Electrochemical and Chemical Behaviors of Titanium in AlCl3-BMIC Melt
2017 Volume 58 Issue 3 Pages 377-382
Details
2017 Volume 58 Issue 3 Pages 377-382
The electrochemical and chemical behaviors of titanium were examined in aluminum chloride-1-butyl-3-methylimidazolium chloride (AlCl3-BMIC) ionic liquid. The reduction of Ti(IV) in AlCl3-BMIC occurs in three consecutive steps: Ti(IV) → Ti(III) → Ti(II) → Ti, and Ti(III) ion can quickly react with Cl− anion to form sparingly soluble TiCl3. However, No elemental titanium can be obtained in either acidic or basic AlCl3-BMIC melt. Metal titanium can be oxidized to Ti(II) by Ti(IV) in AlCl3-BMIC melt. In addition, the anodic oxidation rate of titanium obviously increases in the presence of Ti(IV) and Ti(III) ions, suggesting that the anodic dissolution of titanium involves chemical dissolution. These results indicate that it is difficult to nucleate or stabilize pure titanium in ionic liquid, which makes pure titanium deposition quite difficult.
The electrodeposition of titanium is of great interest for making corrosion resistant layers on a variety of technically important materials. Owing to the higher reactivity of Ti metal, the relevant electrolyte solutions for the electrodeposition of titanium must be aprotic, and titanium electrodeposition is conventionally performed in high-temperature molten salts1,2) and organic solvents3). However, molten salts are highly corrosive and high energy consuming limited by high-temperature operation, and organic solvents are generally volatile and inflammable.
Room-temperature ionic liquids (RTILs) have received growing attention as non-aqueous solvents and been considered as alternative candidates for high-temperature molten salts and organic solvents in the application of active metal electrodeposition owing to their unique physical and chemical properties, such as chemical and thermal stability, low flammability, negligible vapor pressure, high ionic conductivity, and wide electrochemical window4–7). Recently, there were several attempts to electrodeposit titanium at room temperature in RTILs8–16). Mukhopadhyay et al.8,9) have tried to electrodeposit titanium at room temperature in 1-butyl-3-methylimidazolium bis(trifluoromethyl-sulfonyl)amide ([Bmim]Tf2N) ionic liquid containing TiCl4 as a source of titanium. There are some hints on the formation of ultra-thin titanium layers on the electrode surface. However, attempts to obtain titanium deposits with micrometer-thick in our own experience fail completely. Endres et al.10) have investigated the possibility to electrodeposit metal titanium from its halides (TiCl4, TiF4 and TiI4) in different ionic liquids, including 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide ([Emim]Tf2N), 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)amide ([BMP]Tf2N), and trihexyltetradecylphosphonium bis(trifluoromethylsulfonyl)amide ([P14,6,6,6]Tf2N). There is little evidence that elemental titanium can be electrodeposited.
Thermodynamically, Ti deposition should be quite simple in thick layers in ionic liquids, as its electrode potential is −1.63 V vs. SHE. Aluminum, which can be electrodeposited quite easily in different ionic liquids, has practically the same electrode potential for reduction (−1.67 V vs. SHE). What is more, titanium can easily be co-deposited with Al in aluminum chloride based ionic liquids. Tsuda et al.11,12) reported that Al-Ti alloys containing up to ～19% atomic fraction (a/o) titanium could be electrodeposited on copper electrodes from aluminum chloride-1-ethyl-3-methylimidazolium chloride (AlCl3-EMIC) ionic liquid containing TiCl2. And they found that the Al-Mo-Ti ternary alloys could also be electrodeposited in the Lewis acidic AlCl3-EMIC ionic liquid containing (Mo6Cl8)Cl4 and TiCl213). Freyland et al.14) discovered that a distinct two-dimensional nucleation of Al-Ti clusters began to occur at 0.5 V vs. Al/Al3+ exclusively on the Au terraces in an acidic AlCl3-BMIC ionic liquid electrolyte containing TiCl4. Reddy et al.15,16) reported the Ti-Al alloys containing about 15–27 atom% Ti were produced using TiCl4-AlCl3-BMIC (molar ratio 0.019:2:1) electrolytes at cell voltages ranging from 1.5 to 3.0 V and temperatures from 70 to 125℃. But so far, it remains unclear why elemental titanium deposition fails at low temperatures in ionic liquids, whereas clarity about the electrochemical behaviors of titanium has contributed to make it clear.
Herein, we investigated the electrochemical and chemical behaviors of titanium in AlCl3-BMIC ionic liquid.
Preparation of electrolyte, electrochemical measurements and electrodeposition experiments were conducted in an argon-filled glove box (H2O, O2 < 2 ppm). 1-butyl-3-methyl imidazolium chloride (BMIC) ionic liquid was synthesized in our laboratory and described in details elsewhere17). Anhydrous AlCl3 (99.9%, Beijing Chemical Reagents Company) was sublimed under vacuum to yield AlCl3 powders of several millimeters in diameter. TiCl4 (Sinopharm Chemical Reagent Co., Ltd. ≥98%) was used without further purification. The AlCl3 powders were slowly added to BMIC in the dry electrolytic cell with desired molar ratio, and the electrolyte was stirred continuously using a magnetic stirrer to ensure complete dissolution. When the molar ratio of AlCl3 to BMIC is 2:1 and 2:3, respectively, the acidic and basic AlCl3-BMIC ionic liquid was obtained for this experiment. Precise amount of liquid TiCl4 was carefully injected into the AlCl3-BMIC melt by using a glass syringe as TiCl4 is highly hygroscopic and vaporizes quickly.
All electrochemical measurements were performed in a three-electrode cell using a CHI760C Electrochemical Workstation. For experiments in ionic liquids, the working electrode was either a glassy carbon electrode (geometric area = 0.1256 cm2), a Teflon-sheathed glassy carbon rotating disk electrode (geometrical area = 0.099 cm2), or a small disk electrode prepared by sealing a titanium wire (99.98%, 1.0 mm in diameter) in a Pyrex glass tube and cutting off the end of the sealed tube so as to expose the cross section of the wire. The cutting off the end of the sealed tube and transferring of the titanium electrode to cell were carried out in an argon-filled glove box (H2O, O2 < 2 ppm). A graphite plate (Aldrich, SP, 5 mm × 10 mm) was used as the counter electrode and an aluminum wire (99.99%, 1.0 mm in diameter) was used as the reference electrode. This electrode was immersed in a melt with the same composition as the bulk melt and separated from the bulk melt by a porosity E glass frit. The aluminum electrodes were cleaned with a mixture of concentrated aqueous H2SO4-HNO3-H3PO4, rinsed with distilled H2O, and dried under vacuum before use. The glassy carbon working electrode was polished with 0.05 μm alumina, cleaned in an ultrasonic acetone bath, rinsed with distilled water and dried before each measurement.
The electrodeposition experiments were performed on copper sheets (99.99%, 10 mm × 5 mm). The copper electrode was polished successively with increasingly finer grades of emery paper, and finally to a mirror finish with aqueous slurry of 0.05 μm alumina, and rinsed with distilled water. The electrolyte was stirred at a constant speed and heated using a magnetic stirrer with heating (RET Basic). The deposition temperature of 353 K was measured using a thermometer.
The UV-visible spectra of dissolved Ti ions were obtained by using a Hitachi U-3900H spectrometer employing Wilmad no. 107-7 closed-type quartz cells. The path length of these cells was 0.10 cm. The elemental analysis of the deposit was carried out using a Perkin Elmer Optima 3000DV inductively coupled plasma atomic emission spectrometry (ICP-AES).
Ti(IV) was introduced into the AlCl3-BMIC ionic liquid by dissolution of TiCl4 to give bright yellow solutions. Figure 1 shows the typical cyclic voltammograms recorded on glassy carbon electrode in acidic AlCl3-BMIC ionic liquid before and after the addition of TiCl4. Before the addition of TiCl4, the only oxidation and reduction waves appear at potentials close to 0 V (vs. Al(III)/Al) as a result of the deposition and stripping of Al, according to the following well-known reaction4):
| \[4{\rm Al}_2 {\rm Cl}_7^- + 3{\rm e} \rightleftarrows {\rm Al} + 7{\rm AlCl}_4^-\] | (1) |
Cyclic voltammograms recorded at a glassy carbon electrode in acidic AlCl3-BMIC ionic liquid: (
) pure ionic liquid; (
) containing 18.2 mmol L−1 TiCl4. Inset: small scanning range of the CSV recorded in melt containing TiCl4. The sweep rate was 50 mV s−1 and the temperature was 353 K.
After the introduction of TiCl4, two additional reduction waves with the peak potentials, Epc, of about 1.01 V and 0.58 V are apparent. In addition, the wave ascribed to the reduction reaction in Reaction 1 shows a small negative shift, the corresponding oxidation wave is evident at more positive potentials. This stripping wave is attributed to the oxidation of electrodeposited Al-Ti alloy that are more stable toward oxidation than pure Al. Similar behavior was observed in AlCl3-NaCl18) and AlCl3-EMIC11,12) containing Ti(II) species.
The first cathodic wave with a peak potential at 1.01 V (vs Al(III)/Al) is attributed to the reduction of Ti(IV) to Ti(III). The violet TiCl3 precipitate was observed on the electrode surface by electrolysis at potential just negative of the first reduction peak. Besides, UV-visible absorption spectrum of this electrolyte is similar to that of Ti(IV) (see Fig. 2). Our previous investigations also show the reduction of TiCl4 to metallic Ti through a stepwise reduction to Ti(III) and Ti(II) species in ZnCl2-urea deep eutectic solvent19). However, it has been reported by Freyland et al.9) that Ti(IV) in [Bmin]Tf2N ionic liquid was reduced to Ti(II) by one two-electron step, and then Ti(II) reacted with Ti(IV) to form Ti(III). Thus, it needs to be further examined that the reduction process of Ti(IV) corresponds to a wave at 1.01 V. The detailed voltammetric wave at 1.01 V is shown in Fig. 3. The difference in the peak potential and half-peak potential (Epc - Epc/2) for this wave varied between 0.069 and 0.115 V and did not display any obvious dependence on the scan rate over the range extending from 0.005 to 0.500 V s−1. The smaller values of Epc - Epc/2, are reasonably close to the theoretical value of 0.067 V for a one-electron reversible reaction at 353 K, leading us to conclude that the reduction wave in Fig. 3 corresponds to the Ti (IV)/Ti(III) electrode reaction.
UV-visible spectra of the acidic AlCl3-BMIC melt containing 18.2 mmol L−1 TiCl4 (
) before and (
) after electrolysis at 1.0 V and (
) 0.58 V vs. Al(III)Al, (
) containing 18.2 mmol L−1 TiCl2.
Cyclic voltammograms recorded at a glassy carbon electrode in acidic AlCl3-BMIC ionic liquid containing 18.2 mmol L−1 TiCl4. The scan rates were 0.05 V s−1 (
) and 0.50 V s−1 (
).
The reduction of Ti(IV) was also examined as a function of scan rate at a fixed concentration. Figure 4 shows a plot of the peak reduction current (ipc) vs. the square root of the scan rate (v1/2). Each data point in this figure is the average of several measurements. The linearity of this plot indicates that the reduction of Ti(IV) is a diffusion-controlled process. In order to examine the reaction in detail, the peak current ratio (ipa/|ipc|) for this reduction wave and its corresponding oxidation wave was obtained by using Nicholson's semiempirical method20). This ratio is less than one at very slow scan rates but increases to one at faster scan rates (Fig. 4), suggesting that an irreversible homogeneous following chemical step is coupled to the electron-transfer reaction. This coupled chemical step must be due to the precipitation of TiCl3, as was found during the reduction of Ti(IV) in ZnCl2-urea solvent19) and AlCl3-NaCl melt21). Thus, the reduction wave at ca. 1.01 V involves the following sequence of reactions
| \[{\rm Ti}({\rm IV}) + {\rm e} \rightleftarrows {\rm Ti}({\rm III})\] | (2) |
| \[{\rm Ti}({\rm III}) + 6{\rm AlCl}_4^- = {\rm TiCl}_3 {\downarrow} {}+ 3{\rm Al}_2 {\rm Cl}_7^-\] | (3) |
Peak reduction current density and ratio of the anodic-to-cathodic peak currents as a function of sweep rate for a series of voltammograms similar to that shown in Fig. 3.
And also, experiments conducted at a fixed scan rate revealed that the voltammetric peak reduction current (ipc), for the wave at 1.01 V did not always display a linear dependence on the TiCl4 concentration. This behavior was most pronounced at high TiCl4 concentrations and slow scan rates, and it was due to the formation of TiCl3 film on the surface electrode during the reduction of Ti(IV). This TiCl3 film led to a decrease in the active area of the electrode and to many of the other anomalous results that were obtained for this voltammetric wave.
The second reduction wave at 0.58 V (vs Al(III)/Al) we assign to the reduction of Ti(III) to Ti(II). No metallic deposit was obtained except for TiCl3 precipitate after controlled-potential electrolysis at 0.58 V. Moreover, UV-visible absorption spectrum analysis of the electrolyte indicated that Ti(II) ions were presented in solution (Fig. 2). During the reverse scan, the second oxidation peak appears at 0.85 V, which is attributed to the oxidation of Ti(II) to Ti(III).
To eliminate the effect of Al(III) reduction, the cyclic voltammogram of Ti(IV) in basic AlCl3-BMIC melt was also recorded at glassy carbon electrode, as shown in Fig. 5. It is clearly seen from Fig. 5 that the cathodic branch of the cyclic voltammogram exhibits at least two different cathodic reduction processes, suggesting that the electrochemical reduction of Ti(IV) occurs in at least two steps. The first reduction peak is observed at a potential of about 0.89 V, the second much weaker peak is located at a potential of about 0.58 V. The first process at around 0.89 V is attributed to the reduction of titanium(IV) to titanium(III). The violet TiCl3 precipitate was observed on the electrode surface after electrolysis in this basic solution at 0.89 V (vs. Al(III)/Al) and UV-visible absorption spectrum of this solution is also similar to that of Ti(IV). The second process corresponds to the reduction of Ti(III) to a lower valence species. However, there are also no hints for the electrodeposition of elemental titanium down to −0.6 V (vs. Al(III)/Al).
Cyclic voltammograms recorded at a glassy carbon electrode in basic AlCl3-BMIC ionic liquid (
) before and (
) after the addition of TiCl4. The sweep rate was 50 mV s−1 and the temperature was 353 K.
Extended electrodeposition experiments were carried out for the TiCl4 in both acidic and basic AlCl3-BMIC. In the basic AlCl3-BMIC solution containing 18.2 mmol L−1 TiCl4, after electrodeposition for 1 h at different electrode potentials (−0.2 V, −0.4 V, −0.6 V and −0.8 V), there was no noticeable deposition on the surface of working electrode. The color of solution changed during this process, together with UV-Vis spectra analysis of solutions before and after the stages of reduction, indicated that the solution speciation of Ti ions had changed. This is very similar to those reported by F. Endres10) that it is impossible to deposit pure titanium from its halides in different ionic liquids. In the acidic AlCl3-BMIC solution containing 18.2 mmol L−1 TiCl4, after electrodeposition for 1 h at −0.2 V (vs. Al(III)/Al), the Al-Ti alloy deposits containing 22.1 atom% titanium was obtained. The X-ray diffraction (XRD) analysis of this Al-Ti alloy is shown in Fig. 6 and compared with the diffraction pattern of pure aluminum. Although the diffraction pattern of electrodeposited alloy is similar to pure aluminum, a peak shift is observed. The change in lattice parameter is 0.64% which confirms the formation of Al-Ti alloy. These two different results indicate that it is difficult to nucleate or stabilize titanium in the absence of other metals such as aluminum at low temperature.
X-ray diffraction patterns from the Al-Ti alloy electrodeposited at −0.2 V (vs. Al(III)/Al in acidic AlCl3-BMIC containing TiCl4: (a) pure Al deposited at −0.2 V in neat melt, (b) Al77.9Ti22.1.
The CV recorded for the solution containing Ti(IV) shown in Fig. 1 exhibits the possibility of the reduction of Ti(II) to metal titanium. Moreover, as we know, chloride-free deposits containing Ti can be obtained by the electrochemical reduction of Ti(IV) in acidic AlCl3-BMIC melt. This indicates that Ti(II), irrespective of its chemical or physical form, must eventually undergo reduction when the potential approaches 0 V vs. Al(III)/Al. Nevertheless, no pure Ti can be obtained in AlCl3-BMIC melt. Therefore, we carried out a series of experiments to determine if it is possible to remove Ti deposits by the oxidation reaction with higher valence titanium ions, such as Ti(III) and Ti(IV). There are some evidences for the existence of this chance. Tsuda et al have reported that Ti(IV) and Ti(III) can be reduced to Ti(II) by Al or Ti in acidic AlCl3-EMIC ionic liquid12) and AlCl3-NaCl-KCl melt22). The similar way was used to investigate the reaction between metallic Ti with TiCl4 in AlCl3-BMIC melt. In acidic AlCl3-BMIC ionic liquid, Ti(IV) exhibits bands at 287 nm (Fig. 7). The AlCl3-BMIC solution containing 18.2 mmol L−1 Ti(IV) was stirred with several small pieces of Ti metal at 353 K, then the metal Ti dissolved as well as the color of solution changed from bright yellow to yellow-green, and exhibited a spectrum identical to that for Ti(II) (see Fig. 7). Furthermore, the absorbance intensity of the 433 nm band in this spectrum is about two times of the data for 18.2 mmol L−1 Ti(II) solutions shown in Fig. 2, indicating complete conversion of the Ti(IV) to Ti(II) according to the following reaction:
| \[{\rm Ti}({\rm IV}) + {\rm Ti} \rightleftarrows 2{\rm Ti}({\rm II})\] | (4) |
UV-visible spectra of the acidic AlCl3-BMIC melt containing 18.2 mmol L−1 TiCl4 (
) before and (
) after reduction by Ti and (
) by electrodeposited Al-Ti alloy, (
) the basic AlCl3-BMIC melt containing 18.2 mmol L−1 TiCl4 after reduction by Ti.
However, when we repeated these experiments in acidic AlCl3-BMIC ionic liquid by reacting Ti(IV) with Al-Ti alloy deposit obtained from AlCl3-BMIC ionic liquid, Al-Ti alloy deposit did not dissolve and the solution color had little change. Meanwhile the spectrum of the solution after reduction is in consistence with that of Ti(IV) solution (Fig. 7). These results suggest that Ti metal can be easily oxidized and dissolved in AlCl3-BMIC melt containing Ti(IV) ions, while Al-Ti alloy can hardly be oxidized. Thus, it can be concluded that new electrodeposited titanium may easily be removed through the re-dissolution in the presence of high valence titanium ions, while formation of stable Al-Ti alloy could effectively prevent this re-dissolution behavior.
During the chemical reduction of Ti(IV) with Ti in the AlCl3-BMIC melt described previously, the rest potential of a GC-RDE electrode immersed in the solution shifted to a value considerably negative of the Ti(IV)/Ti(III) couple. Figure 8 shows the series of linear scanning voltammograms (LSVs) that were recorded at a GC-RDE at different rotation rates in two solutions that were prepared in acidic and basic melt by the chemical reduction of Ti(IV) with Ti metal. The concentrations of Ti(IV) used to prepare these two solutions were different, and the LSVs were initiated from the rest potential of each solution. Each of the LSVs in Fig. 8 is very similar in appearance to the sampled current voltammogram for Ti(II) reported by Fung and Mamantov23), where two well-defined waves were obtained, and plots of E vs. log[(i − i1)/i] were linear for both waves. Analysis of the intercepts of these plots gave half-wave potentials, E1/2, of 0.46 ± 0.01 and 1.21 ± 0.01 V for the first and second waves, respectively, in acidic AlCl3-BMIC melt, and E1/2 = 0.57 ± 0.01 and 1.07 ± 0.02 V, respectively, for the two waves in basic melt. Thus, E1/2 for these oxidation waves in these LSVs varies considerably with the melt composition. Calculations based on the slopes of these plots yielded n = 0.9 ± 0.1 and 1.0 ± 0.2 for the first and second waves, respectively, indicating that both waves correspond to one-electron reactions. Overall, the LSVs in Fig. 8 are very similar in appearance to those recorded during the oxidation of Ti(II) in acidic AlCl3-EMIC melt12). Taken together, these observations provide good evidence that the species produced during the chemical reduction of Ti(IV) with Ti is Ti(II), and Ti(II) was oxidized through two one-electron steps. The plots of the current density vs. the square root of the electrode rotation rate (Levich plots), for the first oxidation waves were linear and passed through the origins of the plots (see insets of Fig. 8). The diffusion coefficient for Ti(II), DTi(II), in the acidic AlCl3-BMIC melt was estimated from the slopes of Levich plots. Besides, uncertainty about the Ti(II) concentration precluded such calculations for solutions of Ti(II) in the basic melt. The viscosity data reported by Fannin et al.24) were used for these calculations. The estimated value of DTi(II) is 1.32 × 10−7 cm2 s−1, which is in reasonable agreement with those for DTi(II) in the same molten salt.
LSVs recorded at a GC-RDE in the acidic (a) and basic (b) AlCl3-BMIC melt containing Ti(II) ions obtained by the reduction of Ti(IV) with Ti. The sweep rate was 10 mV s−1; the rotation rates were 1000, 1250, 1500, 1750, and 2000 rpm; and the step size was 2 mV. (Inset) Relationship between the limiting current densities and the square root of the angular frequency of rotation.
Figure 9 shows the anodic linear sweep voltammograms that were recorded at a stationary titanium disk electrode in AlCl3-EMIC melt containing Ti ions with different chemical valence. The electrochemical oxidation of metal titanium varied considerably with the composition of the Ti ions. In pure AlCl3-EMIC melt, the potential must be scanned to approximately 0.60 V before dissolution of the titanium electrode commences. After the addition of TiCl2 into the melt, the onset potential of titanium dissolution shifts negatively and current density has a slight change. Since the concentration of the oxidation products, Ti(II) ions, increases, the potential of titanium oxidation increases according to the Nernst equation. The slight change of current density suggests that the presence of Ti(II) in solution has little influence on the anodic dissolution rate of titanium. As TiCl2 was replaced by high valence TiCl3 and higher valence TiCl4, the onset potential of titanium dissolution further negatively shifted and peak current further increased. The increase in current densities is thought to be caused mainly by the fast kinetics for oxidation of metallic titanium. It is to say, the presence of Ti ions with higher valence state, such as Ti(III) and Ti(IV), should promote anodic dissolution reaction of metallic Ti. Furthermore, Ti(IV) in solution has stronger promoting effect on the oxidation of titanium than Ti(III). These results indicate anodic dissolution of titanium in AlCl3-BMIC melt may involve chemical dissolution in the presence of Ti(III) or Ti(IV).
Linear sweep voltammograms recorded at Ti disk electrode in the acidic AlCl3-BMIC melt before and after addition of Ti ions with different valence state. The sweep rate was 10 mV s−1 and the temperature was 353 K.
In this paper, the electrochemical and chemical behaviors of titanium in AlCl3-BMIC ionic liquid have been investigated. In all cases, the cyclic voltammograms indicated possibility of titanium deposition. However, the electrodeposition experiments did not show no any hint of elemental titanium deposition in acidic or basic melt. Chemical reduction experiments revealed that metallic titanium could be readily oxidized and dissolved in AlCl3-BMIC melt containing Ti(IV), while Al-Ti alloy deposits were stable without dissolution. In addition, the anodic oxidation rate of titanium obviously increased in the presence of Ti(IV) and Ti(III) ions, suggesting that the chemical dissolution of titanium involved. These results show that the electrodeposition of elemental Ti is complicated by the precipitation of TiCl3 as well as the oxidation-reduction reactions between metallic titanium and titanium ions at higher valence states, including Ti(IV) and Ti(III) ions. Thus, the electrodeposition of titanium from ionic liquids is extremely difficult. A successful deposition process for titanium in ionic liquids will, in our opinion, requires the formation of stable alloy with other metal.
The authors gratefully appreciate for the support of the Natural Science Foundation of China (Project No. 21263007, 51274108) for this work.
