Galvanic deposition of nanoporous Si onto 6061 Al alloy from aqueous HF

Article abstract of DOI:10.1149/1.3521290

We report galvanic deposition of Si onto 6061 Al alloy from dilute aqueous hydrofluoric acid (HF) at pH 2.5. The overall reaction involves reduction of SiF₆^2- to Si with simultaneous oxidation and dissolution of Al. The Si film is ab

Full text of DOI:10.1149/1.3521290

D68
Journal of The Electrochemical Society, 158 ͑2͒ D68-D71 ͑2011͒
0013-4651/2010/158͑2͒/D68/4/$28.00 © The Electrochemical Society
Galvanic Deposition of Nanoporous Si onto 6061 Al Alloy
from Aqueous HF
Aarti Krishnamurthy,a Don H. Rasmussen,b,c and Ian I. Sunib,c,
,z
*
aDepartment of Chemistry and Biomolecular Science, bDepartment of Chemical and Biomolecular
Engineering, and cMaterials Science and Engineering PhD Program, Clarkson University, Potsdam, New
York 13699-5705, USA
We report galvanic deposition of Si onto 6061 Al alloy from dilute aqueous hydrofluoric acid ͑HF͒ at pH 2.5. The overall reaction
involves reduction of SiF62− to Si with simultaneous oxidation and dissolution of Al. The Si film is about 12 ␮m thick after 6 h of
deposition. High resolution scanning electron microscopy shows that these Si films are nanoporous, with pore sizes ranging from
3 to 8 nm. The nanoporous Si films oxidize rapidly upon sample emersion. Elemental analysis by energy dispersive X-ray
spectroscopy demonstrates that the as-deposited film contains 1–3 atom % Al, 3–6 atom % Cu, and 90–95 atom % Si. We believe
that this is the first report of electrochemical deposition of Si thin films that does not involve organic solvents or molten salt
electrolytes.
© 2010 The Electrochemical Society. ͓DOI: 10.1149/1.3521290͔ All rights reserved.
Manuscript submitted August 3, 2010; revised manuscript received November 4, 2010. Published December 3, 2010.
While the market for photovoltaic cells is currently dominated by
Mn, Ti, and Zn. For some of the galvanic deposition experiments,
99.99% pure Al was purchased from ESPI Metals. For Si electro-
chemistry, B-doped ͑2 ϫ 1019 cm−3͒ degenerate Si͑100͒ wafers
with a resistivity of 0.001–0.005 ⍀ cm were purchased from Uni-
versity Wafer. The electrical connection to the Si wafer’s back side
was made using a Ga–In eutectic.
Voltammetry experiments were controlled with an EG&G PAR
model 273A potentiostat/galvanostat. Impedance measurements
were made by coupling this potentiostat with a Solartron 1250B
frequency response analyzer over the frequency range
0.01 Hz–10 kHz, using an ac probe voltage of 8 mV. The Si film
thickness was measured with a JEOL model 7400F field emission
scanning electron microscope at both 45 and 90°.
thick film Si solar cells, thin film crystalline, polycrystalline, and
amorphous Si solar cells have also been intensively investigated.1
Optical absorption in Si solar cells occurs mainly within the top
micrometer of Si, so the rest of the Si wafer in thick film solar cells
merely provides mechanical support. While thick film Si solar cells
can directly employ Si wafer technology and materials from inte-
grated circuit manufacturing, thin film Si solar cells provide obvious
long-term cost advantages. In addition to photovoltaic applications,
Si thin films are of interest for silicon-on-sapphire complementary
metal oxide semiconductor technology,2 for anode materials in Li
ion batteries,3,4 and for corrosion-resistant coatings.5
Si thin films are typically deposited by expensive vacuum meth-
ods such as chemical vapor deposition and plasma-enhanced chemi-
cal vapor deposition.1 In addition, Si thin films are typically depos-
ited from silane, which is both highly pyrophoric and moisture
sensitive. Electrochemical methods for depositing thin film solar cell
materials are highly advantageous due to their low cost, scalability
to large surface areas, and manufacturability.6 However, Si is a
highly active metal, so the standard reduction potential of SiO2
͓−0.90 V vs normal hydrogen electrode ͑NHE͔͒ is more cathodic
than the standard reduction potential of water ͑−0.83 V vs NHE͒,7
making electrochemical deposition of Si from aqueous electrolytes
notoriously difficult.
Results and Discussion
Figure 1 illustrates a voltammogram of the Al working electrode
rotated at 850 rpm in 10 mM HF + 1 mM HNO3 ͑pH 2.5͒. An ad-
dition of 20 mM Na2SiF6 to this electrolyte had no discernable ef-
fect on the voltammetry results. Figure 1 illustrates that this electro-
lyte is quite corrosive to Al, with anodic currents from Al oxidation
and dissolution observed at all potentials anodic to −1000 mV vs
SCE. This is the reason why such a high scan rate ͑50 mV/s͒ was
employed. Immersion of the Al working electrode rotated at
850 rpm without potential control into 10 mM HF, 1 mM HNO3,
and 20 mM Na2SiF6 for 6 h results in the growth of a Si film about
12 ␮m thick. The open circuit potential was measured during Si
deposition and varied between −700 and −900 mV vs SCE, as
shown in Fig. 2. The as-grown film is dark gray in solution but
changes color to light gray after exposure to laboratory air for 1 h
and then to white upon overnight exposure.
Figures 3 and 4 present scanning electron microscopy ͑SEM͒
images following growth of a 12 ␮m Si film as described above.
Figure 3 illustrates the 12 ␮m Si film ͑middle͒ atop the Al substrate
͑bottom͒ and appears to show a compact Si deposit. However, the
higher resolution image in Fig. 4 shows that the Si deposit contains
nanoscale porosity, with pore sizes ranging from 3 to 8 nm. Other
methods that have been reported for room temperature Si elec-
trodeposition similarly yield porous Si films, as discussed
elsewhere.15 X-ray diffraction studies of our galvanic Si films, both
immediately after deposition and after overnight ambient exposure,
show no diffraction peaks, indicating that our deposits are amor-
phous.
Si electrodeposition from molten salts at elevated temperatures
͑ ജ 750°C͒ has a long history.8-11 However, Si electrodeposition at
room temperature has only relatively recently been achieved from
organic solvents12-17 and from room temperature ionic liquids.18-23
Here we report galvanic deposition of nanoporous Si onto Al from
solutions of dilute aqueous hydrofluoric acid ͑HF͒ at pH 2.5, with
12 ␮m thick Si films grown after 6 h. Energy dispersive X-ray spec-
troscopy ͑EDX͒ measurements show that the as-deposited film con-
tains mainly Si, Cu, and Al.
Experimental
Semiconductor grade 10 wt % HF and concentrated HNO3 were
obtained from J. T. Baker, Na2SiF6 was obtained from Sigma-
Alrdich, and 6061 Al alloy was obtained from McMaster Carr. For
Al electrochemistry, all measurements were performed using a
three-electrode setup with a 12 mm diameter 6061 Al alloy working
electrode rotated at 850 rpm with a rotating disc electrode, Pt spiral
counter electrode, and a reference SCE. A 6061 Al alloy typically
contains 0.8–1.2 wt % Mg, 0.4–0.8 wt % Si, ഛ0.70 wt % Mg,
0.15–0.40 wt % Cu, 0.04–0.35 wt % Cr, and smaller amounts of
The Si deposit is grown atop Al by galvanic deposition, other-
wise known as immersion plating, where a more noble metal is
reduced and deposited onto a less noble substrate that is simulta-
neously oxidized and dissolved. The cathodic ͑1͒, anodic ͑2͒, and
overall ͑3͒ reactions for galvanic deposition of Si onto Al are7
*
Electrochemical Society Active Member.
z E-mail: isuni@clarkson.edu
Downloaded on 2014-06-13 to IP 129.89.24.43 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 ͑2͒ D68-D71 ͑2011͒
D69
3
2
1
0
-1
-2
-3
-1400
-1200
-1000
-800
-600
-400
-200
0
Potential (mV vs. SCE)
Figure 1. Voltammetry of the 6061 Al electrode rotated at 850 rpm in
10 mM HF + 1 mM HNO3 at a scan rate of 50 mV/s.
Figure 4. SEM image of Si deposit at 1,20,000 ϫ magnification, illustrating
nanoscale porosity.
SiF26− + 4e− → Si + 6F− E° = − 1.37 V vs NHE
Al + 6F− → AlF36− + 3e− E° = + 2.07 V vs NHE
͓1͔
͓2͔
3SiF26− + 4Al + 6F− → 3Si + 4AlF36− E° = + 0.70 V ͓3͔
Reaction 2 is written as an oxidation reaction, so the potential given
is the opposite of the standard reduction potential. The standard
potential given for the overall reaction 3 is positive, indicating a
spontaneous reaction. Although reactions 1-3 are most obvious, one
cannot rule out the possibility of intermediate oxidation states of Si
and Al being involved. Reactions similar to those above have been
widely reported for galvanic deposition of a variety of metals onto
Si from aqueous HF solutions, as recently reviewed.24,25 Such pro-
cesses were originally studied to understand contamination of Si
wafers during wet cleaning with HF but have also been proposed for
a variety of technologies involving fabrication of metal nanostruc-
tures and thin films.24,25
Figure 2. ͑Color online͒ Open circuit potential measured as a function of
time at the Al electrode during Si deposition from 10 mM HF, 1 mM HNO3,
and 20 mM Na2SiF6.
However, such methods are not limited to Si substrates, because
our laboratory has reported galvanic deposition of Cu onto Ta from
aqueous HF.26 The present research extends such methods to Al
substrates and to deposition of Si films. The current method for
galvanic deposition of Si may be limited to a few highly active
metal substrates, such as Al, that are anodic to Si in HF.7 Fortu-
nately, Al is the substrate onto which amorphous Si is typically
deposited within thin film solar cells.1 The current process is also
quite similar to the double zincate process for galvanic deposition of
Zn onto Al, as will be described in more detail below.
The EDX spectrum in Fig. 5 illustrates that the deposit contains
Si, Al, Cu, and O. The significant O peak in the EDX spectrum is
believed to arise from sample transfer through air, as has been pre-
viously observed by other research groups.15 Counting only the
heavy elements, EDX analyses of several Si deposits yield the fol-
lowing range of compositions: 1–3 atom % Al, 3–6 atom % Cu, and
90–95 atom % Si. Thus, the Si films contain significant Al, which
probably cannot be avoided, because Al oxidation, dissolution, and
transport through the growing Si film is part of the galvanic Si
deposition process described above.
The Si films also contain significant Cu, which probably arises
from the 6061 Al alloy itself, not from the solution phase, for the
following reasons. First, galvanic Si deposition onto 99.99% pure Al
results in a Si deposit that is removed by convective forces associ-
ated with electrode rotation at 850 rpm, resulting in Si flakes at the
bottom of the cell. EDX analysis of these Si flakes does not show
significant Cu incorporation. Second, atomic absorption spectros-
Figure 3. SEM image at a 45° angle after 6 h of Si deposition from 10 mM
HF, 1 mM HNO3, and 20 mM Na2SiF6. The Al substrate is at the bottom and
the Si film is in the middle.
Downloaded on 2014-06-13 to IP 129.89.24.43 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D70
Journal of The Electrochemical Society, 158 ͑2͒ D68-D71 ͑2011͒
Element
Line
Net
Counts
7074
351
25583
235
Weight %
Weight %
Error
+/- 0.60
+/- 0.08
+/- 0.34
+/- 0.62
Atom %
Atom %
Error
+/- 0.78
+/- 0.06
+/- 0.26
+/- 0.20
O K
47.86
0.69
62.86
0.53
Al K
Si K
46.93
4.53
35.11
1.50
Cu K
Total
100.00
100.00
Figure 5. ͑Color online͒ EDX results after 6 h of Si deposition from 10 mM
HF, 1 mM HNO3, and 20 mM Na2SiF6.
Figure 6. Nyquist plot illustrating the impedance results for degenerate Si at
−800 mV vs SCE in 10 mM HF + 1 mM HNO3.
copy shows that the galvanic deposition bath contains only 140 ppb
Cu, which is likely too low to yield the observed levels of Cu in-
corporation. It is possible that some of Si in the galvanic deposit
arises from the 6061 Al alloy substrate as well.
and Si–O vibrational modes, Chazalviel and co-workers have
closely studied the transition between H- and O-terminated surfaces
as a function of HF concentration and applied potential.32-34 Even in
dilute HF solutions, they observe
a transition from H- to
Si adhesion was not quantitatively studied. Si film adhesion was
good enough to survive convection associated with 850 rpm elec-
trode rotation, but the Si film could be scratched off following
sample emersion. However, the film adhesion was greatly improved
after sample drying. As mentioned above, galvanic Si deposition
from identical solutions onto 99.99% pure Al yielded poorly adher-
ent deposits, so obtaining adequate adhesion depends on the pres-
ence of other elements in the 6061 Al alloy. Cu in this alloy is noble
to both Al and Si, so Cu should either not dissolve or immediately
redeposit in the galvanic deposition bath. We suggest that the Si film
is anchored to Cu atoms within the 6061 Al alloy as Al dissolves
from within the alloy.
O-terminated Si at potentials considerably anodic to those observed
here.32-34 This supports our hypothesis that Si is not oxidized in the
solutions studied here. The observations of Chazalviel and co-
workers, and those of other researchers, have been incorporated into
models of Si etching in HF solutions, where the balance between H-
and O-terminated Si surface coverage is believed to depend on a
complex balance between kinetic and thermodynamic factors.31
Even at more anodic potentials where Si etching begins, the reactiv-
ity of the surface is determined by the number of O-terminated sites,
but the fractional surface coverage of these highly reactive sites is
believed to remain small, with the bulk of the surface still
H-terminated.31
The mechanism of galvanic Si deposition may be similar to that
of galvanic Zn deposition onto Al, which is widely used for double
zincate treatment of Al alloys to prepare the surface for subsequent
electrodeposition or electroless deposition.27-30 Cu pretreatment has
been employed to improve Zn nucleation during the double zincate
process30 and increased Cu concentration at the Al–metal interface
has been reported.28 In addition, variations in the effectiveness of
double zincate treatment with Al alloy content have been reported.30
As mentioned above, we expect that the Si surface deposited in
10 mM HF + 1 mM HNO3 ͑pH 2.5͒ is not oxidized in situ, but is
instead oxidized during sample emersion. Si oxidation in situ would
cause reactions 1-3 to cease, because the surface would be coated
with an electrical insulator. In addition, the color change observed
upon sample emersion is consistent with gradual Si oxidation. One
might expect possible Si oxidation in 10 mM HF + 1 mM HNO3
͑pH 2.5͒ to be well-understood, but this is not the case, because the
Si film is deposited under conditions quite different from those em-
ployed for HF cleaning of Si. Here we report growth of a heavily
Al-doped Si film from solutions with low F− concentrations
͑10 mM͒ at quite cathodic potentials ͑−700 to − 900 mV vs SCE͒.
By comparison, HF cleaning of Si wafers involves much higher F−
concentrations ͑100 mM to 5.0 M͒ of moderately doped n- or
p-type Si. In addition, Si electrochemistry in HF has been studied
primarily at much more anodic potentials, where Si dissolution oc-
curs with porous Si formation or with electropolishing.31
Insight into the nature of the Si-electrolyte surface can also be
obtained from electrochemical studies of highly doped Si, where no
space charge layer forms at either Si interface ͑electrolyte or back
side electrical contact͒. Voltammetry studies of degenerate Si in
10 mM HF + 1 mM HNO3 ͑not shown͒ do not reveal any cathodic
or anodic features, so impedance studies of degenerate Si probe only
the nature of the electrochemical interface. Figure 6 presents a Ny-
quist plot of the results from electrochemical impedance spectros-
copy ͑EIS͒ for degenerate Si in 10 mM HF + 1 mM HNO3 at
−800 mV vs SCE. These data were fit to a Randles equivalent cir-
cuit, with the differential capacitance replaced by a constant phase
element, yielding the best-fit equivalent circuit parameters shown in
Table I.
These results can be compared to the detailed impedance study
of Searson and Zhang of Si in HF at conditions close to those re-
Table I. Best-fit equivalent circuit parameters for impedance
studies of degenerate Si at −800 mV vs SCE in 10 mM HF
+ 1 mM HNO3.
Potential
Rs
Td ϫ 106
Rct
͑mV vs SCE͒ ͑⍀ cm2͒ ͑␮F cm−2 sn−1
͒
nd
͑⍀ cm2͒
Using infrared spectroscopy for direct observation of the Si–H
−800
147 ͑3͒ 3.40 ͑10͒
0.944 ͑5͒ 7.51͑21͒ ϫ 104
Downloaded on 2014-06-13 to IP 129.89.24.43 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 ͑2͒ D68-D71 ͑2011͒
ported here.35 They distinguish between different surface conditions
Acknowledgments
D71
partly by the magnitude of the measured capacitance, which is ex-
pected to be in the nF cm2, ␮F cm−2, and mF cm−2 range for a space
charge layer, Helmholtz layer, and oxide layer, respectively.35,36
Thus, a best-fit capacitance value in the ␮F cm−2 range is considered
as evidence of a Helmholtz layer in potential regimes and at dopant
conditions where the surface is not oxidized and no space charge
layer forms. The results in Table I are consistent with this criterion,
supporting our assertion that Si oxide does not form in the solutions
studied here.
This research was supported by U.S. Army grant no. W911NF-
05-1-0339 and the Center for Advanced Materials Processing
͑CAMP͒ at Clarkson University.
Clarkson University assisted in meeting the publication costs of this ar-
ticle.
References
1. Clean Electricity from Photovoltaics, M. D. Archer and R. Hill, Editors, Imperial
College Press, London ͑2001͒.
In addition, Searson and Zhang observe that the charge transfer
resistance ͑Rct͒ increases from 101 to 103 ⍀ cm2 at potentials where
Si dissolution occurs from 104 to 105⍀ cm2 at more cathodic poten-
tials, such as those studied here, where the surface is likely
H-terminated.35 The charge transfer resistance observed here
͑7.51 ϫ 104 ⍀ cm2͒ is consistent with the results of Searson and
Zhang. These results are consistent with the general observation that
H-terminated Si is much more stable and unreactive than
O-terminated Si, yielding somewhat higher Rct at more cathodic
potentials.31,35
2. R. A. Johnson, P. R. de la Houssaye, C. E. Chang, P. F. Chen, M. E. Wood, G. A.
Garcia, I. Lagnado, and P. M. Asbeck, IEEE Trans. Electron Devices, 45, 1047
͑1998͒.
3. L. B. Chen, K. Wang, X. H. Xie, and J. Y. Xie, Electrochem. Solid-State Lett., 9,
A512 ͑2006͒.
4. T. L. Kulova, A. M. Skundin, Y. V. Pleskov, E. I. Terukov, and O. I. Kon’kov, J.
Electroanal. Chem., 600, 217 ͑2007͒.
5. S. Guruvenket, M. Azzi, D. Li, J. A. Szpunar, L. Martinu, and J. E. Klemberg-
Sapieha, Surf. Coat. Technol., 204, 3358 ͑2010͒.
6. I. M. Dharmadasaz and J. Haigh, J. Electrochem. Soc., 153, G47 ͑2006͒.
7. Standard Potentials in Aqueous Solution, A. J. Bard, R. Parsons, and J. Jordan,
Editors, Marcel Dekker, New York ͑1985͒.
The galvanic Si thin films deposited here can be compared with
Si thin films electrodeposited at room temperature from organic
solvents12-17 and room temperature from ionic liquids ͑RTILs͒.18-23
This is complicated by the wide variation in reported results be-
tween different research groups, which makes comparisons difficult.
However, it has been suggested that compact Si film electrodeposi-
tion from organic solvents is limited to deposits ഛ250 nm thick.20
In addition, it has been suggested that Si films electrodeposited from
organic solvents are always porous and oxidize rapidly upon ambi-
ent exposure.15 Si thin films electrodeposited from RTILs onto flat
substrates appear to have been limited to date to a thickness of
several hundred nanometers. Regardless of the veracity of these gen-
eralizations, thick and compact Si films have not been grown to date
by room temperature electrodeposition. In addition, many of the
reported Si films are contaminated by impurity elements originating
from the solvent. The galvanic Si films reported here suffer from
many of these same issues. The primary advantages of the current
method include its simplicity, the modest cost and toxicity of the
reagents employed, and the ability to grow thick 12 ␮m Si films.
8. G. M. Rao, D. Elwell, and R. S. Feigelson, J. Electrochem. Soc., 127, 1940 ͑1980͒.
9. R. C. de Matthei, D. Elwell, and R. S. Feigelson, J. Electrochem. Soc., 128, 1712
͑1981͒.
10. R. Boen and J. Bouteillon, J. Appl. Electrochem., 13, 277 ͑1983͒.
11. K. L. Carleton, J. M. Olson, A. Kibbler, and R. S. Feigelson, J. Electrochem. Soc.,
130, 782 ͑1983͒.
12. A. K. Agrawal and A. E. Austin, J. Electrochem. Soc., 128, 2292 ͑1981͒.
13. C. H. Lee and F. A. Kroger, J. Electrochem. Soc., 129, 936 ͑1982͒.
14. P. R. Sarma, T. R. R. Mohan, S. Venkatachalam, J. Singh, and V. P. Sundesingh,
Mater. Sci. Eng., B, 15, 237 ͑1992͒.
15. J. P. Nicholson, J. Electrochem. Soc., 152, C795 ͑2005͒.
16. Y. Nishimura and Y. Fukunaka, Electrochim. Acta, 53, 111 ͑2007͒.
17. T. Munisamy and A. J. Bard, Electrochim. Acta, 55, 3797 ͑2010͒.
18. S. Z. el Abedin, N. Borissenko, and F. Endres, Electrochem. Commun., 6, 510
͑2004͒.
19. N. Borissenko, S. Z. el Abedin, and F. Endres, J. Phys. Chem. B, 110, 6250 ͑2006͒.
20. S. Z. el Abedin and F. Endres, Acc. Chem. Res., 40, 1106 ͑2007͒.
21. R. Al-Salman, S. Z. el Abedin, and F. Endres, Phys. Chem. Chem. Phys., 10, 4650
͑2008͒.
22. J. Mallet, M. Molinari, F. Martineau, F. Delavole, P. Fricoteaux, and M. Troyon,
Nano Lett., 8, 3468 ͑2008͒.
23. Y. Nishimura, Y. Fukunaka, T. Nishida, T. Nohira, and R. Hagiwara, Electrochem.
Solid-State Lett., 11, D75 ͑2008͒.
24. S. Yae, N. Nasu, K. Matsumoto, T. Hagihara, N. Fukumuro, and H. Matsuda,
Electrochim. Acta, 53, 35 ͑2007͒.
25. C. Carraro, R. Maboudian, and L. Magagnin, Surf. Sci. Rep., 62, 499 ͑2007͒.
26. S. Sapra, H. Li, Z. Wang, and I. I. Suni, J. Electrochem. Soc., 152, B193 ͑2005͒.
27. D. S. Lashmore, Plat. Surf. Finish., 72, 36 ͑1985͒.
28. F. J. Monteiro and D. H. Ross, Trans. Inst. Met. Finish., 62, 155 ͑1985͒.
29. D. A. Hutt, C. Q. Liu, P. P. Conway, D. C. Whalley, and S. H. Mannan, IEEE
Trans. Compon. Packag. Technol., 25, 87 ͑2002͒.
30. K. Azumi, S. Egoshi, S. Kawashima, and Y. Koyama, J. Electrochem. Soc., 154,
D220 ͑2007͒.
31. K. W. Kolasinski, Phys. Chem. Chem. Phys., 5, 1270 ͑2003͒.
32. C. da Fonseca, F. Ozanam, and J. N. Chazalviel, Surf. Sci., 365, 1 ͑1996͒.
33. A. Belaidi, M. Safi, F. Ozanam, J. N. Chazalviel, and O. Gorochov, J. Electrochem.
Soc., 146, 2659 ͑1999͒.
Conclusions
Galvanic deposition of Si onto 6061 Al alloy occurs from solu-
tions containing 10 mM HF, 1 mM HNO3, and 20 mM Na2SiF6. Si
films about 12 ␮m thick are formed after 6 h of deposition. High
resolution SEM indicates that the Si films are nanoporous, with pore
sizes ranging from 3 to 8 nm. For this reason, the Si film appears to
oxidize rapidly upon sample emersion. However, in situ measure-
ments by EIS on degenerate Si yield capacitance and charger trans-
fer resistance ͑Rct͒ values consistent with an H-terminated Si surface
rather than an oxidized surface. This is consistent with previously
reported studies of Si in dilute HF by infrared spectroscopy, con-
firming that Si is not oxidized in situ. Elemental analysis by EDX
demonstrate that the as-deposited film contains 1–3 atom % Al, 3–6
atom % Cu, and 90–95 atom % Si.
34. J. N. Chazalviel, A. Belaidi, M. Safi, F. Maroun, B. H. Erne, and F. Onazam,
Electrochim. Acta, 45, 3205 ͑2000͒.
35. P. C. Searson and X. G. Zhang, J. Electrochem. Soc., 137, 2539 ͑1990͒.
36. J. C. Lin, C. M. Lai, W. D. Jehng, K. L. Huseh, and S. L. Lee, J. Electrochem. Soc.,
155, D436 ͑2008͒.
Downloaded on 2014-06-13 to IP 129.89.24.43 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).