Synthesis and characterization of porous platinum layers deposited on highly doped n-type porous silicon by immersion plating

Article abstract of DOI:10.1149/1.2337779

In this article, we report on the synthesis and characterization of porous platinum layers on top of highly doped (0.007-0.02 and 0.001-0.003 cm) n-type porous silicon by immersion plating. Porous silicon samples of different morphologies were prepared by electrochemical etching. Plating was performed by simple immersion of the porous samples in HF -containing hexachloroplatinate solutions for some minutes. Scanning electron microscopy showed that the deposits attained up to 8 μm depth. Moreover, the achieved platinum deposits have a random pore structure, whereas the original porous silicon substrates show an organized honeycomb-like parallel pore array. Electron probe microanalysis measurements showed that the deposits have very good in-depth homogeneity, indicating also that almost all silicon originally present was etched away by platinum. The electrochemically active surface area of the platinum deposits was determined to be on the order of 80 m2 cm3 by cyclic voltammetry.

Full text of DOI:10.1149/1.2337779

Journal of The Electrochemical Society, 153 ͑11͒ C741-C746 ͑2006͒
C741
0
013-4651/2006/153͑11͒/C741/6/$20.00 © The Electrochemical Society
Synthesis and Characterization of Porous Platinum Layers
Deposited on Highly Doped n-Type Porous Silicon by Immersion
Plating
,
z
José Geraldo Alves Brito-Neto,* Shintaro Araki, and Masanori Hayase*
Department of Mechanical Engineering, Faculty of Science and Technology, Tokyo University of Science,
Chiba 278-8510, Japan
In this article, we report on the synthesis and characterization of porous platinum layers on top of highly doped ͑0.007–0.02 and
0
.001–0.003 ⍀ cm͒ n-type porous silicon by immersion plating. Porous silicon samples of different morphologies were prepared
by electrochemical etching. Plating was performed by simple immersion of the porous samples in HF-containing hexachloroplati-
nate solutions for some minutes. Scanning electron microscopy showed that the deposits attained up to 8 ␮m depth. Moreover, the
achieved platinum deposits have a random pore structure, whereas the original porous silicon substrates show an organized
honeycomb-like parallel pore array. Electron probe microanalysis measurements showed that the deposits have very good in-depth
homogeneity, indicating also that almost all silicon originally present was etched away by platinum. The electrochemically active
2
3
surface area of the platinum deposits was determined to be on the order of 80 m /cm by cyclic voltammetry.
2006 The Electrochemical Society. ͓DOI: 10.1149/1.2337779͔ All rights reserved.
©
Manuscript submitted March 14, 2006; revised manuscript received July 7, 2006. Available electronically September 6, 2006.
The great variety of pore morphologies that can be obtained with
anodically etched porous silicon, besides its ease of integration with
most standard microelectromechanical systems ͑MEMS͒ fabrication
techniques, makes it a very interesting catalyst support layer for
wavelength dispersive X-ray spectrometer ͑WDS͒. Cyclic voltam-
metry was employed to estimate the electrochemically active area of
the deposits.
1
micro fuel cells. Among the methods eligible for the deposition of
Experimental
metals on porous silicon, the chemical and electrochemical ones
offer a distinct advantage due to their ease of implementation. For
Materials and reagents.— The silicon wafers used in this work
were n-type, ͑100͒-oriented. Two different kinds of samples were
employed: ͑1͒ Sb-doped, 0.007–0.02 ⍀ cm resistivity, 500 ␮m
thick, ͑100͒ oriented, mirror-polished on one side, obtained from
Silicon Technology Corporation ͑from now on, medium-doping wa-
fers for brevity͒ and ͑2͒ As-doped, 0.001–0.003 ⍀ cm resistivity,
1-3
4-12
example, electroplating or immersion plating
of different met-
als from different electrolytes on porous silicon have been objects of
study by several groups.
Immersion plating, where the driving force for metal reduction
comes from the oxidation of silicon in the pore walls, is particularly
interesting. This is due not only to the instrumental simplicity, but
also because it is a self-controlled process. It occurs selectively
where there are available silicon surface sites for reaction and comes
automatically to a halt when the surface is completely covered,
which can lead to more uniform surface coverage and more efficient
use of reagents.
1
00 ␮m thick, ͑100͒ oriented, mirror-polished on both sides, ob-
tained from Chiyoda Koeki ͑from now on, high-doping wafers for
brevity͒.
Hydrofluoric acid, sulfuric acid, and pure ͑99.5%͒ ethanol and
acetone were obtained from Wako Chemicals ͑Japan͒ and used as
2−
received. PtCl6 solutions were prepared by dissolving solid
H PtCl ·6H O, obtained from Wako Chemicals ͑Japan͒. All solu-
2
6
2
One difficult aspect of metal deposition in porous silicon has to
do with the depth attainable by the deposit. Diffusional transport of
reactants from the bulk solution to the deeper parts of the porous
layer imposes a severe restriction on the depth of deposition that can
be obtained. Moreover, as deposition progresses at the upper region
of the porous layer, the transport restriction gets more stringent,
eventually bringing the process to an effective halt. One way to
mitigate this difficulty is to increase the pore diameter in order to
enhance the diffusional flux of material from the surface. This can
be achieved, for example, by increasing the anodization current den-
sity, decreasing the HF concentration in the anodization bath, or by
tions were prepared using ultrapure water ͑Milli-Q, Millipore͒.
Preparation of porous silicon.— The anodization cell used for
the preparation of porous silicon was described in detail elsewhere.
In brief, the cell was built from a poly-trifluoroethylene cylinder
1
͑
diameter 9.2 mm, height 20 mm͒, in which the anodization area is
defined by a Kalrez O-ring ͑inner diameter 9.25 mm͒. A platinum
coil and a platinum wire were used as counter and reference elec-
trodes. A Hokuto-Denko HABF5001 potentiostat interfaced to a PC
computer and controlled by a homemade program was used as the
programmable current source. Anodization was performed at 10°C
inside a refrigerated chamber under ambient illumination. The an-
odization electrolyte was prepared by mixing concentrated hydrof-
luoric acid ͑approximately 46% mass/mass͒, ultrapure water, and
ethanol in a 3:5:2 mass ratio. Prior to anodization, the silicon wafers
were cut into 15 mm square pieces, which were soaked in concen-
trated HF to remove the thermal oxide layer and then thoroughly
washed with water, followed by ethanol, and dried under air. A
copper layer was sputtered onto the back side of the wafers to serve
as electrical contact for anodization.
4
doing anodization in organic solvents.
In this paper, we report on the synthesis and characterization of
porous platinum deposits or top of highly doped n-type porous sili-
2
6
−
con by immersion plating from HF-containing PtCl solutions. Po-
rous silicon samples with different pore size characteristics but with
apparently regular structure were prepared on differently doped
single-crystal substrates. Through plating, a seemingly random po-
rous deposit of metallic platinum appears in place of the original
honeycomb-like structure of porous silicon. By performing immer-
sion plating on such layers, platinum deposits up to 8 ␮m thick
could be obtained with very good in-depth homogeneity.
The morphology of the deposits was analyzed by scanning elec-
tron microscopy. The penetration depth of platinum into the sub-
strate was probed by an electron probe microanalyzer ͑EPMA͒ with
The current program for the medium-doping wafers consisted of
2
an initial 300 mA/cm 20 s high current pulse, followed by 80 s
2
anodization at 130 mA/cm current density. For the high-doping wa-
2
fers, the current program consisted of an initial 300 mA/cm high
current pulse maintained for 10 s, followed by a 30 s descent until
2
2
7
0 mA/cm , 200 s anodization at 70 mA/cm , and a final 150 s de-
scent until 0 current. Overall porosity of the medium-doping porous
silicon samples was found to be close to 54% by gravimetry. For the
high-doping wafers, overall porosity was found to be close to 69%.
*
Electrochemical Society Active Member.
E-mail: jgabrito@rs.noda.tus.ac.jp
z
Downloaded on 2015-05-14 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C742
Journal of The Electrochemical Society, 153 ͑11͒ C741-C746 ͑2006͒
2
and 0.3 cm , was carefully measured with the aid of an optical mi-
croscope for each electrode and used later to calculate current den-
sities. For electrochemical measurements, it was necessary to sputter
a small gold contact layer at the border of the platinum deposits,
because direct electrical contact between the immersion-plated plati-
num layer and silicon does not seem to be good enough, which
causes severe distortions on the voltammograms due to the uncom-
pensated resistance. Figure 1 contains the scheme of the constructed
electrodes. To protect the central area of the deposits a 6 mm wide
hard mask was used during sputtering.
The electrolyte was 0.5 mol/L sulfuric acid. Cyclic voltammo-
grams were recorded at scan rates from 10 up to 200 mV/s between
−
0.35 and 1.35 V vs Ag/AgCl electrode. Electrochemical measure-
ments were performed with agitation by a magnetic stirrer. The elec-
trochemically active surface areas of the deposits were estimated
from the charge associated with the cathodic underpotential deposi-
tion of the hydrogen monolayer on top of platinum during the
negative-going scan at 20 mV/s scan rate. The conventional charge
2
density of 220 ␮C/cm for platinum was used for surface area cal-
Figure 1. Scheme of the assembled electrodes. Above: front view; below:
cross section.
culation. Integration limits were determined using the graphic pro-
cedure proposed by Gilman. Briefly, the low potential limit of
13
charge integration was determined by the intersection of the linear
projections of the line of increasing current corresponding to the
onset of hydrogen evolution and the line of decreasing current cor-
responding to the last peak of the hydrogen monolayer formation
region. The current density measured around 0.2 V vs Ag/AgCl was
used as baseline for charge integration in order to approximately
compensate for the capacitive current.
Anodization conditions were optimized independently for the me-
dium and high-doping samples to maximize the depth attained by
the deposit. In particular for the high-doping samples, the progres-
sive reduction of the current at the end of anodization was employed
to reduce pore size at the bottom of the layer and reduce the prob-
ability of peeling off of platinum layers after deposition.
Immersion plating.— The plating solution consisted of 1 mol/L
Results
sulfuric acid and variable concentrations of PtCl2 and HF. Immer-
−
6
Morphological analysis.— The deposits obtained on medium
and high-doping wafers have quite different visual aspects. The
platinum deposits on medium-doping wafers have a grayish color
and a bright surface finish. For the high-doping wafers, the deposits
are black in color and not shiny. Figures 2 and 3 contain SEM
images of the porous layer right after anodization and of the depos-
ited platinum layers for medium and high-doping samples, respec-
tively. For the medium-doping sample ͑Fig. 2͒, the plating solution
contained 20 mmol/L H PtCl , 1 mol/L H SO , and 400 mmol/L
sion plating was performed right after anodization in the same cell,
also at 10°C, and also under ambient illumination. First, the anod-
ization electrolyte was removed from the cell, the sample was thor-
oughly rinsed with acetone, and the plating solution was quickly
inserted, so that the acetone film could not completely evaporate
from the sample surface. Around 3 mL of plating solution was
poured into the cell. The plating solution was allowed to remain in
contact with the sample for different lengths of time ͑from 5 up to
2
6
2
4
3
0 min͒. Agitation was performed by repeatedly pulling solution out
HF. For the high-doping sample, the plating solution consisted of
0 mmol/L H PtCl , 1 mol/L H SO , and 300 mmol/L HF. In both
and pushing it back into the cell by a Teflon pipette driven by a
motor. After plating the samples were rinsed with acetone, removed
from the cell, and dried under air.
2
2
6
2
4
cases, the plating solution stayed in contact with the porous silicon
layer for 15 min. For the medium-doping sample, the as-anodized
porous layer depth is 10.9 ␮m and average pore diameter is near
100 nm. For the high-doping wafers, 9.7 ␮m deep porous layers
were obtained, but the average pore diameter is smaller, around
40 nm. In both samples, the majority of the pores run parallel to the
direction of current flow, thus being perpendicular to the ͑100͒ sur-
face of the silicon substrate and appear to have few ramifications.
For the medium-doping wafers, pore diameter seems to increase
along the depth of the layer. For the high-doping one, there is a
closing of the pores near the bottom of the layer due to the progres-
sive decrease of the current density towards the end of anodization.
Besides the etching of silicon from the pore walls by the plating
solution, there was an evident filling of the pore space by platinum.
While the porous silicon substrate shows a quite organized
honeycomb-like straight pore array, the platinum deposit shows no
evident systematic organization. The thickness of the platinum de-
posits are 8.3 and 6.5 ␮m for medium and high-doping samples,
respectively. The original porous silicon layer thicknesses were 10.9
and 9.7 ␮m for the medium and high-doping samples, respectively;
thus, there was a ca. 2.6 and 3.2 ␮m reduction, respectively. This
reduction may be due to the smaller molar volume of platinum in
comparison with that of silicon. Some peeling at the upper part of
the platinum layers cannot be ruled out, however. The influence of
Morphological analysis.— A JCM-5100 CarryScope ͑JEOL, Ja-
pan͒ scanning electron microscope ͑SEM͒, operated in secondary
electron imaging ͑SEI͒ mode, was used for morphological analysis
of the porous layers and the platinum deposits. For cross-sectional
imaging, the samples were scratched with a diamond-tipped pen and
cleaved.
Elemental mapping.— A JXA-8900 electron probe microana-
lyzer ͑JEOL, Japan͒ was employed for elemental mapping of the
cross sections of the samples. Platinum and silicon elemental distri-
␣
␣
butions were measured using the platinum M and the silicon K
X-ray emission lines, respectively.
Electrochemical measurements.— The synthesized porous plati-
num layers were used as working electrodes in a conventional three-
electrode cell for electrochemical characterization. A Ag wire with
an anodically deposited AgCl film was used as Ag/AgCl reference
electrode and placed near the working electrode. The counter elec-
trode was a platinum wire placed in a separate chamber, connected
to the main cell by a sintered glass junction. Voltammograms were
recorded under ambient illumination conditions. The same poten-
tiostat used as current source for porous silicon production was em-
ployed for these electrochemical measurements. Adhesive tape ͑3M
plating tape͒ was used to delimit a small portion of the deposit area,
and the delimited small area was the only portion of the sample
exposed to the electrolyte. That area, which was always between 0.2
2
6
−
plating time, PtCl , and HF concentration was systematically inves-
tigated. The influence of the plating time on the deposit obtained is
depicted in Fig. 4 for high-doping wafers. With increasing deposi-
Downloaded on 2015-05-14 to IP 128.122.253.212 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, 153 ͑11͒ C741-C746 ͑2006͒
C743
Figure 2. SEM micrographs for medium-doping wafers: ͑A͒ cross-sectional
view of the porous silicon layer; the dashed line indicates the bottom of the
porous layer; ͑B͒ top view of the porous silicon layer after 30 s plasma
etching to remove the upper microporous layer; ͑C͒ cross-sectional view of
the platinum deposited sample. For cross-sectional views, the silicon sub-
strate is at the left-hand side.
Figure 3. SEM micrographs for high-doping wafers: ͑A͒ cross sectional
view of the porous silicon layer; the dashed line indicates the bottom of the
porous layer; ͑B͒ top view of the porous silicon layer after 30 s plasma
etching to remove the upper microporous layer; ͑C͒ cross-sectional view of
the platinum deposited sample. For cross-sectional views, the silicon sub-
strate is at the left-hand side.
a double-layer structure, with the first few tens of nanometers char-
tion time, progressively deeper regions of the porous layer get filled
with platinum. As the porous platinum deposit develops at the upper
portions of the porous silicon layer, diffusion of reagents to the
deeper regions may become more and more restricted, imposing a
practical limit to the lengthening of the plating time in order to get
14,15
acterized by a microporous layer.
This kind of microporous
layer is also observed in our case. Figure 6 shows SEM images of
the surface of a medium-doping sample right after anodization and
after 30 s of etching in a 50 W capacitively coupled SF plasma.
6
Plasma etching removes that layer and reveals the true honeycomb-
like structure of the deeper macropore array. We have observed that
such etching of the upper regions was not necessary prior to plati-
num deposition. We hypothesize that the microporous layer is etched
away by the plating bath itself. The treatment of the porous sample
after anodization is also of interest. We employed acetone rinsing for
the removal of the anodization electrolyte. Moreover, it is very im-
portant that the plating solution be inserted while the surface is still
wet. It seems that acetone, being not very polar and having a low
surface tension, penetrates more easily than water in the hydropho-
bic pore region. This allows the plating solution to uniformly wet
the pore space, leading to much more reproducible results.
2
−
deeper deposits. Besides, the concentration of PtCl decreases dur-
6
ing plating.
2
6
−
Figure 5 shows the effect of changing the PtCl and the HF
concentrations. We found that it is important that the HF concentra-
tion be at least twice as high as the stoichiometric amount necessary
2
6
−
to dissolve the silicon oxide that gets formed as PtCl is reduced,
2
6
−
i.e., about 12 times the concentration of PtCl . As long as HF
concentration is high enough, it does not seem to have much visible
influence on the shape of the deposit or the extent of deposition. If
2
6
−
HF concentration is lowered while keeping that of PtCl constant,
then there appears a deposition pattern consisting of a platinum film
on the surface and several sparse platinum nuclei at the deeper re-
gions of the porous layer, as can be seen in Fig. 5a. This indicates
that the reaction is being inhibited at the deeper pore regions due to
the lack of fluoride ions. With HF excess, PtCl2− concentration has a
6
Elemental mapping.— Figure 7 contains elemental profiles of
silicon and platinum in the cross section of the platinum-deposited
medium-doping sample. Similar results were obtained for the high-
doping one. Electron probe microanalysis indicates that, except for a
thin transition region at the base of the porous layer, where a more
silicon-rich plateau can be seen, the platinum concentration along
several micrometers of the depth of the deposits is almost uniform.
Moreover, the silicon signal is below the detection limit of the X-ray
spectrometer in the deposit region, indicating a virtually complete
substitution of silicon by platinum by our synthesis procedure.
big influence on the speed of deposition and the depth attainable by
the deposit, although the shape of the deposits does not seem to
change as a function of this variable. Lower PtCl2 concentrations
−
6
require longer plating times to obtain deep deposits and vice versa,
2
6
−
as can be seen in Fig. 5b. When higher PtCl concentration is used,
the deposition process becomes faster all over the porous layer and
a peeling of the platinum deposit, along with the porous silicon
layer, is often observed.
Another interesting point is the question of the influence of the
upper regions of the porous layer. It is not uncommon for anodically
formed macro- or mesopores in highly doped n-type silicon to have
Electrochemical characterization.— Figures 8 and 9 contain cy-
clic voltammograms obtained with some of the assembled platinum
electrodes in 0.5 mol/L H SO solution. The voltammograms in
2
4
Fig. 8 were obtained without employing the contact gold bands
Downloaded on 2015-05-14 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C744
Journal of The Electrochemical Society, 153 ͑11͒ C741-C746 ͑2006͒
Figure 5. Cross-sectional SEM micrographs of platinum deposits obtained
2
−
with medium-doping wafers for ͑A͒ 20 mmol/L ͓PtCl6͔ and 100 mmol/L
2
−
HF after 15 min deposition and ͑B͒ 10 mmol/L ͓PtCl6͔ and 200 mmol/L
HF after 30 min deposition.
the platinum deposits ͑obtained by SEM observation; Figs. 2 and 3͒,
2
3
one gets area/volume ratios equal to 79.75 and 74.83 m /cm ,
respectively.
Figure 4. Cross-sectional SEM micrographs of platinum deposits obtained
with high-doping wafers as a function of the plating time. ͑A͒ 5 min; ͑B͒
1
0 min; ͑C͒ 15 min. The silicon substrate is at the left-hand side.
͑
using the back-side Cu layer as electric contact͒ with a porous
Pt layer synthesized on a medium-doping sample. Similar results
are obtained using high-doping substrates. As can be seen, they
are severely distorted by the uncompensated resistance between
the platinum deposit and the silicon substrate. For the voltammo-
grams shown in Fig. 9, the gold bands were used. The traces of
the typical electrochemical processes happening on platinum elec-
trodes in acid media can be clearly seen. The charge densities
associated with underpotential hydrogen monolayer formation, cal-
culated from the negative-going scans of the voltammograms shown
in Fig. 9 according to the procedure described in the Experimental
section, are −145.6 and −107.0 mC/cm2 for the medium and
high-doping samples, respectively. This gives roughness factors
Figure 6. SEM micrograph of the surface of a medium-doping porous
sample right after anodization; after 30 s plasma etching, it becomes like
Fig. 2b.
͑
real area/projected area͒ equal to 662 and 486, respectively,
for those samples. Taking also into consideration the thickness of
Downloaded on 2015-05-14 to IP 128.122.253.212 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, 153 ͑11͒ C741-C746 ͑2006͒
C745
Figure 9. Cyclic voltammograms obtained with two of the porous-silicon
Figure 7. Elemental mapping of platinum and silicon along the cross section
of the platinum deposited medium-doping sample.
supported platinum electrodes in 0.5 mol/L H SO solution at 20 mV/s scan
2
4
rate. The gold contact bands were used.
Discussion
on medium-doping samples has thicker pore walls, we are led to
believe that the platinum grains that replace it tend to be bigger.
It was observed that HF concentration does not have a big influ-
ence on the deposition behavior as long as it is present in large
excess ͑at least twice the stoichiometric amount necessary to dis-
solve the silica formed by the deposition reaction͒. Under such con-
Although both original porous silicon samples show a rather or-
ganized parallel pore geometry, the platinum deposits grown on
them have a spongy appearance ͑Fig. 2 and 3͒. The original pore
shape is still visible only at the very bottom of the layer, having been
completely clouded by the platinum deposit at all other positions
2
−
2−
along the depth. If the ͓PtCl6͔ concentration is reduced and the
platinum deposit does not reach down to the bottom of the porous
layer, then it becomes possible to see a gradual transition from the
spongy platinum deposit to the original parallel-pore substrate, with
an intermediary region where platinum nuclei can be seen ͑Fig. 5b͒.
This may suggest that deposition in the deeper regions of the porous
layer is limited by the diffusional flux of reactants from the bulk
solution through the pore space. There is a visible difference be-
tween the deposits obtained with medium-doping wafers and those
ones obtained with the high-doping wafers. The first ones seem to
have larger average platinum grain size than the latter ones, inde-
pendently of the concentrations of the reagents in the plating solu-
tion. The reason for this difference in morphology is not clear, but is
probably related to the different pore sizes observed in the medium
and high-doping samples. Because the porous silicon layer obtained
ditions, PtCl6 has a strong influence on both the deposition rate and
the depth attainable by the deposit. From these results, it is evident
that the process consists of at least two steps: a redox one, in which
2
6
−
PtCl is reduced and silicon oxidized, and a dissolution step
2
6
−
Redox step: PtCl ͑aq͒ + Si͑s͒ + 2H O
2
+
−
→
Pt͑s͒ + SiO ͑s͒ + 4H ͑aq͒ + 6Cl ͑aq͒
2
−
+
Dissolution step: SiO ͑s͒ + 6F ͑aq͒ + 4H ͑aq͒
2
2
−
→
SiF ͑aq͒ + 2H O
6
2
Of course, the actual mechanism must be more complex than
this. For example, Harraz et al. proposed that metal nucleation/
growth and silicon oxidation actually occur at different surface
7
sites. HF concentration strongly affects the kinetics of the oxidized
silicon dissolution step, such that, once it is high enough, the first
step becomes the limiting one and the overall reaction rate is con-
2
6
−
trolled by the PtCl concentration.
SEM morphological analysis and elemental mapping indicate
that the platinum deposits reach down to the very bottom of the
porous layer. However, electrical contact between the deposited
platinum layer and the silicon substrate is rather poor, leading to the
necessity of sputtering a gold contact band in order to perform elec-
trochemical measurements with the porous platinum electrodes.
1
6
Kuznetsov et al. performed immersion plating on single-crystal
silicon from HF-containing baths and suggested that a dielectric
layer develops between platinum and silicon in this case, too. The
same might be happening in our case, although further studies will
be necessary to better understand the nature of the contact between
platinum and porous silicon.
The roughness factors of the medium and high-doping samples
obtained from those voltammograms are 662 and 486, respectively,
2
3
which yields area-volume ratios equal to 79.75 and 74.83 m /cm ,
respectively. SEM observation indicates that platinum grain size is
actually smaller for the high-doping samples than for the medium-
doping ones, which suggests a higher surface area and leads one to
expect a higher electrocatalytic activity. The reason for this apparent
discrepancy is under investigation. We suspect that it might be re-
lated to differences in electronic conductance between the samples.
Figure 8. Cyclic voltammograms obtained with a porous platinum electrode
in 0.5 mol/L H SO without the contact gold bands. Medium-doping wafers
2
4
were used.
Downloaded on 2015-05-14 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C746
Journal of The Electrochemical Society, 153 ͑11͒ C741-C746 ͑2006͒
Conclusions also like to thank Tsutomu Iida at Tokyo University of Science for
the electron probe microanalysis measurements.
Porous platinum layers were successfully synthesized by immer-
sion plating performed on porous silicon layers of different pore size
characteristics, but regular geometry, prepared by electrochemical
etching of highly doped n-type silicon substrates. From scanning
electron micrographs we could conclude that a random network of
platinum pores forms in place of the original, rather organized po-
rous silicon substrate. The in-depth elemental composition map of
the samples was determined by electron probe microanalysis. Sili-
con concentration in the deposits is very low and the platinum con-
centration is quite uniform along the depth of the sample. Cyclic
voltammetry experiments were performed with the synthesized elec-
trodes. It was observed that the electronic contact between the plati-
num deposit and the silicon substrate is not good. By performing
electrical contact with the sample from two gold bands sputtered on
top of the deposits, their electrochemically active surface areas
Tokyo University of Science assisted in meeting the publication costs of
this article.
References
1
2
. M. Hayase, T. Kawase, and T. Hatsuzawa, Electrochem. Solid-State Lett., 7, A231
2004͒.
. M. Jeske, J. W. Shultze, M. Thönissen, and H. Münder, Thin Solid Films, 255, 63
͑1995͒.
͑
3. G. D’Arrigo, C. Spinella, G. Arena, and S. Lorenti, Mater. Sci. Eng., C, 23, 13
2003͒.
͑
4
. F. A. Harraz, K. Kamada, K. Kobayashi, T. Sakka, and Y. H. Ogata, J. Electro-
chem. Soc., 152, C213 ͑2005͒.
5. H. Lin, J. Mock, D. Smith, T. Gao, and M. J. Sailor, J. Phys. Chem. B, 108, 11654
͑2004͒.
6
. F. A. Harraz, J. Sasano, T. Sakka, and Y. H. Ogata, J. Electrochem. Soc., 150,
C277 ͑2003͒.
7
. F. A. Harraz, T. Tsuboi, J. Sasano, T. Sakka, and Y. H. Ogata, J. Electrochem. Soc.,
149, C456 ͑2002͒.
2
3
could be determined to be close to 80 m /cm . Such porous silicon
supported platinum layers show promise as catalyst layers to be
employed for MEMS-based micro fuel cells.
8. F. A. Harraz, T. Sakka, and Y. H. Ogata, Electrochim. Acta, 47, 1249 ͑2002͒.
. T. Tsuboi, T. Sakka, and Y. H. Ogata, Appl. Surf. Sci., 147, 6 ͑1999͒.
0. I. Coulthard and T. K. Sham, Appl. Surf. Sci., 126, 287 ͑1998͒.
1. I. Coulthard and T. K. Sham, Solid State Commun., 105, 751 ͑1998͒.
12. I. Coulthard, D.-T. Jiang, J. W. Lorimer, T. K. Sham, and X.-H. Feng, Langmuir, 9,
441 ͑1993͒.
13. S. Gilman, J. Electroanal. Chem., 7, 382 ͑1964͒.
9
1
1
Acknowledgments
3
This study was supported by the industrial technology research
grant program 04A49007 from New Energy and Industrial Technol-
ogy Development Organization ͑NEDO͒ of Japan. J. G. A. Brito-
Neto thanks the Japan Society for the Promotion of Science ͑JSPS͒
for a postdoctoral fellowship ͑number P05338͒. The authors would
1
4. K. Takemoto, Y. Nakamura, and O. Nittono, Jpn. J. Appl. Phys., Part 1, 33, 6432
1994͒.
͑
1
1
5. P. C. Searson, J. M. Macaulay, and F. M. Ross, J. Appl. Phys., 72, 253 ͑1992͒.
6. G. V. Kuznetsov, V. A. Skryshevsky, T. A. Vdovenkova, A. I. Tsyganova, P. Gor-
ostiza, and F. Sanz, J. Electrochem. Soc., 148, C528 ͑2001͒.
Downloaded on 2015-05-14 to IP 128.122.253.212 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).