3 ELECTROCHEMISTRY OF VANADIUM ( II ) AND THE ELECTRODEPOSITION OF ALUMINUM-VANADIUM ALLOYS IN THE ALUMINUM CHLORIDE-1-ETHYL-3-METHYLIMIDAZOLIUM CHLORIDE MOLTEN SALT

The electrochemical behavior of vanadium(II) was examined in the 66.7-33.3 mole percent aluminum chloride-1-ethyl-3-methylimidazolium chloride molten salt containing dissolved VCl2 at 353 K. Voltammetry experiments revealed that V(II) could be electrochemically oxidized to V(III) and V(IV). However, at slow scan rates the V(II)/V(III) electrode reaction is complicated by the rapid precipitation of V(III) as VCl3. The reduction of V(II) occurs at potentials considerably negative of the Al(III)/Al electrode reaction, and Al-V alloys cannot be electrodeposited from this melt. However, electrodeposition experiments conducted in VCl2-saturated melt containing the additive, 1-ethyl-3-methylimidazolium tetrafluoroborate, resulted in Al-V alloys. The vanadium content of these alloys increased with increasing cathodic current density or more negative applied potentials. X-ray analysis of Al-V alloys that were electrodeposited on a rotating copper wire substrate indicated that these alloys did not form or contain an intermetallic compound, but were non-equilibrium or metastable solid solutions. The chloride-pitting corrosion properties of these alloys were examined in aqueous NaCl by using potentiodynamic polarization techniques. Alloys containing ~10 a/o vanadium exhibited a pitting potential that was 0.3 V positive of that for pure aluminum.


Introduction
Single-phase binary aluminum alloy films that are supersaturated with transition metal solutes exhibit enhanced resistance to chloride-induced pitting corrosion compared to pure aluminum [1].Examples of these alloys include Al-Cr, Al-Mn, Al-Mo, Al-Nb, Al-Ti, Al-W, and Al-Zr.In most cases, the corrosion resistance of these alloys increases with increasing solute concentration [2], provided that the solute does not form a second-phase precipitate [3].Because the normal equilibrium solubility of transition metals in aluminum is rarely more than about 1 % atomic fraction (a/o), it is necessary to resort to non-equilibrium alloying methods such as melt spinning, ion implantation, reactive plasma spraying, sputter deposition, and thermal evaporation to prepare these metastable alloys.Isothermal electrodeposition from chloroaluminate melts, such as those obtained by combining anhydrous aluminum chloride with sodium chloride, mixtures of sodium chloride and potassium chloride, 1-(1-butyl)pyridinium chloride (BuPyCl), or 1-ethyl-3-methylimidazolium chloride (EtMeImCl), offers another route to these non-equilibrium materials.The literature in this area has been reviewed extensively [4].Some pertinent examples of corrosion-resistant metastable aluminum alloys that have been electrodeposited from chloroaluminate melts include Al-Cr, Al-Mn, Al-Nb, and Al-Ti [11][12][13].
Corrosion-resistant metastable Al-V alloys have been prepared by sputter deposition [2].The chloride pitting potential of these aged alloys (about 2 years) increases monotonically with their vanadium content, rising to about 1.2 V positive of pure aluminum for alloys containing ~20 a/o vanadium.However, there is a paucity of reports describing the electrodeposition of these alloys from any solvents, including chloroaluminate molten salts.In an early investigation, Verdieck and Yentma [14] attempted to electrodeposit vanadium from the Lewis acidic 66-20-14 m/o AlCl 3 -NaCl-KCl melt containing V 2 O 5 or K 2 VO 2 F 3 at 429 K by using controlled current techniques, but the resulting aluminum deposits did not contain an appreciable quantity of vanadium, and the deposition potential was the same as that for aluminum.These results have been taken to indicate that the electrodeposition of vanadium from chloroaluminate molten salts takes place at potentials negative of the potential of the Al(III)/Al couple.Ryan et al. [15] examined the electrochemistry of several vanadium oxides and oxyhalides, including V 2 O 5 , Na 3 VO 4 , NaVO 3 , and NH 4 VO 3 , in the 55.0-45.0m/o AlCl 3 -EtMeImCl melt at room temperature.However, none of these compounds could be reduced to vanadium metal or Al-V alloys within the electrochemical window of this solvent.
In this article, we describe the electrochemistry of vanadium solutes in the 66.7-33.3m/o AlCl 3 -EtMeImCl melt at room temperature as it pertains to the electrodeposition of Al-V alloys.In addition, we report the electrodeposition of Al-V alloys from this molten salt after it has been modified with EtMeImBF 4 , and the chloride pitting corrosion properties of these alloys.

Chemicals
Anhydrous VCl 2 (Aldrich, 95 %) was used as received.The procedures employed for the synthesis and re-crystallization of 1-ethyl-3-methylimidazolium chloride (EtMeImCl) were identical to those described in a previous article [16].The re-crystallized EtMeImCl was melted under vacuum in a special vacuum flask for 24 h to remove traces of solvent.Anhydrous AlCl 3 (Fluka, puriss) was purified by triple sublimation under vacuum in a large reaction vessel.The Lewis acidic room temperature molten salt that resulted from the combination of a molar excess of AlCl 3 with EtMeImCl was purified by constant-current electrolysis between Al electrodes for a period of one week.Finally, the AlCl 3 -EtMeImCl melt was filtered through a porosity C Pyrex glass frit (Ace Glass) and evacuated to less that 1×10 -3 torr for 24 h to remove proton impurities [17].
The preparation of the EtMeImBF 4 molten salt was carried out by reacting EtMeImCl with NH 4 BF 4 in dry acetonitrile by using a procedure roughly similar to that described by Fuller et al. [18].For a typical reaction, 30 g (0.205 mol) of EtMeImCl was placed in a 500-ml flask fitted with a Pyrex glass plug and dissolved in 200 ml of acetonitrile.To this solution, a 10 % molar excess of NH 4 BF 4 was added.The flask was then sealed and allowed to stir at room temperature for a week.After completion of the anion exchange process, the precipitate, which consisted of NH 4 Cl and unreacted NH 4 BF 4 , was removed by filtration through a porosity E glass frit (Ace Glass).The molten salt solution was heated to ~343 K under vacuum (1×10 -3 torr) for 6 hours to remove traces of water and/or acetonitrile.After a week, very small amounts of precipitate, which were not eliminated by the filtering process, were separated by decantation.The final product was colorless and transparent.
The preparation of the EtMeImPF 6 molten salt was carried out by reacting EtMeImCl with KPF 6 in dry acetonitrile [19].In a typical synthesis, 30 g (0.205 mol) of EtMeImCl was placed in a 500-ml round-bottom flask fitted with a Pyrex glass plug and dissolved in 200 ml of acetonitrile.To this solution, a 10 % molar excess of KPF 6 was added.The flask was fitted with a reflux condenser, heated to 343 K, and stirred for one week.After the completion of the anion exchange process, the solution was heated to 373 K under vacuum (1×10 -3 torr) for 6 hours to remove acetonitrile and/or water.Unreacted KPF 6 and the solid KCl reaction product were removed by filtration through a porosity E glass frit.The final molten salt product was nearly colorless and perfectly transparent at 373K.

Apparatus, measurements, and analysis
Electrochemical experiments were carried out in a three-electrode cell.A Pine Instruments Teflon-sheathed Pt rotating disk electrode (geometrical area=0.099cm 2 ) was used as the working electrode.Coils of 0.10-cm diameter Al wire (Alfa Aesar, 99.999 %) were used for the counter and reference electrodes.These electrodes were immersed in the melt with the same composition as the bulk melt, but were separated from the bulk melt by a porosity E glass frit.The aluminum electrodes were cleaned with a mixture of concentrated H 2 SO  Staircase voltammetry experiments were carried out with an EG&G Model 273 potentiostat/galvanostat controlled with EG&G M270 software running on a PC with a Pentium III processor.The electrodeposition of Al-V alloy samples was performed with an EG&G PARC Model 173 potentiostat/galvanostat equipped with a Model 179 digital coulometer plug-in module.Bulk alloy samples of approximately 10-µm thickness were deposited onto a length of 1.25-mm diameter rotating copper wire under constant-current or controlled potential conditions.These electrodeposits were washed successively with distilled water and acetone and then dried under vacuum before they were subjected to further study.
Potentiodynamic pitting corrosion measurements were carried out on bulk alloy samples at room temperature in a 0.1 mol/l aqueous solution of NaCl in distilled H 2 O.This solution was thoroughly deaerated with nitrogen gas before each experiment.The reference electrode for these measurements was a sodium-saturated calomel electrode (SSCE), and the counter electrode was a large surface area platinum wire coil.A known length of alloy plated Cu wire was exposed to the NaCl solution by using a heat-shrink tubing mask.The sample was then scanned at 0.5 mV s -1 by using cyclic staircase voltammetry.
The solubility of VCl 2 in the melt was measured by using atomic absorption spectroscopy (Perkin-Elmer Model 2380).Standard methods were employed for these measurements.Surface morphology and elemental analysis of alloy samples were carried out at the University of Mississippi with a JEOL JSM-6100 scanning electron microscope, which was equipped with a Link energy-dispersive X-ray spectrometer (EDS) or with a Hitachi S-2600H scanning electron microscope system at Kyoto University, Kyoto, Japan.The Al-V crystal structure was probed with a Rigaku Multi Flex X-ray diffractometer in 2θ scan mode (incident angle, θ=1°) at Kyoto University.The resulting X-ray data were compared with the appropriate JCPDS card.

Oxidation of vanadium (II)
VCl 2 dissolved in the Lewis acidic 66.7-33.3mole percent AlCl 3 -EtMeImCl melt, designated hereafter as "66.7 m/o melt", to produce a dark-brown-colored solution.Atomic absorption spectrophotometric analysis of the vanadium content of a saturated solution prepared at 353 K indicated that the VCl 2 solubility was 29.5 mM. Figure 1 shows cyclic staircase voltammograms recorded at a stationary platinum disk electrode as a function of scan rate in melt containing 6.8 mM VCl 2 .The step size for these voltammograms was 2 mV, which is small enough to permit the analysis of these voltammograms with the conventional theories developed for linear scan voltammetry [20].Each voltammogram shows two oxidation (A and B) and two reduction (C and D) waves.A voltammetric scan that was reversed at about 1.7 V confirms that wave C arises from the reduction of the product of the oxidation reaction that gives rise to wave A. Therefore, wave D must correspond to the reduction of the product of wave B.
To investigate waves A and C in more detail, a series of cyclic staircase voltammograms were recorded as a function of scan rate in VCl 2 -saturated 66.7 m/o melt.The resulting voltammograms are shown in Fig. 2. Data derived from these voltammograms are collected in Table 1.The data in this table show that the anodic peak current function, i pa /ν 1/2 , is approximately independent of scan rate.However, the peak current ratio, i pc/ i pa , which was estimated by using Nicholson's semi-empirical method [21], is less than one at slow scan rates, but increases toward one as the scan rate increases (Table 1 and Fig. 3).Thus, the electrode reaction corresponding to wave A (Fig. 1 and 2) involves an oxidation reaction with a coupled irreversible chemical step [22].At slow scan rates, the peak potential-half peak potential separation for wave A, E pa -E pa/2 approaches the 0.067 V value, expected for a one-electron reversible electrode reaction at 353 K.
However, E pa -E pa/2 increases slightly as the scan rate increases, suggesting that the electron transfer reaction may not be completely reversible.If this is the case, then both the coupled chemical step and the quasi-reversible nature of the electrode reaction may drive the positive shift in E pa with increasing scan rate that is seen in Table 1 [23].If the potential of the Pt electrode was scanned through wave A and held at about 1.6~1.7 V, then the electrode became covered with a violet-colored film, which was identical to VCl 3 .Like TiCl 3 , VCl 3 was found to be completely insoluble in the 66.7 m/o melt.Figure 1, taken together with the voltammetric results presented above, suggest that wave A arises from the one-electron quasi-reversible oxidation of solvated vanadium(II), V(II) solv , to solvated vanadium(III), V(III) solv , followed by precipitation of VCl 3 (s) in an irreversible chemical step: (1) The oxidation of vanadium(II) was also examined with staircase voltammetry at a rotating Pt disk electrode in the same 6.8 mM solution of VCl 2 described above.The results of these experiments are shown in Fig. 4. Two oxidation waves are evident in these Pt-RDE voltammograms.These waves correspond to waves A and B in Fig. 1.Only wave A in Fig. 4 exhibits a well-defined limiting current.Furthermore, the limiting current for this wave was only observed when the scan rate was 50 mV s -1 , and the vanadium (II) concentration was less than about 10 mM.At slower scan rates and higher concentrations, the electrode became fouled with a passivating film of VCl 3 , and no limiting current could be observed.Because of the insolubility of vanadium(III), it was not possible to obtain much useful information about waves B and D by using either voltammetric technique.However, the overall voltammetric signature for the oxidation of vanadium(II) is very similar to that for titanium(II) in this melt at this temperature [24].That is, the voltammetric oxidation of titanium(II) results in two quasi-reversible one-electron oxidation waves; and the oxidation of titanium(II) to titanium(III) is complicated by the insolubility of TiCl 3 .On this basis, we propose that wave B corresponds to the one electron oxidation of vanadium(III) to vanadium(IV), whereas wave D reflects the reverse of this process.
A Levich plot, i.e., a plot of the limiting current density, i l , as a function of the square root of the electrode rotation rate, ω 1/2 , that was constructed from the data in Fig. 4 for wave A is shown in the inset of this figure.The slope of this plot was used with the Levich equation and the viscosity data given by Fannin et al. to calculate the diffusion coefficient, D V(II) , and the Stokes-Einstein product, D V(II) η/T for V(II) solv .The former was found to be 2.41×10 -7 cm 2 s -1 at 353 K, and latter was estimated to be 3.3×10 -11  g cm s −2 K -1 .The corresponding values for Ti(II) solv at this temperature in this same melt are 2.22×10 -7 cm 2 s -1 and 3.1×10 -11 g cm s -2 K -1 [24], in excellent agreement with the values for V(II) solv .However, it should be noted that the diffusion coefficient for Ti(II) solv varies considerably with concentration due to the formation of a polymeric diffusing species.The value quoted here was that measured in the most dilute solution examined during the previous investigation.

Reduction of vanadium(II)
Figures 5a and 5b show cyclic staircase voltammograms recorded at a Pt-RDE in the pure 66.7 m/o melt and in the same melt after the dissolution of 29.5 mM VCl 2 , respectively.In pure melt (Fig 5a), the reduction and oxidation waves centered at around 0 V correspond to the well known electrode reaction involving the electrodeposition and stripping of aluminum metal [26]: (3) The dissolution of VCl 2 results in little change in the voltammetric signature for the deposition and stripping of aluminum (Fig. 5b).Thus, there is no obvious evidence for (Inset) Relationship between the limiting current densities, i l , and square root of the angular rotation frequencies, ω 1/2 .4Al 2 Cl 7 -+ 3e - 7AlCl 4 -+ Al ↔ the co-deposition of vanadium with aluminum from this melt.In order to test for the presence of vanadium in the aluminum electrodeposits, galvanostatic electrolysis experiments were performed in these solutions over the cathodic current range extending from -10 to -100 mA cm -2 at 353 K.A rotating Cu-wire substrate was used for all experiments.However, EDS analysis of the resulting electrodeposits revealed that they consisted of pure aluminum (Table 2).That is, no appreciable vanadium was detected in these deposits, suggesting that the equilibrium potential of the V(II)/V electrode reaction is negative of that for the Al(III)/Al reaction.
Previous research in this laboratory indicated that the fluoride-based roomtemperature molten salt, EtMeImBF 4 , may be a beneficial additive for inducing the codeposition of vanadium with aluminum in the 66.7 m/o melt containing dissolved VCl 2 .Thus, galvanostatic and potentiostatic experiments were conducted at rotating Cu-wire substrates following the introduction of this additive into the plating bath.The results of these experiments are shown in Table 2, and they indicate that this additive leads to the electrodeposition of Al-V alloys.However, these experiments also indicate that significant amounts of this additive must be dissolved in the plating solution before its full effect is realized.In this case, the vanadium content of the resulting Al-V alloys is dependent on either the deposition current density or the applied potential, depending on the plating method that was employed.Attempts to prepare Al-V deposits at larger cathodic current densities or at more negative potentials than those listed in the table resulted in powdery, non-adherent coatings that could not be analyzed.
In order to gain some understanding about the mode of action of the EtMeImBF 4 additive, cyclic staircase voltammetry experiments were conducted at a Pt-RDE in plating baths containing this ionic compound.The results of these experiments are shown in Fig. 6.A comparison of Fig. 5a with Fig. 6a indicates that adding EtMeImBF 4 to pure 66.7 m/o melt results in a ~200 mV negative shift in the potential at which the onset of the aluminum deposition reaction is observed.The dissolution of both EtMeImBF 4 and VCl 2 produces an additional negative shift and decreases the rate of the aluminum deposition process (Fig. 6b).The latter conclusion is inferred from the observation that the Pt-RDE must be scanned to ca. -0.78 V to attain approximately the same current density that is observed at -0.25 V in Fig. 6a.
The related ionic fluoride compound, EtMeImPF 6 , which is also a low melting salt, was tested to determine if it was a beneficial additive for inducing the co-deposition of vanadium with aluminum.To test the effects of this additive, galvanostatic and potentiostatic bulk deposition experiments similar to those described above for the EtMeImBF 4 additive were carried out at a rotating Cu-wire substrate over the same range of current densities and potentials.However, EDS analysis of the resulting electrodeposits indicated that no appreciable vanadium was co-deposited with aluminum from this solution, even when the concentration of the EtMeImPF 6 additive was identical to that used for experiments with EtMeImBF 4 (0.76 M).Fig. 6c shows cyclic staircase voltammograms recorded at a Pt-RDE in 66.7 m/o melt containing both dissolved VCl 2 and EtMeImPF 6 .Qualitatively, these voltammograms are similar to those recorded in plating baths containing VCl 2 and EtMeImBF 4 (Fig. 6b).However, there is one obvious difference: the addition of EtMeImPF 6 does not suppress the kinetics of the aluminum deposition reaction as effectively as does EtMeImBF 4 .That the negative shift in the potential for the onset of the aluminum deposition reaction is smaller in plating baths containing the former additive compared to baths containing the latter compound.The forgoing voltammetric experiments suggest that the mode of action of the EtMeImBF 4 additive is to facilitate the electrodeposition of vanadium at the expense of the electrodeposition of aluminum.Because the V(II)/V electrode reaction lies negative of Al(III)/Al, this additive allows more negative potentials to be accessed before the onset of aluminum deposition.At the same time, it greatly suppresses the rate of the aluminum deposition reaction.Thus, for a given applied potential or current density, the partial current density for vanadium deposition would be a larger fraction of the total current density, enhancing the amount of vanadium in the electrodeposits.In view of these results, it is curious that the addition of EtMeImPF 6 to plating baths containing dissolved VCl 2 does not lead to similar results.Although both BF 4 -and PF 6 -are often thought to be relatively inert, non-coordinating anions, the former may not be as inert as supposed when added to the strongly Lewis acidic 66.7 m/o melt.For example, the conjugate acid of the latter species, PF 5 , is a stronger Lewis acid than AlCl 3 , whereas BF 3 is a weaker Lewis acid than AlCl 3 [27].Thus, the beneficial effects of the EtMeImBF 4 additive that are reported herein may be due to alteration of the melt chemistry and/or anionic speciation as a result of a reaction between the Lewis base, .Such reactions either do not take place or are less extensive for the less basic PF 6 -anion.However, an investigation of the chemistry of BF 4 -and PF 6 -in the AlCl 3 -EtMeImCl melt using techniques such as 27 Al NMR and Raman spectroscopy is needed in order to fully understand why and how BF 4 -facilitates the electrodeposition of Al-V alloys, whereas PF 6 -does not.Unfortunately, such an investigation is outside the scope of this study.

Characterization of Al-V electrodeposits
Figures 7 and 8 show SEM photographs illustrating the surface morphology of bulk Al-V samples electrodeposited on Cu-wire substrates using galvanostatic and potentiostatic techniques as described above.In all cases, the Al-V alloys appear to nucleate along the striations of the drawn copper wire surface.The surface morphology is clearly dependent on the applied current density.For example, the sample prepared at -20 mA cm -2 (Fig. 7a), which is the lowest current density that was investigated, shows a fine, leaf-like structure.As the cathodic current density is increased to -50 mA cm -2 , the resulting electrodeposit assumes a more compact, nodular structure (Fig. 7b), whereas those prepared at the highest practical current density, -100 mA cm -2 , are by comparison, relatively smooth and compact (Fig. 7c).Thus, grain refinement is apparently driven by the incorporation of vanadium into the aluminum deposit because an increase in the deposition current density would ordinarily lead to an increase in grain size.Similarly, both of the samples that resulted from potentiostatic electrodeposition experiments exhibited a reasonably smooth, compact surface morphology (Figs.8a and 8b).EDS elemental compositional maps were prepared for the Al-V electrodeposit with the highest vanadium content (Fig. 8b) and are shown in Fig. 9.These maps reveal a very uniform microdistribution of aluminum and vanadium in this sample.EDS analysis of the various Al-V alloy samples prepared during the course of this study indicated that all were free from chloride contamination.Figure 10 shows an XRD pattern collected for the as deposited Al-V sample containing 9.1 a/o V.All XRD peaks were identified either with aluminum or with the copper substrate.Reflections for α-Al 10 V, an intermetallic compound that might be   expected based on the Al/V elemental ratio of this deposit, are shown at the bottom of this figure for comparison.The lack of any reflections that can be assigned to discrete Al-V intermetallic compounds suggests that vanadium has alloyed substitutionally with aluminum in this deposit to form a non-equilibrium or metastable solid solution.Similar results were found for Al-Ti alloys prepared in the Lewis acidic AlCl 3 -EtMeImCl melt containing dissolved TiCl 2 [24].

Pitting potential measurements
The corrosion resistance of the Al-V alloys described in Table 2 was investigated by recording potentiodynamic anodic polarization curves in N 2 -saturated aqueous 0.1 M NaCl.These polarization curves are shown in Fig. 11.As noted for Al-Mn alloys examined under similar conditions [8], Al-V alloys display a stable passive region characterized by a very small potential-independent current followed by a sudden rise in current at the pitting potential.This same result was also noted for electrodeposited Al-Ti alloys [24].The variation of the Al-V pitting potential with alloy composition is shown in Fig. 12.The addition of ~10 a/o vanadium increases the pitting potential of the alloy by about +0.3 V.This increase is only about 50 % that seen for Al-V alloys with the same composition that were prepared by sputter deposition [2], but it is comparable to the improvement in pitting potential seen for Al-Ti alloys containing ~10 a/o titanium that were prepared in this same molten salt by using similar techniques.One possible explanation for the difference in the pitting potentials of the electrodeposited and sputterdeposited Al-V alloys is the fact that the latter were aged for about 2 years before they were tested, whereas the pitting potentials of the former were measured within a few weeks of their preparation.

Conclusions
Al-V alloys containing ~10 a/o vanadium can be electrodeposited from the 66.7-33.3m/o AlCl 3 -EtMeImCl molten salt containing dissolved VCl 2 , provided that high concentrations of the additive, EtMeImBF 4 , are present in the plating solution.However, the  2.
addition of similar amounts of the related fluoride salt, EtMeImPF 6 , to the plating bath does not lead to the co-deposition of vanadium with aluminum.Al-V alloys that were electrodeposited from the AlCl 3 -EtMeImCl melt were dense and compact.They appear to be non-equilibrium or metastable solid solutions in which vanadium has substituted into the aluminum lattice.The chloride pitting potential of the Al 90.9 V 9.1 alloy was about +0.3 V greater than that of pure aluminum and was comparable to the pitting potential of the Al 90 Ti 10 alloy that was electrodeposited from 66.7-33.3m/o AlCl 3 -EtMeImCl containing TiCl 2 .However, the pitting potential of the electrodeposited Al 90.9 V 9.1 alloy sample prepared during this investigation was inferior to that of aged samples of similar composition that were prepared by sputter deposition.

Figure 3 .
Figure 3. Plot of the ratios of the cyclic voltammetric anodic to cathodic peak currents as a function of scan rate.

Figure 4 .
Figure 4. Staircase voltammograms recorded at a Pt-RDE in the solution described in Fig. 1.The scan rate was 50 mV s -1 ; the step size was 2 mV; and the angular rotation frequencies were 105, 131, 157, 183, and 209 rad s -1 .(Inset) Relationship between the limiting current densities, i l , and square root of the angular rotation frequencies, ω 1/2 .

BF 4 - 7 -.
, and the principal Lewis acidic species in the melt, Al 2 Cl The fact that a relatively high concentration of EtMeImBF 4 is needed to effect the electrodeposition of Al-V alloy tends to support an equilibrium reaction between BF 4 -and Al 2 Cl 7 -

Figure 9 .
Figure 9. SEM and EDS images of the Al 90.9 V 9.1 alloy shown in Fig. 8b: (a) SEM image, (b) aluminum EDS map, and (c) vanadium EDS map.

Figure 12 .
Figure 12.Pitting potentials as a function of alloy composition for the electrodeposited Al-V alloys in Fig.11and Table2.
4 , HNO 3 and H 3 PO 4 ; rinsed with distilled H 2 O ; and dried under vacuum before use.All experiments were carried out in a nitrogen gas-filled glove box (VAC Atmospheres NEXUS system) with an O 2 and H 2 O content <5 ppm.

9 Table 1 .
Voltammetric data for the oxidation of V(II) at a stationary Pt electrode