Deposition of Group 6A-derived inorganic semiconductor films as studied by quartz crystal microgravimetry

Article abstract of DOI:10.1016/S0013-4686(00)00472-2

This paper focuses on the use of QCM for the study of semiconductor film deposition processes. Specifically the electrosynthesis of metal chalcogenides (In₂S₃, CdS, and CdTe) is considered. A brief background is first given for electrodeposition as a process candidate for semiconductor film preparation. Previous studies are reviewed on the use of QCM (and specifically EQCM) in this area. New combined voltammetry-EQCM data are presented for the oxidative deposition of sulfur on polycrystalline Au surfaces from alkaline sulfide baths. The anodic growth of CdS films and the cathodic electrosynthesis of In₂S₃ are studied by the combined approach. Finally, data are presented on the cathodic electrosynthesis of CdTe films using EQCM in a rotating disc electrode (RDE) configuration. From an instrumental perspective, the presented data illustrate the virtues of combining the QCM technique with cyclic voltammetry, coulometry, and hydrodynamic (RDE) voltammetry for studies of semiconductor film deposition.

Full text of DOI:10.1016/S0013-4686(00)00472-2

Electrochimica Acta 45 (2000) 3749–3756
www.elsevier.nl/locate/electacta
The deposition of Group 6A-derived inorganic
semiconductor films as studied by quartz crystal
microgravimetry
a,
a
b
b
c
N. Myung *, S. Kim , D. Lincot , C. Lepiller , N.R. de Tacconi ,
c,1
K. Rajeshwar
a
Department of Applied Chemistry, Konkuk Uni6ersity Chungju Campus, Chungju, Chungbuk 380-701, South Korea
Laboratoire d’Electrochimie et de Chimie Analytique (CNRS), Ecole Nationale Superieure de Chimie de Paris,
b
1
1 rue Pierre et Marie Curie, 75 231 Paris Cedex 5, France
c
Department of Chemistry and Biochemistry, The Uni6ersity of Texas at Arlington, Arlington, TX 76019-0065, USA
Received 2 August 1999; received in revised form 1 October 1999
Abstract
This paper focuses on the use of QCM for the study of semiconductor film deposition processes. Specifically the
electrosynthesis of metal chalcogenides (In S , CdS, and CdTe) is considered. A brief background is first given for
2
3
electrodeposition as a process candidate for semiconductor film preparation. Previous studies are reviewed on the use
of QCM (and specifically EQCM) in this area. New combined voltammetry-EQCM data are presented for the
oxidative deposition of sulfur on polycrystalline Au surfaces from alkaline sulfide baths. The anodic growth of CdS
films and the cathodic electrosynthesis of In S are studied by the combined approach. Finally, data are presented on
2
3
the cathodic electrosynthesis of CdTe films using EQCM in a rotating disc electrode (RDE) configuration. From an
instrumental perspective, the presented data illustrate the virtues of combining the QCM technique with cyclic
voltammetry, coulometry, and hydrodynamic (RDE) voltammetry for studies of semiconductor film deposition.
©
2000 Elsevier Science Ltd. All rights reserved.
Keywords: Quartz crystal microgravimetry; Semiconductor film deposition processes; Electrosynthesis; Metal chalcogenides
1
. Introduction
adlayers and surface reconstruction; and kinetics of
electrochemical processes. The early observation of
Sauerbrey [1] that changes in the resonant frequency of
a piezoelectric crystal such as quartz are proportional
to mass changes, led logically to the extensive use of the
electrochemical QCM (EQCM) counterpart for metal
deposition and stripping analyses [2–4]. Many similar
studies were also performed on the electrodeposition
(and electrodissolution) of conducting polymer films [5].
However, instances where QCM or EQCM has been
utilized for the study of inorganic semiconductor films
are more scarce. This topic, therefore, constitutes the
focus of this paper.
As also exemplified by the other articles in this
Special Issue, quartz crystal microgravimetry (QCM)
has been successfully used in the study of many electro-
chemical problems, including trace analysis; mass trans-
port in redox and conducting polymers; double layer
properties and depletion layer effects at electrode/solu-
tion interfaces; surface roughness and liquid properties;
*
Corresponding author.
Also corresponding author.
1
0
013-4686/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0013-4686(00)00472-2

3
750
N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
2
+
aq)
+
−
Semiconductor films derived from Group 6A (16)
Cd( +H SeO
+4H +6e “CdSe(s)+3H O
2
2
3(aq)
elements (O, S, Se, Te) have applicability in many
optoelectronic devices ranging from solar cells to xero-
graphic systems. Depending on their chemical composi-
tion and stoichiometry, these materials exhibit
interesting optoelectronic properties, including photo-
conductivity and both n- and p-type semiconductor
behavior. Specific examples of binary and ternary com-
(1)
Variants of this cathodic scheme have been developed,
including the use of non-aqueous deposition baths [10].
Another innovation begins with the initial deposition of
a chalcogen (e.g. sulfur) multilayer on a gold substrate
followed by cathodic reduction of this layer to increase
the chalcogenide ion activity at the electrode/solution
interface. Subsequent dosing of the bath with the requi-
site metal ions generates an insoluble metal chalco-
genide layer at the substrate/solution interface. This
approach has been demonstrated for iron sulfide [11],
indium sulfide [12], and cadmium selenide [13]. Further
data are presented below.
pounds include CdS, ZnO, CdTe, CdSe, CuInSe , etc.
2
1
.1. Preparation of thin films
Many electronic and optoelectronic applications ne-
cessitate the deposition of the aforementioned materials
in thin film form on targeted substrates. High tempera-
ture (e.g. evaporation) and ultra-high vacuum (e.g.
sputtering) methods have been extensively employed for
the preparation of semiconductor thin films. Indeed,
QCM has served a useful function in many of these
cases as a film thickness monitor of the gas phase
deposition process. However, low-temperature film de-
position approaches have many attractive features. For
example, problems associted with interlayer diffusion
and preferential evaporation that plague high-tempera-
ture deposition processes are absent here. Electrodepo-
sition [6,7] and chemical bath deposition (CBD) [8,9]
are the two major process candidates for low-tempera-
ture film deposition approaches. Both these processes
are low cost and readily amenable to scale-up. Relative
to electrodeposition, CBD does offer the advantage of
not requiring a conductive substrate. On the other
hand, electrodeposition, because of its electrical nature,
offers precise process control. Both processes are com-
patible with irregular substrate shapes.
Interestingly, metal oxide films can also be electrode-
posited cathodically using soluble precursor species.
Two categories of cathodic oxide deposition processes
can be recognized. The first (as in the anodic approach
above) involves a change in the oxidation state, as
exemplified by the cathodic electrosynthesis of Cu O
2
from Cu(II) precursor species [14]. The second category
is not based on a change of the metal oxidation state
but on an electrochemical reaction involving oxygen or
other additives (e.g. nitrate). The net effect is electro-
−
chemical generation of OH ions and subsequent pre-
cipitation of the oxide/hydroxide phase [15–18].
The CBD process can be illustrated by the classical
scheme for deposition of metal sulfide (MS) films [9]:
2
+
−
MLn +SC(NH ) +2OH
2 2
“
MS+2H O+nL+CN H
2
(2)
2
2
where L is a ligand for the metal ion. Sulfide com-
pounds are by far the most studied category under
CBD but an increasing number of metal selenides have
been synthesized as well [9]. Other than thiourea (see
Eq. (2) above), many chalcogenide precursors have
been used; these have been reviewed recently [8,9].
1
.2. Film deposition chemistry and electrochemistry
Electrodeposition of metal oxide/chalcogenide films
1.3. Prior studies of inorganic semiconductor film
can be done either anodically or cathodically [10]. In
the former approach, a metal anode is exposed to an
deposition by QCM
2
−
alkaline medium containing X
X=O, S, Se, Te). The main difficulty with anodic film
precursor species
Table 1 contains a listing of studies where QCM (or
EQCM) has been employed in studies of inorganic
semiconductor film deposition. The scope of this tech-
nique is enhanced when combined with other electro-
chemical probes such as cyclic voltammetry and
coulometry. Of course this feature is common to other
test systems (e.g. metals, conducting polymers) as well.
Nonetheless, chalcogenide film deposition involves elec-
trochemical reaction steps intertwined with chemical
processes that are much easier to interpret with a
combined approach. We illustrated this notion with
data on Te many years ago [19].
(
growth is that associated with mass transport limita-
tions. That is, sustained film growth necessitates the
flux of metal ions ‘outward’ to the solution boundary
2
−
and/or X
species ‘inward’ toward the metal/film
boundary. Thus, the (growing) semiconductor film pre-
sents a barrier to film growth that ultimately becomes
self-limiting.
In the cathodic approach, soluble precursor species
are co-deposited onto a suitable (inert) substrate such
as conductive glass, glassy carbon, gold, etc. This is
exemplified by the overall reaction scheme for the depo-
sition of CdSe:
More recently, EQCM has been used in conjunction
with coulometry to probe changes in film stoichiometry

N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
3751
Table 1
dt (obtained from QCM) with predictions from nucle-
ation/growth models (usually elaborated with the aid of
chronoamperometric experiments) provides new mecha-
nistic insights into the film growth process itself
Examples of the use of QCM (or EQCM) in studies of
inorganic semiconductor film deposition
Entry no.
System(s) studieda
Ref. no(s).
[9,28,33].
1
2
3
4
5
6
7
8
9
0
1
2
3
Te
Cd–Se
Cd–Te
Cu–S, Cu–In–S
Cu–Se
Fe–S
In–S
Zn–O
Cu–O
Cd–S (CBD)
Zn–S (CBD)
Cd–Se (CBD)
Se
[19]
[13,20–22]
[23]
[24,25]
[26]
[11]
[12]
[16]
[27]
[28,29]
[30]
2
. Experimental
Details of the electrochemical instrumentation and
the EQCM set-up are given elsewhere [22,26,34]. Either
a 5 or a 9 MHz gold-overlaid AT-cut quartz crystal was
used as the mass sensor and as the working electrode.
All potentials herein are quoted with respect to either
Ag/AgCl/3 M KCl (or 3 M NaCl) or SCE reference.
All chemicals were from commercial sources and were
of the highest purity available; they were used as re-
1
1
1
1
[31]
[32]
ceived. Deaerated (ultrapure N or argon purged) and
2
a
double-distilled water was used in all the experiments.
All data below pertain to the laboratory ambient tem-
perature unless otherwise noted.
Films were deposited electrochemically in all cases unless
otherwise indicated in parentheses.
Experiments on CdTe deposition were carried out at
70–80°C using a custom-made rotating EQCM setup.
The solution was made of dissolved cadmium sulfate
in the Cu–S system [24] with increasing anodization
3
+
and In
insertion dynamics in the ternary Cu–In–S
system [25]. A rotating microbalance has been designed
to combine the virtues of a rotating disk electrode
configuration with the QCM probe and used for the
study of the Cu–Se binary system [26]. The EQCM
−
3
(
0.5 M) and tellurium dioxide (10
M) at pH 1.4.
3. Results and discussion
technique has been applied to the study of Cu/Cu O
2
layered nanostructures [27]. Flow injection has been
combined with EQCM for the study of CdSe film
deposition [22]. The use of a dual electrochemical cell
with selective stripping and EQCM analyses, has facili-
tated the compositional assay of metal chalcogenide
films [20,21,23]. Finally, QCM has been combined with
AC impedance spectroscopy for the study of CBD of
CdS [28]. Comparison of the temporal derivative, dm/
3
.1. Surface modification of gold with sulfur
Oxidative deposition of sulfur on gold has been
studied by previous authors (including one of us, N.R.
de T) using in situ spectroscopy and/or voltammetry
[35–38]. Since this process is an important component
of both the anodic deposition of metal chalcogenides
and the cathodic electrosynthesis of these compounds,
we present below new data on the study of sulfur
deposition on gold by combined voltammetry-EQCM.
Fig. 1 contains a cyclic voltammogram (Fig. 1a) and
an accompanying EQCM trace (Fig. 1b) for an initially
bare polycrystalline gold surface in an 1 M NaOH/10
mM Na S electrolye. The potential scan was initiated in
2
the positive direction from ca. −1.5 V. The barely
discernible anodic voltammetric feature at : −0.90 V
is accompanied by mass gain that can be identified as
the oxidative underpotential deposition of sulfur. This
process has been shown by other authors [38] to corre-
spond to a fractional monolayer of sulfur on the gold
surface. Further negative potential excursion culminates
in a prominent anodic wave peaking at : −0.2 V with
accompanying mass gain. This corresponds to multi-
layer deposition of sulfur. There is a slight mass loss at
the tail of this peak corresponding to a weaker anodic
feature at :0 V.
Fig. 1. Combined cyclic voltammetry (Fig. 1a) and EQCM
(Fig. 1b) profiles for a polycrystalline Au surface in 1 M
NaOH/10 mM Na S. Each scan began at −1.5 V in the
positive direction. The potential scan rate was 100 mV/s.
Sulfur deposition continues on the return cycle as
manifest in the EQCM trace in Fig. 1b up to : −0.6
2

3
752
N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
well with the original rest potential of the medium
−0.6 V).
(
The sulfur deposition process is further amplified in
the data in Fig. 2. A combined potentiodynamic-poten-
tiostatic waveform was impressed on the current-mass
measurements. Starting from a potential of −0.9 V,
the gold electrode was anodically scanned up to 0 V.
The potential was then switched and arrested at −0.22
V for a period of 2 min. A similar protocol was
employed for metal sulfide film growth (see below).
3.2. Electrosynthesis of metal sulfide films on gold
One can either start with a metal layer on the native
gold surface and anodically grow a sulfide film on it
[
[
7,36] or, alternatively as we have shown more recently
11,12], begin with a sulfur layer on the gold and
cathodically grow a metal sulfide film on it. Both these
complementary approaches are illustrated by combined
voltammetry-EQCM data in this section.
Fig. 3 contains data pertaining to the classical anodic
deposition approach for CdS [39–46]. A cadmium layer
was initially deposited on the gold surface from a 10
mM CdSO /0.5 M Na SO medium as described else-
where [22]. Starting at a potential of −1.8 V, this
electrode was then scanned in the positive direction in a
Fig. 2. Voltammetry (Fig. 2a), EQCM (Fig. 2b), and potential
waveform (Fig. 2c) for the oxidative deposition of sulfur on
the Au surface. The electrolyte was 0.05 M Na S/0.1 M
2
Na SO . The combined potentiodynamic/potentiostatic experi-
2
4
4
2
4
ment began at −0.9 V, then scanned in the positive direction
up to 0 V; the scan direction was switched at this point and a
cathodic scan initiated to −0.22 V. The potential was held at
1
M NaOH/10 mM Na S medium. The corresponding
2
−
0.22 V for 2 min. The potential scan-rate was 20 mV/s.
voltammetry and EQCM profiles are contained in Fig.
a and b, respectively. The initial mass gain at poten-
3
V. Subsequently two cathodic features are accompanied
by mass loss corresponding to the electro-reduction of
the deposited sulfur layer. Soluble sulfide and poly-
sulfide species are produced by this process and can be
utilized for the deposition of metal sulfides as shown in
the next section below. The ‘minimum’ in the EQCM
trace (Fig. 1b) and the termination of the anodic wave
on the return cycle (Fig. 1a) appear to coincide very
tials between : −1.5 and −1.1 V corresponds to the
growth a few monolayers of CdS on the gold surface.
The accompanying voltammogram features are barely
discernible in Fig. 3a (at the current sensitivity shown)
although we and others [40] have observed a sharp
anodic feature at more sensitive scale settings.
A more gradual mass (Fig. 3b) increase, at potentials
positive of −1.1 up to :0.2 V, is accompanied by a
region in the corresponding voltammetry profile in
which there is very little current flow. This ca. 1.3 V
voltage span corresponds to the ‘passivation regime’ in
CdS film growth [40,41,45]. At potentials higher than
:
0.2 V, the mass uptake rate abruptly increases (Fig.
3b) accompanied by an anodic wave in Fig. 3a. On the
return cycle, the mass continues to increase up to :0.2
V followed by minor changes that culminate in a
pronounced mass loss at : −1.7 V. The sulfur in CdS
presumably is cathodically stripped leaving behind the
residual Cd layer on the gold surface at −1.8 V. This
latter process is accompanied by a cathodic feature in
the voltammogram at : −1.7 V in Fig. 3a.
The new combined voltammetry-EQCM data on the
anodic Cd–S system (presented in Fig. 3) is in good
accord with the mechanistic picture gained earlier from
voltammetric investigations [40,41]. Following the ini-
tial formation of a CdS adlayer of monomolecular
Fig. 3. Cyclic voltammetry (Fig. 3a) and EQCM (Fig. 3b)
profiles for a Cd layer pre-deposited on the Au surface (see
text). The electrolyte and scan conditions were as in Fig. 1.
dimensions, a barrier layer grows to :50 A thickness
,

N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
3753
4b) and the stripping of the cadmium oxide (formed on
the forward scan) manifests as a sharp cathodic peak
with accompanying mass loss at : −1.0 V (Fig. 4).
Indeed, examination of the EQCM profile provides a
sensitive in situ indicator of the extent to which the
metal sulfide film is contaminated by metal oxide/hy-
droxide phases. Voltammetry alone would be insensi-
tive in this regard.
As mentioned earlier, an alternate metal sulfide
growth strategy is to start with an initial sulfur layer on
the gold surface built up as shown in Fig. 2. This is
illustrated in Fig. 5. After the sulfur layer is grown on
the Au surface to a requisite thickness (Fig. 2), the
electrode is transferred (under potential control at −
3
+
0
.22 V) to an In -ion containing medium. Further
Fig. 4. As in Fig. 3 but for a 1 M NaOH electrolyte instead.
scan in the negative direction results in the reductive
deposition of indium (as the sulfide) with concomitant
mass gain as illustrated in Fig. 5b. The initial mass loss
seen at : −0.6 V can be assigned to the nascent
reduction of sulfur (as soluble sulfide species, Fig. 1).
Subsequent precipitation of these species as indium
sulfide result in the rapid mass gain seen after this
initial feature (Fig. 5b).
over a very wide potential span. This initial layer has
been shown to be compact and uniform [40,41]. Subse-
quently, the kinetics abruptly change and a thick,
porous layer grows on top of the initial film — this is
the process seen in Fig. 3 beginning at :0.2 V.
It is worth noting that none of the features in Fig. 3
correspond to parallel oxide/hydroxide phase formation
involving Cd. This can be assessed from a comparison
of the voltammetry-EQCM profiles obtained for a Cd
film in 1 M NaOH (i.e. in the absence of sulfide species)
Fig. 6 contains more voltammetry data on the In-S
system illustrating the influence of thickness of the
initial sulfur layer on subsequent assimilation. In Figs.
6a, b and c, the sulfur layers were grown potentiostati-
cally (at −0.22 V, Fig. 2) until they were approxi-
mately 15, 30, and 50 monolayers thick. The cathodic
stripping peaks are shown in each case in Figs. 6a, b
and c, respectively. When these sulfur layers are trans-
ferred (under potential control) to a HCl medium (pH
(
Fig. 4). The mass uptake pattern on the forward scan
in the two cases is clearly different (c.f. Fig. 3b and Fig.
:
2) containing 0.01 M In (SO ) and scanned in the
2 4 3
negative direction as before, the stripping pattern is
drastically altered as seen in Figs. 6a , b and c . The
1
1
1
cathodic feature at more negative potentials (seen espe-
cially in Figs. 6b and c ) presumably corresponds to
1
1
proton reduction.
Finally, referring back to Fig. 5, the subsequent mass
gain and the cathodic wave at : −0.75 V seen on the
reverse scan (after switching the scan direction at −1.0
V) have been discussed earlier [12]. These features arise
from the deposition of indium atop the In2S3 layer.
Interestingly, this deposition is catalyzed by the indium
sulfide layer; no deposition of indium is seen at these
potentials when the sulfide layer is absent on the gold
surface.
3.3. Electrodeposition of CdTe
Since the pioneering work by Kr o¨ ger et al. [47,48] on
Fig. 5. Cyclic voltammetry (Fig. 5a) and EQCM (Fig. 5b)
profiles for the cathodic deposition of In2S3 and indium on a
sulfur-modified Au surface (see Fig. 2). The solution was 0.01
M In (SO ) acidified to pH :2 with HCl. The potential
binary II–VI and III–V semiconductor compounds,
CdTe represents the most widely studied and a model
system for semiconductor electrodeposition. Elucida-
tion of the deposition mechanism continues to remain
2
4 3
scan-rate was 20 mV/s.

3
754
N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
Fig. 6. Frames (a), (b), and (c) show the cathodic stripping of sulfur layers grown respectively to 130, 260, and 460 mC thickness
on the polycrystalline Au surface (as prescribed in Fig. 2, see text). The electrolyte was 0.1 M Na SO /0.05 M Na S and the potential
2
4
2
scan-rate was 20 mV/s. The voltammograms in frames (a ), (b ), and (c ) were the same electrodes as above but the sulfur stripping
1
1
1
was done in a pH 2 0.01 M In (SO ) electrolyte. The stripping of sulfur now proceeds concomitantly with the growth of In S layers
2
4 3
2 3
(see Fig. 5).
an elusive goal in this system. The EQCM technique
has been employed before, both by us and by other
authors [19,23,49] for this purpose.
One interesting aspect of CdTe deposition is that the
overall mechanism involves basically two successive
electrochemical steps (the possibility of other parallel
reactions also exist):
forward and back potential excursions is a striking
point to study. Combining the quantity of electricity
with the mass variation it appears that during the
cathodic scan mostly tellurium is formed whereas dur-
ing the reverse anodic sweep predominantly cadmium
telluride tends to be formed. This can be associated
with changes in the nature of the substrate: during the
forward scan, the substrate is pristine gold but during
the reverse scan the substrate is composed of both
tellurium and cadmium formed below −0.66 V versus
SCE. Because of rather different work functions, cad-
mium provides an ohmic contact for the growth of
−
Te(IV)+4e “Te
(2a)
(2b)
2
+
−
Cd +Te+2e “CdTe
where the formation of CdTe takes place in Reaction
b which can be considered as an underpotential depo-
2
sition of cadmium in the presence of Te. To obtain
stoichiometric CdTe it is usual to have Reaction 2a
under diffusion control with low Te(IV) concentration,
and to increase the rate of Reaction 2b by high cad-
mium concentration and as negative a deposition po-
tential as possible. An elegant tool for investigating in
situ the competition between the two reactions is to
manipulate the hydrodynamic regime since virtually
only Reaction 2a is affected.
Fig. 7 shows combined voltammetric and gravimetric
measurements made under well-defined hydrodynamic
condition by using rotating EQCM. We clearly see on
the voltammogram characteristic electrochemical fea-
tures of the CdTe system. The deposition of CdTe/Te
films is associated with the presence of well-defined
reduction waves between −0.20 and −0.65 V (vs.
SCE). Many pieces of information can be extracted
from such curves. As an example, the difference be-
tween the amplitude of the deposition current between
Fig. 7. Combined cyclic voltammetry (solid line) and EQCM
(
dotted line) data using a rotated EQCM set-up (electrode
rotation speed=200 RPM). The potential scan-rate was 5
mV/s and the bath conditions were as follows: T=70°C,
−
3
[Te(IV)]=10
M, [Cd(II)]=0.5 M, pH=1.4.

N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
3755
References
[
[
[
[
1] G. Sauerbrey, Z. Phys. 155 (1959) 206.
2] R. Schumacher, Angew Chem. 29 (1990) 329.
3] D.A. Buttry, Electroanal. Chem. 17 (1991) 1.
4] K.K. Kanazawa, O.R. Melroy, IBM J. Res. Devel. 37
1993) 157.
(
[
5] K. Doblhofer, K. Rajeshwar, in: T.A. Skotheim, R.L.
Elsenbaumer, J.R. Reynolds (Eds.) Handbook of Con-
ducting Polymers, second ed., Marcel Dekker, New York,
1
998, p. 531 (Review).
[
6] G. Hodes, in: I. Rubinstein (Ed.), Physical Electrochem-
istry, Marcel Dekker, New York, 1995, p. 515 (Chapter
1
1).
[
[
7] K. Rajeshwar, Adv. Mater. 4 (1992) 23.
8] S. Gorer, G. Hodes, in: P.V. Kamat, D. Meisel (Eds.),
Semiconductor Nanoclusters — Physical, Chemical, and
Catalytic Aspects, Elsevier, Amsterdam, 1997, p. 297.
9] D. Lincot, M. Froment, H. Cachet, Adv. Electrochem.
Sci. Eng. 6 (1999) 165.
Fig. 8. Plots of charge versus mass for a set of films grown at
0.6 V (vs. SCE). The parameter is the electrode rotation
speed.
−
[
[
10] K. Rajeshwar, N.R. de Tacconi, in: P.V. Kamat, D.
Meisel (Eds.), Semiconductor Nanoclusters — Physical,
Chemical, and Catalytic Aspects, Elsevier, Amsterdam,
n-type CdTe, which is not the case for gold, promoting
the deposition of Te to the detriment of CdTe. This
illustrates the possibility of using EQCM to analyze the
ohmicity of the substrate/layer interface which is very
important for thin film applications.
1
997, p. 321.
[
[
11] N.R. de Tacconi, O. Medvedko, K. Rajeshwar, J. Elec-
troanal. Chem. 379 (1994) 545.
12] N.R. de Tacconi, K. Rajeshwar, J. Electroanal. Chem.
444 (1998) 7.
Finally, Fig. 8 contains mass versus charge plots
obtained at a constant deposition potential (−0.6 V vs.
SCE). These data illustrate the virtue of combining
QCM with coulometry experiments. Three sets of data
are shown at variant hydrodynamic conditions (0, 200,
and 500 RPM) corresponding to current densities of
[13] N. Myung, N.R. de Tacconi, K. Rajeshwar, Electrochem.
Commun. 1 (1999) 42.
[14] T.D. Golden, M.G. Shumsky, Y. Zhou, et al., Chem.
Mater. 8 (1996) 2499.
[
[
15] S. Peulon, D. Lincot, Adv. Mater. 8 (1996) 166.
16] S. Peulon, D. Lincot, J. Electrochem. Soc. 145 (1998) 864.
−
2
[17] Z.H. Gu, T.Z. Fahidy, J. Electrochem. Soc. 146 (1999)
56.
0.3, 1.2 and 2.15 mA cm , respectively. The decrease
1
of the slope is related to the enrichment in elemental
tellurium, relative to the result at 0 RPM where the
deposit should be close to stoichiometry (as deduced
from EDX measurements). The excess of Te, deduced
from the experimental slopes with respect to the value
for 0 RPM is 20% at 200 RPM and 40% at 500 RPM.
These conditions illustrate the case of a boundary ki-
netic regime between Reactions 2a and 2b above and,
more generally, how EQCM can be useful for gaining
compositional insights. A systematic study of these
aspects, including the influence of deposition potential
is in progress.
[
[
18] J.A. Switzer, Am. Ceram. Bull. 66 (1987) 1521.
19] E. Mori, C.K. Baker, J.R. Reynolds, et al., J. Electroanal.
Chem. 252 (1988) 441.
[
[
20] C. Wei, C.S.C. Bose, K. Rajeshwar, J. Electroanal. Chem.
3
27 (1992) 331.
21] N. Myung, C. Wei, K. Rajeshwar, J. Electroanal. Chem.
356 (1993) 281.
[22] N. Myung, J.H. Jun, H.B. Ku, Microchem. J. 62 (1999)
15.
[23] C. Wei, N. Myung, K. Rajeshwar, J. Electroanal. Chem.
3
47 (1993) 223.
[
[
24] N.R. de Tacconi, K. Rajeshwar, R.O. Lezna, J. Phys.
Chem. 100 (1996) 18234.
25] M.I. Schimmel, N.R. de Tacconi, K. Rajeshwar, J. Elec-
troanal. Chem. 453 (1998) 187.
26] A. Marlot, J. Vedel, J. Electrochem. Soc. 146 (1999) 177.
27] E.W. Bohannan, L.-Y. Huang, F.S. Miller, et al., Lang-
muir 15 (1999) 813.
Acknowledgements
[
[
D.L. and C.L. acknowledge support from the Eu-
ropean Union (Program No. JOR 3-CT-97-0150). N.R.
de T and K.R. thank the US Department of Energy
[
[
28] D. Lincot, R. Ortega-Borges, J. Electrochem. Soc. 139
(
1880) 1992.
(
Division of Basic Energy Sciences) for partial financial
29] R. Ortega-Borges, D. Lincot, J. Electrochem. Soc. 140
(1993) 3464.
support of this research. Last, but not least, we thank
Gloria Madden for assistance in the preparation of this
manuscript.
[30] B. Mokili, M. Froment, D. Lincot, J. Phys. IV 5 (1995)
26.

3
756
N. Myung et al. / Electrochimica Acta 45 (2000) 3749–3756
[
[
31] H. Cachet, M. Froment, G. Maurin, J. Electroanal. Chem.
06 (1996) 239.
32] H. Cachet, H. Essaaidi, M. Froment, et al., J. Electroanal.
Chem. 396 (1995) 75.
33] C. Wei, K. Rajeshwar, J. Electroanal. Chem. 375 (1994) 109.
34] C.S.C. Bose, K. Rajeshwar, J. Electroanal. Chem. 333
[41] L.-S.R. Yeh, P.G. Hudson, A. Damjanovic, J. Appl.
Electrochem. 12 (1982) 153.
[42] V.I. Birss, L.E. Kee, J. Electrochem. Soc. 133 (1986) 2097.
[43] S.B. Saidman, J.R. Vilche, A.J. Arvia, Electrochim. Acta
32 (1987) 1153.
4
[
[
[44] M. Krebs, M.I. Sosa, K.E. Heusler, J. Electroanal. Chem.
301 (1991) 101.
(1992) 235.
[
[
35] A.N. Buckley, I.C. Hamilton, R. Woods, J. Electroanal.
Chem. 216 (1987) 213.
36] R.O. Lezna, N.R. de Tacconi, A.J. Arvia, J. Electroanal.
Chem. 283 (1990) 319.
[45] D. Ham, K.K. Mishra, K. Rajeshwar, J. Electrochem.
Soc. 138 (1991) 100.
[46] N.R. de Tacconi, K. Rajeshwar, J. Phys. Chem. 97 (1993)
6504.
[
[
37] X. Gao, Y. Zhang, M.J. Weaver, Langmuir 8 (1992) 668.
38] L.P. Colletti, D. Teklay, J.L. Stickney, J. Electroanal.
Chem. 369 (1994) 145.
39] B. Miller, A. Heller, Nature 262 (1967) 680.
40] L.M. Peter, Electrochim. Acta 23 (1978) 165.
[47] M.P.R. Panicker, M. Knaster, F.A. Kr o¨ ger, J. Elec-
trochem. Soc. 125 (1978) 566.
[48] F.A. Kr o¨ ger, J. Electrochem. Soc. 125 (1978) 2028.
[49] J.G.N. Matias, J.F. Juliao, D.M. Soares, et al., J. Elec-
troanal. Chem. 431 (1997) 163.
[
[
.