Nucleation behavior in electroless displacement deposition of metals on silicon from hydrofluoric acid solutions

Article abstract of DOI:10.1016/j.electacta.2007.04.058

We investigate the nucleation behavior in the electroless displacement deposition of metal particles (Pt, Rh, Pd, Cu, Ag, and Au) onto n-Si wafers from a metal-salt solution containing HF. The particle density of metals varies widely from 10⁶ (Pt) to 10¹¹ (Au) cm^-2, depending on the kind of metal. Deposited metals can be classified into two types of nucleation behavior. One consists of the platinum group elements, including Pt, Rh, and Pd, which display lower particle densities than elements of the other group and depend on the type of pretreatment of the n-Si wafer, and thus the surface conditions of Si. The second group consists of the copper group elements, including Cu, Ag, and Au, which display higher particle density than the first group and are independent of pretreatment. The size of deposited particles decreases from hundreds nm to tens nm as the particle density increases. Moreover, the displacement deposition of the Pt and Ag particles onto n-Si are in progressive and instantaneous nucleation modes, respectively.

Full text of DOI:10.1016/j.electacta.2007.04.058

Electrochimica Acta 53 (2007) 35–41
Nucleation behavior in electroless displacement deposition
of metals on silicon from hydrofluoric acid solutions
Shinji Yaea,b,∗,1, Noriaki Nasua, Kohei Matsumotoa, Taizo Hagiharaa,
Naoki Fukumuroa, Hitoshi Matsudaa
a Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo,
2167 Shosha, Himeji, Hyogo 671-2280, Japan
b CREST, JST, 4-1-8 Honcho, Kawaguchi, Saitama 322-0012, Japan
Received 30 November 2005; received in revised form 25 December 2006; accepted 13 April 2007
Available online 25 April 2007
Abstract
We investigate the nucleation behavior in the electroless displacement deposition of metal particles (Pt, Rh, Pd, Cu, Ag, and Au) onto n-Si
wafers from a metal-salt solution containing HF. The particle density of metals varies widely from 106 (Pt) to 1011 (Au) cm−2, depending on the
kind of metal. Deposited metals can be classified into two types of nucleation behavior. One consists of the platinum group elements, including
Pt, Rh, and Pd, which display lower particle densities than elements of the other group and depend on the type of pretreatment of the n-Si wafer,
and thus the surface conditions of Si. The second group consists of the copper group elements, including Cu, Ag, and Au, which display higher
particle density than the first group and are independent of pretreatment. The size of deposited particles decreases from hundreds nm to tens nm
as the particle density increases. Moreover, the displacement deposition of the Pt and Ag particles onto n-Si are in progressive and instantaneous
nucleation modes, respectively.
© 2007 Elsevier Ltd. All rights reserved.
Keywords: Noble metal deposition; Metal nanoparticles; n-Si; Photoelectrochemistry; Surface states
1. Introduction
wafer cleaning solutions containing fluoride [19–22]. Recently,
we reported that electroless displacement deposition produced
fine metal particles suitable for the formation of porous silicon
using metal-particle-enhanced HF etching and for the creation
of efficient photoelectrochemical solar cells [23–26]. This study
investigates the nucleation behavior of electroless displacement
deposition.
The electrochemical formation of fine metal lines or con-
tacts on Si wafers has been applied to the electronics industry to
replace aluminum with copper as a conducting material. Much
attention has been directed to the electrodeposition of metals
onto Si [1–9]. The electroless displacement deposition of met-
als onto silicon is another electrochemical method. It is a simple
process involving only the immersion of silicon wafers into
metal-salt solutions containing HF, and various kinds of metals
can be deposited on Si using this method [10–15]. It can be used
for the metallization of silicon and the catalyzation of silicon
surfaces for autocatalytic metal deposition [14,16–18]. It is also
important as a mechanism of metal impurity deposition from
2. Experimental
Single-crystal n-Si wafers (Yamanaka Semiconductor, CZ
(100), ca. 1 ꢀ cm) were cut into pieces 0.8 cm × 0.8 cm in area
and used as substrates. We used four types of pretreatments for n-
Si: washed ultrasonically with acetone for five min and etched
with CP-4A (a mixture of HF, HNO3, CH3COOH and H2O
(3:5:3:22 in volume)) for three min and a 7.3 M (M = mol dm−3
)
∗
Corresponding author at: Department of Materials Science and Chemistry,
HF solution for two min (hereafter, we call this pretreatment A);
washed by the RCA method (i.e., immersing in a mixture of
NH4OH, H2O2 and H2O (1:1:5 in volume) at 353 K for 15 min,
and then in a mixture of HCl, H2O2 and H2O (1:1:5 in volume)
Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji,
Hyogo 671-2280, Japan. Tel.: +81 79 267 4911; fax: +81 79 267 4911.
E-mail address: yae@eng.u-hyogo.ac.jp (S. Yae).
1
ISE member.
0013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2007.04.058

36
S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
at 353 K for 15 min) and etched with a 7.3 M HF (pretreatment
B); immersed in 14 M HNO3 at 353 K for 30 min after pre-
treatment A or B (pretreatments C and D, respectively). The
electroless displacement deposition of these metals was carried
out by immersing the n-Si substrates in a 1.0 mM metal-salt
(H2PtCl6, AgNO3, HAuCl4, CuCl2, PdCl2 or RhCl3) solution
containing 0.15 M HF at 313 K. Unless otherwise stated, the
immersion time of n-Si in the deposition solution was 120 s.
The n-Si substrates were immersed in ultra pure water after
each pretreatment and deposition step. The immersion poten-
tial of n-Si was measured using a Ag/AgCl electrode as a
reference electrode and a potentiogalvanostat (Nikko Keisoku,
NPGS-2501, potentiometer mode) under room-light condition
(0.01–0.1 mW cm−2, measured with a thermopile (Neoark, PM-
335)). The surface inspection of the n-Si substrates and the
measurement of particle density of metals were performed with
a high-resolution scanning electron microscope (SEM, Hitachi,
S-900, 20 kV).
deposition solution. The particle density of Pt decreased by two
orders of magnitude from 3 × 108 to 5 × 106 cm−2, and the size
of particles increased from 20–300 to 150–600 nm in pretreat-
ments A–D. On the other hand, the particle density of Ag was
nearly constant between 3 and 5 × 1010 cm−2, independent of
pretreatment.
The particle densities of the various metal particles deposited
by displacement on the n-Si substrates were plotted accord-
ing to the methods of pretreatment in Fig. 3. Each point in
Fig. 3 indicates the average value of three different positions
of each sample. Deposited metals can be classified into two
types by the value of particle density and dependence on pre-
treatment. One type consists of Pt, Rh, and Pd, which display
lower particle densities than the other group and are depen-
dent on pretreatment. These metals are platinum group elements
(groups 9 and 10). The second group consists of Cu, Ag, and
Au, which display high particle densities (1010–1011 cm−2) and
are independent of the pretreatment type. Furthermore, each
of these metals, which are copper group elements (group 11),
showslower-densitydispersionvaluesthanthoseoftheplatinum
group elements. Pretreatment D yielded a wide range of parti-
cle density, from 3 × 106 for Pt to 1 × 1011 cm−2 for Au; this
difference between Pt and Au is larger than four orders of mag-
nitude. The particle density of Pt particles for pretreatment A
3. Results
Fig. 1 shows typical SEM images of n-Si substrates (pretreat-
ment A) after immersion into the metal-salt solutions containing
HF. Fine metal particles were deposited on the n-Si substrates.
The particle density of metals varied according to the kind of
metal, ca. 3 × 108 (Pt), 7 × 108 (Rh), 1 × 1010 (Pd), 2 × 1010
(Cu), 4 × 1010 (Ag) and 1 × 1011 cm−2 (Au). The size of metal
particles decreased from 20–300 nm for Pt to 5–70 nm for Au
as the particle density increased. The particle density of Pt is a
similar value of 3 × 108 cm−2 for Pt particles electrodeposited at
−0.7 V versus SCE from a 5 mM H2PtCl6 aqueous solution on
the n-Si substrate pretreated by the same method as A [9]. Fig. 2
shows the SEM images of n-Si wafers that were pretreated by
method B, C, or D and immersed in the Pt or Ag displacement
was scattered in a wide range, from 8 × 107 to 1 × 109 cm−2
.
The particle density for pretreatment B was in the range of
3–6 × 107 cm−2, that is, one order lower and with a lower
dispersion than pretreatment A. The n-Si substrates pretreated
using method A or B have no oxide layer caused by final HF
etching, while those for methods C and D have similar oxide
layers caused by HF etching and subsequent nitric acid oxi-
dation. These results show that silicon oxide layers on n-Si
substrates lower the particle density for displacement deposition
of Pt and that the different cleaning methods of silicon involving
Fig. 1. SEM of n-Si substrates (pretreatment A) after immersion into 1.0 mM metal-salt solutions containing 0.15 M HF for 120 s at 313 K.

S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
37
Fig. 2. SEM of n-Si substrates after immersion into displacement deposition solution of Pt or Ag. B, C, and D indicate the n-Si substrates pretreated by methods
B–D, respectively.
CP-4A etching and the RCA method give different Pt particle
densities.
immersion time of 120 s (Fig. 2), the first nucleation and growth
of metal particles were slower than those for pretreatment A
(Fig. 4). The particle densities of Pt increased with immersion
time and then maintained nearly constant values in both pre-
treatments A and D (Fig. 5). These results indicate that the Pt
nucleation is in a progressive mode. The nucleation and growth
of Pt particles proceed simultaneously in the early stages of dis-
placement deposition, which causes wide variation in Pt particle
size. The immersion potential of the n-Si pretreated by method
A was nearly constant during the deposition, but the potential
for pretreatment D had a large negative peak in the early stages
of deposition (Fig. 6). Pretreatment B showed similar changes
in particle density and potential to pretreatment A, but the par-
ticles deposited for a short immersion time (<20 s) were more
homogeneously scattered than those for pretreatment A. Pre-
treatment C gave similar results to pretreatment D with different
particle density values, as shown in Figs. 2 and 3. The displace-
ment deposition of Ag indicated different nucleation behavior
(Figs. 7 and 8). In the case of pretreatment A, fine particles,
1–30 nm in size, were formed by a 1-s immersion in the deposi-
tion solution (Fig. 7). Its particle density of 1.4 × 1011 cm−2 was
higher than the 4 × 1010 cm−2 value found for the 120 s immer-
sion case (Fig. 1). In the case of 10 s immersion in the deposition
solution, the size of the deposited particles, 1–50 nm, was larger
and the particle density, 0.8 × 1011 cm−2, was lower than those
for the 1-s immersion case (Fig. 7). Immersion longer than about
10 s gave a nearly constant or only a slightly decreasing value of
the particle density (Figs. 7 and 8). Immersion of the n-Si sub-
strates pretreated by method D for 10 s formed larger and more
scattered-sized particles, from 5 to 90 nm, and with a one-order
lower particle density, 0.1 × 1011 cm−2, than those for method
A (Fig. 7). The particle density increased with immersion time
until 30 s and then stayed nearly constant or decreased slightly
Typical SEM images of n-Si substrates, the particle densi-
ties of deposited particles, and the immersion potential of n-Si
in the initial stage of Pt displacement deposition are shown
in Figs. 4–6, respectively. For pretreatment A, Pt particles of
15–150 nm in size were deposited by immersion of the n-Si
substrate in the deposition solution for 5 s (Fig. 4). There were
some areas in which no Pt particle was visible (with the same
magnification), even in the sample shown in Fig. 4. The parti-
cle density of Pt deposited for an immersion time shorter than
20 s spatially varied in each sample using pretreatment A. In
spite of the large particle formation for pretreatment D at an
Fig. 3. Particle density (D) of various metals deposited by displacement on n-
Si pretreated by methods A–D. Marks represent the following samples: white
circles: Pt, white triangles: Rh, crosses: Pd, white diamonds: Cu, white squares:
Ag, and black triangles: Au.

38
S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
Fig. 4. SEM of n-Si substrates (pretreatment A or D) after immersion in displacement deposition solution of Pt for 5–20 s. A and D indicate n-Si pretreated by
methods A and D, respectively. Numbers after A or D indicate immersion time. For example, A-5s indicates n-Si substrate pretreated by method A and immersed in
deposition solution of Pt for 5 s.
(Figs. 7 and 8). Pretreatments B and C showed Ag deposition
behaviors similar to pretreatments A and D, respectively.
and Pd, which display lower particle densities than the other
group and are dependent on pretreatment. The second group
consists of Cu, Ag, and Au, which display high particle densi-
ties and are independent of the pretreatment type. The metals
of each group have common properties as follows. The metals
of the former group are platinum group elements (groups 9 and
10) that form metal silicide compounds with silicon [29]. The
metals of the latter group are copper group elements (group 11)
that form a eutectic or peritectic alloy with silicon [29]. In each
group, the particle density increased as the standard electrode
potential of metals increased (Pt 0.744, Rh 0.76, Pd 0.915, Cu
0.340, Ag 0.799 and Au 1.002 V versus SHE [28]). Therefore,
there are two kinds of mechanisms determining the particle den-
sity. One classifies the nucleation behavior into two types, and
4. Discussion
The particle density of metal particles deposited by displace-
ment on the n-Si substrates varied widely among the different
kinds of metals. The particle density increased in the order of
Pt, Rh, Pd, Cu, Ag and Au. We could not find any physical or
thermodynamic property, such as melting point [27], electroneg-
ativity [27], or standard electrode potential [28], of these metals
that changes by this order. Deposited metals can be classified
into two types of nucleation behavior. One consists of Pt, Rh,
Fig. 5. Particle density (D) of deposited Pt particles on n-Si substrates as a
function of immersion time (t) in deposition solution. Black circles and white
squares represent samples pretreated by methods A and D, respectively.
Fig. 6. Immersion time (t) dependence of potential (E) of n-Si in displacement
deposition solution of Pt. Solid line and dashed line indicate n-Si substrates
pretreated with methods A and D, respectively.

S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
39
Fig. 7. Same as Fig. 4 except that n-Si substrates were immersed in Ag deposition solution.
the other determines the particle densities of the metals in each
group. These mechanisms are not clear at present, and clarifying
them will require further investigation. The particle density of
Pt clearly depends on the type of pretreatment. The sensitivity
of Pt nucleation on the Si surface condition is caused by the fact
that the activity of Pt deposition on a Si surface is much lower
than that on a metal surface. This difference in activity was indi-
cated by the following result. Fig. 9a shows an n-Si substrate
(pretreatment A) that was immersed in the Ag deposition solu-
tion for 20 s and then in the Pt solution for 60 s. Comparing
Figs. 1 and 9a indicates that a high particle density of Pt was
obtained using fine Ag particles as seeds. The deposition of Pt
on Si requires strong nucleation sites, such as the Ag particles.
Pretreatment A gave higher and more scattered particle den-
sity values than those for pretreatment B (Fig. 3). It has been
reported that the nucleation of metal displacement deposition
on Si wafers is influenced by introducing defects onto Si, such
Fig. 8. SameasFig. 5exceptthatn-SisubstrateswereimmersedinAgdeposition
solution.
Fig. 9. SEM of n-Si substrates (pretreatment A) immersed in Ag deposition solution for 20 s and then in Pt solution for 60 s (a), and that of Pt solution for 20 s and
Ag for 60s (b).

40
S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
as ion implantation [30] and nanoindentation [31]. The CP-4A
etching used in pretreatment A is a wet chemical polish etching
[32,33]. At present, the RCA method used in pretreatment B is
the most common and suitable cleaning method of Si wafers
[32,33]. Thus, pretreatment B is expected to clean Si surfaces
that have a lower density of Pt nucleation sites such as sur-
face states, defects, impurities, and roughness than those that
can be cleaned by pretreatment A. The formation of a silicon
oxide layer lowers the particle density of Pt. This result has two
possible reasons. One is that the density of the interface states
between the Si and Si oxide is very low, especially in the case
of an oxide layer formed by chemical oxidation using nitric
acid [34]. The other possible reason is that the Si oxide layer
affects the nucleation of Pt particles. The thickness of the Si
oxide layer is estimated to be 1.2 nm [34], which takes 0.8 min
to be etched with 0.15 M HF at room temperature [35]. The
immersion potential change of n-Si pretreated by method D
indicates the dissolution of Si oxide continued for ca. 30 s at
313 K (Fig. 6). Since the etching of Si oxide with HF is a chem-
ical reaction, no electrochemical charge transfer between Si and
the solution occurs. The displacement deposition occurs by a
local cell mechanism that consists of a local cathode deposition
of metal particles with the injection of holes into the valence
band of the n-Si, and the local anode dissolution of silicon by
the injected holes and HF. Consequently, the deposition of Pt
particles proceeds only on a bare Si surface. Non-uniformity of
the thickness or etching reaction of the Si oxide layer produces
a partially bare Si surface in the early stage of immersion in
the deposition solution. The nucleation site density of oxide-
covered Si (pretreatments C and D) is lower than that of the
entire bare Si surface (pretreatments A and B). The above expla-
nation is consistent with the initial deposition behavior shown
in Figs. 4 and 5.
Silvershowedvastlydifferentbehaviorindisplacementdepo-
sition from Pt. This can be explained by the high activity of Ag
deposition on the Si surface. Fig. 9b shows the reverse case of
Fig. 9a. The n-Si substrate was immersed in the Pt displace-
ment deposition solution followed by the Ag solution. Fine Ag
particles were deposited between the Ag-covered Pt particles
(Figs. 4A-20s and 9b). This shows that the nucleation of Ag
particles on Si has much higher activity than that for Pt. Thus,
the particle densities of Ag are high and independent of the type
of pretreatment. The Ag nucleation on the n-Si substrate pre-
treated by method A is in the instantaneous mode (Fig. 7A-1s).
After rapid nucleation, the deposited particles are dissolved or
detached from the Si surface by the Si anode dissolution accom-
panying the Ag cathode deposition. Then, the Ag particles grow,
and the increase in particle size attaches the particles to each
other. Accordingly, the distribution density decreases in the first
10 s and then stays nearly constant or decreases slightly. More
particle attachment was shown in the Au deposition (Fig. 1). In
the case of pretreatment D, the nucleation of Ag particles and
dissolution of the Si oxide layer simultaneously proceeded in
the first 30 s. The particle density increased as the area of bare
Si increased. After dissolution of all Si oxide, the growth of Ag
particles mainly occurs in cases similar to that of pretreatment
A. Therefore, the nucleation of Ag with pretreatment D seems
to be in the progressive mode, and the particle size varies widely
after 30 s deposition.
In conclusion, we revealed in the present work that (1) the
nucleation of electroless displacement deposition of metals on n-
Siwafersdependsonthekindofdepositedmetalsandthesurface
conditions of Si wafers; (2) deposited metals can be classified
into two types of nucleation behavior: one consisting of platinum
group elements, including Pt, Rh, and Pd, that display lower
particle densities than the other group and that are dependent
on pretreatment, and a second group consisting of copper group
elements, including Cu, Ag, and Au, that display higher particle
density than the first group and that are independent of pre-
treatment; (3) the nucleation of Pt and Ag on n-Si are in the
progressive mode and the instantaneous mode, respectively.
Acknowledgements
The present work was partly supported by Grants-in-Aid
for Scientific Research (C) (17560638) from Japan Society for
the Promotion of Science, and for education and research from
Hyogo Prefecture through the University of Hyogo.
References
[1] X.G. Zhang, Electrochemistry of Silicon and its Oxide, Kluwer Aca-
demic/Plenum Pub., New York, NY, 2001, Ch. 6.3.
[2] G. Oskam, J.G. Long, A. Natarajan, P.C. Searson, J. Phys. D: Appl. Phys.
31 (1998) 1927.
[3] P. Garrido, E. Go´mez, E. Valle´s, J. Electroanal. Chem. 441 (1998) 147.
[4] K. Morisawa, M. Ishida, S. Yae, Y. Nakato, Electrochim. Acta 44 (1999)
3725.
[5] M. Cerisier, K. Attenborough, J. Fransaer, C. Van Haesendonck, J.-P. Celis,
J. Electrochem. Soc. 146 (1999) 2156.
[6] R.M. Stiger, S. Gorer, B. Craft, R.M. Penner, Langmuir 15 (1998) 790.
[7] B. Rashkova, B. Guel, R.T. Po¨tzschke, G. Staikov, W.J. Lorenz, Elec-
trochim. Acta 43 (1998) 3021.
[8] J.C. Ziegler, R.I. Wielgosz, D.M. Kolb, Electrochim. Acta 45 (1999) 827.
[9] S. Yae, M. Kitagaki, T. Hagihara, Y. Miyoshi, H. Matsuda, B.A. Parkinson,
Y. Nakato, Electrochim. Acta 47 (2001) 345.
[10] L.A. Nagahara, T. Ohmori, K. Hashimoto, A. Fujishima, J. Vac. Sci. Tech-
nol. A 11 (1993) 763.
[11] P. Gorostiza, J. Servat, J.R. Morante, F. Sanz, Thin Solid Films 275 (1996)
12.
[12] P. Gorostiza, P. Allongue, R. D´ıaz, J.R. Morante, F. Sanz, J. Phys. Chem.
B 107 (2003) 6454.
[13] P. Gorostiza, M.A. Kulandainathan, R. D´ıaz, F. Sanz, P. Allongue, J.R.
Morante, J. Electrochem. Soc. 147 (2000) 1026.
[14] M. Chemla, T. Homma, V. Bertagna, R. Erre, N. Kubo, T. Osaka, J. Elec-
troanal. Chem. 559 (2003) 111.
[15] M.V. ten Kortenaar, J.J.M. de Goeij, Z.I. Kolar, G. Frens, P.J. Lusse, M.R.
Zuiddam, E. van der Drift, J. Electrochem. Soc. 148 (2001) C28.
[16] V.M. Dubin, J. Electrochem. Soc. 139 (1992) 1289.
[17] M.K. Lee, J.J. Wang, H.D. Wang, J. Electrochem. Soc. 144 (1997) 1777.
[18] L. Magagnin, R. Maboudian, C. Carraro, Electrochem. Solid-State Lett. 4
(2001) C5.
[19] H. Morinaga, M. Suyama, T. Ohmi, J. Electrochem. Soc. 141 (1994)
2834.
[20] L. Mouche, F. Taradif, J. Derrien, J. Electrochem. Soc. 142 (1995) 2395.
[21] O.M.R. Chyan, J.-J. Chen, H.Y. Chien, J. Sees, L. Hall, J. Electrochem.
Soc. 143 (1996) 92.
[22] T. Homma, C.P. Wade, C.E.D. Chidsey, J. Phys. Chem. B 102 (1998)
7919.

S. Yae et al. / Electrochimica Acta 53 (2007) 35–41
41
[23] S. Yae, Y. Kawamoto, H. Tanaka, N. Fukumuro, H. Matsuda, Electrochem.
Comm. 5 (2003) 632.
[24] S. Yae, H. Tanaka, T. Kobayashi, N. Fukumuro, H. Matsuda, Phys. Stat.
Sol. (c) 2 (2005) 3476.
[25] S. Yae, T. Kobayashi, T. Kawagishi, N. Fukumuro, H. Matsuda, in: P.
Schmuki, al. et (Eds.), Pits and Pores III: Formation, Properties and
Significance for Advanced Materials, The Electrochemical Society Pro-
ceedings Series 2004-19, The Electrochemical Society, Pennington, 2006,
p. 141.
[29] T.B. Massalski, H. Okamoto, P.R. Subramanian, L. Kacprzak (Eds.),
Binally Alloy Phase Diagrams, 2nd ed., Plus Updates, ASM International,
Materials Park, OH, 1996.
[30] P. Gorostiza, J. Servat, F. Sanz, J.R. Morante, Inst. Phys. Conf. Ser. 149
(1996) 293.
[31] T. Homma, N. Kubo, T. Osaka, Electrochim. Acta 48 (2003) 3115.
[32] X.G. Zhang, Electrochemistry of Silicon and Its Oxide, Kluwer Aca-
demic/Plenum Pub, New York, NY, 2001, Ch. 7.
[33] S.A. Campbell, The Science and Engineering of Microelectronic Fabri-
cation, 2nd ed., Oxford Univ. Press, New York, NY, 2001, Ch. 11.1 and
14.1.
[26] S. Yae, T. Kobayashi, T. Kawagishi, N. Fukumuro, H. Matsuda, Sol. Energy
80 (2006) 701.
[27] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 85th ed., CRC
Press, Boca Raton, FL, 2004.
[34] H. Kobayashi, Asuha, O. Maida, M. Takahashi, H. Iwasa, J. Appl. Phys.
94 (2003) 7328.
[28] A.J. Bard, R. Parsons, J. Jordan (Eds.), Standard Potentials in Aqueous
Solution, IUPAC, Marcel Dekker, New York, 1985.
[35] V. Lehmann, Electrochemistry of Silicon, Willy-VCH, Weinheim, 2002,
Ch. 2.