Electrochemistry of titanium and the electrodeposition of Al-Ti alloys in the Lewis acidic aluminum chloride-1-ethyl-3-methylimidazolium chloride melt

Article abstract of DOI:10.1149/1.1554915

The chemical and electrochemical behavior of titanium was examined in the Lewis acidic aluminum chloride-1-ethyl-3-methylimidazolium chloride (AlCl₃-EtMeImCl) molten salt at 353.2 K. Dissolved Ti(II), as TiCl₂, was stable in the 66.7-33.3% mole fraction (m/o) composition of this melt, but slowly disproportionated in the 60.0-40.0 m/o melt. At low current densities, the anodic oxidation of Ti(0) did not lead to dissolved. Ti(II), but to an insoluble passivating film of TiCl₃. At high current densities or very positive potentials, Ti(0) was oxidized directly to Ti(IV); however, the electrogenerated Ti(IV) vaporized from the melt as TiCl₄(g). As found by other researchers working in Lewis acidic AlCl₃-NaCl, Ti(II) tended to form polymers as its concentration in the AlCl₃-EtMeImCl melt was increased. The electrodeposition of Al-Ti alloys was investigated at Cu rotating disk and wire electrodes. Al-Ti alloys containing up to ～19% atomic fraction (a/o) titanium could be electrodeposited from saturated solutions of Ti(II) in the 66.7-33.3 m/o melt at low current densities, but the titanium content of these alloys decreased as the reduction current density was increased. The pitting potentials of these electrodeposited Al-Ti alloys exhibited a positive shift with increasing titanium content comparable to that observed for alloys prepared by sputter deposition.

Full text of DOI:10.1149/1.1554915

Electrochemistry of Titanium and the Electrodeposition of
Al-Ti Alloys in the Lewis Acidic Aluminum Chloride −
1-Ethyl-3-methylimidazolium Chloride Melt
Tetsuya Tsuda, Charles L. Hussey, Gery R. Stafford and John E. Bonevich
J. Electrochem. Soc. 2003, Volume 150, Issue 4, Pages C234-C243.
doi: 10.1149/1.1554915
Receive free email alerts when new articles cite this article - sign up in the
box at the top right corner of the article or click here
Email alerting
service
To subscribe to Journal of The Electrochemical Society go to:
http://jes.ecsdl.org/subscriptions
© 2003 ECS - The Electrochemical Society
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

C234
Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
0013-4651/2003/150͑4͒/C234/10/$7.00 © The Electrochemical Society, Inc.
Electrochemistry of Titanium and the Electrodeposition
of Al-Ti Alloys in the Lewis Acidic Aluminum
Chloride–1-Ethyl-3-methylimidazolium Chloride Melt
a, ,z
Tetsuya Tsuda,a, Charles L. Hussey,
*
Gery R. Stafford,b,
*
*
and John E. Bonevichb
aDepartment of Chemistry and Biochemistry, The University of Mississippi, University,
Mississippi 38677, USA
bNational Institute of Standards and Technology, Materials Science and Engineering Laboratory,
Gaithersburg, Maryland 20899, USA
The chemical and electrochemical behavior of titanium was examined in the Lewis acidic aluminum chloride-1-ethyl-3-
methylimidazolium chloride (AlCl3-EtMeImCl) molten salt at 353.2 K. Dissolved Ti͑II͒, as TiCl2 , was stable in the 66.7-33.3%
mole fraction ͑m/o͒ composition of this melt, but slowly disproportionated in the 60.0-40.0 m/o melt. At low current densities, the
anodic oxidation of Ti͑0͒ did not lead to dissolved Ti͑II͒, but to an insoluble passivating film of TiCl3 . At high current densities
or very positive potentials, Ti͑0͒ was oxidized directly to Ti͑IV͒; however, the electrogenerated Ti͑IV͒ vaporized from the melt as
TiCl4(g). As found by other researchers working in Lewis acidic AlCl3-NaCl, Ti͑II͒ tended to form polymers as its concentration
in the AlCl3-EtMeImCl melt was increased. The electrodeposition of Al-Ti alloys was investigated at Cu rotating disk and wire
electrodes. Al-Ti alloys containing up to ϳ19% atomic fraction ͑a/o͒ titanium could be electrodeposited from saturated solutions
of Ti͑II͒ in the 66.7-33.3 m/o melt at low current densities, but the titanium content of these alloys decreased as the reduction
current density was increased. The pitting potentials of these electrodeposited Al-Ti alloys exhibited a positive shift with increas-
ing titanium content comparable to that observed for alloys prepared by sputter deposition.
© 2003 The Electrochemical Society. ͓DOI: 10.1149/1.1554915͔ All rights reserved.
Manuscript submitted July 23, 2002; revised manuscript received September 30, 2002. Available electronically February 28, 2003.
The resistance of aluminum to chloride-induced pitting corrosion
effects of the melt composition, Ti͑II͒ concentration, and hydrody-
namic transport rate on the Al-Ti alloy composition and to determine
the best route for introducing Ti͑II͒ into the electroplating bath.
can be improved significantly by alloying with transition metals
such as Cr, Cu, Mo, Mn, Nb, Ta, Ti, V, and Zr and rare earths such
as Er and Ce. However, improved corrosion resistance is not ob-
served unless these solutes are present at concentrations greatly ex-
ceeding their normal equilibrium solubilities, which in most cases is
less than 1% atomic fraction ͑a/o͒. This is because the pitting poten-
tial of aluminum is often related directly to the solute concentration,
with greater concentrations leading to increased corrosion
resistance.1,2 However, if the solute metal forms a precipitate, then
the corrosion resistance of the aluminum is reduced because the
precipitated solute forms microgalvanic corrosion cells.3 Thus, non-
equilibrium alloying methods such as rapid solidification or melt
spinning.4,5 ion implantation,6-9 reactive plasma spraying,10,11 sput-
ter deposition,3,12-16 and thermal evaporation17 are needed in order
to prepare single-phase materials. However, a more advantageous
method for producing these alloy films is electrodeposition because
this approach can lead to alloy coatings of uniform composition and
structure.
Titanium is one of several transition elements that have been
found to improve the corrosion resistance of aluminum.13 The elec-
trochemistry of titanium has been studied extensively in Lewis
acidic chloroaluminate melts,18-24 and it has been determined that
Al-Ti alloys can be electrodeposited from solutions of Ti͑II͒ in these
melts.20-24 In 66.7-33.3% mole fraction ͑m/o͒ AlCl3-NaCl melt at
423 K, Al-Ti alloys containing up to 28 a/o titanium were obtained
by varying the Ti͑II͒ concentration and/or the applied current
density.20,21 However, when the Ti͑II͒ concentration was raised
above 150 mmol LϪ1 in this melt at this temperature, the titanium
content of the alloy became virtually independent of the current
density, and TiAl3 , with an ordered L12 crystal structure was pro-
duced.
Experimentala
The procedures used for the synthesis of EtMeImCl, the purifi-
cation of AlCl3 by sublimation, and the preparation and purification
of the AlCl3-EtMeImCl molten salt were identical to those de-
scribed in previous articles.25,26 Anhydrous titanium ͑II͒ chloride
͑Aldrich, 99.98%͒ and anhydrous titanium ͑III͒ chloride ͑Aldrich,
99.999%͒ were used as received. Titanium ͑IV͒ chloride ͑Matheson,
Coleman, & Bell, 99.5%͒ was used after dehydration. All molten
salt experiments were carried out in a nitrogen gas-filled glove box
Ͻ
͑VAC Atmospheres NEXUS system͒ with O2 and H2O
5 ppm.
Molten salt voltammetry experiments were conducted by using a
Pine Instruments Co. model AFCBP1 bipotentiostat controlled with
PineChem for Windows software running on a microcomputer. The
anodic polarization of a titanium wire working electrode and the
electrodeposition of Al-Ti alloy were performed with an EG&G
PARC model 173 potentiostat/galvanostat equipped with a model
179 digital coulometer plug-in module. Potentiodynamic pitting cor-
rosion measurements were carried out with an EG&G PARC model
263 potentiostat/galvanostat.
All electrochemical experiments were conducted in three-
electrode cells. For experiments in the molten salt, the working elec-
trode was either a Pine Instruments Teflon-sheathed platinum rotat-
ing disk electrode ͑geometrical area ϭ 0.099 cm2), a length of 0.5
mm diam titanium wire ͑Alfa Aesar, 99.98%͒, or a small disk elec-
trode prepared by sealing a length of the aforementioned titanium
wire in a Pyrex glass tube and cutting off the end of the sealed tube
so as to expose the cross section of the wire. Coils of 1.0 mm diam
aluminum wire ͑Alfa Aesar, 99.999%͒ were used for the counter and
reference electrodes. These electrodes were immersed in a melt with
the same composition as the bulk melt, but were separated from the
bulk melt by a porosity E glass frit ͑Ace Glass͒. The aluminum
electrodes were cleaned with a mixture of concentrated H2SO4 ,
In this article, we report the electrochemistry of titanium in the
Lewis
acidic
AlCl3-1-ethyl-3-methylimidazolium
chloride
(AlCl3-EtMeImCl) molten salt at 353.2 K as it pertains to the elec-
trodeposition of Al-Ti alloys. The aim of this study is to clarify the
* Electrochemical Society Active Member.
z E-mail: chclh@chem1.olemiss.edu
a Certain trade names are mentioned for experimental information only; in no case
does it imply a recommendaion or endorsement by NIST.
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
C235
HNO3 , and H3PO4 , rinsed with distilled H2O, and dried under
vacuum before use.
Table I. Solubility of Ti„II… in acidic AlCl3-EtMeImCl.
Alloy samples of approximately 10 ␮m thickness were electro-
plated from solutions of Ti͑II͒ in the AlCl3-EtMeImCl molten salt
onto a length of 1.25 mm diam copper wire or a copper disk ͑geo-
metrical area ϭ 0.80 cm2). The elemental composition of each
sample was determined by energy dispersive X-ray ͑EDX͒ spectros-
copy. Potentiodynamic pitting corrosion measurements were carried
out on these alloy samples at room temperature in a 0.1 mol LϪ1
solution of NaCl in distilled H2O. This solution was deaerated with
nitrogen gas before each experiment. The reference electrode for
these measurements was a sodium-saturated calomel electrode
͑SSCE͒, and the counter electrode was a large surface area platinum
wire coil. A known length of the plated Cu wire was exposed to the
NaCl solution by using a heat-shrink Teflon tubing mask, and the
sample was scanned at 0.5 mV sϪ1 by using cyclic staircase volta-
mmetry.
AlCl3-EtMeImCl
AlCl3-NaCl
423
Temperature ͑K͒
353
60.0
60 Ϯ 1a
66.7
170 Ϯ 2b
66.7
mol % AlCl3
Solubility of Ti͑II͒
32521
͑mmol LϪ1
͒
a
b
Color of solutions: dark green, dark reddish-yellow.
EXAFS results of Dent et al.,28 may be inadequate under these con-
ditions. By contrast to TiCl2 , TiCl3 was found to be insoluble in all
compositions of the acidic AlCl3-EtMeImCl melt.
A small amount of purple precipitate identical to solid TiCl3 was
present in the 60.0 m/o melt after the dissolution of TiCl2 , but was
not obvious in the 66.7 m/o melt. The presence of TiCl3 in solutions
of TiCl2 could arise through several different routes, including slow
disproportionation of the dissolved Ti͑II͒ as shown in Eq. 2
UV-visible spectroscopy was performed with a Hewlett Packard
model 8543 diode-array spectrophotometer or a Varian CARY 5
spectrophotometer employing Wilmad no. 107-7 closed type 0.10
cm path length quartz cells. The solubility of Ti͑II͒ in the melt was
measured by using atomic absorption spectroscopy ͑Perkin-Elmer
model 2380͒. Standard methods were employed.
3 Ti AlCl ͒ Ϫ ϩ 3AlClϪꢀ2TiCl s͒ ϩ Ti ϩ 6Al ClϪ ͓2͔
͓
͑
͔
͑
3
4
3
2
4
7
The surface morphology and elemental analysis of alloy samples
were investigated with a JEOL JSM-6100 scanning electron micro-
scope ͑SEM͒ equipped with a Link energy dispersive X-ray spec-
trometer at the University of Mississippi SEM/EDX facility located
in the Department of Mechanical Engineering or with a Hitachi
S-4700 field-emission scanning electron microscope at the National
Institute of Standards and Technology ͑NIST͒. The electrodeposits
were also examined by X-ray diffraction ͑XRD͒ by using a Siemens
D-500 diffractometer with Cu K␣ radiation. The lattice parameters
of the deposits were refined using the Cu substrate reflections as an
internal standard. Transmission electron microscopy ͑TEM͒ was
performed on selected deposits prepared for both plan view and
cross-sectional analysis. The plan view films were detached from the
planar Cu substrate and affixed to a Mo support ring. Cross sections
of Al-Ti, deposited onto Cu wires, were formed by first overplating
with Cu by using a cyanide Cu strike followed by bright Cu from a
CuSO4 solution.27 Cross-sectional cuts, 0.75 mm thick, were then
dimple polished down to 20 ␮m thickness. Plan view and cross
sectional samples were subsequently milled with 4.5 kV Ar ions at 5
degrees incidence. Portions of the samples with electronic transpar-
ency were examined at 300 kV by using a JEOL 3010 TEM.
or by oxidation of the Ti͑II͒ through reaction with adventitious im-
purities in the melt or with oxygen in the glove box atmosphere.
Sørlie and Øye29 reported that Ti͑II͒ tends to disproportionate in
AlCl3-KCl melts containing less than 60.0 m/o AlCl3 , where the
AlClϪ4 concentration is large, but is stable in strongly Lewis acidic
melts.
In order to examine the stability of dissolved TiCl2 , the room-
temperature UV-visible absorption spectra of Ti͑II͒ in both 60.0 and
66.7 m/o AlCl3-EtMeImCl were recorded as a function of concen-
tration ͑Fig. 1͒. Dilute solutions of Ti͑II͒ appeared pale green in the
former melt and pale yellow-green in the latter melt. The principal
absorption bands in these spectra are summarized in Table II. In the
60.0 m/o melt, Ti͑II͒ exhibits bands at 344 and 444 nm, whereas the
spectra of Ti͑II͒ in 66.7 m/o melt show bands at 339 and 433 nm and
a shoulder at 407 nm. In addition, the Ti͑II͒ spectra in these two
melts differed considerably at wavelengths greater than 500 nm. As
shown in Table II, the absorption bands recorded in the acidic
AlCl3-EtMeImCl melt are in general shifted to higher energies com-
pared to those observed in acidic AlCl3-KCl melt.29 A shift in the
wavelengths of the absorption maxima was also observed with
changes in the acidity of the AlCl3-EtMeImCl melt. This shift is
probably caused by changes in ligand field strength brought about
by changes in the anionic speciation of the melt as the melt compo-
sition is changed.
Results and Discussion
Dissolution of TiCl2.—The dissolution of solid TiCl2 was inves-
tigated as
a
method to supply Ti͑II͒ in Lewis acidic
Figure 2 shows the relationship between the Ti͑II͒ concentration
and the absorbance at 444 nm in the 60.0 m/o melt and 433 nm in
the 66.7 m/o melt. These wavelengths correspond to the main bands
in the spectra shown in Fig. 1. The plot constructed with absorbance
data in the 66.7 m/o melt shows good adherence to Beer’s law over
the composition range that was examined. However, the absorbance
of the 444 nm band in 60.0 m/o melt was generally smaller than that
for the corresponding band in the 66.7 m/o melt, and most signifi-
cantly, it did not obey Beer’s law. There was no obvious evidence of
TiCl3 or titanium metal in the dilute solutions of Ti͑II͒ in 60.0 m/o
melt that were employed for this spectroscopic investigation. How-
ever, the spectra recorded in the 60.0 m/o melt exhibit a large
wavelength-independent background absorbance ͑Fig. 1a-d͒. This
phenomenon was most apparent at wavelengths above 500 nm and
was absent from Ti͑II͒ spectra recorded in the 66.7 m/o melt. This
large background absorbance is attributed to light scattering by
finely dispersed particulate matter in these solutions. We conclude
from these observations and the nonlinearity of the Beer’s law plot
that Ti͑II͒ slowly disproportionates in the 60.0 m/o melt. This im-
plies that the potential of the Ti͑II͒/Ti couple is very close to that for
the Ti͑III͒/Ti͑II͒ couple. Thus, in view of the uncertainty about the
AlCl3-EtMeImCl melts for the electrodeposition of Al-Ti alloys.
TiCl2 dissolved slowly in these melts at 353.2 K. The solubility of
TiCl2 in the two melt compositions used for this investigation, 60.0-
40.0 and 66.7-33.3 m/o AlCl3-EtMeImCl, designated hereafter as
the 60.0 and 66.7 m/o melts, respectively, was estimated by prepar-
ing saturated solutions of TiCl2 in both melts at 353.2 K, dissolving
a weighed portion of the supernatant liquid from these saturated
solutions in distilled water, and determining the Ti͑II͒ concentration
with atomic absorption spectroscopy using standard methods. The
results of these experiments are given in Table I. Because the solu-
bility of TiCl2 increases with increasing melt acidity, i.e., with in-
creasing Al2ClϪ7 concentration, the TiCl2 dissolution reaction must
be
2Al2ClϪ ϩ TiCl s͒ → Ti AlCl ͒ Ϫ ϩ AlCl4Ϫ
͓1͔
͑
͓
͑
͔
3
2
4
7
However, as is shown later in this article, the Ti͑II͒ species is likely
to be polymeric or aggregated, especially when the Ti͑II͒ is present
at high concentrations. Thus, the simple six-coordinate structure for
dissolved Ti͑II͒ given in Eq. 1, which can be inferred from the
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

C236
Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
Figure 2. Beer’s law plot for Ti͑II͒ in the AlCl3-EtMeImCl melt prepared
from the data in Fig. 1: ͑᭹͒ 444 nm, 60.0 m/o melt, ͑᭺͒ 433 nm, 66.7 m/o
melt, and ͑᭡͒ the solution in Fig. 1j at 433 nm.
AlCl3-EtMeImCl melt by reacting TiCl3 with Ti or Al metal. No
UV-vis spectrum was observed for Ti͑III͒ in the supernatant liquid
of melts containing solid TiCl3 , attesting further to its virtually
complete insolubility in this melt. However, if sufficient powdered
TiCl3 was added to a known volume of melt to produce a 4.86 mmol
LϪ1 solution and if this suspension of TiCl3 was stirred with a small
piece of aluminum metal at 353.2 K, then the solution became
yellow-green and exhibited a spectrum identical to that for Ti͑II͒
͑Fig. 1j͒. Furthermore, the absorbance of the 433 nm band in this
spectrum is in excellent agreement with the data for Ti͑II͒ solutions
shown in Fig. 2, indicating complete conversion of the TiCl3 to
Ti͑II͒. This may be an important result because aluminum and tita-
nium metal, as well as TiCl3 , are much less expensive than TiCl2 .
Thus, the in situ reduction of TiCl3 to soluble Ti͑II͒ by titanium or
aluminum metal may be a useful route for introducing Ti͑II͒ into an
organic chloroaluminate plating bath.
Figure 1. UV-vis absorption spectra of Ti͑II͒ in the AlCl3-EtMeImCl melt:
͑top͒ 60.0 m/o melt, Ti͑II͒ concentrations: ͑a͒ 8.21, ͑b͒ 4.03, ͑c͒ 2.08, ͑d͒
1.00 mmol LϪ1; ͑middle͒ 66.7 m/o melt, Ti͑II͒ concentrations: ͑e͒ 8.16, ͑f͒
4.06, ͑g͒ 2.03, and ͑h͒ 1.01 mmol LϪ1; ͑bottom͒ 66.7 m/o melt, ͑i͒ after
stirring with sufficient TiCl3 to prepare a 4.86 mmol LϪ1 solution, ͑j͒ after
the addition of Al metal to the solution in ͑i͒ and stirring for several hours,
͑k͒ after the anodic oxidation of a Ti electrode at 0.60 V, and ͑l͒ after the
addition of Al metal to the solution in ͑k͒ and stirring for several hours. All
measurements were carried out in 0.10 cm path length quartz cells.
Ti͑II͒ concentration and stability, we report herein only qualitative
data for experiments carried out in the 60.0 m/o composition of the
AlCl3-EtMeImCl melt.
During previous investigations conducted in AlCl3-NaCl-KCl,
Ti͑II͒ was introduced into the melt by the direct reduction of TiCl3
or TiCl4 with Al metal.22 This method was investigated in 66.7 m/o
Electrochemical oxidation of Ti(II).—The electrochemistry of ti-
tanium has been investigated extensively in the 65.0 and 66.7 m/o
AlCl3-NaCl melts18,30 and in the 66.7 m/o AlCl3-EtMeImCl melt.19
However, somewhat different approaches were used in these studies.
Table II. Summary of absorption spectral data.
Ionic species
Solvent
Temperature ͑K͒
␭ ͑nm͒
Ref.
Ti͑II͒
ϳ298
ϳ298
562 ϳ 654
471 ϳ 629
ϳ298
344a,444
340,407a,433
357,453,702
464,702
—
—
—
465,690
288
287
This work
This work
29
AlCl3-EtMeImCl ͑60.0-40.0 m/o͒
AlCl3-EtMeImCl ͑66.7-33.3 m/o͒
AlCl3-KCl ͑51.0-49.0 m/o͒
29
AlCl3-KCl ͑66.7-33.3 m/o͒
Ti͑III͒
This work
This work
29
AlCl3-EtMeImCl ͑60.0-40.0 m/o͒
AlCl3-EtMeImCl ͑66.7-33.3 m/o͒
AlCl3-KCl ͑51.0-49.0 m/o͒
ϳ298
746 ϳ 894
479 ϳ 553
ϳ298
ϳ298
ϳ298
29
AlCl3-KCl ͑66.7-33.3 m/o͒
Ti͑IV͒
This work
This work
19
AlCl3-EtMeImCl ͑60.0-40.0 m/o͒
AlCl3-EtMeImCl ͑66.7-33.3 m/o͒
AlCl3-EtMeImCl ͑60.0-40.0 m/o͒
290
a Shoulder.
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
C237
Figure 3. Voltammograms recorded at a Pt-RDE at 353.2 K in the 66.7 m/o
AlCl3-EtMeImCl melt at 2000 rpm. The Ti͑II͒ concentrations were ͑a͒ 12.6,
Figure 4. Voltammograms recorded at a Pt-RDE at 353.2 K in the 60.0 m/o
AlCl3-EtMeImCl melt at 2000 rpm. The Ti͑II͒ concentrations were ͑a͒ 8.30,
͑b͒ 16.7, and ͑c͒ 52.5 mmol LϪ1. The sweep rates were 0.01 V sϪ1. Inset:
Levich plot constructed from limiting current data at 0.75 V.
͑b͒ 25.2, ͑c͒ 50.3, and ͑d͒ 170 mmol LϪ1. The sweep rates were 0.01 V sϪ1
.
Inset: Levich plot constructed from limiting current data at 0.75 V.
For example, experiments in former ionic solvent were carried out
with dissolved TiCl2 or electrogenerated Ti͑II͒ as the starting mate-
rial at elevated temperatures ͑Ͼ423 K͒, whereas experiments in-
volving the latter melt were carried out at ambient temperature with
TiCl4 as the titanium solute. Regardless of the oxidation state of the
starting species, all of these investigations revealed that the electro-
chemistry of titanium is dominated by the two one-electron, quasi-
reversible electrode reactions shown below
31
formation of a polymeric or aggregated species, Ti(II)
,
͖
n
in the
͕
melt as the Ti͑II͒ concentration was increased, which leads to a
diffusing entity with a larger solvodynamic radius and correspond-
ingly smaller diffusion coefficient. This electroinactive polymeric
species was postulated to be in rapid equilibrium with the electro-
active monomeric Ti͑II͒ species. The data presented above indicate
that the formation of Ti͑II͒ polymers must also take place in the
Lewis acidic AlCl3-EtMeImCl melt. Furthermore, the Ti͑II͒ aggre-
gation process seems to be dependent on the melt acidity, becoming
more extensive as the melt is made more acidic.
Ti IV͒ ϩ eϪꢀTi III͒
͓3͔
͓4͔
͑
͑
Ti III͒ ϩ eϪꢀTi II͒
͑
͑
Diffusion coefficients for Ti͑II͒, DTi(II)
, in the 66.7 m/o
AlCl3-EtMeImCl melt at 353.2 K were estimated from the slopes of
Levich plots at each of the formal Ti͑II͒ concentrations. ͓Uncertainty
about the Ti͑II͒ concentration precluded such calculations for solu-
tions of Ti͑II͒ in the 60.0 m/o melt͔. The viscosity data reported by
Fannin et al.32 were used for these calculations. The experimental
limiting current densities, i1 , and the estimated values of DTi(II) are
given in Table III along with the Stokes-Einstein products,
DTi(II)␩/T, where ␩ and T are the absolute viscosity of the melt and
the absolute temperature, respectively. As shown in Table III, the
value of DTi(II) decreases as the formal concentration of Ti͑II͒ in-
creases in concordance with previous findings in acidic AlCl3-NaCl
melt.18,21,30
Furthermore, at slow voltammetric scan rates or long times, both the
oxidation of Ti͑II͒ and the reduction of Ti͑IV͒ are complicated by
the insolubility of Ti͑III͒, which precipitates as red-violet or purple
TiCl3 .
The electrochemical behavior of Ti͑II͒ varied considerably with
the melt composition. Figures 3 and 4 show Pt-RDE voltammo-
grams of Ti͑II͒ dissolved in the 66.7 and 60.0 m/o AlCl3-EtMeImCl
melts, respectively, at 353.2 K. These voltammograms were re-
corded at different Ti͑II͒ concentrations. The voltammograms in
these figures are very similar in appearance to the sampled current
voltammogram for Ti͑II͒ reported in the classic article by Fung and
Mamantov18 in that they exhibit two waves that can be ascribed to
the reactions given in Eq. 3 and 4. The first waves in Fig. 3 and 4,
which correspond to the reverse of Eq. 4, exhibit well-defined lim-
iting currents in the 0.6 to 1.0 V potential range at all of the Ti͑II͒
concentrations that were studied. Levich plots, i.e., plots of the lim-
iting current or current density vs. the square root of the electrode
rotation rate, for this wave were linear and passed through the ori-
gins of the plots ͑see insets of Fig. 3 and 4͒. However, the limiting
current for the first wave at a fixed rotation rate did not scale linearly
with the formal Ti͑II͒ concentration as predicted by the Levich equa-
tion. This deviation was more obvious in the more acidic 66.7 m/o
melt than in the 60.0 m/o melt. A similar deviation between the
cyclic voltammetric peak current for the oxidation of Ti͑II͒ at a
platinum electrode and the formal Ti͑II͒ concentration was also
noted during experiments carried out in the Lewis acidic
AlCl3-NaCl melt.18,21,30 This unusual behavior was attributed to the
The second oxidation waves shown in Fig. 3 and 4 exhibit a
distorted appearance. In some cases, the limiting currents for these
waves are larger than the limiting currents for the first oxidation
Table III. Summary of electrochemical data for Ti„II… at 353.2 K
Ti͑II͒ formal
Mol %
AlCl3
concentration
il
108 DTi(II)
1012 ␩DTi(II) /T
͑mmol LϪ1
͒
͑mA cmϪ2
͒
͑cm2 sϪ1
͒
͑g cm sϪ2 KϪ1
͒
66.7
12.6
25.2
50.3
95.4
170
0.72
0.94
1.78
3.21
3.40
22.2
11.8
10.9
9.97
5.47
30.5
16.2
15.0
13.7
7.51
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

C238
Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
Figure 6. UV-vis absorption spectra recorded in 60.0 m/o AlCl3-EtMeImCl
melt: ͑a͒ 1.57 mmol LϪ1 TiCl4 at room temperature, ͑b͒ after anodic polar-
ization of a Ti electrode in neat melt at 0.05 mA cmϪ2 at 353.2 K, and ͑c͒
after the solution in ͑a͒ was heated to 353.2 K for 12 h.
Figure 5. Linear sweep voltammograms recorded as a function of tempera-
ture in the 66.7 m/o AlCl3-EtMeImCl melt: ͑a͒ Ti disk electrode, room
temperature; ͑b͒ Ti disk electrode, 353.2 K; ͑c͒ Ti disk electrode, 373.2 K;
and ͑d͒ Pt disk electrode at room temperature. The potential sweep rates were
0.01 V sϪ1
.
large oxidation overpotential is also consistent with the observed
temperature dependence of the potential at which the onset of tita-
nium oxidation is observed.
waves, suggesting that the oxidation of Ti͑III͒ to Ti͑IV͒ is a complex
process. Additional experiments beyond the scope of the present
investigation are needed in order to deduce the mechanism leading
to these results.
To further investigate the oxidation of titanium metal in the
AlCl3-EtMeImCl melt, the anodic polarization of a titanium elec-
trode with a large surface area was carried out in the 60.0 and 66.7
m/o melts at 353.2 K by using galvanostatic and potentiostatic meth-
ods. During galvanostatic electrolysis at low current density ͑ϳ0.05
mA cmϪ2͒, the potential was approximately 0.60 V at the outset of
the experiment, but quickly increased to 1.15 V. At the conclusion of
this electrolysis experiment, the color of the melt was pale yellow
and no TiCl3 precipitate was present in the cell. From the relation-
ship between the charge passed and weight loss of the titanium wire,
the number of electrons involved in the reaction was calculated from
Faraday’s law to be approximately 4, suggesting that the galvano-
static oxidation of titanium at the modest current density of this
experiment produces TiCl4 . The direct oxidation of Ti metal to
Ti͑IV͒ at large positive potentials without the formation of a passive
film of solid TiCl3 was noted by Stafford and Moffat30 during pre-
vious work carried out in 66.7 m/o AlCl3-NaCl.
Figure 6 compares the UV-visible spectrum of the 60.0 m/o melt
containing 1.57 mmol LϪ1 TiCl4 and a spectrum of the solution
resulting from the galvanostatic polarization of Ti metal as described
above. the spectrum of the TiCl4 solution is in excellent agreement
with that reported by Carlin et al.19 and consists primarily of a
single band centered at 288 nm ͑Table II͒. However, only a weak
band was found at this wavelength in the melt at the conclusion of
the experiment. When a melt solution containing dissolved TiCl4
was heated to 353.2 K with stirring, the spectrum of Ti͑IV͒ gradu-
ally disappeared as the TiCl4 vaporized from the melt. A cyclic
voltammogram acquired at the conclusion of the electrolysis experi-
ment gave a similar result. Taken together, the spectroscopic and
voltammetric results indicate that the Ti͑IV͒ produced by the anodic
oxidation of titanium at 353.2 K evaporates from the melt during
and after the electrolysis as TiCl4(g). These results suggest that
galvanostatic electrolysis is not a good method for introducing Ti͑II͒
because of the difficulty of controlling the oxidation potential.
To circumvent the direct oxidation of Ti metal to Ti͑IV͒, poten-
tiostatic electrolysis was carried out in the 66.7 and 60.0 m/o melts
at an applied potential of 0.60 V. However, the electrolysis current
Anodic dissolution of titanium metal.—Anodic linear sweep vol-
tammograms that were recorded at a stationary titanium disk elec-
trode in the 66.7 m/o melt as a function of temperature are shown in
Fig. 5. At room temperature, the titanium electrode must be scanned
to approximately 0.60 V before dissolution commences. As the tem-
perature of the melt was raised to 353.2 K and higher, the potential
for the onset of titanium dissolution decreased. At all of the tem-
peratures that were investigated, oxidation of the titanium metal
electrode produced only a single well-defined wave and proceeded
until the electrode was covered with a passive film of TiCl3(s),
which blocked the electrode. At elevated temperatures, e.g., 353.2
and 373.2 K, this film breaks down when the potential is scanned
above about 1.0 V as signaled by a sharp rise in the anodic current
and the production of Ti͑IV͒ in agreement with the voltammograms
in Fig. 3. However, if the potential of the titanium electrode is
scanned above 1.0 V at room temperature, a small current flows, but
there is no appreciable breakdown of the TiCl3 passive film that
would signal the generation of Ti͑IV͒. Because Ti͑II͒ is a stable
oxidation state of titanium in the 66.7 m/o melt ͑vide supra͒, it is
unlikely that the first oxidation wave in the voltammograms shown
in Fig. 5 arises from the direct, single-step, three-electron oxidation
of Ti͑0͒ to TiCl3 . Instead, the first wave probably reflects the suc-
cessive reactions shown below
Ti 0͒ ϩ 3AlClϪꢀ Ti AlCl ͒ Ϫ ϩ 2eϪ
͓5͔
͑
͓
͑
͔
3
4
4
Ti AlCl ͒ Ϫ ϩ 3AlClϪꢀTiCl s͒ ϩ 3Al ClϪ ϩ eϪ ͓6͔
͓
͑
͔
͑
3
4
3
2
4
7
Because only a single oxidation wave is observed in the 0.5 to 1.0 V
potential region, the potentials for these reactions must be closely
spaced, and/or a large overpotential must be required to initiate the
reaction in Eq. 5, which shifts the potential of this reaction to a
potential that overlaps with the reaction in Eq. 6. The presence of a
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
C239
Table IV. Al-Ti alloy composition as a function of the electrode
rotation rate.
Rotation rate ͑rpm͒
500 1000 1500 2000
9.9 11.2
Mol %
AlCl3
0
250
Titanium
content ͑a/o͒
60.0
66.7
2.4
10.9
—
—
11.2
15.1
11.7 14.4 15.7
13.9
trode as it was formed. However, if this solution was stirred with Ti
or Al metal for several hours at 353.2 K, the TiCl3 precipitate dis-
appeared, and the absorption spectrum of the solution became iden-
tical to that for Ti͑II͒ ͑Fig. 1͒.
Electrodeposition of Al-Ti alloys.—Figure 7 shows cyclic volta-
mmograms for stationary and rotating Pt disk electrodes in the 60.0
m/o melt with and without added Ti͑II͒. It was demonstrated during
a previous investigation in acidic AlCl3-NaCl melt that the potential
of the Al stripping wave shifts positively after the addition of
Ti͑II͒.21 This behavior occurs because Ti metal co-deposits with Al
metal to form Al-Ti alloy, which is more difficult to oxidize than
pure Al. Similar behavior was observed in AlCl3-EtMelmCl con-
taining Ti͑II͒. Furthermore, the magnitude of this shift increased
with increasing Ti͑II͒ concentration. Virtually identical results were
observed in the 66.7 m/o AlCl3-EtMeImCl melt. However, the ad-
dition of Ti͑II͒ to acidic AlCl3-EtMeImCl melt did not increase the
overpotential for Al deposition as found for Ti͑II͒ solutions in
AlCl3-NaCl. In fact, a small positive shift was evident.
Figure 7. Cyclic voltammograms recorded at a ͑top͒ stationary and ͑bottom͒
rotating Pt-RDE ͑2000 rpm͒ in the 60.0 m/o AlCl3-EtMeImCl melt at 353.2
K: ͑–––͒ neat melt, ͑
͒ 40.4 mmol LϪ1 Ti͑II͒, and ͑
͒ 59.5 mmol
LϪ1 Ti͑II͒. The potential sweep rates were 0.01 V sϪ1
.
In order to investigate the formation of bulk Al-Ti alloys, elec-
trodeposition was performed by using dc galvanostatic methods at a
current density of Ϫ10 mA cmϪ2 in a Ti͑II͒-saturated melt. The
substrate for these experiments was a copper rotating disk electrode
͑Cu-RDE͒. As shown in Fig. 8 and Table IV, the Ti metal content of
the Al-Ti alloys depends on the electrode rotation rate at small ro-
tation rates, but becomes independent of a rotation rate above 500
rpm. Therefore, all deposition experiments were conducted at rota-
tion rates exceeding 500 rpm.
diminished to nearly zero, and the electrode became coated with a
reddish-purple film of TiCl3 shortly after the electrolysis com-
menced. An absorption spectrum of this solution showed that no
Ti͑II͒ was obtained ͑Fig. 1k͒. Vigorous stirring did not lead to a
different result, but merely dislodged the solid TiCl3 from the elec-
In the 66.7 m/o AlCl3-NaCl melt at 423.2 K at Ti͑II͒ concentra-
tions greater than 150 mmol LϪ1, the Ti metal content of the Al-Ti
alloy became virtually independent of the current density, and TiAl3
with an ordered L12 crystal structure was the predominant alloy
produced.21,30 However, different results were obtained in the 66.7
m/o AlCl3-EtMeImCl melt at 353.2 K. Although solutions contain-
ing more than 150 mmol LϪ1 Ti͑II͒ can be prepared in this melt
without difficulty ͑Table I͒, the Ti metal content of bulk Al-Ti alloys
prepared in these solutions was found to increase as the applied
current density is decreased ͑Table V and Fig. 8͒. This result is to be
expected if the reduction potential of the Ti͑II͒/Ti couple is positive
of that for Al͑III͒/Al because at low reduction current densities and
correspondingly less negative potentials, the partial current density
for the electrodeposition of Ti metal would be a larger fraction of the
total current density. At larger reduction current densities or more
negative potentials, the partial current for Ti metal probably reaches
its mass-transport-limited value, whereas the partial current density
for the deposition of Al metal continues to increase relative to the Ti
metal deposition current density, becoming a larger fraction of the
total current and resulting in alloy deposits with a smaller fraction of
Ti metal. The results of this investigation indicate that the titanium
content of Al-Ti alloys electrodeposited from the AlCl3-EtMeImCl
Table V. Electrodeposited Al-Ti alloys.
Electrodeposition current density ͑mA cmϪ2
͒
Figure 8. Relationship between the electrode rotation rate, the current den-
sity, and the titanium content of bulk Al-Ti alloys deposited from Ti͑II͒-
saturated AlCl3-EtMeImCl melts at 353.2 K: ͑᭹͒ 60.0 m/o melt, Ϫ10 mA
cmϪ2, ͑᭺͒ 66.7 m/o melt, Ϫ10 mA cmϪ2, and ͑᭝͒ 66.7 m/o melt as a
function of the current density at a rotation rate of 2000 rpm.
mol %
AlCl3
Ϫ2.5
Ϫ5.0
Ϫ10.0
Ϫ15.0
Ϫ20.0
12.4
Titanium
content ͑a/o͒
66.7
19.1
17.5
15.1
12.8
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

C240
Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
Figure 9. Field-emission scanning electron micrographs of Al-Ti alloys electrodeposited at 353.2 K from 66.7 m/o AlCl3-EtMeImCl melt containing 170 mmol
LϪ1 Ti͑II͒. The current density and alloy compositions were ͑a͒ Ϫ20.0 mA cmϪ2, 7.0 a/o Ti; ͑b͒ Ϫ5.00 mA cmϪ2, 11.4 a/o Ti; ͑c͒ Ϫ1.00 mA cmϪ2, 15.6 a/o;
and ͑d͒ Ϫ0.50 mA cmϪ2, 18.4 a/o Ti.
the aluminum lattice parameter over this range of composition. This
is to be expected as the smaller Ti atoms substitute for Al in the fcc
lattice. The second change was significant broadening of the X-ray
reflections with increasing Ti metal content, suggesting a decrease in
the deposit grain size. This latter result supports the SEM analysis
that showed a decrease in grain size with an increase in the Ti metal
content ͑Fig. 9͒.
melt is limited by the solubility of TiCl2 and the minimum practical
current density that can be used for the timely preparation of such
deposits.
Characterization of Al-Ti electrodeposits.—The morphologies of
four electrodeposits are shown in Fig. 9. In this figure, the current
density is decreased from Ϫ20 mA cmϪ2 in Fig. 9a to Ϫ0.50 mA
cmϪ2 in Fig. 9d. This corresponds to an increase in the Ti metal
content from 7.0 to 18.4 a/o. In all cases, the deposits were nodular,
compact, and dense. Many of the nodules showed crystallographic
faceting, although no symmetry characteristic of a particular growth
direction was observed. It is clear that the nodule size decreases with
decreasing current density and increasing titanium content. Typi-
cally one would expect an increase in grain size as the deposition
current density or overpotential is decreased. However, in this case,
the grain refinement is driven by the incorporation of Ti metal into
the alloy rather than the deposition overpotential. This nicely dem-
onstrates the impact that impurities and alloying additions have on
deposit morphology. A high magnification view of the 18.4 a/o Al-Ti
deposit ͑Fig. 9d͒ is shown in Fig. 10. The step faceting is clearly
seen in each nodule, confirming that the nodules are single crystals.
It also suggests that although the macrodeposit grows by three-
dimensional nucleation and growth, each nodule grows primarily by
two-dimensional, layered growth.
The chemically disordered fcc structure obtained in this study
from the room temperature melt is a significant departure from the
ordered L12 crystal structure deposited from AlCl3-NaCl at 423.2
K.20,33 Figure 11 shows the XRD patterns of Al-Ti films electrode-
posited at 353.2 K from AlCl3-EtMeImCl and at 423.2 K from
AlCl3-NaCl. For comparison, a calculated diffraction pattern for
fully ordered (L12) Al3Ti is shown.34 The diffraction pattern of the
423.2 K deposit clearly shows the 100 and 110 superlattice reflec-
tions of the ordered L12 structure. High-resolution TEM indicates
that the L12 domains are on the order of 5-10 nm in size.33 This
explains the significant line broadening seen in the superlattice re-
flections for the higher temperature electrodeposit ͑Fig. 11b͒. The
superlattice reflections are clearly absent from the diffraction pattern
of the alloy deposited at 353.2 K. In addition, the 111 fundamental
reflection is shifted to lower values of 2␪, reflecting the larger lattice
parameter as a result of the substoichiometric Ti metal content. Ex-
tensive broadening of the superlattice reflections might make them
difficult to detect by XRD. A report in the literature describing a
glancing angle X-ray examination of Al-Ti alloys electrodeposited
from eutectic AlCl3-NaCl-KCl containing Ti͑III͒ also shows the
fundamental reflections for fcc Al for compositions containing up to
27.5 a/o Ti metal. However, the authors were unable to confirm the
Electrodeposits containing 7.0 to 18.4 a/o Ti metal were exam-
ined by X-ray diffraction ͑XRD͒ methods. All of the diffraction
patterns were indexed to a disordered face-centered cubic ͑fcc͒
structure and were very similar to that of pure aluminum. Two dis-
tinct changes in the diffraction patterns with increased Ti metal con-
tent of the alloy were observed. The first was a 0.33% decrease in
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
C241
Figure 10. Field-emission scanning electron micrograph with increased magnification of the alloy sample shown in Fig. 9.
of all electrodeposits examined is on the order of 0.5 to 5 ␮m.
Deposits formed at 353.2 K are chemically disordered whereas those
formed at 423.2 K have L12 domains measuring 5-10 nm in size.
Previous work suggests that these domains appear to have grown
through a first-order nucleation and growth process, independent of
the electrocrystallization process.33 This type of domain structure is
quite common in rapidly solidified alloys wherein the disordered
phase produced by the solidification process transforms to the equi-
librium ordered phase quite rapidly during solid-state cooling to
room temperature.35 Although this often occurs at temperatures that
do not allow for significant bulk diffusion, the activation barrier to
the nucleation of ordered domains is rather small because both the
nucleus and matrix have essentially the same crystal structure and
composition.36 Consequently, nucleation tends to be homogeneous
and independent of lattice defects. These results strongly suggest
that the Al-Ti alloy deposits in the disordered state and that ordering
occurs in the solid state, subsequent to the charge-transfer step and
adatom incorporation into the lattice. The degree of chemical order-
disorder is directly linked to the deposition temperature.
L12 structure due to the lack of superlattice reflections in the XRD
pattern.22
In order to confirm the chemical disorder in the Al-Ti deposited
from AlCl3-EtMeImCl melt, electron diffraction analysis of selected
electrodeposits in both plan view and cross section was performed
͑Fig. 12͒. A typical diffraction pattern taken on the ͓001͔ zone axis
from a deposit containing 18.4 a/o Ti metal is shown in Fig. 12a.
The pattern is consistent with single-phase fcc Al-Ti having a lattice
parameter close to 0.405 nm. No superlattice reflections were found
in any of the diffraction patterns indicating that the detected phase
has a random ordering of Al and Ti atoms. For comparison, the
diffraction patterns from films containing 16.0 a/o Ti and 24.0 a/o
Ti, both deposited from AlCl3-NaCl at 423.2 K are shown in Fig.
12b and c, respectively.20 The 100 and 110 superlattice reflections
are clearly visible, and their intensities grow progressively relative
to the fundamental 200 spots as the titanium concentration is in-
creased. The presence of superlattice reflections indicates that the
crystal structure of the 423.2 K deposits is not disordered fcc be-
cause the 100 and 110 reflections are forbidden for the fcc lattice. It
is also clear from Fig. 12 that the chemical disorder of the room
temperature melt deposit is not simply due to the low Ti content.
The 16.0 a/o Ti alloy deposited from AlCl3-NaCl clearly shows the
superlattice reflections indicative of L12 ordering, Fig. 12b, even
though the composition is less than stoichiometric Al3Ti.
Pitting potential measurements.—Potentiodynamic anodic polar-
ization curves recorded in deaerated aqueous NaCl for Al-Ti alloys
electrodeposited on copper wire electrodes are shown in Fig. 13. As
noted for Al-Mn alloys examined under similar conditions,37 Al-Ti
alloys display a stable passive region characterized by a very small
The electrodeposited Al3Ti described above displays crystallo-
graphic features of distinctly different length scales. The grain size
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

C242
Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
Figure 11. ͑a͒ Calculated XRD pattern for fully ordered (L12) Al3Ti34 and experimental XRD patterns ͑Cu K␣͒ from as-deposited Al-Ti alloys containing ͑b͒
26.3 a/o Ti that was deposited from 66.7 m/o AlCl3-NaCl at 423.2 K, and ͑c͒ 18.4 a/o Ti that was deposited from 66.7 m/o AlCl3-EtMeImCl at 353.2 K.
potential-independent current followed by a sudden rise in current at
the pitting potential. The variation of the pitting potential with alloy
composition is shown in Fig. 14. The addition of ϳ8 a/o Ti metal
increases the pitting potential of the alloy by about 0.30 V; however,
alloys containing ϳ15 a/o Ti metal show only a 0.45 V increase in
pitting potential relative to pure Al. Thus, the greatest benefit is
realized by the addition of a relatively modest amount of Ti metal.
This same result was noted for electrodeposited Al-Mn alloys, ex-
cept that the formation of a two-phase fcc ϩ glass structure for
alloys containing more than 5 a/o Mn led to a decrease in the pitting
Figure 12. Selected area diffraction patterns of Al-Ti alloys taken on the ͓001͔ zone axis from as-deposited Al-Ti alloys containing ͑a͒ 18.4 a/o Ti deposited
from 66.7 m/o AlCl3-EtMeImCl at 353.2 K, ͑b͒ 16.0 a/o Ti deposited from 66.7 m/o AlCl3-NaCl at 423.2 K, and ͑c͒ 24.0 a/o Ti deposited from 66.7 m/o
AlCl3-NaCl at 423.2 K.20
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org

Journal of The Electrochemical Society, 150 ͑4͒ C234-C243 ͑2003͒
Acknowledgments
C243
This research was supported by the Air Force Office of Scientific
Research, Grant no. F49620-00-1-0123.
The University of Mississippi assisted in meeting the publication costs of
this article.
References
1. G. S. Frankel, R. C. Newman, C. V. Jahnes, and M. A. Russak, J. Electrochem.
Soc., 140, 2192 ͑1993͒.
2. J. R. Davis, Corrosion of Aluminum and Aluminum Alloys, ASM International,
Materials Park, OH ͑1999͒.
3. W. C. Moshier, G. D. Davis, and G. O. Cote, J. Electrochem. Soc., 136, 356 ͑1989͒.
4. P. Furrer and H. Warlimont, Mater. Sci. Eng., 28, 127 ͑1977͒.
5. M. Fass, D. Itzhak, D. Eliezer, and F. H. Froes, J. Mater. Sci., 6, 1227 ͑1987͒.
6. A. H. Al-Saffar, V. Ashworth, A. K. O. Bairamov, D. J. Chivers, W. A. Grant, and
R. P. M. Procter, Corros. Sci., 20, 127 ͑1980͒.
7. M. V. Zeller and J. A. Kargol, Appl. Surf. Sci., 18, 63 ͑1984͒.
8. P. M. Natishan, E. McCafferty, and G. K. Hubler, J. Electrochem. Soc., 135, 321
͑1988͒.
9. A. S. Edelstein, R. K. Everett, J. H. Perepezko, and M. H. da Silva Bassani, J.
Mater. Res., 12, 385 ͑1997͒.
10. K. Honda, A. Hirose, and K. F. Kobayashi, Mater. Sci. Eng., A, 222, 212 ͑1996͒.
11. H. C. Chen and E. Pfender, Thin Solid Films, 280, 188 ͑1996͒.
12. W. C. Moshier, G. D. Davis, J. S. Ahearn, and H. F. Hough, J. Electrochem. Soc.,
134, 2677 ͑1987͒.
13. G. S. Frankel, M. A. Russak, C. V. Jahnes, M. Mirzamaani, and V. A. Brusic, J.
Electrochem. Soc., 136, 1243 ͑1989͒.
14. G. D. Davis, W. C. Moshier, T. L. Fritz, and G. O. Cote, J. Electrochem. Soc., 137,
422 ͑1990͒.
15. B. A. Shaw, G. D. Davis, T. L. Fritz, B. J. Rees, and W. C. Moshier, J. Electrochem.
Soc., 138, 3288 ͑1991͒.
Figure 13. Potentiodynamic polarization curves for Al and Al-Ti alloys in
deaerated 0.1 M aqueous NaCl: ͑a͒ Al ͑99.999%͒, ͑b͒ Al92.4Ti7.6 , ͑c͒
Al88.6Ti11.4 , and ͑d͒ Al84.3Ti15.7 . The potential sweep rates were 0.5 mV sϪ1
.
16. R. Banerjee, R. Ahuja, and H. L. Fraser, Phys. Rev. Lett., 76, 3778 ͑1996͒.
17. M. Fahoume, M. El Khamlichi, C. Baltzinger, and C. Burggraf, Mater. Res. (Zug.
Switz), 1-2, 391 ͑1994͒.
18. K. W. Fung and G. Mamantov, J. Electroanal. Chem., 35, 27 ͑1972͒.
19. R. T. Carlin, R. A. Osteryoung, J. S. Wilkes, and J. Rovang, Inorg. Chem., 29, 3003
͑1990͒.
potential.37 The increase in the pitting potential observed for elec-
trodeposited Al-Ti alloys is comparable to that found for Al-Ti al-
loys prepared by sputter deposition.13
20. G. M. Janowski and G. R. Stafford, Metall. Mater. Trans. A, 23, 2715 ͑1992͒.
21. G. R. Stafford, J. Electrochem. Soc., 141, 245 ͑1994͒.
22. J. Uchida, H. Seto, and A. Shibuya, J. Surf. Finishing Soc. of Jpn, (Hyomen Gi-
jutsu), 46, 1167 ͑1995͒.
23. H.-Y. Hsu, D.-L. Chen, H.-W. Tsaur, and C.-C. Yang, in Molten Salts XII,’’ P. C.
Trulove, H. C. De Long, G. R. Stafford, and S. Deki, Editors, PV 99-41, p. 585,
The Electrochemical Proceedings Series, Pennington, NJ ͑1999͒.
24. T. Takenaka, A. Hoshikawa, and M. Kawakami, in Molen Salt Chemistry and
Technology ͑1993͒ M.-L. Saboungi and H. Kojima, Editors, PV 93-9, p. 184, The
Electrochemical Society Proceedings Series, Pennington NJ ͑1993͒.
25. B. J. Tierney, W. R. Pitner, J. A. Mitchell, C. L. Hussey, and G. R. Stafford, J.
Electrochem. Soc., 145, 3110 ͑1998͒.
26. X. H. Xu and C. L. Hussey, J. Electrochem. Soc., 140, 1226 ͑1993͒.
27. L. J. Durney, Electroplating Engineering Handbook, Van Nostrand Reinhold Co.,
New York ͑1984͒.
28. A. J. Dent, K. R. Seddon, and T. Welton, J. Chem. Soc. Chem. Commun., 1990, 315
͑1990͒.
29. M. Soerlie and H. A. Oeye, Inorg. Chem., 20, 1384 ͑1981͒.
30. G. R. Stafford and T. P. Moffat, J. Electrochem. Soc., 142, 3288 ͑1995͒.
31. J. Brynestad, S. Von Winbush, H. L. Yakel, and G. P. Smith, Inorg. Nucl. Chem.
Lett., 6, 889 ͑1970͒.
32. A. A. Fannin, Jr., D. A. Floreani, L. A. King, J. S. Landers, B. J. Piersma, D. J.
Stech, R. L. Vaughn, J. S. Wilkes, and L. J. Williams, J. Phys. Chem., 88, 2614
͑1984͒.
33. G. R. Stafford, L. A. Bendersky, and G. M. Janowski, Editors, Defect Structure,
Morphology, and Properties of Vapor and Electrodeposits, The Mineral, Metals
and Materials Society, Warrendale, PA ͑1984͒.
34. D. K. Smith, M. C. Nichols, and M. E. Zolensky, A Fortran IV Program for
Calculating X-Ray Powder Diffraction Patterns-Version 10, Department of Geo-
sciences, The Pennsylvania State University, University Park, PA ͑1983͒.
35. W. J. Boettinger, L. A. Bendersky, J. A. Cline, J. A. West, and M. J. Aziz, Mater.
Sci. Eng., A, 133, 592 ͑1991͒.
Figure 14. Pitting potentials as a function of the Ti content of the Al-Ti
alloys in Fig. 13: ͑a͒ Al ͑99.999%͒, ͑b͒ Al92.4Ti7.6 , ͑c͒ Al88.6Ti11.4 , and ͑d͒
36. D. A. Porter and K. E. Easterling, Phase Transformations in Metals and Alloys, Van
Nostrand Reinhold, Co, Ltd., Berkshire, UK ͑1981͒.
Al84.3Ti15.7
.
37. T. P. Moffat, G. R. Stafford, and D. E. Hall, J. Electrochem. Soc., 140, 2779 ͑1993͒.
Downloaded on 2012-10-05 to IP 128.143.23.241 address. Redistribution subject to ECS license or copyright; see www.esltbd.org