 8 shows the XRD patterns of the electrodeposited Al coatings. All the strong diffraction peaks in the XRD patterns can be referred to the Al coatings and the Cu substrate. In total, four obvious diffraction peaks, (111), (200), (220), and (311), can be found from the obtained aluminum deposits. It means that the Al coatings are constituted of fcc phase of Al metal. In order to study the effect of the two additives on the orientation and grain size of Al coatings, the texture coefficient (TC) and the grain size were adopted and calculated via XRD reflections [39], and the results are listed in Table 1. As the additive was added, the grain size of the Al decreased significantly, which is in accordance with the variation tendency of particle size. The addition of TAMCl and NaI could contribute to the formation of a preferred crystallographic orientation along (220) plane and refine the grain size effectively. Figure 8. The XRD patterns of the electrodeposition Al coating obtained on a copper foil in AlCl3-NaCl-KCl (80-10-10 wt.%) molten salts electrolyte, without additive, 1% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI, at 150 °C.  Figure 8. The XRD patterns of the electrodeposition Al coating obtained on a copper foil in AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salts electrolyte, without additive, 1% TMACl, 1% TMACl + 5% NaI and 1% TMACl + 10% NaI, at 150 • C.

Corrosion Behaviors
The potentiodynamic polarization was performed so as to test the corrosion resistance of Al coatings obtained from AlCl 3 -NaCl-KCl molten salts electrolyte with varied concentration of additives. A cast pure Al plate was also tested as a parallel control experiment. Figure 9 presents a series of typical potentiodynamic polarization curves of the electrodeposited Al coatings with different additives, and other detail parameters calculated by the Tafel extrapolation are shown in the Table 2. The potentiodynamic polarization was performed so as to test the corrosion resistance of Al coatings obtained from AlCl3-NaCl-KCl molten salts electrolyte with varied concentration of additives. A cast pure Al plate was also tested as a parallel control experiment. Figure 9 presents a series of typical potentiodynamic polarization curves of the electrodeposited Al coatings with different additives, and other detail parameters calculated by the Tafel extrapolation are shown in the Table 2. It can be seen that the Al coatings electrodeposited from AlCl3-NaCl-KCl molten salts electrolyte without additive present the poor corrosion resistance with the highest icorr of 40.161 μA. After 1% TMACl was added in to the electrolyte, the potentiodynamic polarization parameters did not change much. It is reasonable, since the Al coatings were not dense and compact enough, the loose Al particles made the coatings easily corroded. Further, this also gives an explanation as to why there is no passive region found in these two curves. In contrast, as the 1% TMACl + 10% NaI was added, the corrosion current icorr of the electrodeposited Al coatings declined dramatically to 3.625 μA, which is quite close to that of pure Al. A conspicuous passive region was observed in the potentiodynamic polarization curve, indicating that the Al coating exhibits a great corrosion resistance. It also proved that the addition of TMACl could not affect the quality of Al coatings, while the addition of TMACl and NaI could improve the Al coatings prominently. The EIS tests were performed at the OCP to further evaluate the corrosion resistance of the electrodeposition Al coatings obtained from AlCl3-NaCl-KCl molten salts electrolyte with additive and the results of are presented in Figure 10. It is shown that the shapes of Nyquist plots of pure Al, Al It can be seen that the Al coatings electrodeposited from AlCl 3 -NaCl-KCl molten salts electrolyte without additive present the poor corrosion resistance with the highest i corr of 40.161 µA. After 1% TMACl was added in to the electrolyte, the potentiodynamic polarization parameters did not change much. It is reasonable, since the Al coatings were not dense and compact enough, the loose Al particles made the coatings easily corroded. Further, this also gives an explanation as to why there is no passive region found in these two curves. In contrast, as the 1% TMACl + 10% NaI was added, the corrosion current i corr of the electrodeposited Al coatings declined dramatically to 3.625 µA, which is quite close to that of pure Al. A conspicuous passive region was observed in the potentiodynamic polarization curve, indicating that the Al coating exhibits a great corrosion resistance. It also proved that the addition of TMACl could not affect the quality of Al coatings, while the addition of TMACl and NaI could improve the Al coatings prominently.
The EIS tests were performed at the OCP to further evaluate the corrosion resistance of the electrodeposition Al coatings obtained from AlCl 3 -NaCl-KCl molten salts electrolyte with additive and the results of are presented in Figure 10. It is shown that the shapes of Nyquist plots of pure Al, Al coating electrodeposited without additive, with 1% TMACl and with 1% TMACl + 10% NaI were very similar, revealing that the corrosion process is mainly controlled by the charge-transfer [40]. In the Nyquist plots, the larger semicircular arc diameter means the greater resistance. Bode and Bode-phase results have proved that the Al coating achieved with 1% TMACl + 10% NaI presented the highest /Z/. Hence, the Al coating electrodeposited from AlCl 3 -NaCl-KCl molten salts electrolyte with 1% TMACl + 10% NaI exhibit a great corrosion resistance.
The equivalent circuit in accordance with the Nyquist curves is shown in Figure 10d. R L is the resistance of the AlCl 3 -NaCl-KCl molten salts electrolyte at a high frequency limit. R 1 is the polarization resistance related to the charge transfer, and Z determines the double layer capacitance. The simulation values obtained by Z-view software for the equivalent elements are shown in Table 3. Generally, a higher R 1 reveals a lower corrosion resistance. According to Table 3, it is obvious that the Al coating electrodeposited from AlCl 3 -NaCl-KCl molten salts electrolyte with 1% TMACl + 10% NaI presented the best corrosion resistance.
coating electrodeposited without additive, with 1% TMACl and with 1% TMACl + 10% NaI were very similar, revealing that the corrosion process is mainly controlled by the charge-transfer [40]. In the Nyquist plots, the larger semicircular arc diameter means the greater resistance. Bode and Bode-phase results have proved that the Al coating achieved with 1% TMACl + 10% NaI presented the highest /Z/. Hence, the Al coating electrodeposited from AlCl3-NaCl-KCl molten salts electrolyte with 1% TMACl + 10% NaI exhibit a great corrosion resistance. The equivalent circuit in accordance with the Nyquist curves is shown in Figure 10d. RL is the resistance of the AlCl3-NaCl-KCl molten salts electrolyte at a high frequency limit. R1 is the polarization resistance related to the charge transfer, and Z determines the double layer capacitance. The simulation values obtained by Z-view software for the equivalent elements are shown in Table 3. Generally, a higher R1 reveals a lower corrosion resistance. According to Table 3, it is obvious that the Al coating electrodeposited from AlCl3-NaCl-KCl molten salts electrolyte with 1% TMACl + 10% NaI presented the best corrosion resistance.

Conclusions
In order to effectively reduce the volatility of the AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salts electrolyte, increase electrodeposition efficiency and improve the quality of electrodeposition Al coatings, two different additives, TMACl and NaI were introduced and investigated. The addition of TMACl could inhibit the volatilization of AlCl 3 , extending the service life of electrolyte. High quality Al coatings were obtained by electrodeposition from AlCl 3 -NaCl-KCl molten salts electrolytes at 150 • C with varied concentration of NaI and TMACl. A uniform, compact and smooth layer of electrodeposited Al films can be obtained by adding 1% TMACl + 5% NaI and 1% TMACl + 10% NaI. The electrodeposited Al coatings achieved great corrosion resistance, close to that of pure Al plate, with a corrosion current of 3.625 µA. The average particle size of the Al deposits was approximately 2 ± 1 µm. The average thickness of the electrodeposition Al coatings was approximately 7 ± 2 µm. The electrodeposition of Al on a Cu electrode in AlCl 3 -NaCl-KCl (80-10-10 wt.%) molten salts electrolyte proceeded via a three-dimensional instantaneous nucleation-the process of nucleation and growth exhibits irrelevance with NaI or TMACl. The addition of TMACl could not affect and improve the electrodeposition effectively. However, the addition of TMACl and NaI could intensify the cathodic polarization and inhibit the electrodeposition of Al. The combined effort of the two additives could increase the conductivity and facilitate to refine the particle size, contributing to the formation of a continuous, homogeneous and uniform Al coatings. They can be used as effective additives in molten salts electrolytes.