Deposition of selenium thin layers on gold surfaces from sulphuric acid media: Studies using electrochemical quartz crystal microbalance, cyclic voltammetry and AFM

Article abstract of DOI:10.1016/j.electacta.2009.10.008

In this paper we report here new considerations about the relationship between the mass and charge variations (m/z relationship) in underpotential deposition (UPD), bulk deposition and also in the H₂Se formation reaction. Nanogravimetric experiments were able to show the adsorption of H₂SeO₃ on the AuO surface prior to the voltammetric sweep and that, after the AuO reduction, 0.40 monolayer of H₂SeO₃ remains adsorbed on the newly reduced Au surface, which was enough to gives rise to the UPD layer. The UPD results indicate that the maximum coverage with Se_ads on polycrystalline gold surface corresponds to approximately 0.40 monolayer, in good agreement with charge density results. The cyclic voltammetry experiments demonstrated that the amount of bulk Se obtained during the potential scan to approximately 2 Se monolayers, which was further confirmed by electrochemical quartz crystal microbalance (EQCM) measurements that pointed out a mass variation corresponding of 3 monolayers of Se. In addition, the Se thin films were obtained by chronoamperometric experiments, where the Au electrode was polarized at +0.10 V during different times in 1.0 M H₂SO₄ + 1.0 mM SeO₂. The topologic aspects of the electrodeposits were observed in Atomic Force Microscope (AFM) measurements. Finally, in highly negative potential polarizations, the H₂Se formation was analyzed by voltammetric and nanogravimetric measurements. These finding brings a new light on the selenium electrodeposition and point up to a proposed electrochemical model for molecule controlled surface engineering.

Full text of DOI:10.1016/j.electacta.2009.10.008

Electrochimica Acta 55 (2010) 1184–1192
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Deposition of selenium thin layers on gold surfaces from sulphuric acid media:
Studies using electrochemical quartz crystal microbalance, cyclic voltammetry
and AFM
Murilo Feitosa Cabral , Valber A. Pedrosa , Sergio Antonio Spinola Machadoa
a
b,∗
a
Instituto de Química de São Carlos, Universidade de São Paulo, C.P. 780, 13566-590 São Carlos, SP, Brazil
Department of Materials Engineering, Samuel Ginn College of Engineering, Auburn University, Auburn, AL 36849, USA
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 July 2009
Received in revised form
In this paper we report here new considerations about the relationship between the mass and charge
variations (m/z relationship) in underpotential deposition (UPD), bulk deposition and also in the H2Se
formation reaction. Nanogravimetric experiments were able to show the adsorption of H2SeO3 on the
AuO surface prior to the voltammetric sweep and that, after the AuO reduction, 0.40 monolayer of H2SeO3
remains adsorbed on the newly reduced Au surface, which was enough to gives rise to the UPD layer. The
UPD results indicate that the maximum coverage with Seads on polycrystalline gold surface corresponds
to approximately 0.40 monolayer, in good agreement with charge density results. The cyclic voltammetry
experiments demonstrated that the amount of bulk Se obtained during the potential scan to approxi-
mately 2 Se monolayers, which was further confirmed by electrochemical quartz crystal microbalance
(EQCM) measurements that pointed out a mass variation corresponding of 3 monolayers of Se. In addi-
tion, the Se thin films were obtained by chronoamperometric experiments, where the Au electrode was
polarized at +0.10 V during different times in 1.0 M H2SO4 + 1.0 mM SeO2. The topologic aspects of the
electrodeposits were observed in Atomic Force Microscope (AFM) measurements. Finally, in highly neg-
ative potential polarizations, the H2Se formation was analyzed by voltammetric and nanogravimetric
measurements. These finding brings a new light on the selenium electrodeposition and point up to a
proposed electrochemical model for molecule controlled surface engineering.
2
6 September 2009
Accepted 3 October 2009
Available online 13 October 2009
Keywords:
Selenium
EQCM
Underpotential deposition
AFM
©
2009 Elsevier Ltd. All rights reserved.
1
. Introduction
on several semiconductor substrates, which exhibit an intriguing
thickness-dependent stability, because this phenomenon occurs at
a potential more positive than the equilibrium potential [5]. Under-
potential deposition (UPD) provides a straightforward approach for
controllably and reproducibly modifying structure and composi-
tion of metal electrodes with a layer of dissimilar metal [6,7].
Moreover, the underpotential deposition has been applied in
activating noble metals [8], formation of arrays of nanoparticles
[9], deposition of chalcogenides using the electrochemical atomic
layer epitaxy [10], and a template to characterize the formation of
self-assembled layers [11,12].
Several investigators have used the quartz microbalance
(EQCM), often in combination with cyclic voltammetric, with AFM
of semiconductor films proving to be a powerful tool for conser-
vation in situ and investigation of surface roughness effects during
electrochemical metal depositions and dissolution [13–15]. Under-
standing such selenium special behavior in thin films growth is a
major aim in several electrochemical studies.
Over the past years, much research has been devoted to explor-
ing the potential of semiconductors for future electronic devices.
Therefore, development of new methods to control surface prop-
erties is an area of great interest. Recently, several inorganic
compounds were extensively tested in order to obtain endurable,
uniform and compact metal coatings for semiconductor thin films
prepared at constant room temperature and pressure conditions
[
1]. Among them, selenium deposition plays a distinctive role as
one of the most practical technique to obtain precursor films to be
doped for semiconducting layers in several applications.
For the purpose of surface modification, electrodeposition
method has been shown to reduce costs remarkably when com-
pared with most physical methods [2,3]. It was already known that
the mechanical properties of electrodeposits are affected by the
initial stages of the electrodeposition process [4]. Recently, there-
fore, researchers have used an underpotential deposition process
Selenium ad-atom deposit has been the focus of a large body of
research [16–22]. The electrochemical reduction of Se4+ in aqueous
solutions is a complex subject, which has been studied on dif-
ferent surfaces [18], including single crystals [[19] and references
∗ Corresponding author. Tel.: +1 334 844 3404; fax: +1 334 844 3400.
E-mail address: vzp0002@auburn.edu (V.A. Pedrosa).
0
013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.10.008

M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
1185
®
therein]. The electroreduction of H SeO3 is a complex process that
RQCM Maxtek Inc . The appropriate softwares were employed
2
®
depends on several factors such as (i) sensitivity to the electrodic
surface; (ii) underpotential deposition; (iii) semiconductor com-
pound formation; and (iv) coupled chemical reactions. The most
efficient electrochemical way to produce crystalline and amor-
phous films of selenium is by cathodic reduction of selenious
ions in an acid medium [19]. Some authors previously postu-
lated that the electrodeposition of Se occurs following Reaction (1)
from AUTOLAB and Maxtek Inc , respectively.
The working electrode (Au-EQCM) was a 5 MHz AT-cut quartz
crystal disc with a diameter of 12.5 mm. Both sides of the quartz
crystal were covered with thin gold films on a Ti adhesion layer,
but only one of these faces (a disc of 1.37 cm2 geometric area) was
exposed to the electrolyte. The electrochemical area was evalu-
2
ated as 4.9 cm based on measurements of the voltammetric charge
[
21]:
related to the reduction of a complete monolayer of gold oxide, fol-
lowing the Trasatti and Petrii procedure [23,24]. This procedure
gives a roughness factor (R) value of 3.6 for the Au-EQCM elec-
trode. All charge densities or surface coverage referred to in this
work is related with such electrochemical area. The reference elec-
trode was an Ag/AgCl (3 M KCl) and the auxiliary electrode was a
H SeO + 4H + 4e− → Seads + 3H O
+
(1)
2
3
2
This reaction was studied by Santos and Machado [21] using the
EQCM-Pt electrode in acid media, which showed that the anodic
responses of mass variations were divided in three well-defined
potential regions. It showed that the third region between 1.1 and
2
Pt foil with geometric area of 1 cm .
1
.5 V could be attributed to the Se + H O dissolution peak and the
2
Pt oxide formation. However, Solaliendres et al. [22] have reported
that the selenium underpotential processes involved the formation
2.3. Preparation of Se thin films
of H Se species onto Au electrode in perchloric acid solution. An
EQCM combined with voltammetric techniques was employed to
The Se thin films were deposited onto Au-EQCM by potentio-
static polarization (+0.10 V) at different times (t = 30, 60, 120, 300,
400 and 600 s) in 1.0 M H2SO4 + 1 mM SeO2 (Eeq. = +0.70 V; t = 2 s),
without magnetic stirring.
2
elucidate the H Se reduction mechanism. Some authors suggested
2
that at more negative potentials, specifically, the formation of the
4+
2−
H Se species for the direct reduction of Se to Se , occur with
2
the transference of six electrons in accordance with Reaction (2)
2.4. Atomic Force Microscopy
[
16,17]:
H SeO + 6H + 6e− → H Se + 3H O
+
The AFM experiments were carried out with an Explorer model
(2)
2
3
2
2
from Digital Instruments® model Multimade. To minimize the sur-
Up to now, the electrochemical behavior of selenium has not
face deformation and material removal, the experiments were
performed in intermittent non-contact mode, using a silicon can-
been yet fully investigated. In this way, the aim of this work is to
report electrochemical results concerning the selenium deposition
on quartz crystal gold electrodes from a sulphuric acidic solu-
tion. With that, the mechanisms the several processes concerning
selenium electrodeposition have been investigated using several
techniques, such as electrochemical quartz crystal microbalance,
chronoamperometry, and cyclic voltammetry. All these results
were integrated with the nanogravimetric results. In previous stud-
ies, the electrochemical behavior of selenium was investigated with
−1
tilever with a spring constant of 70 N m at a scan rate of 1 Hz.
All measurements were taken under air at room temperature,
employing the root mean square roughness (Rrms) to compare the
surfaces quantitatively and evaluate the topological changes on the
plates. As the roughness value depends on the observation scale,
all experiments were carried out on a scale of 1 ␮m. After sele-
nium deposition, all electrodes were rinsed with Millipore water
before AFM imaging. We ensured that the electrodes were perfectly
clean before each experiment by successive cleaving, rising with
Millipore water, and imaging in air.
emphasis in partial processes like UPD, bulk deposition or H Se for-
2
mation or exploring aspects related to the adsorption of anions and
hydrogen reduction reaction as a stage of intermediates formation
[
5,7,16–22].
3. Results and discussion
In this way, the objective of this work is to perform simulta-
neous voltammetric and nanogravimetric measurements in order
to obtain the respective mass/charge relationships. In an extensive
potential window, this study can bring more details in the elec-
trodeposition of Se on polycrystalline Au surfaces, from an acidic
electrolyte. It will be proposed an electrochemical model for such
complex electrode reaction.
3
.1. Voltammetric responses during the selenium deposition
process
The voltammetric behavior for the Au-EQCM electrode
immersed in 1.0 M H SO4 and 1.0 mM SeO2 electrolyte is shown
in Fig. 1. The electrode potential was swept between 1.60 V (initial
2
−
1
and final potential) and −0.55 V (inverse potential) at 0.10 V s
.
2
. Experimental
The dotted line showed the steady-state voltammetric profile for
Au-EQCM in the absence of SeO (Fig. 1). Based on previous reports
2
2.1. Chemicals
[22], the assignment of these voltammetric peaks (R1, R2, R3, R4)
is as follows: reduction of the Au oxide (R ), underpotential depo-
1
SeO2 (99.99% purity) and sulphuric acid (Suprapur) were
sition of Se ad-atoms (R2), deposition of bulk selenium (R3), and
formation of H2Se (R4) in a direct reduction by 6 electrons of Se4+
.
obtained from Merck and used as received. Millipore-Q puri-
fied water was used to prepare all solutions, which, previously
to the experiments, were deaerated with high-purity N2 (White
Martins SS) and the experiments were carried out at room
temperature.
In the reverse scan, the observed peaks are usually attributed to the
bulk Se oxidation (Ox1), the Se ad-atoms oxidation (Ox2), and the
Au surface, or Au–Se alloy as described by several authors, oxida-
tion (Ox3) [17–22,25–27]. All these peaks are the main subject of
the present paper and will be explored in the following section.
2.2. Equipments and apparatus
3.2. Se UPD process
The electrochemical instrumentation consisted of an Autolab
potentiostat (PGSTAT30, Ecochimie, The Netherlands) linked to a
PC-AMD K6-II microcomputer and to a quartz crystal microbalance
Although the UPD Se had already been studied, it certainly is an
extremely interesting electrochemical process and deserves addi-

1
186
M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
way, using the molar mass of Se as 78.96 g mol−1, it is possible to
−
2
find a theoretical value of 87.0 ng cm , considering that each Seads
occupies two Au sites. This value is then associated to the reduction
of Se4+ from the solution and the formation of a full monolayer of
Seads on the Au surface. The experimental value obtained from Fig. 2
−
2
was 36.0 ng cm , which is in agreement with the charge density
from the voltammetric experiments of Fig. 1 for 0.4 monolayer of
Se ad-atoms. Such coverage values for Se ad-atoms on Au surface
have already been reported by Furuya and Motoo [28]. However, in
our work, the experimental value is not obtained as a loss of mass,
but instead as a expected gain of mass. Thus, the simple scheme
4
+
−
Se + 4e → Se
does not justify the experimental behavior.
ads
Consequently, some interesting aspects in the massogram curve
remain unclear and justify a deeper analysis. In this way, the next
step is to develop an empirical model that allows us to quantita-
tively analyze the massogram showed in Fig. 2. From previously
published work, it has been established that at 1.6 V, the initial
potential of the massogram, the electrode surface is fully oxidized
Fig. 1. Cyclic voltammogram for the Au-EQCM electrode in 1.0 mM SeO2 + 1.0 M
−
1
H2SO4. The dotted line represents the response in 1.0 mol L
H2SO4. Scan
−1
rate = 0.10 V s
.
and covered with a sub-monolayer of adsorbed H SeO3 molecules,
2
which was originated from the dissolution of SeO in the acid elec-
2
trolyte [32].
Controversy can be found in the literature concerning the
adsorption of H SeO3 on gold surfaces. Some authors previously
2
concluded that such species do not seem to adsorb on Au or AuO
surfaces [33]. In order to clarify such possibility, we performed
an adsorption experiment, in which the quartz crystal microbal-
ance gold electrode was polarized at 1.60 V until stabilization of
the mass profile and then 1 mM SeO2 was added to the electrolyte
−
1.0 M H SO ). A total of 120.0 ng cm positive mass variation was
2 4
2
(
obtained up to mass profile stabilization, as presented in Fig. 3. This
mass value corresponds to 0.85 monolayers of H SeO adsorbed
2
3
on AuO surfaces (each molecule occupying two surface sites). In
this way, our experiment showed, without doubt, that H SeO3
2
does adsorb on the AuO surface, which justifies the discussions
below.
Scanning towards the negative direction, no mass change is
detected until the potential values reach approximately 1.0 V,
where AuO begins to be reduced. Such transformation should
Fig. 2. Voltammetric profile for the Se UPD process. Cyclic voltammogram (full line)
and massogram (dotted line) profiles for the Se atomic layer obtained in 1.0 mM
−1
−2
SeO2 + 1.0 M H2SO4. Scan rate = 0.10 V s
.
promote a loss of mass of approximately 35.0 ng cm (same con-
siderations as above for Seads, but using the molar mass of oxygen,
−
1
tional investigations [21,22]. We performed experiments in which
the initial potential was set to 1.60 V and swept in the negative
direction, showing that the UPD Se starts at about 0.35 V, while
the respective re-oxidation of the monolayer occurs at approxi-
mately 0.70 V. The charge density for Seads oxidation obtained from
16.0 g mol ). In the experimental result showed in Fig. 2, a loss
−
2
of mass of 75.0 ng cm
between 1.0 and 0.80 V, much larger than the expected one. This
behavior is not observed in the blank solution (with only H SO4
in the electrolyte), thus suggesting that it cannot be associated to
can be detected in the potential range
2
−
2
−
the voltammetric profile, presented in Fig. 2, was 162 ␮C cm
.
HSO4 adsorption. Clearly, the excess of mass should be attributed
This value corresponds to the charge required for the formation
of 0.4 monolayer of Se ad-atoms on the Au surface, considering
1
that a complete oxygen monolayer on gold (AuO ), where each
−
2
Oads atom occupies one Au surface atom, involves 390 ␮C cm
23,24,26–28]. In this way, taking into account that each Se ad-
[
atom occupies two Au surface atoms with the transference of 4
electrons per ad-atom, the amount of Seads on Au surface is corre-
spondent to 0.4 monolayer [29]. To confirm such results, we carried
out EQCM experiments to determine the mass variation associated
to the monolayer formation in the potential window between 0.75
and 0.35 V. The results can be seen as the massogram (highlighted
line) in Fig. 2.
The theoretical mass variation associated with the formation of
a complete Seads monolayer can be calculated assuming that, on
polycrystalline Au surface, there are available, as adsorbing sites,
−
9
2
.16 × 10 mol of Au atoms per square centimeter [30,31]. In this
Fig. 3. Mass–time gravimetric curve obtained with the Au-EQCM electrode before
and after the addition of an aliquot of SeO2 in 1.0 M H2SO4. Final concentration of
SeO2 equals to 1 mM. Electrode potential fixed at +1.60 V.
1
AuO does not correspond to the phase oxide layer. AuO indicated a full mono-
layer of oxygen ad-atoms on gold surface.

M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
1187
Scheme 1.
Scheme 3.
The theoretical mass variation shown in Scheme 2 corresponds
to the transformation of 0.4 monolayer of selenic acid adsorbed
in the Au surface (as discussed before) in Seads. The experimen-
−
2
tal value obtained (23.0 ng cm ), highlighted in the massogram of
Fig. 2, is in full agreement with the proposition of only 0.4 mono-
layer of Seads and is formed, thus confirming the voltammetric
charge measurement.
At 0.35 V, in the end of Seads process, the potential sweep is
switched to the positive direction. Up to +0.80 V was observed as
−
2
only a slightly rising in mass (12.0 ng cm ) and can probably be
associated to water (or anions) adsorption on top of the Seads layer.
At 0.80 V, the voltammogram in Fig. 2 showed the beginning of
Seads oxidation, with a well-defined voltammetric peak at 0.89 V.
This process is normally associated with a loss of mass of approx-
−
2
imately 87.0 ng cm . However, in Fig. 2 it is possible to observe,
−
2
in this potential region, a mass gain of approximately 40.0 ng cm
−
2
associated to an anodic voltammetric charge of 162 ␮C cm . This
unexpected behavior should be associated to Scheme 3.
Thus, the voltammetric oxidation peak does not promote the
redissolution of adsorbed Se except its oxidation, followed by oxy-
gen incorporation. This causes the rise in the mass by a value similar
to the value obtained through calculation.
Fig. 4. Mass–time gravimetric plots obtained with the Au-EQCM electrode before
and after the addition of SeO2 to the 1.0 M H2SO4 electrolyte. Final concentration of
SeO2 equals to 1 mM. Electrode polarized at +0.70 V.
to a sub-monolayer of H SeO3 adsorbed that leaves the electrode
surface together with the oxygen ad-atoms during the AuO reduc-
tion as follows:
2
The adsorbed SeO blocks the electrode surface and inhibits the
2
formation of AuO. In this way, in the voltammetric response, the
onset of surface oxide formation is shifted to potentials more posi-
tive than 1.20 V, while in the massogram there is a plateau, meaning
that no change in the electrode mass is occurring. After 1.20 V,
the Au surface begins to oxidize, resulting in the gain of mass of
The amount of mass left to the H SeO desorption is around
2
3
−2
4
0.0 ng cm
, much less than expected for a full coverage
(
Scheme 1). Most likely included in this balance is the re-adsorption
of a fraction of selenic acid on the Au newly reduced surface, after
some kind of rearrangement. This adsorbed selenic acid species
should provide the 0.4 monolayer of Seads, as found in the voltam-
metric experiments, which will be justified in the next paragraph.
−
2
approximately 50.0 ng cm following the scheme below:
One important aspect of this general scheme is that the cycle
closes and the electrode returns to the same mass it presented in
the beginning of the potential scan.
To complete the balance, the initial coverage with H SeO3 should
2
To perform a deeper investigation on the nature of the UPD Seads
oxidation peak at 0.90 V, we performed the experiment presented
in Fig. 5, which shows the ꢀm/ꢀq plot. This relationship is a rep-
resentation of Faraday’s law, obtained by calculating the charge
variation from the voltammogram and the mass variation from the
massogram for each potential step. The slope of each linear seg-
ment of such plot is related with the molar mass of the electroactive
species. An analysis of Fig. 5 shows a straight line obtained between
be taken as 0.7 monolayer, a little less than the maximum coverage
showed in Fig. 3 (0.85 monolayer). However, we have to consider
that the adsorption time on polarized electrode should be less than
the 20 min took to obtain the maximum coverage in Fig. 3.
Aiming to investigate the possible adsorption of H SeO3 on
2
reduced gold surfaces, we performed the quartz crystal microbal-
ance experiment and reported the results in Fig. 4. Here the Au
electrode was kept polarized at 0.70 V during 750 s for the mass
stabilization, followed by the addition of 1 mM SeO2 to the elec-
0
.80 and 0.95 V, followed by an exponential section up to 1.0 V. This
last section can be roughly approximated by a “linear” fitting. The
trolyte (again, 1.0 M H SO ). The initial rising of approximately
2
4
−2
1
50.0 ng cm can be justified by the adsorption of several disor-
deredmonolayersof H SeO onthe surface. Afterthat, the electrode
2
3
coverage begins to organize until, finally, only a sub-monolayer of
adsorbed selenic acid remains on the surface. The total mass varia-
tion, between the SeO2 addition and this stabilized film formation
−
2
–
around 50 min of immersion – is about 50.0 ng cm , or approxi-
mately 0.4 monolayer of H SeO . So, it is also possible to postulate
2
4
that the selenic acid can adsorb on reduced Au surface generating
coverages as high as 0.4 monolayers.
The next electrode process that occurs on this newly reduced
surface is the formation of the Se ad-atoms monolayer, originated
from the 0.4 monolayer of adsorbed H SeO , which begins at 0.75 V.
2
3
Here the general scheme should be:
Fig. 5. ꢀm vs ꢀq plots for Se UPD oxidation. The slope value for the linear region is
included.
Scheme 2.

1
188
M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
Scheme 4.
slope of the first section is calculated as 1.5 × 10−4 g C−1
.
nFꢀm
ꢀ
Q = nFꢀN =
(3)
(4)
M
nFꢀm
M =
= nF slope
ꢀQ
Using the Faraday equation, a value of molar mass (M) of
−1
1
4.5 g mol is obtained, considering the number of electrons (n)
equals one. This corresponds to the reaction involving oxygen
atoms adsorbed on two sites of the Au surface. This fits quite well
in the scheme shown in Eqs. (5) and (6). The last section in Fig. 5,
from 0.95 to 1.0 V, corresponds to a deep increase in mass with-
out some considerable variation in charge. This process yielded an
exponential dependence that can be attributed the adsorption of
Fig. 7. ꢀm vs ꢀq plots for selenium bulk oxidation. The slope values for the linear
region is included.
tion, we consider the mass increase for each Seads complete layer
−
2
water and anions on the Au Se(OH)2 surface as an approximation
as (0.4 × 85) ng cm , the ꢀm value found for bulk deposition cor-
responds to approximately 3.5 layers, which must be handled with
care due to the lack of epitaxy with the growing of some layers.
So, this value is satisfactory with the voltammetric one (2.3 layers
observed).
2
of the processes depicted in Scheme 4.
3.3. Se bulk process
For bulk Se dissolution, the mass profile shows a large decrease
towards more positive potential, as the amount of Se decreases,
The first cyclic voltammetric response for bulk Se deposition
is presented in the highlighted region of Fig. 6, showing that Se
bulk deposition starts immediately after the underpotential depo-
sition region, as previously reported on polycrystalline gold [22].
Furthermore, the integrated area under peak O1 (associated with
the bulk Se redissolution) was calculated to be approximately
−
2
from Fig. 6. The observed loss of mass is about 100.0 ng cm , mean-
−
2
ing that, at least 20.0 ng cm
of Seads remain in the electrode
surface, possibly due to a closest package of the first monolayer.
It must be emphasized that similar results have already been
obtained for the bulk electrochemical behavior of Se on polycrys-
talline gold [23,32–35].
In Fig. 7, the nature of this electrochemical process is further
analyzed based in the ꢀm/ꢀq plot related with the potential region
for bulk Se oxidation (0.53–0.80 V in Fig. 6). As before, the slope
of the straight line obtained is related with the molar mass of
the reagent the Faraday’s law (Eqs. (3) and (4)). Here, we found
−
2
3
74 ␮C cm . Such bulk deposition must occur on top of the pre-
viously formed UPD monolayer, whose charge was found to be
−
2
1
62 ␮C cm
in the previous section. If a few bulk layers follow
a certain epitaxial behavior, the charge previously determined
for bulk Se should correspond to approximately 2.3 layers. Of
course, in this case, the deposition/dissolution potentials shifted
to the Nernstian value, since no more Se–Au interactions were in
course.
−
4
−1
a value of 2.47 × 10 g C for the bulk Se oxidation. By multiply-
−
1
ing by the Faraday constant, i.e., 96,487 C mol and by the number
of electrons transferred in the process (4) a molar mass value of
The mass variations that occurred during the potential scanning
are also presented in Fig. 6 as the dot line. In the bulk deposition
region (between 0.35 and 0.0 V in the negative potential scanning
and 0.0–0.2 V in the positive direction, after the potential sweep
inversion at 0.0 V), a sharp increase in mass is observed, whose
−
1
9
5.0 g mol can be obtained, which is close to the theoretical value
4+
−1
for Se (79.0 g mol ). As the electrodeposition of bulk Se proceeds
through a very complex multi-step process, the ꢀm/ꢀq slope value
obtained can be considered in good agreement with the atomic
mass for Se.
−2
value is approximately 120.0 ng cm . If, as in the previous sec-
3
.4. Electrochemical formation of H Se
2
Further expanding the negative potential window of the Au/Se
electrode up to −0.50 V gives rise to a new cathodic process
described by some authors as the direct reduction of Se4+ to Se
2−
(
in a six electron process) and by others as the initial reduction of
+4
2−
Se to Se followed by Se to Se in two half-reactions [5,16,22,36].
At room temperature, H Se is a gas. Thus, it nucleates, grows
2
forming bubbles, and detaches from the electrode surface. Such
gas evolution consumes the bulk Se phase, and consequently, the
bulk Se redissolution should generate a smaller electrochemical (or
nanogravimetric) response.
In this work, we have studied the H Se formation process as
2
the last segment of the electrode potential sweep towards nega-
tive direction. Here, the H Se is generated as the reduced product
2
of few monolayers of bulk Se deposited during the scanning, i.e.,
approximately three monolayers.
Fig. 6. Voltammetric behavior of Se bulk process. Cyclic voltammogram (full line)
In the potential range related to the H Se nanogravimetric
2
and massogram (dotted line) profiles for Se in 1.0 mM SeO2 + 1.0 M H2SO4. Scan
−1
response, shown highlighted in Fig. 8, we observed a loss of mass
rate = 0.10 V s
.

M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
1189
Fig. 8. Cyclic voltammogram and massogram profiles for selenium electrochemical
−
1
processes on Au-EQCM in 1.0 mM SeO2 + 1.0 M H2SO4. Scan rate = 0.10 V s
.
of approximately 170.0 ng cm−2, which suggests that bulk Se is
leaving the electrodic surface. This mass variation value is higher
than that discussed in the previous section for bulk Se, because,
as the potential scanning is now conducted up to more negative
values, the amount of bulk Se obtained on the electrode surface is
also higher than in the previous case. Scanning between 0.40 and
Fig. 9. ꢀm vs ꢀq plots for selenium bulk reduction and H2Se formation showed in
Fig. 8. The slope values for the linear region are included.
as discussed in the item before). This lower value can also be a
consequence of the formation of the Au–Se surface alloy.
−
2
−
0.25 V allows the deposition of approximately 200.0 ng cm of
bulk Se. After −0.25 V, the reduction of such bulk Se occurs, gener-
−
3
.5. Mass variations to electrochemical processes involving 4e
−
or 6e
ating H Se and causing the Se dissolution, as in the scheme below:
2
AuSeads + 2H+ + 2e− → H Se
The m/z analysis of the electrochemical process during the elec-
(5)
(6)
2
troreduction of Se4+ in the H Se potential region is shown in Fig. 9.
2
H SeO + 6H + 6e− → H Se
+
2
3
2
In accordance with Kazacos and Miller, a study of the reduction
of selenious acid and in H SO4 solution has shown that the ini-
2
There is some controversy in literature about the extension of
the contribution from Reactions (3) or (4) for the cathodic peak in
the voltammogram shown in Fig. 8 [5,16,22,36].
tial cathodic reaction involves an overall transfer of six electrons
[
16,17]. Two molecules of the newly formed H Se may react in
2
the presence of H SeO3 from the bulk solution, yielding elemen-
2
These electroreduction processes are associated with a voltam-
tal selenium. The rate of the chemical reaction is dependent on the
concentration of H SeO3 and is sufficiently rapid at high H SeO3
−
2
metric cathodic peak at −0.35 V and a charge value of 600 ␮C cm
.
2
2
As before, considering that a full monolayer of H Se would involve a
2
concentrations.
−
2
charge density of 185 ␮C cm (two electrons transferred and two
The analysis of m/z relationship has been performed in the
potential range between 0.32 and −0.6 V as detailed in Fig. 9 (exper-
imental data obtained from the voltammogram and the massogram
presented in Fig. 8). We consider a table involving three electro-
chemical reactions (Table 1).
active sites occupied by each H Se molecule) and a loss of mass
2
−
2
somewhat higher than (0.4 × 87.5) ng cm , the values reported
above suggest a loss of approximately three monolayers of Seads,
due to its electroreduction to H Se. Finally, the hydrogen evolution
2
reaction promotes the cathodic currents observed around −0.55 V.
In the reverse scan, after the hydrogen evolution reaction,
at potentials more positive than −0.40 V, the cathodic current
observed in Fig. 8 shows diffusion controlled behavior due to the
bulk Se deposition until the electrode potential reaches approx-
imately 0.30 V. The mass variation associated to such deposition
Initially, the electrodeposition process started with the trans-
ference of four electrons, in the potential region between 0.32 and
−
0.22 V (bulk Se deposition). The theoretical m/z relationship for
−
1
−1
,
this reaction is 33.2 g mol thus, experimental data, 28.0 g mol
presents an excellent agreement. In the next potential range (−0.2
to −0.31 V), in the beginning of H Se formation, it is observed that
−2
2
was found to be approximately 60.0 ng cm , and the amount of Se
deposited should be added to the remaining Seads after the poten-
the reduction of Se0 to Se
2−
predominates with a transference
of two electrons, which seems to be confirmed by the experi-
mental and theoretical m/z values (experimental m/z value was
tial incursion towards the H Se formation region, to generate the
2
Se bulk phase that will be oxidized at 0.70 V. At this potential value,
the anodic peak observed is due to the dissolution of Se bulk and has
−
1
−1
−
35.4 g mol
and the theoretical one is −40.5 g mol ). Finally,
in the potential region between −0.31 and −0.40 V, another very
−2
a loss of mass of 55.0 ng cm associated with it. This is a mass value
lower than that expected by the amount of Seads in the bulk phase,
which can be associated to some intermetallic Au–Se compound,
stable at this potential window, as already reported [16,25]. As pre-
viously observed, the oxidation of the UPD Se, at 0.90 V, promotes
a gain of mass indicating that Se4 does not leave the electrode
surface, but instead forms the corresponding oxygenated species
good agreement between theoretical and experimental m/z rela-
−1
−1
respectively) points
tionships (−22.5 g mol
and −23.7 g mol
to a transference of six electrons, probably as a consequence of
+
Table 1
Reactions and m/z theoretical values of selenium in aqueous
solution.
(
SeO32+).
Finally, the mass variation observed in Au oxidation potential
Reactions
H2SeO3 + 4H + 4e → Se + 3H2O
m/ztheoretical
−2
+
−
0
−1
−1
−1
region was approximately 25.0 ng cm as showed in Fig. 8. This
value is very close to that calculated for the adsorption of a com-
plete monolayer of oxygen onto gold electrode (i.e., 32.0 ng cm
33.24 g mol
−40.48 g mol
−22.49 g mol
0
+
−
Se + 2H + 2e → H2Se
+
−
−
2
H2SeO3 + 6H + 6e → H2Se + 3H2O

1
190
M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
a mass increase of 250.0 ng cm 2 is obtained during the growth of
−
the Se thin film, which corresponds to approximately 8.0 monolay-
−
2
ers (considering that each Se monolayer requires 31.69 ng cm ).
This result corroborated the chronoamperometric experiments and
the epitaxial growth up to this film thickness. The film stops to grow
significantly after around 400 s of deposition. This effect has been
related to the fact that Se has low p-type conductivity and cuts off
the cathodic current [39].
The cyclic voltammogram of an Au-EQCM electrode coated with
the selenium thin film deposited, as discussed above in 1.0 M H SO
2
4
−
1
at 0.10 V s in the H Se formation region, is shown in Fig. 11A. The
2
voltammetric profile for this film was presented in previous work,
in similar experimental conditions [42]. Here, the voltammetric
profile showed marked differences in relation with that discussed
in the previous section, with SeO2 in the electrolyte. During a neg-
ative scan between 0.50 and −0.60 V, the reduction peak related
to the H Se formation is observed, as well as its equivalent anodic
2
Fig. 10. Gravimetric and potentiostatic current transient for the electrodeposition
of H2SeO3 on Au electrode at +0.10 V. The dotted line corresponds to the stationary
current for Se deposition, while the full line represents the corresponding mass
variation during the deposition time. The grid lines are included to illustrate the
theoretical mass variation values corresponding to the formation of 2, 6.5 and 8
monolayers of deposited bulk Se.
peak in the reverse scan. This couple of peaks is attributed to the
2−
reduction of Se to Se in the negative scan, as discussed above, and
the reaction with H+ generating gaseous H Se species, which dif-
2
fuses away from the electrodic surface. The oxidation of H Se to Se
2
promotes the anodic peak at approximately −0.18 V in the positive
scan.
the high overpotential applied in the electrode. At more negative
The mass decrease observed during the H Se formation, in the
2
−
1
−2
potentials, only the HER (m/z theoretical = −2.0 g mol and exper-
negative scan, shown in Fig. 11A, was −124.54 ng cm sufficient to
imental = −1.5 g mol−1) can be found.
propose the dissolution of four monolayers of Se from the surface
−
2
(
31.69 ng cm per layer, as before). In the reverse scan, during the
3
.6. Characterization of selenium thin films
potential range of −0.60 to 0.50 V, we observed a positive mass vari-
−
2
ation of +66.6 ng cm (2.1 monolayers), which corresponds then
0
In the preparation of thin films, it is important to control film
to the re-oxidation of H2Se to Se . In Fig. 11B the potential scan
properties such as homogeneity and porosity in order to obtain
high-quality coverages. Electrochemistry has been proven to be a
very helpful tool since it provides information about several pro-
cesses taking place during the film formation [37–39].
for a range more positive than 0.50 V is presented, in which two
electrochemical processes are observed at 0.90 and 1.0 V, related
to the oxidation of bulk Se and Se ad-atoms oxidation. The first
electrochemical process in 0.90 V corresponds to about 2.7 mono-
layers of Se, with a loss of mass of −88.0 ng cm , very close to the
results described in the voltammetric analysis of Se thin film on the
Au-EQCM surface. The oxidation process at 1.0 V is related to the
oxidation of Seads. Apparently, this process is not affected by the
evolution of H2Se gaseous species.
The Atomic Force Microscopy images for Se film prepared under
potentiostatic conditions are collected in Fig. 12. The electrode-
positions were conducted at different times, as explained above.
The gold electrodes used in this experiment were obtained from
commercial CD-R gold surfaces. This is a simple, reproducible, and
inexpensive way for the ex situ analysis with reproducible areas,
and as the CD-R surfaces are obtained by chemical vapor depo-
sition, the surface has morphologic aspects very similar to the
−
2
In this work, Se thin films were obtained by chronoampero-
metric experiments where the Au-EQCM electrode was polarized
at +0.10 V during 600 s in an electrolyte composed of 1.0 M
H SO + 1.0 mM SeO . The correspondent deposition charge was
2
4
2
evaluated from the area under the curve, as depicted in Fig. 10.
After several experiments, the average deposition charge was
−
2
1
.38 mC cm . Considering that: (a) the theoretical charge for
−
2
one Se monolayer on gold surface is 162 ␮C cm (0.41 × 390, as
detailed before), (b) each Se atom occupies two atoms on the
gold surface, and (c) the growing of the Se monolayers occurs in
an epitaxial mode, the calculated charge should be equivalent to
approximately 8.5 monolayers [40,41]. It can also be observed in the
massogram correspondent to the chronoamperogram (Fig. 10), that
Fig. 11. (A) Cyclic voltammogram of the Au-film EQCM coated with Se thin film in 1.0 M H2SO4. (B) Massogram profiles for Se thin. Scan rate = 0.10 V s−1
.

M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
1191
Fig. 12. Ex situ AFM images of CD-R gold surface, including the growth of a Au–Se thin film, in 1.0 mM SeO2 + 1.0 M H2SO4 at different times of (A) 30 s, (B) 60 s, (C) 120 s, (D)
3
00 s, (E) 400 s and (F) 600 s. Electrodeposition potential = 0.10 V. Scan size = 5 ␮m × 5 ␮m.
Table 2
4. Conclusions
Values of roughness (Rrms) of CD-R gold surface and
selenium thin films.
The cyclic voltammetry in simultaneous measurement with the
electrochemical quartz crystal microbalance allowed an extensive
discussion of every stage of Se electrodeposition on polycrystalline
Au surfaces.
Au-EQCM/Se coated
Rrms
Au-EQCM
3.10
Se coated (t = 30 s)
Se coated (t = 60 s)
Se coated (t = 120 s)
Se coated (t = 300 s)
Se coated (t = 400 s)
Se coated (t = 600 s)
3.54
4.62
6.71
7.20
8.50
13.75
The analysis of charge (from the voltammograms) and mass
(nanogravimetric measurements) variations have indicated a quite
complex electrochemical behavior of H SeO4 on gold polycrys-
2
talline electrode from the first monolayer up to the deposition of
thin films.
In the first stage, at positive potentials, the nanogravimetric
experiments showed, beyond doubts that H SeO3 adsorbs on the
2
EQCM electrodes [43]. AFM studies were performed to observe the
topologies of the electrodeposited catalyst films. The films were
deposited from different times. The Se films topology presents
regular pyramidal structures (three-dimensional) and is similar to
characteristics observed in the literature. The principal feature of
this topology is a composition of medium islands of densely pack
grains, indicating homogeneous deposition. Such morphological
aspects are entirely distinct from the epitaxy observed in the pre-
vious items. However, it has to be considered that the amount of
electrodeposited bulk Se here is much larger than before (about 50
monolayers or more), thus justifying the lost of epitaxy and the for-
mation of the observed morphology. The Rrms values for different
deposition times are presented in Table 2 and were estimated using
AuO surfaces, generating a complete layer on the electrode sur-
face. Here, we have shown for the first time, the mass variation
obtained from the generation of a full monolayer of AuO(H SeO ) in
2
3
a simultaneous chronopotentiometric/nanogravimetric measure-
ment. This is a very important point in this work, since its existence
was an object of controversy with previously published papers.
From such precursor layers, the underpotential deposition occurs
followed by the bulk deposition.
Another interesting conclusion is that, in UPD conditions,
Seads does not leave the electrode surface after electrooxidation
2−
reactions, but the reduction/oxidation occurs from/to (SeO3 )ads
instead. This means that in the reduction step, where an increase
in mass value should be expected due to the formation of Seads
monolayer, we instead observed a loss of mass while in the oxida-
tion step; an increase of mass has been evidentiated and not the
expected loss of mass by Seads redissolution.
®
the Nanoscope Digital Software by the average of five different
sites for the Se films samples. So, this estimative suggests the “true”
roughness factor of the films and can be useful in characterization
of electrochemical area in such thin films of Se.

1
192
M.F. Cabral et al. / Electrochimica Acta 55 (2010) 1184–1192
During the bulk Se deposition/redissolution, the electrochemi-
[11] M. Wanunu, A. Vaskevich, I. Rubinstein, J. Am. Chem. Soc. 126 (2004) 5569.
[
[
[
12] K. Shimazu, T. Kawaguchi, T. Isomura, J. Am. Chem. Soc. 124 (2002) 652.
13] X.H. Zhang, Phys. Chem. Chem. Phys. 10 (2008) 6842.
14] A. Bund, O. Schneiderb, V. Dehnkec, Phys. Chem. Chem. Phys. 4 (2002) 3552.
cal response was the expected one, and in both voltammetric and
chronoamperometric depositions, we were able to show that an
epitaxial (or almost) behavior occurred until the film grew to up to
layers. What is important to obtain is the homogeneous coverage
of the surface.
[15] J.O’M. Bockris, S. Fletcher, R.J. Gale, S.U.M. Khan, D.M. Kolb, D.J. Mazur, K. Uosaki,
N.L. Weinberg, Annu. Rep. Prog. Chem., Sect. C: Phys. Chem. 92 (1995) 23.
8
[
[
16] M.Z. Kazacos, B. Miller, J. Electrochem. Soc. 127 (1980) 869.
17] M.Z. Kazacos, B. Miller, J. Electrochem. Soc. 127 (1980) 2378.
Finally, the H Se formation was studied under different condi-
[18] I. Nandhakumar, J.M. Elliott, G.S. Attard, Chem. Mater. 13 (2001) 3840.
[19] B.M. Huang, T.E. Lister, J.L. Stickney, Surf. Sci. 392 (1997) 27.
2
tions, and we showed that, depending on the applied potential, the
reaction proceeded with the transference of four or six electrons.
This was also an important point that was possible to be clarified
by the nanogravimetric/voltammetric association.
[
[
20] C. Wei, N. Myung, K. Rajeshwar, J. Electroanal. Chem. 375 (1994) 109.
21] M.C. Santos, S.A.S. Machado, J. Electroanal. Chem. 567 (2004) 203.
[22] M.O. Solaliendres, A. Manzoli, G.R. Salazar-Banda, K.I.B. Equiluz, S.T. Tanimoto,
S.A.S. Machado, J. Solid State Electrochem. 12 (2008) 679.
[
[
23] S. Trasatti, O.A. Petrii, J. Electroanal. Chem. 327 (1992) 352.
24] M. Tian, G. Pell, B.E. Conway, Electrochim. Acta 48 (2003) 2675.
Acknowledgements
[25] R.W. Andrews, D.C. Johnson, Anal. Chem. 47 (1975) 294.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
26] T.E. Lister, J.L. Stickney, J. Phys. Chem. 100 (1996) 19568.
27] B.M. Huang, T.E. Lister, J.L. Stickiney, Surf. Sci. 392 (1997) 27.
28] N. Furuya, S. Motoo, J. Electroanal. Chem. 102 (1979) 155.
The authors thank FAPESP (proc. 04/09906-3), CNPq and CAPES
for financial assistance during the execution of this work.
29] M. Alanyalioglu, U. Demir, C. Shannon, J. Electroanal. Chem. 561 (2004) 21.
30] N. van Nga c´ , O. Vittori, G. Quarin, J. Elecroanal. Chem. 167 (1984) 227.
31] M. Volaire, O. Vittori, M. Porthault, Anal. Chim. Acta 71 (1974) 185.
32] A. Manzoli, S.A.S. Machado, Thin Solid Films 515 (2007) 6860.
33] A. B a˘ nic a˘ , A. Cule t¸ u, F.G. B a˘ nic a˘ , J. Electroanal. Chem. 599 (2007) 100.
34] Z. Fijaek, A. Lozak, K. Sarna, J. Electroanal. Chem. 10 (1998) 846.
35] M. Kemell, H. Saloniemi, M. Ritala, M. Leskelä, Electrochim. Acta 45 (2000) 3737.
36] R.W. Andrews, D.C. Johnson, Anal. Chem. 48 (1976) 1056.
References
[
[
[
[
[
1] D. Lincot, Thin Solid Films 487 (2005) 40.
2] I.M. Dharmadasa, J. Haigh, J. Electrochem. Soc. 153 (2006) G47.
3] T.S. Olson, P. Atanassov, D.A. Brevnov, J. Phys. Chem. B 109 (2005) 1243.
4] M.O. Kostin, O.H. Sheikina, Mater. Sci. 38 (2002) 898.
5] R. Vaidyanathan, S.M. Cox, U. Happek, D. Banga, M.K. Mathe, J.L. Stickney, Lang-
muir 22 (2006) 10590.
37] M.F. Cabral, Ph.D. Thesis, University of São Paulo, 2008.
38] F. Sittner, W. Ensinger, Thin Solid Films 515 (2007) 4559.
39] E.A. Streltsov, S.K. Poznyak, N.P. Osipovich, J. Electroanal. Chem. 518 (2002)
[
[
6] R. Vidu, N. Hirai, S. Hara, Phys. Chem. Chem. Phys. 3 (2001) 3320.
7] M.J. Esplandiu, M.A. Schneeweiss, D.M. Kolb, Phys. Chem. Chem. Phys. 1 (1999)
103.
[40] N. Furuya, S. Motoo, Mater. Chem. Phys. 22 (1989) 309.
[41] B.W. Gregory, J.L. Stckiney, J. Electroanal. Chem. 300 (1991) 543.
[42] M.F. Cabral, H.B. Suffredini, V.A. Pedrosa, S.T. Tanimoto, S.A.S. Machado, Appl.
Surf. Sci. 254 (2008) 5612.
4
847.
[
[
8] L. Xiao, L. Zhuang, Y. Liu, J. Lu, H.D. Abru n˜ a, J. Am. Chem. Soc. 131 (2009) 602.
9] M.D. Lay, J.L. Stickney, J. Am. Chem. Soc. 125 (2009) 1352.
10] L.P. Colletti, B.H. Flowers Jr., J.L. Stickney, J. Electrochem. Soc. 145 (1998) 1442.
[
[
43] D. Lowinsohn, E.M. Ritcher, L. Angnes, M. Bertotti, Electroanalysis 18 (2006) 89.