Different behavior in immersion plating of nickel on porous silicon from acidic and alkaline fluoride media

Article abstract of DOI:10.1149/1.1562595

Immersion plating of nickel (Ni) on a porous silicon (PS) layer has been investigated in concentrated ammonium fluoride (NH₄F) and dilute hydrofluoric acid (HF) solutions containing Ni²⁺. When the PS sample was exposed to the bath containing 5 M NH₄F and 50 mM Ni²⁺ at pH 8, metallic Ni was clearly observed at the open-circuit potential without using a reducing agent or any activation treatment at room temperature. However, no deposition was detected when the sample was immersed in a 0.5 wt% HF solution containing the same concentration of Ni²⁺ at pH 2. The different deposition behavior is discussed on the basis of mixed potential theory, changes in the stability of Si-H bonds of PS as indicated in the Fourier transform infrared spectra, and the different state of Ni complex formation as obtained from the UV-vis spectra. The position of the Ni redox levels in both solutions with respect to the bandedges of Si was also determined and the results revealed a nearly similar energetic situation. Results from current-potential curves showed that a slow Ni deposition from the alkaline solution occurred in the dark and different possible reaction mechanisms were proposed. They also revealed the enhancement of the deposition under illumination, indicating that the deposition mechanism is electron transfer from the conduction band. X-ray photoelectron spectroscopy detected no silicon oxides on the PS surface after being immersed in the alkaline solution, giving merit that a binary PS/Ni nanostructure without Si oxides can be successfully achieved.

Full text of DOI:10.1149/1.1562595

Journal of The Electrochemical Society, 150 ͑5͒ C277-C284 ͑2003͒
C277
0
013-4651/2003/150͑5͒/C277/8/$7.00 © The Electrochemical Society, Inc.
Different Behavior in Immersion Plating of Nickel on Porous
Silicon from Acidic and Alkaline Fluoride Media
z
Farid A. Harraz, Junji Sasano, Tetsuo Sakka,* and Yukio H. Ogata*
Institute of Advanced Energy, Kyoto University, Uji, Kyoto 611-0011, Japan
Immersion plating of nickel ͑Ni͒ on a porous silicon ͑PS͒ layer has been investigated in concentrated ammonium fluoride (NH F)
4
2
ϩ
and dilute hydrofluoric acid ͑HF͒ solutions containing Ni . When the PS sample was exposed to the bath containing 5 M NH4F
and 50 mM Ni at pH 8, metallic Ni was clearly observed at the open-circuit potential without using a reducing agent or any
activation treatment at room temperature. However, no deposition was detected when the sample was immersed in a 0.5 wt % HF
2
ϩ
2
ϩ
solution containing the same concentration of Ni at pH 2. The different deposition behavior is discussed on the basis of mixed
potential theory, changes in the stability of Si-H bonds of PS as indicated in the Fourier transform infrared spectra, and the
different state of Ni complex formation as obtained from the UV-vis spectra. The position of the Ni redox levels in both solutions
with respect to the bandedges of Si was also determined and the results revealed a nearly similar energetic situation. Results from
current-potential curves showed that a slow Ni deposition from the alkaline solution occurred in the dark and different possible
reaction mechanisms were proposed. They also revealed the enhancement of the deposition under illumination, indicating that the
deposition mechanism is electron transfer from the conduction band. X-ray photoelectron spectroscopy detected no silicon oxides
on the PS surface after being immersed in the alkaline solution, giving merit that a binary PS/Ni nanostructure without Si oxides
can be successfully achieved.
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1562595͔ All rights reserved.
Manuscript submitted July 22, 2002; revised manuscript received November 18, 2002. Available electronically March 18, 2003.
Porous silicon ͑PS͒ is usually prepared on the surface of crystal-
line silicon ͑c-Si͒ by chemical or electrochemical etching in HF-
based solutions. The structure of the PS consists of highly branched
sition of less-noble metals to the illumination area, and hence im-
mersion plating is also a promising avenue for patterned metalliza-
tion.
The tendency of noble metals ͑e.g., Ag, Pt͒ to deposit onto the
surface of PS by immersion plating has been realized since the early
works of PS/metal deposition. In contrast, the deposition of less-
1
,2
or spongy nanopores. The surface properties of the fabricated
layer are much different from those of bulk silicon and highly de-
pendent on the type and doping level of substrate, surface orienta-
tion, and etching conditions.2 Originally the material was studied
,3
noble metals including Ni was not available in aqueous solution
4
,5
12,20,21
for its application as an insulating material in microelectronics.
containing simple metal salts.
Deposition of less-noble metal
Later on, it was observed that the porous layer shows a remarkable
luminescence at room temperature, which opened new opportuni-
on PS often has a different technological interest and wide applica-
tion in the industry than noble ones. In Ni, the process may contrib-
ute to the field of magnetic recording media. In addition, if such
less-noble metals are introduced inside the porous layer, there is a
6
ties in some technologically important areas and presented the pos-
sibility for applications in optoelectronics and display
7
,8
22
technologies.
possibility of forming silicides by subsequent heat-treatment. Con-
An existing challenge in fabricating PS devices rests with estab-
lishing electrical contact to the PS structure. The achievement of
good contact is a serious target for full realization of the potential
application in microelectronics, in addition to improving device per-
formance.
sequently, deposition of Ni on PS by simple immersion plating is a
challenging subject from both basic and practical viewpoints.
Metal deposition on semiconductor from fluoride-based solutions
has received comprehensive study by many workers in electronics
industry to evaluate silicon surface contamination and to reveal dis-
2
3,24
location and defects in silicon wafer surfaces.
HF͒ solution is frequently used as an effective etching solution
during anodization to form PS. Treatment of silicon in 40% NH4F is
Hydrofluoric acid
Different approaches have been realized to deposit metals on a
PS surface. The deposition of metallic thin films is widely per-
formed in vacuum from the vapor phase by evaporation or
͑
9
25,26
sputtering. However, wet processes including electrodeposition,
reported to produce atomically flat surfaces.
Metal deposition in
electroless plating, and immersion plating could represent an alter-
native and efficient approach for metal deposition on the PS surface
the presence of such fluoride species is expected to eliminate the
formation of an interfacial oxide layer that simultaneously formed
during deposition in aqueous solutions without fluoride ions. The
formation of such silicon oxide was observed to yield a ternary
instead of a simple binary system.
10-16
as reported by a number of workers.
We refer here to a recent
related work in which the different methods of deposition are com-
pared and summarized.17
Of particular importance in this study is the immersion plating
method owing to the following advantages: ͑i͒ In contrast to elec-
troplating, metal deposition can be achieved by simply dipping the
PS substrate in a solution containing appropriate metal ions without
any externally applied electrical bias to the PS. (ii) While conven-
tional electroless plating requires surface activation to catalyze the
deposition as well as a reducing agent in solution, no prior activation
of the surface is needed for immersion plating. (iii) The process is
usually operated at room temperature with high applicability for
mass production. In addition to the aforementioned attractive fea-
tures, immersion plating can produce deposits in the form of metal
Several researchers investigated metal deposition from fluoride
solution focusing on the deposition mechanism and aiming to evalu-
ate the contamination effect caused by various metals
2
3,27
͑Ag,Cu,Ni͒.
Most of these studies were carried out in HF-
containing electrolytes. As a recent work of Kuznetsov et al., a Pt
deposition on Si from HF solution was presented concentrating on
measuring the electrical properties of the produced structure.2 Ta-
8
Ϫ
kano et al. presented an alkaline aqueous solution without F ions
29
to deposit Ni in a conventional NiP electroless plating bath. They
also observed Ni deposition in a simple solution of Ni2 and
(NH4)2SO4 containing no reducing agent and operated at 80°C.
They could successfully fabricate minute Ni dots on silicon from
ϩ
islands that 1 could catalyze charge-transfer processes in
8,19
30
photocatalysis.
Furthermore, it is possible to localize the depo-
this alkaline bath. Homma et al. investigated the silicon surface
morphology during the initial nucleation stage of Cu deposition in
an NH F containing solution using scanning tunneling microscopy
4
3
1
͑
STM͒. Li et al. reported an electrochemical behavior of silicon in
*
Electrochemical Society Active Member.
E-mail: farid@iae.kyoto-u.ac.jp
z
dilute and buffered HF ͑HF mixed with NH F) solutions contami-
4
Downloaded on 2014-11-28 to IP 155.33.16.124 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C278
Journal of The Electrochemical Society, 150 ͑5͒ C277-C284 ͑2003͒
nated with traces of Cu2ϩ in the presence and absence of illumina-
tion by total reflection X-ray fluorescence ͑TRXRF͒
32
measurements. In a recent report, Ye et al. employed attenuated
total reflectance-Fourier transform infrared ͑ATR-FTIR͒ spectros-
copy and X-ray photoelectron spectroscopy ͑XPS͒ to investigate Cu
deposition on silicon surface from NH F and HF-containing solu-
4
33
tions to evaluate the surface chemistry of silicon.
Concerning the mechanism of metal deposition on semiconduc-
tor surfaces, generally, the deposition process can proceed through
3
4,35
the conduction band, valence band, or through surface states.
The conduction band mechanism occurs with a redox couple having
a negative equilibrium potential ͑e.g., transition metal/metal ion͒
and the charge carriers are electrons. For a redox couple with a
positive equilibrium potential ͑e.g., noble metal/metal ion͒ the depo-
sition process likely occurs via the injection of holes into the va-
lence band.
In this paper we report the results of a systematic investigation to
characterize the behavior of a PS layer on immersion plating in
2
ϩ
dilute acidic HF and alkaline NH F solutions containing Ni ions
4
at pH 2 and 8, respectively. We expect that changing the pH level of
solutions may affect the solution chemistry as well as the reactivity
of the Si surface, which in turn may influence the metal deposition
during immersion plating. The deposition behavior in both solutions
is compared and the reaction mechanism is proposed.
Experimental
Preparation of PS sample.—Single-side polished ͑100͒-oriented
boron-doped p-type silicon wafers ͑10-20 ⍀ cm in resistivity͒ were
utilized in this work. Prior to conducting experiments, the Si wafers
were rinsed in acetone to remove organic contaminants and subse-
quently the native oxide was etched in 5 wt % HF. The electro-
chemical anodization cell was made of trifluoroethylene resin with
an O-ring at the bottom that allowed an active electrode area of
2
0
.785 cm . The surfaces were anodized galvanostatically in a 1:1.7
in volume of 47 wt % HF and 99.5 wt % pure ethanol for 20 min at
2
a current density of 2 mA/cm at room temperature. After etching,
the PS samples were dried under a stream of dry Ar gas.
Figure 1. FESEM micrographs of as-prepared PS layer anodized using
p-Si͑100͒, 10-20 ⍀ cm, at 2 mA/cm in 22 wt % HF in EtOH for 20 min; ͑a͒
top view ͑b͒ cross-sectional view.
2
Plating solutions.—Two different kinds of plating solutions were
used in this study. Alkaline bath at pH 8 was prepared from 50 mM
NiSO •6H O as Ni salt in 5 M NH F aqueous solution. The other
4
2
4
plating solution was acidic and contained the same concentration of
2
ϩ
Ni ions in 0.5 wt % HF ͑ϳ0.3 M͒ giving pH 2. All solutions were
prepared from reagent-grade chemicals and Millipore deionized wa-
ter ͑18 M ⍀ cm resistivity͒. The immersion plating experiments
were operated by dipping the PS substrates in a 10 mL electrolytic
bath for various times at room temperature. After immersion, the PS
sample was rinsed in distilled water and dried by Ar gas.
saturated KNO -agar salt bridge made of a polyethylene tube. A
3
He-Ne laser ͑633 nm, 4 mW͒ was used as the light source. The scan
Ϫ1
rate was 10 mV s . All measurements were carried out under stag-
nant conditions. The ohmic contact on the rear side between the PS
and a Cu current collector was achieved by rubbing an InGa eutectic
alloy. Capacitance measurements were performed for Si electrodes
using a Solarton 1255A impedance/gain-phase analyzer combined
with a Princeton 273A potentiostat/galvanostat at a frequency of 5
kHz under different electrode potentials in the range of 0 to Ϫ1.5 V
͑vs. Ag/AgCl͒. The nature and quality of the back contact are of
main are importance during measurements of the potential depen-
dence of the space-charge region ͑SCR͒ capacitance Csc using the
Mott-Schottky plot. Poor or nonohmic back contact can deviate the
correct determination of Csc and can negatively affect the perfor-
mance of the method. Therefore, Al back contact was used instead
of In-Ga back contact in these experiments. The measurements were
conducted in the dark and under stagnant conditions. The solutions
were deoxygenated by bubbling Ar gas for 20 min prior to each
experiment.
Surface characterization.—FTIR spectra were recorded ex situ
using a Nicolet Avatar 360 spectrometer. Each spectrum is the aver-
age of 64 scans so as to improve the signal-to-noise ratio, with a
Ϫ1
resolution of 4 cm . The spectra of Si surfaces preliminarily im-
mersed in dilute HF solution were used as reference. XPS measure-
ments were performed using a Shimadzu ESCA-750 spectrometer
with a Mg K␣ excitation source. The peak positions were calibrated
on the basis of the C 1s line as 284.5 eV. The surface morphology of
the as-prepared sample and the samples after immersion plating
were also inspected using an Elionix ERA-800 field emission-type
scanning electron microscope ͑FESEM͒.
Electrochemical measurements.—Open-circuit potential ͑OCP͒
experiments were carried out in the dark for PS exposed to the
plating baths using a potentiostat/galvanostat model 273A, Princeton
Applied Research with Ag/AgCl as a reference electrode. Current-
potential curves, in the dark and under illumination utilizing a PS as
working electrode, were recorded using an automatic polarization
system HZ-3000, Hokuto Denko. A three-electrode cell with a Plexi-
glass window was used. A Pt spiral served as a counter electrode.
The reference electrode was an Ag/AgCl connected to the cell by a
Results and Discussion
Morphology of as-anodized PS.—The surface morphology of a
freshly prepared PS layer was examined using FESEM imaging.
Typical micrograph observations are presented in Fig. 1. The top-
view image reveals some pores in the range of 10-20 nm diam Fig.
1a. The roughness in the image seems to arise from pores whose
Downloaded on 2014-11-28 to IP 155.33.16.124 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, 150 ͑5͒ C277-C284 ͑2003͒
C279
Figure 3. OCP vs. immersion time, measured in the dark, for ͑a͒ PS in 0.5
wt % HF of pH 2, ͑b͒ a clean Ni wire in HF of pH 2, ͑c͒ a clean Ni wire in
5
M NH F of pH 8, and ͑d͒ PS in NH F of pH 8. Ni ion concentration in
4 4
solutions was 50 mM as NiSO4 .
nm. The film thickness can be determined to be 120 nm. On pro-
longed immersion we expect that the growth of the nuclei will con-
tinue until they completely coalesce into a continuous film.
According to the above results, it is obvious that Ni deposition
2
ϩ
proceeds very differently in NH F and HF solutions containing Ni
4
ions. During immersion plating, metal deposition is usually accom-
2
0
panied by the oxidation of PS. At pH 2 and 8 the equilibrium
Figure 2. FESEM image of Ni deposits on a PS substrate after 2 min im-
mersion in 50 mM NiSO •6H O ϩ 5 M NH F aqueous solution of pH 8;
potential of SiO /Si is calculated to be, respectively, Ϫ0.97 and
Ϫ1.32 V vs. the standard hydrogen electrode ͑SHE͒ at 25°C. Thus,
2
4
2
4
͑
a͒ top view and ͑b͒ cleaved cross section.
thermodynamically it is expected that the deposition of Ni should
2
ϩ
occur in both solutions since the Ni /Ni couple has a less negative
dimensions are less than 10 nm in PS samples prepared from p-type
silicon. A cleaved cross-sectional sample, Fig. 1b, exhibits a spongy
redox potential than the SiO /Si at both pH values. However, this is
2
3
in contradiction with the obtained results in which Ni deposition was
possible only in the alkaline bath. The measurement of OCP as a
function of immersion time may provide considerable insight on the
different deposition behaviors.
structure of the porous layer. The uppermost part of the image shows
about 50 nm Au thin layer that was evaporated to make the surface
conductive. The image allows a determination of the porous layer
thickness of about 1.5 ␮m. The porosity obtained from gravimetric
3
6
OCP.—We measured the OCP of the PS in contact with these
analysis is 75-80%.
deposition baths. The values of OCP are subject to different factors
3
8
27
Ni deposition in both solutions.—We investigated the influence
of the two plating baths on PS samples under open-circuit condi-
tions. When the PS substrate was exposed to the alkaline solution
such as electrolyte composition, different solution contaminants,
3
9
the presence of dissolved oxygen, whether measurements are car-
40
ried out in the dark, under illumination, or normal room light. In
addition to these factors, the potential is strongly affected by the
roughness of the substrate surface that directly affects the surface
reactivity during immersion. The measurement was conducted in the
dark and without removing the dissolved oxygen.
2
ϩ
containing 5 M NH F with 50 mM Ni ions at pH 8, immediate
4
deposition of metallic Ni was clearly observed on the surface. We
found that Ni deposition began on the PS sample in this alkaline
bath as soon as they brought in contact. The measurements of XRD
patterns and XPS spectra confirmed the presence of Ni. In contrast
to the above observation, we detected no deposition, even after 20
min, for all samples immersed in the dilute HF electrolytic solution
containing the same concentration of Ni2 at pH 2 under the same
experimental conditions.
Figure 3 shows the time-dependent OCP responses of PS sample
immersed in the two plating baths. The potential of PS in the dilute
2
ϩ
HF containing Ni ions, Fig. 3a, stays at a more positive value,
ϩ
compared to the NH F solution, and almost leveled at a constant
4
potential of about Ϫ0.380 V ͑vs. Ag/AgCl͒ during 20 min of mea-
suring time. The potential also remains constant after about 1 h.
More positive values of potential may be related to the surface pas-
The deposited Ni from the alkaline solution was examined by
FESEM imaging at different exposure times. At the initial stage of
immersion, many Ni nuclei distributed in a form like islands were
detected on the PS surface as we reported recently.37 Top-view and
cross-sectional micrographs of PS after 2 min immersion are shown,
as examples, in Fig. 2a and b, respectively. A Ni thin film forms
uniformly on the PS surface with coalescence of the nuclei ͑Fig. 2a͒.
The FESEM image for the cross-sectional view shows that the Ni
film is not continuous and has grain sizes on the order of 100-200
3
8
sivation of the PS by hydrogen termination. The surface of PS is
stable in HF solution and does not change.
However, the OCP of PS in contact with the alkaline solution
started initially at much more negative potential and gradually
shifted toward the positive direction before reaching a leveling po-
tential, Fig. 3d. The shift of the potential is likely generated by the
charge transfer between Ni ions in solution and the PS electrode
Downloaded on 2014-11-28 to IP 155.33.16.124 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C280
Journal of The Electrochemical Society, 150 ͑5͒ C277-C284 ͑2003͒
surface. Thus, the difference in the potential vs. time responses may
be explained by the progressive Ni nucleation and the PS oxidation
at the beginning of immersion. As the two processes proceeded, Ni
particles gradually covered the PS surface, which is reflected on the
positive shift of the potential. Once the PS surface was completely
covered by Ni, the potential leveled at a constant value of about
Ϫ0.52 V vs. Ag/AgCl. This value is close to the calculated value of
the redox potential for Ni2 /Ni ͑Ϫ0.492 V vs. Ag/AgCl͒ with 50
mM Ni ion concentration, which is also confirmed experimentally
by measuring the rest potential of a clean Ni wire in the same solu-
tion. The potential observed for the Ni wire was Ϫ0.49 V and de-
picted in Fig. 3c. The potential obtained for the Ni wire immersed in
the acidic solution was similar as in the alkaline solution, see Fig.
ϩ
2
ϩ
3b. This result reveals that the equilibrium potential of the Ni/Ni
in the acidic solution is less noble ͑more negative͒ than the OCP of
PS in the same solution ͑compare a and b͒. Consequently, the depo-
sition of Ni is now not expected to occur in HF solution of pH 2.
The leveled potential in the alkaline solution, Fig. 3d, indicates a
full coverage of PS surface with Ni. The existence of such behavior
has been observed during Cu deposition in a previous work. Under
this condition, the deposition process is expected to be self-ceased
as the PS surface is covered by the metal film. It is not clear why the
potential shifted toward more negative direction as we changed the
4
1
Figure 4. Mott-Schottky plots for p-Si in ͑a͒ 0.5 wt % HF deposition solu-
tion at pH 2 and ͑b͒ 5 M NH F deposition solution at pH 8. The electrode
4
was previously electropolished as described in the text. The plots are linear
and the values of flatband potential are obtained from the intercept with the
potential axis.
plating bath from the acidic HF to the alkaline NH F solution. Prob-
4
ably it may stem from a different complex formation with Ni ions.
Our observation reveals that the addition of NH F to the NiSO
4
4
deoxygenated 5 wt % HF solution was 0.23 V while the value for PS
measured in 20 wt % HF was 0.36 V vs. Ag/AgCl.
According to the above results, there is a pH-dependence of EFB
and the shift obtained is Ϫ73 mV/pH, contrasted to the Nernstian
solution resulted in generation of Ni ammine complex. We infer the
complex formation from the color change of Ni ion solutions from
green to blue. This complex was also confirmed by measuring the
UV-vis spectra in which an absorption peak related to Ni ammine
2
ϩ
dependence on pH ͑Ϫ60 mV/pH͒ which is generally observed in
complex, in the form of ͓Ni͑NH ) (H O) ] , was detected. In the
3
3
2
3
46
many semiconductors. Other authors reported a dependence of
HF solution the solution color with Ni ions remained green and the
UV-vis spectra revealed only the presence of Ni aqua-complex.
Because metal deposition involves the charge transfer across the
semiconductor/solution interface, the band structure of the semicon-
ductor may play an important role in determining deposition behav-
ior. In the next section the relative energetic position between the
silicon bandedges and the Ni redox levels in both electrolytes is
determined using the values of flatband potential and OCP.
4
7
Ϫ45 mV/pH. Here a question arises on the origin of the EFB shift.
It is generally accepted that this shift in E is due to the adsorption
FB
ϩ
Ϫ 46
of H /OH . As we show later in the FTIR section, the dissolution
rate of Si increases with pH. Under this condition and according to
Ref. 48, the substitution reaction of Si-H → Si-OH proceeds much
faster as pH increases from 2 to 8. Since Si-OH is considered as
weak acid, deprotonation of the Si-OH bond according to
Ϫ
ϩ
Si-OH ↔ Si-O ϩ H is suggested to occur. Accordingly, the shift
Mott-Schottky plots and evaluation of flatband potential.—We
used the Mott-Schottky method to determine the flatband potential
of flat silicon (EFB) by measuring the potential dependence of the
SCR capacitance CSC at 5 kHz and applying the Mott-Schottky
relation
in E is probably related to the formation of Si-OH that can be
FB
deprotonated.
Energy band diagram of the silicon/solution interface.—The flat-
band potential is now estimated from Mott-Schottky plots. The OCP
of Si in both solutions is known from experimental measurements;
the values were taken during the initial stage of immersion. The
2
1
/C ϭ ͑2/qN ␧ ͒͑E Ϫ EFB͒
͓1͔
SC
d
0
2
ϩ
redox couple of Ni /Ni is also measured by immersing a Ni wire in
the plating solutions. In contrast to our expectation, the equilibrium
where q is the electron charge, N the dopant density, ␧ the dielectric
d
2
ϩ
constant of semiconductor, ␧ the permittivity of vacuum, and E is
potential for Ni /Ni redox couple is almost the same in both depo-
sition baths, Fig. 3b and c, although the form of the complex state is
0
the electrode potential.
2
For obtaining a good linearity between 1/CSC and E, electropol-
different. The energetic position of the valence bandedge (E ) is
v
ishing for silicon electrodes was carried out before measurements as
reported by Allongue et al.42 and described in Ref. 43. The elec-
tropolishing pretreatment was carried out potentiostatically at ϩ4.0
calculated by considering that the valence band is located about 0.24
15
V below the Fermi level (E ) for p-Si ͑10 ⍀ cm, dopant density 10
f
Ϫ3
49
cm ͒; i.e., E Ϫ E ϭ 0.24 V. In addition, the band bending (Vb)
f
v
V ͑vs. Ag/AgCl͒ in a solution of 1 M NH F at pH 4.5 for 15 min.
can be estimated from the direct subtraction of OCP and flatband
4
The observed current was 15 mA. A subsequent oxide layer was
removed by dipping the samples in dilute HF solution. Typical Mott-
Schottky plots measured for Si in both plating solutions are shown
in Fig. 4. The plots are linear over the entire potential region and its
extrapolation should cross the potential axis at the flatband potential.
We obtained EFB ϭ 0.48 V in the acidic bath of pH 2 whereas the
value in the alkaline bath of pH 8 was 0.04 V ͑vs. Ag/AgCl͒. The
potential (V ϭ OCP-EFB). The experimental values of EFB
,
b
2ϩ
OCP, E
, and V are summarized in Table I. Using the above
b
Ni/Ni
Table I. OCP, EFB , ENi2¿ , and V „V ϭ OCP Ϫ E … values
ÕNi
b
b
FB
measured for a Si electrode in both plating solutions. All values
are in V and referred to AgÕAgCl. The values of OCP were taken
at the early immersion.
dopant density N , determined from the slope of the plots, ranged
d
1
5
Ϫ3
between 4.9 and 5.2 ϫ 10 cm , being in good agreement with
4
4
the reported data. The values of flatband potential for p-type Si in
wt % HF solution with different dissolved oxygen content and for
PS measured in 20 wt % HF in the presence of different surfactants
Solution
OCP
EFB
ENi2ϩ/Ni
V ϭ OCP Ϫ E
b
FB
5
Bath at pH 2
Bath at pH 8
Ϫ0.41
Ϫ0.76
0.48
0.04
Ϫ0.49
Ϫ0.49
Ϫ0.89
Ϫ0.80
were reported previously.4
5,39
The potential obtained for p-Si in
Downloaded on 2014-11-28 to IP 155.33.16.124 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, 150 ͑5͒ C277-C284 ͑2003͒
C281
Figure 6. Transmission FTIR spectra measured for ͑a͒ as-anodized PS, ͑b͒
after 20 min immersion of PS in 0.5 wt % HF solution containing 50 mM Ni
ions at pH 2, and ͑c͒ after immersion in 5 M NH F solution at pH 8 con-
4
taining the same concentration of Ni ions for 10 min.
Figure 5. Energy band diagrams of the silicon interface with ͑a͒ the alkaline
solution at pH 8 and ͑b͒ the acidic solution at pH 2. The position of the
bandedges and the redox potential for Ni /Ni couple are shown. The ex-
perimental data used in constructing the diagram are summarized in Table I.
corded using FTIR spectroscopy. The large surface area of PS makes
it a good substrate for investigation by transmission mode of IR
spectroscopy. Great care should be taken to limit exposure of the
surface to air during measurements.
2
ϩ
The spectra of as-anodized PS and the sample after immersion in
both plating baths are shown in Fig. 6. The peak assignment in Fig.
information, the energy diagrams of the silicon/solution interface are
constructed for both cases and the result is depicted in Fig. 5. Using
flat Si instead of PS during capacitance measurements permits one
to obtain more reliable results for EFB . Our previous results during
potential dependence of Csc using PS substrate revealed a deviation
51-53
6
follows Ogata et al.
First, the spectrum of a freshly prepared
Ϫ1
PS, Fig. 6a, exhibits SiH stretching modes at 2090-2150 cm . The
x
spectra also show the Si-H bending mode at 630 cmϪ1, the SiH
2
Ϫ1
wagging mode at 680 cm , and the SiH2 scissors mode at 910
Ϫ1
Ϫ1
cm . In addition, a small peak can be seen at 810 cm related to
3
9
from the linearity and the results were not very reproducible. Even
so, a small shift obtained in EFB on changing PS to Si electrode has
no significant influence in the present discussion. Furthermore, the
OCP usually shows more negative value ͑less noble direction͒ when
we changed flat Si to PS substrates.2 As a result, the energy band
diagram is expected to shift toward the negative direction without
affecting the valence and the conduction band relative position and,
hence, we believe that the data received from utilizing Si electrode
can be used to describe the energy band diagram of our system
qualitatively.
It can be seen that the nickel redox level in the alkaline bath of
pH 8, Fig. 5a, lies relatively closer to the silicon valence bandedge
in comparison with the acidic bath of pH 2, Fig. 5b. However, there
is still no significant overlap between the valence band and the re-
dox levels in alkaline solution. This result suggests that the charge
exchange process with the valence band appears to be not probable
in both solutions.
SiF . Moreover, the presence of a small absorption peak around
x
Ϫ1
1100 cm , which is assigned to the Si-O-Si antisymmetric stretch-
ing, indicates that the PS layer is partially oxidized during sample
transfer.
1,50
Figure 6b and c display the spectra of the PS after immersion in
2
ϩ
HF and NH F solutions containing Ni ions, respectively. As can
4
be seen, the immersion in HF solution does not remove any species
on the surface and the IR spectra related to SiH species remain
x
almost unchanged after 20 min treatment. This result indicates that
the Si-H bonds of PS are stable in the acidic plating bath. However,
marked change in the spectra is observed for the PS immersed in the
alkaline bath; the absorption peaks related to the characteristic sur-
face hydrides are severely diminished after only 10 min exposure.
This result indicates that the PS layer is dissolved much faster in an
alkaline bath of pH 8. The acceleration of Si dissolution therefore
enhances the rate of Ni deposition. This observation is consistent
with the reported results that the hydrogen-terminated Si surface is
3
4
FTIR spectra and change in characteristic surface species.—In
any deposition process, the morphology and the surface chemistry
play important roles in determining the nature of deposits. Because
mobile charge carriers are involved during immersion plating, it is
expected that the properties of PS surface in the deposition solution
should affect the deposition process. Thus, the change in the char-
acteristic surface species of PS before and after immersion was re-
stable under open-circuit conditions at low pH values.
Based on the above observations, it is realized that the dissolu-
tion rate of PS in both solutions during the immersion plating is
different and this should influence the Ni deposition behavior sig-
4
7
nificantly. According to Allongue et al., Si dissolution is faster
with increasing pH and the etching rate at pH 8 was about four times
greater than that at pH 2. This variation of the etch rate with pH
Downloaded on 2014-11-28 to IP 155.33.16.124 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C282
Journal of The Electrochemical Society, 150 ͑5͒ C277-C284 ͑2003͒
Figure 7. Si 2p XPS spectra of PS samples after 2 min immersion in 50 mM
NiSO of ͑a͒ 5 M NH F solution of pH 8 and ͑b͒ 0.5 wt % HF solution of
pH 2.
Figure 8. Current-potential curves for PS electrodes measured in contact
with ͑a͒ blank 5 M NH F solution of pH 8, ͑b͒ 5 M NH F solution contain-
4
4
4
4
ing 50 mM Ni2 in the dark, and ͑c͒ the same as ͑b͒ but under laser illumi-
ϩ
nation. Scanning was starting after 10 min deposition at the OCP in both
Ϫ1
cases. The scan rate was 10 mV s
.
reflects the different stability of Si-H bonds in both solutions as
revealed by FTIR spectra.
In the acidic bath of pH 2 it is expected that the hydrogen evo-
lution reaction ͑HER͒ is a competitive reaction to metal deposition.
immersion. This result adds merit for the alkaline plating bath that a
binary PS/Ni nanostructure without Si oxides can be achieved suc-
cessfully.
This agrees well with the thermodynamics where the equilibrium
ϩ
potential of the H /H couple is Ϫ0.12 V ͑vs. SHE͒ which is less
2
Effect of illumination on Ni deposition.—Figure 8 shows typical
current-potential curves for a PS electrode in 50 mM Ni2 ions with
NH4F solution at pH 8 in the dark and under laser illumination. The
negative than that of Ni ion reduction. Under this condition, Ni
deposition is suppressed and the mixed potential of the whole pro-
cess can be achieved through the contribution of proton reduction.
Experimentally, hydrogen bubbles were observed on the PS surface
during immersion. This is the possibility that may be considered to
explain the absence of deposition from HF solution.
ϩ
curve for a blank solution, i.e., 5 M NH F containing no Ni ions, in
4
the dark is also shown. The measurements were recorded at a po-
Ϫ1
tential scan rate of 10 mV s . Before starting the scan in each case
However, much more hydrogen bubbles were observed during
the potential was held at the OCP for 10 min to deposit Ni. Then
scanning were started to the negative direction.
In the blank solution in the dark, curve a, the forward and reverse
scan show almost no current response and this is a typical diode
behavior of the interface. It was observed also that the porous layer
immersion in the NH F alkaline solution, although the equilibrium
4
ϩ
potential of the H /H couple in this solution ͑Ϫ0.48 V vs. SHE͒ is
2
more negative than that in the HF solution. The reason for this can
ϩ
be attributed to the reduction of H and/or H O being catalyzed by
2
is chemically dissolved in NH F solution during measurement. In
the initially deposited Ni particles. Once a first population of Ni
nuclei exists, they can act as cathodic sites on which HER is pro-
moted because the reduction of protons at the Ni/solution interface
occurs much more pronounced than the reduction at the Si/solution
interface due to the lower hydrogen overpotential. In addition, this
great amount of H2 evolution may stem from the dissociation of
surface hydrides in the alkaline bath at pH 8, even in the absence of
added Ni ions, as confirmed by IR spectra measurements. Hydrogen
gas evolution was also observed and confirmed by gas chromatog-
raphy and weight loss measurements for Si during etching in 40%
4
solution containing Ni ions in the dark, voltammogram b, we cannot
observe appreciable reduction current in the negative scan, although
Ni was deposited as revealed from the anodic stripping at a peak
potential of about Ϫ0.15 V ͑vs. Ag/AgCl͒. The deposition in the
dark is slow as revealed from the relatively small stripping peak, in
comparison with peak c.
Under illumination, voltammogram c, the photopotential shifts
the OCP to the positive value and the amount of deposited Ni is
clearly increased as seen from the higher intensity of the stripping
peak. Therefore, illumination could enhance Ni deposition. Table II
collects the values of OCP in the dark and under illumination at two
different times.
54
NH F solution at OCP.
4
XPS spectra in the region of Si 2p.—Figure 7 shows the XPS
spectra in the region of Si 2p for the PS sample that was immersed
Deposition mechanism from the alkaline bath.—Two possibili-
ties are considered herein to explain the reaction mechanism. The
first mechanism is that the Ni deposition takes place through hole
injection from the metal ions in solution into the valence band.
for 2 min in the NH F solution containing Ni ions, and immersed for
4
2
min in the HF solution containing Ni ions under the same condi-
tions. The spectra were recorded immediately after immersion. The
XPS spectra of sample a and b show only one peak at around 99.0
and 99.2 eV, respectively, which is attributed to the elemental Si in
the Si 2p photoemission line.55 The peak observed after immersion
in the HF bath, b, is sharper and shows about four times higher
intensity than that corresponding to a. This is because no Ni is
deposited on sample b. In addition, the peak, which usually appears
Table II. Dark and illuminated OCP values for a PS electrode
2
¿
measured in contact with 5 M NH4F containing 50 mM Ni at 0
and 10 min immersion.
around 103.0 eV related to SiO , when metal is deposited, is not
2
OCP, V ͑vs. Ag/AgCl͒
0 min
10 min
observed in sample a, confirming that the PS surface after the im-
mersion process is free from Si oxides. The result of XPS is consis-
tent with IR spectra in which no oxide remains on the PS after
Dark
Illuminated
Ϫ0.83
Ϫ0.51
Ϫ0.52
Ϫ0.25
Downloaded on 2014-11-28 to IP 155.33.16.124 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, 150 ͑5͒ C277-C284 ͑2003͒
C283
These injected holes are consumed in the PS oxidation process.
photoexcited carriers. These photoelectrons can be transferred to the
solution via the conduction band and thereby the Ni ion reduction is
promoted. Moreover, the difference between the OCP values, in the
dark and under illumination, shown in Table II results in a reduction
in the band bending. The decrease in band bending can be viewed as
a shift of the Si bands with respect to the redox levels in solution.
Ϫ
56
Taking into account that F ions are predominant at pH Ͼ 6, the
deposition process through the valence band may be summarized as
follows
2
ϩ
0
ϩ
͓
Ni͑NH ͒ ͑H O͒ ] ͑aq͒ → Ni ϩ 3NH ͑aq͒ ϩ 3H O ϩ 2h
3
3
2
3
3
2
͓
2͔
Si ϩ 6F ͑aq͒ ϩ 4h → SiF 6Ϫ 2͑aq͒ E° ϭ Ϫ1.20 V vs. SHE
3͔
Conclusions
0
Ϫ
ϩ
In this paper, the different behaviors of Ni deposition on a PS
͓
layer by simple immersion plating in alkaline NH F and acidic HF
4
solutions were investigated. The metal deposition proceeded very
differently in both solutions; deposition was observed only from the
alkaline solution. Comparing the energetic position of the valence
bandedge of silicon with the Ni redox level in both solutions, it was
shown that the charge-transfer process with valence band in both
solutions appears to be not probable.
E° of Reaction 2 should be located between Ϫ0.26 and Ϫ0.48 V vs.
SHE for aqua and hexaammine complexes, respectively. Reactions
2, 3 are described as valence band processes ͑the holes in the right
side are being injected into the valence band͒. In this mechanism, it
is anticipated that the Ni redox levels should have a good overlap
with the valence band. However, this is in contradiction with the
energetic situation shown in Fig. 5a. Thus, it is doubtful whether this
mechanism can be responsible for the deposition reaction.
The result from UV-vis spectra revealed that Ni is predominantly
present as a Ni-aqua-complex, while in the NH F solution, the
4
dominant Ni species is the ammine complex. The formation of such
an ammine complex may be responsible for the shift in the OCP of
PS toward more negative direction, which indicates the higher fea-
sibility of Ni deposition.
The second mechanism we considered is that the reduction of
Ni2 proceeds by the galvanic displacement of PS in the alkaline
bath. Nickel ions gain the electrons generated from the electro-
chemical oxidation of PS to be reduced and deposit. Then, Si oxides
are readily dissolved in the presence of fluoride species. A similar
ϩ
Absence of Ni deposition in the HF solution is both due to the
stability of Si-H bonds in solution, as revealed from the IR spectra,
as well as the fact that HER is the predominant reaction so that the
mixed potential may be achieved without contribution from Ni ion
reduction. In addition, the OCP of PS was located at more noble
reductive deposition of Cu and Ag on PS from aqueous solutions has
been reported previously by our group and others.2
0,57,58
The gal-
vanic process can be described by the following set of reactions
Anodic reaction ͑oxidation͒
2ϩ
potential than that of the Ni /Ni couple and hence the deposition is
ϩ
Ϫ
not favored.
Si ϩ 2H O → SiO ϩ 4H ϩ 4e
E° ϭ Ϫ0.85 V vs. SHE
2
2
An interesting observation from the alkaline solution is that in
the presence of fluoride species the PS oxidation product was readily
dissolved; no Si oxides were detected on the surface as confirmed by
XPS and IR measurements. Consequently, utilizing this alkaline
bath appears promising for obtaining a binary PS/Ni nanostructure
instead of a ternary system that usually forms from aqueous solution
without fluoride species.
͓4͔
and
ϩ
Ϫ
SiH ϩ 2H O → SiO ϩ ͑4 ϩ x͒H ϩ ͑4 ϩ x͒e
͓5͔
x
2
2
Cathodic reaction ͑reduction͒
In addition, it was observed that under laser illumination the
deposition of Ni from the alkaline bath was enhanced. We show also
that Ni deposition can occur both in the dark with different possible
mechanisms and by electron extraction from the conduction band
under the effect of light.
2
ϩ
Ϫ
0
͓
Ni͑NH ͒ ͑H O͒ ] ͑aq͒ ϩ 2e → Ni ϩ 3NH ͑aq͒ ϩ 3H O
3
3
2
3
3
2
͓6͔
Dissolution of the oxidation product
ϩ
Ϫ
Ϫ2
SiO ϩ 4H ϩ 6F ͑aq͒ → SiF6 ͑aq͒ ϩ 2H O
͓7͔
2
2
Acknowledgments
Another alternative reaction route for the anodic oxidation of Si
shown in Reactions 4 and 5, is the one step direct etching of Si in
presence of F into the soluble complex SiF6 as described by
The authors thank Y. Mizutani and T. Nagayama of Kyoto Mu-
nicipal Institute for Industrial Research for FESEM observations, T.
Morii and S. Satoh of Kyoto University for UV-vis spectra measure-
ments. This research was supported partly by a grant-in-aid from the
Ministry of Education, Culture, Sports, Science and Technology, Ja-
pan, and partly by the Sasakawa Scientific Research Grant from The
Japan Science Society. F.A.H. acknowledges support from the Japa-
nese government in the form of a Monbusho fellowship.
Ϫ
Ϫ2
0
Ϫ
Ϫ2
6
Ϫ
Si ϩ 6F ͑aq͒ → SiF ͑aq͒ ϩ 4e
E° ϭ Ϫ1.20 V vs. SHE
8͔
͓
Absence of cathodic current in the negative scan, Fig. 8b, indicates
that the released electrons from the PS oxidation are spontaneously
consumed for the reduction of Ni2 , and hence the whole process
can be seen as a simple displacement deposition and the electron
transfer is through the conduction band. In view of the consider-
ations mentioned above, the second possibility is highly probable,
though the first mechanism ͑hole injection to the valence band͒ still
can hold because the band bending usually changes during the im-
mersion plating event. However, at this moment we do not have
evidence whether holes are being injected or not. Further investiga-
tion is needed in this area for complete understanding.
Kyoto University assisted in meeting the publication costs of this article.
ϩ
References
1
. M. A. Tischler, R. T. Collins, J. H. Stathis, and J. C. Tsang, Appl. Phys. Lett., 60,
375 ͑1992͒.
1
2. A. G. Cullis, L. T. Canham, and P. D. Calcott, J. Appl. Phys., 82, 909 ͑1997͒.
3. P. C. Searson, in Advances in Electrochemical Science and Engineering, Vol. 4, H.
Gerischer and C. W. Tobias, Editors, VCH, Weinheim ͑1995͒.
4
5
. T. Unagami, Jpn. J. Appl. Phys., 19, 231 ͑1980͒.
. F. Otoi, K. Anzai, H. Kitabayashi, K. Uchiho, and Y. Mizokami, J. Electrochem.
Soc., 131, C319 ͑1984͒.
An important piece of information concerning the effect of laser
light on the deposition pathway is highlighted. We observed that Ni
deposition, which can proceed via the conduction band, is enhanced
under illumination. Deposition of Ni via the conduction band under
the effect of illumination was also suggested by Jeske et al. from
6. L. T. Canham, Appl. Phys. Lett., 57, 1046 ͑1990͒.
7
. F. Namavar, H. P. Marusha, and N. M. Kalkhoran, Appl. Phys. Lett., 62, 3159
1992͒.
. M. J. Sailor and E. J. Lee, Adv. Mater., 9, 783 ͑1997͒.
͑
8
9. D. G. Anderson, N. Anwar, B. J. Aylett, L. G. Earwaker, M. I. Nasir, J. P. G. Farr,
K. Stiebahl, and J. M. Keen, J. Organomet. Chem., 437, C7 ͑1992͒.
10. S. Steiner, F. Kozlowski, and W. Lang, Thin Solid Films, 255, 49 ͑1995͒.
Ϫ 12
aqueous solution without F . The photon energy of the He-Ne
1
1
1. S. Dhar and S. Chakrabarti, Appl. Phys. Lett., 68, 1392 ͑1996͒.
2. M. Jeske, J. W. Schultze, M. Thonissen, and H. Munder, Thin Solid Films, 255, 63
͑1995͒.
laser used as a light source is 1.8 eV, which is greater than the
bandgap of Si. This leads to the generation of electron-hole pairs as
¨
¨
Downloaded on 2014-11-28 to IP 155.33.16.124 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C284
Journal of The Electrochemical Society, 150 ͑5͒ C277-C284 ͑2003͒
1
3. D. Andsager, J. Hilliard, L. H. AbuHassan, M. Plisch, and M. H. Nayfeh, J. Appl.
Phys., 74, 4783 ͑1993͒.
4. J. E. Hilliard, H. M. Nayfeh, and M. H. Nayfeh, J. Appl. Phys., 77, 4130 ͑1995͒.
5. T. Tsuboi, T. Sakka, and Y. H. Ogata, Appl. Surf. Sci., 147, 6 ͑1999͒.
6. L. V. Belyakov, D. N. Goryachev, and O. M. Sreseli, Semiconductors, 34, 1334
37. F. A. Harraz, T. Sakka, and Y. H. Ogata, Phys. Status Solidi A, In press.
38. V. Bertagna, R. Erre, F. Rouelle, and M. Chemla, J. Electrochem. Soc., 146, 83
͑1999͒.
1
1
1
39. T. Tsuboi, T. Sakka, and Y. H. Ogata, Electrochim. Acta, 46, 1013 ͑2001͒.
40. V. Bertagna, F. Rouelle, and M. Chemla, J. Appl. Electrochem., 27, 1179 ͑1997͒.
41. F. A. Harraz, T. Sakka, and Y. H. Ogata, Electrochim. Acta, 47, 1249 ͑2002͒.
42. P. Allongue, C. H. de Villeneuve, L. Pinsard, and M. C. Bernard, Appl. Phys. Lett.,
67, 941 ͑1995͒.
͑2000͒.
1
1
7. R. H e´ rino, Mater. Sci. Eng., B, 69-70, 70 ͑2000͒.
8. P. Allongue, E. Souteyrand, and L. Allemand, J. Electrochem. Soc., 136, 1027
͑1989͒.
43. P. Gorostiza, R. D ´ı az, M. A. Kulandainathan, F. Sanz, and J. R. Morante, J. Elec-
1
2
2
9. G. Oskam, D. Vanmaekelbergh, and J. J. Kelly, Electrochim. Acta, 38, 301 ͑1993͒.
0. T. Tsuboi, T. Sakka, and Y. H. Ogata, J. Appl. Phys., 83, 4501 ͑1998͒.
1. F. A. Harraz, T. Tsuboi, J. Sasano, T. Sakka, and Y. H. Ogata, J. Electrochem. Soc.,
troanal. Chem., 469, 48 ͑1999͒.
44. S. M. Sze, Physics of Semiconductor Devices, 2nd ed., p. 32, Wiley, New York
͑1981͒.
1
49, C456 ͑2002͒.
45. Y. H. Ogata, T. Ikeda, T. Sakka, and T. Kobayashi, Electrochim. Acta, 45, 2219
͑2000͒.
46. S. R. Morrison, Electrochemistry at Semiconductor and Oxidized Metal Electrodes,
Plenum, New York ͑1980͒.
2
2
2. T. Ito, A. Yamama, and H. Hiraki, Appl. Surf. Sci., 41Õ42, 301 ͑1989͒.
3. O. M. R. Chyan, J. J. Chen, H. Y. Chien, J. Sees, and L. Hall, J. Electrochem. Soc.,
1
43, 92 ͑1996͒.
2
2
2
2
2
2
4. I. Teerlinck, P. W. Mertens, H. F. Schmidt, M. Meuris, and M. M. Heyns, J.
Electrochem. Soc., 143, 3323 ͑1996͒.
5. G. S. Higashi, Y. J. Chabal, G. W. Trucks, and K. Raghavachari, Appl. Phys. Lett.,
47. P. Allongue, V. Kieling, and H. Gerischer, Electrochim. Acta, 40, 1353 ͑1995͒.
48. P. Allongue, C. H. de Villeneuve, S. Morin, R. Boukherroub, and D. D. M. Wayner,
Electrochim. Acta, 45, 4591 ͑2000͒.
49. S. M. Sze, Physics of Semiconductor Devices, 2nd ed., p. 26, Wiley, New York
͑1981͒.
50. Y. H. Ogata, J. Sasano, J. Jorne, T. Tsuboi, F. A. Harraz, and T. Sakka, Phys. Status
Solidi A, 182, 71 ͑2000͒.
51. Y. Ogata, H. Niki, T. Sakka, and M. Iwasaki, J. Electrochem. Soc., 142, 195 ͑1995͒.
52. Y. Ogata, H. Niki, T. Sakka, and M. Iwasaki, J. Electrochem. Soc., 142, 1595
͑1995͒.
5
6, 656 ͑1990͒.
6. H. E. Hessel, A. Feltz, U. Memmert, and R. J. Behm, Chem. Phys. Lett., 186, 275
1991͒.
7. V. Bertagna, F. Rouelle, G. Revel, and M. Chemla, J. Electrochem. Soc., 144, 4175
1997͒.
͑
͑
8. G. V. Kuznetsov, V. A. Skryshevsky, T. A. Vdovenkova, A. I. Tsyganova, P. Gor-
ostiza, and F. Sanz, J. Electrochem. Soc., 148, C528 ͑2001͒.
9. N. Takano, N. Hosoda, T. Yamada, and T. Osaka, J. Electrochem. Soc., 146, 1407
53. Y. H. Ogata, F. Kato, T. Tsuboi, and T. Sakka, J. Electrochem. Soc., 145, 2439
͑1998͒.
͑1999͒.
3
3
3
0. N. Takano, D. Niwa, T. Yamada, and T. Osaka, Electrochim. Acta, 45, 3263 ͑2000͒.
1. T. Homma, C. P. Wade, and C. H. D. Chidsey, J. Phys. Chem. B, 102, 7919 ͑1998͒.
2. G. Li, E. A. Kneer, B. Vermeire, H. G. Parks, and S. Raghavan, J. Electrochem.
Soc., 145, 241 ͑1998͒.
54. H. Fukidome, Ph.D. Thesis, Graduate School of Engineering, Osaka University,
Osaka, Japan ͑2000͒.
55. C. D. Wagner, W. M. Riggs, L. E. Davis, J. F. Moulder, and G. E. Mullenberg,
Handbook of X-Ray Photoelectron Spectroscopy, Perkin-Elmer, Eden Prairie, MN
͑1979͒.
56. G. V. Kuznetsov, V. A. Skryshevsky, T. A. Vdovenkova, A. I. Tsyganova, P. Goro-
tiza, and F. Zans, J. Electrochem. Soc., 148, C528 ͑2001͒.
57. T. K. Sham, I. Coulthard, J. W. Lorimer, A. Hiraya, and M. Watanabe, Chem.
Mater., 6, 2085 ͑1994͒.
58. I. Coulthard, R. Sammynaiken, S. J. Naftel, P. Zhang, and T. K. Sham, Phys. Status
Solidi A, 182, 157 ͑2000͒.
3
3
3. S. Ye, T. Ichihara, and K. Uosaki, J. Electrochem. Soc., 148, C421 ͑2001͒.
4. G. Oskam, J. G. Long, A. Natarajan, and P. C. Searson, J. Phys. D, 31, 1927
͑
1998͒.
5. P. M. Hoffmann, A. Radisic, and P. C. Searson, J. Electrochem. Soc., 147, 2576
2000͒.
6. D. Hamm, J. Sasano, T. Sakka, and Y. H. Ogata, J. Electrochem. Soc., 149, C331
2002͒.
3
3
͑
͑
Downloaded on 2014-11-28 to IP 155.33.16.124 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).