Electrodeposition of Aluminum in the 1-Ethyl-3-Methylimidazolium Tetrachloroaluminate Ionic Liquid

: The electrodeposition of Al was investigated in an ionic liquid (IL), with 1-ethyl-3-me-thylimidazolium tetrachloroaluminate ([EMIm]AlCl 4 ) as the electrolyte with AlCl 3 precursor. The [EMIm]AlCl 4 electrolyte exhibited a wide and stable electrochemical window from 3.2 to 2.3 V on a glassy carbon electrode when temperature was increased from 30 °C to 110 °C. The addition of AlCl 3 into [EMIm]AlCl 4 generated significant well-developed nucleation growth loops, and new coupled reduction and oxidation peaks in cyclic voltammograms corresponding to the Al deposition and dissolution, respectively. A calculation model was proposed predicting compositions of anions in AlCl 3 /[EMIm]AlCl 4 system, and [Al 2 Cl 7 ] − was found to be the active species for Al deposition. In AlCl 3 /[EMIm]AlCl 4 (1:5), the reduction rate constants were 1.18 × 10 −5 cm s −1 and 3.37 × 10 − 4 cm s − 1 at 30 °C and 110 °C, respectively. Scanning electron microscope (SEM), energy dispersive spectroscope (EDS), and X-ray diffraction (XRD) microscope results showed that the metallic Al film had been successfully deposited on glassy carbon electrodes through constant-potential cathodic reductions. The [EMIm]AlCl 4 was a promising electrolyte directly used for Al deposition.

The AlCl3-to-IL ratio determines the distribution of various Al anion species and viscosity of the system, which controls reaction kinetics and diffusion process for Al deposition. The AlCl3/[BMIm]Cl mixtures with 2:1 molar ratio resulted in a dull Al deposit whereas the 1.5:1 molar ratio mixture led to a bright Al surface because the AlCl3/[BMIm]Cl at 1.5:1 is more viscous with less aluminum complex ions [4]. Al deposition is the direct result from electrochemical reduction of chloroaluminate complexes [14,15].
In general, the chloroacidity determines reactivity and electrochemistry in the ionic liquid electrolyte. The Al deposition chemistry is complicated because of the chemical nature of different chloroaluminate complexes species including chemical equilibria and interconversion [16,17]. The equilibrium composition was highly dependent on the amount of AlCl3 added to the ionic liquid. The main anions are [AlCl4] − , [Al2Cl7] − , and Cl − in the alkylimidazolium chloride and AlCl3 solutions [18][19][20]. When AlCl3 concentration is high, more complicated anions such as [Al3Cl10] − , [Al4Cl13] − , and [Al2OCl5] − will be formed, which are confirmed by Infrared (IR) spectroscopy and the Nuclear Magnetic Resonance (NMR) spectroscopy [21].
[EMIm]AlCl4 can simplify the preparation process for electrochemical deposition with better control on mixing and heating. Furthermore, it is commercially available on industrial scale and makes Al deposition possible directly at lower temperatures. However, there is only some computational study available for exploring the unique physical properties of the tetrachloroaluminate systems recently [38,39]. However, due to the sensitivity of moisture and oxygen in the experimental performance in ionic liquids, the operations need to be conducted in protective atmosphere such as an inert gas environment. It makes the promotion and scale-up of the process difficult in larger scales [3,30]. The electrochemical characteristics of these systems have not received attention and their potential as a medium for the Al deposition has not been examined.
In this work, the electrochemical window of the [EMIm]AlCl4 will be defined and thermodynamic models for AlCl3/IL systems at 30 °C and 110 °C will be developed. By employing AlCl3/[EMIm]AlCl4 as the electrolyte, electrodepositions will be carried out on glass carbon substrates in the open air, and the products will be characterized for the morphology, composition, and phase information. These results will be of academic and industrial interest in not only the development of metal deposition processes, but also employing the electrochemistry of the [EMIm]AlCl4 as a model system to deepen our understanding of the related process chemistry and electrode reactions.
[EMIm]AlCl4 is a light-yellow transparent liquid at room temperature. All chemicals were used as received. Glassy carbon disk electrodes were used as working electrode (GC, 1 mm diameter, eDAQ, Colorado Spring, CO, USA). Aluminum wires were used as counterelectrode and reference electrode (1 mm diameter, 99.9995% metal basis, Alfa Aesar). Techne Dri-Block ® Digital Block Heater was used for temperature control (±1 °C). A 3 mL glass vial (eDAQ Inc. Colorado Spring, CO, USA) was used for electrodeposition cell. VersaSTAT 4 Potentiostat (Princeton Applied Research, Oak Ridge, TN, USA) with VersaStudio was used for all electrochemical measurements and data acquisition.

Procedure and Methodology
To determine the stability of [EMIm]AlCl4, cyclic voltammetry (CV) was measured for [EMIm]AlCl4 in the three-electrode electrochemical cell at a series of temperatures including 30 °C, 50 °C, 70 °C, 90 °C, and 110 °C at a scan rate 100 mV s −1 .
AlCl3/[EMIm]AlCl4 mixtures were prepared by adding portion of AlCl3 powders into the [EMIm]AlCl4 liquid at the room temperature and heated up to 110 °C for 2 h until homogeneity was reached. As the electrical conductivity for AlCl3/IL (exclusive of Al species) started to decrease when the ratio was higher than 1 [25], a small amount of AlCl3 was added and the final ionic liquid mixture was AlCl3/[EMIm]AlCl4 at a molar ratio of 1:5. In order to determine the Al 3+ reduction potential, CVs in AlCl3/[EMIm]AlCl4 (1:5) were measured at various scan rates including 10, 20, 50, 100, and 150 mV s −1 at 30 °C and 110 °C, respectively. The nucleation mechanism was studied by short-time chronoamperometry (CA) measurements. The current-time transients were measured by applying a series of constant cathodic potentials in the kinetic regime. After each current-time transient measurement, a constant potential 1.0 V was applied on the glassy carbon disk working electrode for 5 min to clean the electrode surface and remove the Al layer on the working electrode. All potentials were reported versus the Al reference electrode.
The Al electrodepositions in AlCl3/[EMIm]AlCl4 (1:5) were carried out by long-time CA measurements. The constant potential depositions were controlled a little higher than the Al/Al 3+ potential. A polished and clean Al wire working electrode was used for each deposition. The Al deposition samples were washed thoroughly in acetonitrile first and then acetone. The samples were left in air at room temperature for 24 h before any material characterization.

Chloroaluminate Complexes Distribution Calculation
(1) The total mole fraction for all species was unity: where was the modified Temkin ion fraction [40] for the species i, = , and was the total molar amount of anions: In addition, the charge balance also met the equation: By assuming a complete dissociation of the IL, it gave [ ] + = . The materials balance for Al species was 3 was introduced as the initial molar ratio of AlCl3: Combining the mass balance equations and charge balance equation, the following equation was established: Equations (1)-(8) were solved simultaneously using numerical method within EX-CEL for 3 in the range of 0 and 1 at interval 0.0001.

Stability Window of [EMIm]AlCl4 at Various Temperatures
CVs in [EMIm]AlCl4 at different temperatures were measured to determine the stability window (Figure 1). The CVs displayed wide flat zones between the fast-growing oxidation and reduction currents. The onset potentials for oxidation-current growth were very close to 1.5 V which is the oxidation of chloride anions [41,42]. In contrast, the onset potentials for the reduction-current growth were strongly dependent on temperatures, characteristic of a positive shift with increasing temperatures. At the negative potential limits, the cathodic peaks were attributed to the reduction of imidazolium ring [41]. Based on the onset potential differences for the oxidation and reduction current growth, the electrochemical windows of [EMIm]AlCl4 on the GC electrode were determined. The potential windows decreased from 3.2 to 2.3 V as the temperature was increased from 30 °C to 110 °C. These results are consistent with literature data measured under similar conditions [41,43]. No oxidation peak was observed in the electrochemical windows in the CVs indicating that the Al anion [AlCl4] − dissociated from [EMIm]AlCl4 was a very stable Lewis neutral anion and could not lead to Al electrodeposition [44][45][46].     Figure 2a,b, respectively. At 30 °C, the onset potentials for Al deposition became higher from −0.25 V to −0.3 V as the scan rates increased from 10 mV s −1 to 150 mV s −1 . At 110 °C, the onset potentials for Al deposition at different scan rates were concentrated at approximately −0.14 V, much lower than those in CVs at 30 °C. The potential range for CVs at 30 °C was from OCP to −0.9 V, and the current intensity at 150 mV s −1 was responded at 12.5 to −15.2 mA cm −2 , the highest current range. For CVs at the same scan rate 150 mV s −1 at 110 °C, the current response was much higher in a range of −63.8 to 87.2 mA cm −2 . At both temperatures, the re-oxidation charges were smaller than the reduction charges. At 30 °C, the Al dissolution was around 26.5% to 36.4% of the charge used in the reduction reaction, compared with the proportion of 61.8% to 79.2% in CVs at 110 °C.   The CVs in AlCl3/[EMIm]AlCl4 (1:5) ( Figure 2) at both temperatures exhibited the cathodic and anodic peaks potentials were more separated at higher scan rates. The separations of the Al reduction and re-oxidation peaks were more than 330 mV, which was larger than the critical peak splitting of 282 mV ( 212 mV, n is the number of electrons involved in the charge-transfer step, n = 0.75) [53,54], suggesting an irreversible system. For an irreversible electrode reaction, the rate constant (k 0 ) for the Al deposition was estimated from the shift of the cathodic peak potential (Epc) with the scan rates in CVs, based on Equation (10) where E 0 is the standard potential (E 0 = 0 for the current system), F is the Faraday's constant 96,485 C mol −1 , R is the gas constant 8.314 J mol −1 K −1 , D is the diffusion coefficient of active species [Al2Cl7] − , and α is the transfer coefficient. Figure 4a showed linear relations of Epc and lnv 1/2 . The values of could be derived from the slopes of lines at different temperatures, which were 0.38 and 0.74 for reactions at 30 °C and 110 °C, respectively. The diffusion coefficients were calculated from the Randles-Sevcik equation, based on the relationship between the cathodic peak current densities (jpc) correlating with scan rate (v) for an irreversible system in Equation (11)   The Al deposits and the nucleation mechanisms can be achieved from the chronoamperometry studies. Al deposition potentials at both 30 °C and 110 °C were chosen as the values achieved at 200 mV s −1 in CVs. In this way, it guaranteed the constant potentials in CA measurements were higher than the Al deposition potentials. The current-time profiles showed an initial slight decrease in current density followed by a sharp increase and gradual decrease to a plateau (Figure 5a,b). The initial decrease was due to the charging of the double layer and a decayed current during the nucleation process [50,57,58]. The current increase was the result of independent nucleus size growing and consequently the increase of total electroactive Al surface area. The following current decrease was because of the overlap of the diffusion zone leading to the formation of one diffusion layer [50]. As the applied potential being more significant, the current density peak became sharper with higher intensity and reached a higher plateau.
Initial stages of metal deposition were usually associated with a three-dimensional (3D) nucleation. For the diffusion controlled 3D nucleation, the instantaneous and progressive nucleation mechanisms were normally expressed in Equations (12) and (13), respectively [59].
where j is the current density at time t, t0 is the induction time, and tm is the time at the maximum current density jm. Figure 5c,d gives a graphic analysis of the Al nucleation mechanism with the data extracted from Figure 5a,b for the Al deposition onto the GC electrode at 30 °C and 110 °C, respectively. The theoretical model curves were generated from Equations (12) and (13). At 30 °C, the nucleation process correlated with the progressive nucleation mechanism in the initial stage. After reaching the maximum current, the deposition was close to the progressive nucleation curve, but the process may be affected by other factors. At 110 °C, the phenomenon was more complicated. It followed the progressive nucleation mechanism for a short time, but the data lines fell on progressive and instantaneous nucleation mechanisms, and in the region in between. This nucleation kinetic was different from literature results for the Al deposition from AlCl3/[EMIm]Cl which exhibited a better fit with the 3D instantaneous nucleation [3,9]. It is possibly the results of electrode surface changes during the depositions. Initially, the electrode surface was flat and small. Next, the surface gradually grew to a hemispheric shape with a large surface area. The changes happened in such a short time that the mass transfer and nucleation kinetics were complicated and could not be simply explained with only one mechanism. In addition, the existence of side reactions during the nucleation, such as the reduction of moisture and oxygen dissolved in the ionic liquid mixture, made the electroplating processes more complicated [60]. Al-contained species would precipitate due to the hydrogen extraction in water electrolysis and oxidized by oxygen generated. Charges were consumed not only by the Al deposition from [Al4Cl7] − , but also by the side reaction discussed above. Therefore, the process resulted in a lower current efficiency and aluminum extraction ratio. Al electrodepositions were studied at 110 °C on the glassy carbon substrates. Constant potentials were applied at the Al deposition potential defined with CV measurements. Figure 6a,b shows the SEM images for two Al deposits under constant-potential polarization after the charge reached 2.9 C cm −2 and 14.5 C cm −2 , respectively. The deposit with less charges exhibited both bright and black regions. Further growth of the deposit layer with more charges lead to the complete coating of the substrate, accompanied by the formation of minor cracks. The elemental analysis (Figure 6c) of the bright zone and the dark zone in Figure 6a confirmed that the polyhedral particles were 100% Al, and the black zone corresponded to the GC substrate. The XRD pattern of the Al layer deposited in Figure 6b was shown in Figure 6d. The peaks were sharp and matched well with the is the shape factor at 0.9; is the X-ray wavelength at 1.5406 Å; β is the line broadening at half the maximum intensity in radians, 0.008378; and θ is the Bragg angle. The size of Al particles was calculated at 8.77 nm. Therefore, the result strongly supported the deposition of nano-size metallic Al during the cathode polarization.

Conclusions
In this work, we explored the potential of the [EMIm]AlCl4 as the ionic liquid electrolyte and AlCl3 as the precursor for the electrodeposition of Al. Because of its wide electrochemical window and low melting point, the [EMIm]AlCl4 is a prospective ionic liquid for the electrodeposition of Al. Thermodynamic models were established to show the composition of Al anions in AlCl3/IL mixtures at 30 °C and 110 °C. It was demonstrated that nano-sized Al was successfully deposited on glassy carbon after AlCl3 was added to the tetrachloroaluminate. The results from this work prove that the AlCl3/[EMIm]AlCl4 mixture is a promising electrodeposition both in developing the process chemistry and improving the control of Al coating.

Data Availability Statement:
No data is available for report.