Studies of stoichiometry of electrochemically grown CdSe deposits

Article abstract of DOI:10.1016/j.electacta.2010.08.024

The proper deposition bath composition for electrochemical synthesis of the CdSe deposit in the hexagonal structure of the right elemental stoichiometry, and photoreacting as an n-type semiconductor which can be used as a stable photoanode is investigated. The deposits were prepared by a cyclic potentiodynamic technique and the concentration of Cd²⁺ and SeO ₃^2- in the deposition baths varied from 10^-4 M to 0.1 M, and from 10^-5 M to 10^-3 M, respectively. The electrochemical, the X-ray diffraction (EDS and XRD), and the photoactivity studies of a number of deposits have shown that application of the solution composition following Cd:Se = 5:1 results in deposition of the stoichiometric CdSe. The detected ratio of reagents is explained on the base of reaction mechanism and necessary excess of cadmium ions preventing CdSe deposit dissolution. The procedure of CdSe electrosynthesis was developed to yield of a direct semiconductor in the hexagonal structure. The necessity for cadmium cations excess is explained on the basis of the mixed electrochemical/chemical deposition mechanism.

Full text of DOI:10.1016/j.electacta.2010.08.024

Electrochimica Acta 55 (2010) 8908–8915
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Studies of stoichiometry of electrochemically grown CdSe deposits
∗
Krzysztof Bie n´ kowski, Marcin Strawski, Bartosz Maranowski, Marek Szklarczyk
Laboratory of Electrochemistry, Department of Chemistry, Warsaw University, ul. Pasteura 1, 02-093 Warsaw, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2010
Received in revised form 5 August 2010
Accepted 5 August 2010
Available online 14 August 2010
The proper deposition bath composition for electrochemical synthesis of the CdSe deposit in the hexago-
nal structure of the right elemental stoichiometry, and photoreacting as an n-type semiconductor which
can be used as a stable photoanode is investigated. The deposits were prepared by a cyclic potentiody-
2+
2−
−4
namic technique and the concentration of Cd and SeO3 in the deposition baths varied from 10 M
−5
−3
to 0.1 M, and from 10 M to 10 M, respectively. The electrochemical, the X-ray diffraction (EDS and
XRD), and the photoactivity studies of a number of deposits have shown that application of the solution
composition following Cd:Se = 5:1 results in deposition of the stoichiometric CdSe. The detected ratio of
reagents is explained on the base of reaction mechanism and necessary excess of cadmium ions prevent-
ing CdSe deposit dissolution. The procedure of CdSe electrosynthesis was developed to yield of a direct
semiconductor in the hexagonal structure. The necessity for cadmium cations excess is explained on the
basis of the mixed electrochemical/chemical deposition mechanism.
Keywords:
Cadmium selenide
Deposition
Stoichiometric ratio
Cyclic voltammetry
EDS
XRD
© 2010 Published by Elsevier Ltd.
1
. Introduction
Photoelectrochemistry started with Becquerel’s studies on the
[19,20]. It drove the quest for the ways of preparation of nanostruc-
tured semiconductive materials, which exhibit new electronic and
optical properties. But most attention was drawn to the synthesis of
II–VI semiconductive nanocrystals and nanofilms, and among them
cadmium selenide (CdSe) still remains one of the most thoroughly
studied nanocrystal systems.
The CdSe band gap for bulk material is 1.7 eV, a threshold wave-
length 729 nm [21], and depending on nanostructure radius, it
might be tuned toward higher energy values (blue shift) up to
2.8 eV, a threshold wavelength 420 nm, for 1.2 nm nanocrystal [22].
Additionally, CdSe is a direct type of semiconductor and its electron
affinity is in the range of 3.8–4.7 eV depending on nanostructure
radius [23]. CdSe, because of its photoelectrochemical and photo-
physical properties is well known as a compound semiconductor
suitable for solar energy conversion, optoelectronic devices, and
for the preparation of antivirial drugs. It may also have a positive
impact on the environment, as it was shown that it might be used
photoeffects of solar illumination on metal electrodes in 1839 [1].
Much later it was understood that the most pronounced photo-
effects were to be expected for semiconductive materials, while
Brattain and Garrett observed large photocurrents during expos-
ing a semiconductor electrode to the action of light of energy
exceeding the electrode material band gap [2]. In the next years,
photoelectrochemistry advanced rapidly [3–7]. The possibility of
practical application of semiconductor electrodes was discussed
by Williams as early as in 1960 [8]. A spark that ignited rapid
research in electrochemistry of semiconductors was the observa-
tion of Fujishima and Honda [9,10]. These authors discovered a
water photosplitting process at the irradiated TiO2 electrode. The
expectation for inexpensive production of hydrogen fuel resulted in
an enormous increase in research effort in photoelectrochemistry
field, which was additionally emphasised by the world oil crises
to sequester CO [24].
2
(
e.g. Refs. [11–17]). With time, because of some difficulties (e.g.
Methods for CdSe preparation can be divided into two major
groups: dry methods carried out in the gas phase (vacuum evapora-
tion and chemical vapor deposition) and wet methods (liquid phase
synthesis and electrochemical deposition). While the gas phase
techniques offer a high controllability of the growth of the struc-
tures and the feasibility to obtain a pure material, they, without
exceptions, require an expensive high vacuum and high temper-
ature equipment, and additionally, the gaseous waste material is
another serious problem of these techniques.
low efficiency of photogalvanic cells) the interest in electrochem-
istry of semiconductors dropped significantly. A new attempt to
develop these fields (electrochemistry of semiconductors and pho-
toelectrochemistry) started with Brus’s [18] demonstration that
a quantum confinement of photocreated electron–hole pairs led
to size dependent optical properties of semiconductor materials
The wet methods of CdSe nanostructures production are rela-
tively inexpensive and simple to be carried out. Among a variety
of the wet techniques, the electrochemical ones are the only
∗ Corresponding author. Tel.: +48 22 822 0211; fax: +48 22 822 5996.
E-mail address: szklarcz@chem.uw.edu.pl (M. Szklarczyk).
0
013-4686/$ – see front matter © 2010 Published by Elsevier Ltd.
doi:10.1016/j.electacta.2010.08.024

K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
8909
techniques which can be used for a deposition of films on large and
irregular surfaces. In comparison with other wet methods during
electrodeposition, a strict control of an oxidation state of selenium
is not required, the deposition occurs closer to an equilibrium
state, and because the process is electrical in nature, it can be
controlled precisely.
On the other hand, the electrochemical methods require a
precise definition of experimental conditions such as potentials,
reactant concentration, supporting electrolyte, deposition time,
and the material of the working electrode is of great importance as
well. The discussion on the theoretical thermodynamic conditions
to be met to achieve cathodic deposition of stable binary semi-
conductor II–VI compounds have been developed by Kröger [25].
He distinguished two classes of deposition taking into account the
difference in the deposition potential, the differences in the rate
constants and exchange currents of the individual components. On
this basis he tried to predict the stability of the deposits in depen-
dence of the ions concentration in the bulk of the solution, and has
found it to be proportional. Finally, he formulated the relations for
the rest potentials in dependence on a discharge rate constant and
ionic concentration. The deposition of CdSe he ascribed to class I,
which means that this deposition is determined by a less noble
selenium.
In previous years the CdSe was electrodeposited from the
solutions of pH = 0 to pH = 14 [26–31]. The use of the reagent con-
centrations and their ratios varied widely, the Cd:Se ratio varied
from 200:1 to 1:200 [26,29,31,32]. The potentials studied ranged
from −1.5 V to 2.0 V [26,30]. The substrates used were Au, Pt, Ag,
Cu, amalgamated Au [26,33], Ti [34], ITO [35] and HOPG [31]. A
wide range of studied experimental conditions was examined in
order to determine the best way of electrosynthesis of the material
with the proper elemental stoichiometric ratio and of the suitable
dimensions. Recently an excellent review on electrodeposition of
semiconductive materials has been published by Lincot [36]. In this
paper the two steps synthesis of semiconductive films is discussed,
as well as the method of obtaining the precursor layer and after
deposition annealing treatment. The influence of the deposition
conditions on the free energy of the deposit formation is shown
as an essential factor for the quality of the deposits.
Fig. 1. The CV dependencies of the Pt electrode placed in 0.5 M H2SO4 solution (thick
−
1
line) and 0.1 M Na2SO4 solution, (thin line). Sweep rate 50 mV s
.
ues were recalculated to a normal hydrogen electrode scale (NHE)
through this paper. Both counter and reference electrodes were
placed in the compartments separated by a glass frit.
Electrochemical activity of the working electrode surface and
the purity of the solutions were checked by a cyclic voltammetry
(CV), at first in an acidic solution (Fig. 1, thick line), and then in the
solution of pH = 3 (Fig. 1, thin line). Obtaining of well defined CV
dependencies for the Pt electrode attesting a measure of solution
high quality was a desirable condition to start deposition exper-
iments. CdSe deposition was carried out by a CV technique. The
deposition time was determined by a number of CV cycles reg-
−1
istered at 50 mV s . A synthesized CdSe deposit was examined
for its photoactivity by registering a CV dependence in the basic
2+
2−
electrolyte solution free of Cd and SeO3 ions. For the study of
stoichiometry of the deposits the electrode was taken out of the cell,
carefully washed out with Millipore-Q water, then dried and placed
in a chamber of a secondary electron microscope (SEM) equipped
with EDS device. The confirmation of the CdSe compound formation
and detection of its crystallographic structure was done by an X-ray
diffraction technique. In order to keep the electrode material con-
stant, the XRD experiments were carried out for the CdSe deposits
electrosynthesized on platinum substrate, which introduced a high
background due to much higher absorption of X-ray energy by plat-
inum than by graphite, which is usually employed as a substrate in
such studies.
The aim of the present studies was to investigate the influence
of deposition solution composition on the elemental stoichiometry
and photoreactivity of the CdSe surface structures. The deposits
were prepared by a potentiodynamic technique, the concentra-
2
+
tion of Cd varied from 0.1 M to 0.2 M, and the concentration of
2
−
−5 −3
SeO3
varied from 10 M to 10 M. The elemental stoichiome-
During the experiment, solutions were deoxygenated by bub-
bling Ar through an electrochemical cell and all experiments were
try, compound formation and crystal structures were analyzed by
X-ray energy dispersive spectroscopy (EDS) and X-ray diffraction
spectroscopy (XRD), respectively. The photoelectrochemical activ-
ity of the CdSe structures was determined and the magnitude and
direction of photocurrents were correlated with a structure and
elemental stoichiometry.
◦
carried out at an ambient temperature of 22 ± 2 C.
2.2. Solutions
All solutions were prepared from Millipore-Q water and reagent
grade 3CdSO ·8H O, SeO , Na SO , and 98% H SO . The support-
4
2
2
2
4
2
4
ing electrolyte solution, pH = 3, was prepared from 0.5 M Na SO
solution, and pH was adjusted with 0.5 M sulfuric acid. The deposi-
tion baths (B) of five different Cd:Se ratios, 200:1 (BA), 5:1 (BB), 2:1
2
4
2
. Experimental
2.1. Electrodes and experimental procedures
(
BC), 1:1 (BD), and 1:5 (BE) were prepared by mixing CdSO solu-
4
2+
tion with SeO solution. The Cd concentration range varied from
1
1
2
A three-electrode and a three-compartment electrochemical
−
−
4
3
2−
−5
0
0
M to 0.1 M, while SeO3 concentration varied from 10 M to
M. The composition of the deposition baths is given in Table 1.
cell with a quartz window were used. Working electrodes were
made of a polycrystalline platinum disk. The working area was
circular in shape, ca. 9 mm in diameter. The surface of the elec-
trode was polished with alumina, down to 0.05 ␮m, and then it
was cleaned with Millipore-Q water in an ultrasonic bath.
2.3. Equipment
Counter electrodes were made of Pt/Pt gauze or wire, and an
A three-electrode cell with a quartz window was used in all
experiments. The current–voltage behavior was monitored with a
programmed potentiostat (EG&G 263A, U.S.A.).
electrode potential was measured against an Hg|HgSO reference
4
electrode placed in saturated Na SO solution. The potential val-
2
4

8
910
K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
Table 1
Chemical composition of deposition baths.
No.
Solution composition
Cd:Se ratio in the
bulk of solution
−
3
2+
−5
−3
2−
2−
BA1
BA2
2 × 10 M Cd + 10 M SeO3
200:1
−1
2+
2 × 10 M Cd + 10 M SeO3
−
4
2+
−4
2−
2−
BB1
BB2
5 × 10 M Cd + 10 M SeO3
−3 2+ −3
5:1
5 × 10 M Cd + 10 M SeO3
−
4
2+
−4
−3
2−
2−
BC2
BC2
2 × 10 M Cd + 10 M SeO3
2:1
−3
2+
2 × 10 M Cd + 10 M SeO3
−
4
2+
−4
−3
2−
2−
BD1
BD2
1 × 10 M Cd + 10 M SeO3
1:1
−3
2+
1 × 10 M Cd + 10 M SeO3
−
4
2+
−4
−3
2−
2−
BE1
BE2
10 M Cd + 5 × 10 M SeO3
1:5
−3
2+
10 M Cd + 5 × 10 M SeO3
The source of light was a 1 kW xenon lamp placed in a lamp
housing (LSH 521, LOT ORIEL, Germany). A water filter was placed
between the cell and the source of light to filter out the long-
wavelength part of radiation in order to avoid excessive heating
of the solution.
SEM imaging was done with a LEO 435 VP microscope (Ger-
many). The device used a low sample current to give the highest
possible resolution and a detailed structure imaging. Under such
conditions, the effect of charging up was diminished. The chamber
−6
pressure was 10 Torr and pictures were registered for 7 keV or
5 keV electron beam energy to obtain the best quality images. The
1
samples prior to being placed in vacuum, were washed out with
Millipore-Q water and then dried.
Elemental analysis measurements were carried out with multi-
channel EDS device (Röntec, model M1, Germany). The pressure in
−6
the chamber was 10 Torr. EDS spectra were measured at 15 keV
electron beam energy. For each sample four places of the diame-
ter ca. 1 mm were studied and in the paper the average of these
−
3
2
Fig. 2. The CV dependencies of the Pt electrode placed in: (A) 5 × 10
M
−
3
CdSO4 + 10 M SeO2 + 0.1 M Na2SO4 solution, pH = 3. First and steady state runs are
denoted by thin and thick lines, respectively. Sweep rate 50 mV s ; (B) thick solid
line was registered for 5 × 10 M CdSO4 + 0.1 M Na2SO4 solution, pH = 3; thin solid
measurements is quoted.
−1
The XRD analysis was carried out with an X-ray diffractometer
−3
Bruker D8 Discover). The radiation source was Cu K␣, ꢀ = 1.5406 A˚ .
line was registered for 10 M SeO2 + 0.1 M Na2SO4 solution, pH = 3.
−
3
(
Measurements were done in a sample free orientation mode, and
the total acquisition time was 15,000 s.
the first cycle the Pt electrode surface is still free of any deposits and
this potential range is rather too anodic for selenium ions reduc-
tion or CdSe depositon. The following peak, C6, can be ascribed to
two processes, i.e. cadmium bulk deposition and a CdSe formation,
reactions 1 and 17, Table 2. The shift of this peak with cycling can
be explained by the decreasing energy of CdSe formation on the
surface covered with some Cd/CdSe even if the film is not continu-
ous. Such a deposit might be considered as a nucleation center for
further deposition. The current increase above −0.88 V we ascribe
to the onset of hydrogen evolution reaction. On the anodic side,
in the first run, the peak at ca. −0.42 V, A1, is due mainly to the
Cd oxidation process (reaction 2, Table 2) and possibly to the CdSe
dissolution (reaction 18, Table 2). Two well pronounced maxima
at 0.87 V, A3, and 1.23 V, A4, can be attributed to the oxidation of
selenium to selenite (reaction 12 Table 2), and to the oxidation of
CdSe to selenite, respectively (reaction 19, Table 2).
Going cathodic, the peaks at 0.55 V, C1, and 0.35 V, C2, are
formed. Viewing the potential values they do not correspond to any
reactions in Table 2. To identify these processes, it is necessary to
compare the CV dependencies presented in Fig. 2A with those pre-
sented for the supporting electrolyte solution in Fig. 1, solid line,
and a CV dependence for this solution but registered over wider
potential range (Fig. 3). On both figures a cathodic peak at ca. 0.55 V
is observed. This peak is due to the desorption of adsorbed oxy-
gen layer from the Pt electrode surface. An additional argument for
identification of C1 peak, Fig. 2A as an oxygen desorption process
is a current magnitude of this maxima. In all cases (Fig. 1, solid line,
3
. Results and discussion
.1. Electrochemistry of Cd2+ + SeO3 solution
2−
3
The CV dependence for Pt electrode placed in BB2 solution reg-
istered for the wide potential range, −1.0 to 1.5 V, is presented in
Fig. 2A. Two CV runs are presented, the initial run followed by the
steady state run. Various peaks and plateaus can be distinguished.
To assign these features it is necessary to take into account the
deposition and dissolution of CdSe itself and additionally the possi-
ble redox processes of Cd2 and SeO3 ions. Furthermore, because
the general goal is to produce either nano/micro structures or thin
films, it is always possible that some parts of the working electrode
material free of deposit is exposured to the solution, and its elec-
troactivity can also be observed. The most important reactions are
listed in Table 2. The strong dependence of the reaction potential
on electrode material and solution composition can easily be noted.
Such a situation can easily result in misleading reaction identifica-
tion on the basis of the shape of CV dependence, and correct process
classification requires cautious approach.
The dependencies presented in Fig. 2A were started to be reg-
istered in the cathodic direction from ca. 0.1 V. After five cycles a
steady-state curve was registered (Fig. 2A, thick line). During the
first cycle (Fig. 2A, thin line) C4–C7 features were distinguished. The
C4 and C5 maxima are likely to represent cadmium under potential
deposition (UPD), reaction 3, Table 2, because at the beginning of
+
2−

K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
8911
Table 2
The list of reactions and observed potentials at electrodes placed in solutions containing Cd or/and SeO3 ions.
2
+
2−
No
1
Reaction
Potential (V)
Conditions
Refs
[27]
Cd2+ + 2e− → Cd
Onset at −0.88
CdSe single crystal electrode in 0.1 M
Na2SO3, pH = 7, SCE
Onset at −0.75
Ti electrode with CdSe deposited film,
[29]
[30]
[29]
0.25 M CdSO4, pH = 2.5, SCE
−
0.75
Ag electrode covered with Se, 1 mM
CdSO4, pH = 8.5, Ag/AgCl
Ti electrode covered with CdSe
deposited film, 0.25 M CdSO4, pH = 2.5,
SCE
2
Cd → Cd2+ + 2e−
−0.48
−
0.76
HOPG, 0.08 M CdSO4, pH = 3, SCE
[31]
[30]
−
+
−
3
4
Cd2+ + e → Cd + e → Cd (UPD process)
−0.41 to −0.69
Ag electrode covered with Se, 1 mM
CdSO4, pH = 8.5, Ag/AgCl
SeO32− + 8H+ + 6e− → H Se + 3H O
−0.38
HOPG, 2 × 10 M SeO2, pH = 3, SCE
Au electrode, 0.5 mM Na2SeO3 + 0.1 M
CdSO4 + 0.25 M Na2SO4, pH = 0, SCE
Au electrode, 0.1 M
−
3
[31]
[26]
2
2
−
0.5
SeO32 + 6H+ + 6e → Se2− + 3H2O
−
−
5
−1.05
[37]
(
NH4)2SeO3 + 2.15 M (NH4)2SO3 + 4.5 M
NH3, EDTA, pH = 9.4, SCE
HOPG electrode, 0.8 mM
6
7
8
H2SeO3 + Cd2+ + 4H+ + 6e → CdSe + 3H2O
−
Onset at −0.6
0.29
[31]
[38]
[29]
H2SeO3 + 0.08 M CdSO4, pH = 3, SCE
Pt electrode, 0.5 mM SeO2 + 0.1 M
NaClO4, SCE
Ti electrode covered with CdSe
deposited film, 0.25 mM
HSeO3 + 6H+ + 6e → HSe− + 3H2O
−
−
SeO32 + 6H+ + 4e → Se + 3H2O
−
−
−0.45
H2SeO3 + 0.25 M CdSO4, pH = 2.5, SCE
Ag electrode, 0.5 mM Se (IV), pH = 8.5
Se4+ + 4e → Se
−
Onset at −0.6
−0.92
[30]
[32]
9
1
0
Se + 2e− → Se2−
Ti electrode, 0.01 M H2SeO3 + 1 M
K2SO4, pH = 2.4, NHE
−
1.0
Ti electrode covered with CdSe
deposited film, 0.25 mM
[29]
H2SeO3 + 0.25 M Na2SO4, pH = 2.5, SCE
CdSe single crystal electrode, 0.1 M
Na2SO3 solution, pH = 7, SCE
Ti electrode, 0.01 M H2SeO3 + 1 M
K2SO4, pH = 2.4, NHE
HOPG electrode, 1 mM H2SeO3 + 0.12 M
CdSO4, pH = 2.7, SCE
Au electrode,0.1 M (NH4)2SeO3 + 2.15
M (NH4)2SO3 + 4.5 M NH3, EDTA
pH = 9.4, SCE
Onset at −1.1
−0.51
[27]
[32]
[31]
[37]
−
1
1
1
1
2
3
Se + H+ + 2e → HSe−
Se + 3H2O → H2SeO3 + 4H + 4e−
+
0.8
Se2 + 3H2O → SeO32− + 6H+ + 6e−
−
−0.65
+
1
4
5
H2Se → Se + 2H + 2e−
HSe + PtO + H+ → PtSe + H2O
−0.74
Au electrode, pH = 0, SCE
[26]
[39]
−
1
0.2
Pt electrode, 0.24 mM H2SeO3 + 0.1 M
H2SO4, SCE
Au electrode, 0.1 M
2−
1
6
7
Cd2+ + Se → CdSe
Onset at −0.65
Onset at −0.72
[37]
[29]
(
NH4)2SeO3 + 2.15 M (NH4)2SO3 + 4.5 M
NH3, EDTA, pH = 9.4, SCE
Ti electrode covered with CdSe
deposited film, 0.25 mM
H2SeO3 + 0.25 M CdSO4 + 0.25 M
Na2SO4, pH = 2.5, SCE
Cd2+ + SeO3 + 6H+ + 6e → CdSe + 3H2O
2−
−
1
1
1
8
9
CdSe → Cd2+ + Se2−
−0.25
Au electrode, 0.1 M
[37]
[29]
(
NH4)2SeO3 + 2.15 M (NH4)2SO3 + 4.5 M
NH3, EDTA pH = 9.4, SCE
Ti electrode covered with CdSe
deposited film, 0.25 mM
H2SeO3 + 0.25 M CdSO4 + 0.25 M
Na2SO4, pH = 2.5, SCE
HOPG electrode, 1 mM H2SeO3 + 0.12 M
CdSO4, pH = 2.7, SCE
−
0.50
CdSe + 3H2O → Cd + H2SeO3 + Cd2+ + 4H+ + 6e
2+
−
1.3
[31]
Fig. 2A and Fig. 3) its magnitude is almost the same, 0.5 mA. A simi-
lar explanation was proposed by Simk u¯ nait e˙ et al. [40]. The second
this maximum could be a formation of the surface selenide, PtSe in
agreement with reactions 7 and 15 [38,39].
ˇ
cathodic peak, 0.35 V (C2, Fig. 2A) can be recognized on the basis of
similar analysis as a peak due to the hydrogen adsorption process,
which is represented in Fig. 1, and Fig. 3 by a cathodic current in
the potential range 0.0–0.25 V. The weak point of such an explana-
tion is that the potential of this peak is too anodic and the potential
difference between peaks C1 and C2, Fig. 2A, is only 0.22 V while in
Fig. 3 it is ca. 0.4 V. An alternative to hydrogen adsorption could be
either reaction 4 or 9 (Table 2), for which the potentials reported
There are few differences between the peak spectrum observed
in the first CV cycle and in the subsequent ones (Fig. 2A). The peak
at 0.05 V, C3, followed by a plateau up to −0.4 V can be attributed
to reaction 8, Table 2, i.e. to the selenium deposition reaction. Some
authors identify this potential range as the sum of 8, 10, 16 reac-
tions [41]. Another important difference is the anodic shift of the
C6 peak. It can be explained by a decrease in the energy of the
2+
Cd reduction and CdSe formation on the existing Cd/CdSe film in
comparison to the Pt electrode surface. A similar explanation can
[
30,31] are fairly close. The third explanation for the appearance of

8
912
K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
disappeared. The potential plateau observed above 0.76 V might
be ascribed to the platinum surface oxide formation. The potential
peak at 1.13 V observed for SeO32 solution, Fig. 2B, thin solid line,
cannot be ascribed to the oxidation of CdSe as it was identified as
peak A4 in Fig. 2A. This maximum is proposed here to be identified
−
with the oxidation of Se deposit to H SeO process, i.e. reaction 12,
2
3
Table 2. It means that this process is shifted anodically by ca. 0.3 V
in comparison with its potential in the ion mixture solution (peak
A3, Fig. 2A). Apparently the energy of oxidation of Se co-deposited
with CdSe is lower than that of pure Se deposit and can indicate the
occurrence of catalytic properties of CdSe towards Se oxidation.
The main differences in the cathodic branches of the CV depen-
dencies registered either in the mixture of ions (Fig. 2A thin solid
line) or in SeO32 solution (Fig. 2B, thin solid line) are the absence
−
of the C3 potential peak in SeO32 solution, and the occurrence of
−
two new peaks at −0.26 V and −0.43 V in this solution. The absence
of the C3 potential peak in SeO32 solution indicates that the iden-
tification of this process proposed above as the sum of reactions
8, 10 and 16, Table 2 is false and that this peak can be ascribed to
reaction 16 alone. The new peaks at −0.26 V and −0.43 V (Fig. 2B,
thin solid line) can be ascribed to the formation of Se deposit (reac-
−
Fig. 3. The CV dependency of the Pt electrode placed in 0.1 M Na2SO4 solution,
−1
pH = 3. Sweep rate 50 mV s
.
−
be applied to the anodic shifts of A3 and A4 peaks. An interest-
ing feature is the formation of the plateau, A2, on the anodic side
of the A1 peak. We ascribe this phenomenon to the splitting cad-
mium (A1) oxidation from the dissolution of the CdSe film (A2). It
has to be noted that the electrode potentials of the reactions pro-
posed above differ in some cases from those listed in Table 2. We
explain these discrepancies as the differences in the electrochemi-
cal activity of the electrode material surface. Platinum is known for
being one of the most electroactive electrode materials, especially
towards adsorption of different species. These exclusive properties
of platinum might be the explanation for the reaction potential dif-
ferences between potentials observed in this work and those listed
in Table 2, especially A3, A4, C5, C6 that show the lower energy
reaction at Pt electrode.
tion 8, Table 2) and then its reduction to HSe anion (reaction 11,
Table 2).
3.2. Influence of solution composition on CdSe stoichiometry and
its photoactivity
On the basis of the studies of Cd2+ + SeO32− solution electro-
chemistry the potential range 0.4 V to −0.9 V was chosen for the
studies of the CdSe deposit preparation. The CV curves are pre-
sented in Fig. 4A–E. The first observation is that for the baths BA
(Cd:Se ratio 200:1), BD (Cd:Se ratio 5:1), BC (Cd:Se ratio 2:1) the
maxima for CdSe and Cd deposition are well defined, while for the
bath BD (Cd:Se ratio 1:1) it is not the case, and the oxidation peak for
CdSe/Cd is not observed, which indicates that the deposit is either
extremely thin, it is in an island form, or it is not formed at all. For
Two cathodic, −0.2 V, −0.7 V, and two anodic maxima, −0.65 V,
−3
−
0.15 V are detected (Fig. 3) in a cathodic potential range, both
BD2 baths, i.e. for a higher concentration, 10 M, it starts to mani-
fest itself slightly. This comparison confirms the idea that the Cd:Se
ratio in the bulk solution might be much more important for the
deposit formation and its stoichiometry than the magnitude of the
concentration itself.
2+
2−
are not detected in the solution containing Cd and SeO3 ions
Fig. 2A). The proposed explanation for this observation has to take
(
into account the process of the reduction of the supporting elec-
trolyte, because there is nothing else in this solution. We propose
that following hydrogen adsorption the evolution starts at −0.05 V,
but then the current decreases because reduction of sulfates and
adsorption of the product of this reaction take place at such a nega-
tive potential. The adsorbed product blocks the hydrogen evolution
process. The further cathodic change of the electrode potential
results in a hydrogen evolution below −0.7 V on the new electrode
surface. The anodic peak at −0.65 V is probably due to a hydrogen
desorption, and the peak observed at −0.15 V is due to the oxidation
of adsorption layer.
The examples of SEM image of CdSe deposit and EDS spec-
trum (insert) are presented in Fig. 5. The SEM picture shows that
electrosynthesized deposit is in an dendrite island form. The EDS
spectrum was obtained by focusing an X-ray beam on the CdSe
structures (white species in Fig. 5).
In order to check if these structures are the CdSe in the chemical
compound form rather than the elemental mixture, a XRD tech-
nique was employed and a typical result is presented in Fig. 6. The
corresponding line spectra for Pt, CdSe in cubic (zinc blende) form
and CdSe in hexagonal (wurzite) form provided by the JCPDS (Joint
Committee on Powder Diffraction Standards) are presented in this
figure. As expected, the CdSe signal is strongly overlaid by the Pt
signals. Independently of this difficulty, clear signals for the CdSe
can be distinguished in the presented diffraction pattern, in the
For comparison the CV dependencies for Pt electrode placed in
2
+
2−
either Cd (thick solid line) or SeO3 (thin solid line) solutions
for the same concentration as that in the mixture (Fig. 2A), are
shown in Fig. 2B. Note that the CV dependence for the mixture
of both ions (dotted line) is not an over imposition of the indi-
vidual dependencies. Some differences in the peak potentials and
◦
◦
◦
◦
◦
angular region between 23 and 30 , at 42 , 49.8 , and 71.8 . The
current magnitude are easily seen. The current due to Cd2 reduc-
tion and Cd deposit oxidation strongly decreases in the mixture
of the ions (cf. Fig. 2B thick solid line and Fig. 2A thin solid line),
while the currents in the potential range where the Se and CdSe
deposit oxidation take place are much larger than those registered
in the mixture of the ions (cf. Fig. 2B thin solid line and Fig. 2A thin
solid line). As expected the anodic peak A1, Fig. 2A, due to the cad-
+
complex maximum between 23 and 30 indicates the presence of
a wurzite structure, but the single maximum at 49.8 indicates a
◦
◦
◦
zinc blende structure. For the hexagonal structure a triplet maxi-
mum (2 0 0, 1 1 2, 2 0 1 peaks) is expected [42] around this angle,
but since the 1 1 2 peak is much more pronounced in most cases
than the two sides peaks, the presence of the single peak can be
considered as the signal from the hexagonal structure, too [43,44].
mium oxidation is much larger in the pure solution of Cd2 cations.
+
The peak at 42 can be tied up with either a cubic or a hexago-
◦
◦
The potential peak A3 (Fig. 2A), which was identified as reaction
nal structure, but the peak at 71.8 is ascribed only to a hexagonal
1
2 (Table 2) in the absence of Cd2+ cations in the bulk of solution,
structure. It has to be stressed that the wurzite form of CdSe was

K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
8913
Fig. 4. The CV dependencies of the Pt electrode placed in: (A) BA1 and BA2 solutions denoted by thin and thick lines, respectively; (B) BB1 and BB2 solutions denoted by thin
and thick lines, respectively; (C) BC1 and BC2 solutions denoted by thin and thick lines, respectively; (D) BD1 and BD2 solutions denoted by thin and thick lines, respectively;
−
1
and (E) BE1 and BE2 solutions denoted by thin and thick lines, respectively. pH = 3. Sweep rate 50 mV s
.
obtained with no use of the thermal treatment often necessary for
transformation of the cubic into hexagonal form [42]. Summarizing
the XRD data, one may state that the hexagonal form of CdSe was
deposited, but the inclusions of a cubic form are probably present,
as well.
ratio, because for the lower concentration the deposit was not mas-
sive enough to obtain a signal intensity that could be analyzed.
From the data presented in Table 3 it is clear that with the decreas-
ing Cd:Se ratio in the bulk of solution the elemental Cd:Se ratio
decreases, and below the ratio of 5:1 the Se amount in deposit dra-
matically increases. The stoichiometric deposits were obtained for
the high Cd:Se ratios in the bulk of the solution. Similar results
The EDS analysis results are given in Table 3. The EDS data were
obtained only for the more concentrated solution at a given Cd:Se
Table 3
Elemental composition of deposits obtained by EDS technique.
Bath
BA2
BB2
BC2
BD2
BE2
Solution composition
Cd:Se ratio in the bulk of solution
Cd:Se ratio in deposit
10 M Cd2+ + 10 M SeO3
2−
−
1
−3
200:1
5:1
1:13
1:1
5 × 10− M Cd2+ + 10 M SeO3
3
−3
2−
2−
2 × 10− M Cd2+ + 10 M SeO3
3
−3
2:1
1:1.9
1:3
10 M Cd2+ + 10 M SeO3
2−
−
3
−3
1:1
10 M Cd2+ + 5 × 10−3 M SeO3
−3
2−
1:5
0:1

8
914
K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
Fig. 5. The images of CdSe deposits taken by SEM technique. The insert is an
EDS spectrum registered exclusively for the structure observed. The deposit was
obtained from BB2 solution. The electrode was taken out of the solution after five
cycles between −0.9 V and 0.35 V. The white bar at the bottom represents 10 ␮m.
were obtained for Ti electrode [29,38], HOPG electrode [31] and Ni
electrode [38].
To confirm the results obtained by EDS technique, which is a sta-
tistical method yielding the results strongly dependent on the place
of the surface chosen for analysis, the photoactivity of the produced
deposits was studied. The results are presented in Fig. 7. For the
deposit obtained from the BB solution, for which the stoichiomet-
ric elemental ratio was detected as 1:1 (Table 3), the photocurrent
observed is n-type (Fig. 7A) as expected for CdSe material. The sec-
Fig. 7. The V–iph dependencies for CdSe deposit obtained from (A) BB1 solution, (B)
BC2 solution and (C) BE2 solution. (A and C) Photo and dark currents are denoted by
thin and thick lines, respectively. (B) The cathodic runs are shown. The dark current
is denoted by dotted line, the first and the last photocurrent runs are denoted by
dashed and solid lines respectively. Arrows point the direction of the change of the
−
1
photocurrent with cycling. pH = 3. Sweep rate 50 mV s
.
ond deposit tested was the one obtained from BC solution for which
a small excess of selenium was observed (Table 3). In this case
the decreasing n-type and increasing p-type photo response were
observed (Fig. 7B), proving the presence of some Se for which the
p-type response is expected. The time dependence of the photoef-
fect (Fig. 7B) indicates that the deposit obtained from the solution
of the Cd:Se ratio equal to 2:1 is unstable, and decomposes to cad-
mium ion and selenium. For deposit obtained from BE solution a
stable p-type photoeffect was detected (Fig. 7C) as it is expected
Fig. 6. XRD spectra. The uppermost spectrum was registered for CdSe deposit
electrosynthesized on Pt electrode from BB2 solution. The line spectra are JCPDS
reference. (a) CdSe spectrum in hexagonal (wurzite) form (spectrum number 75-
5
680), (b) CdSe in cubic (zinc blende) form (spectrum number 65-2891), and (c) Pt
surface (spectrum number 65-2868).

K. Bie n´ kowski et al. / Electrochimica Acta 55 (2010) 8908–8915
8915
for a pure selenium, and it is in agreement with the data presented
in Table 3.
very grateful to Dr Mikołaj Donten for his help in assistance during
EDS measurements.
2+
The determined requirement for the excess of Cd concentra-
2
−
tion over SeO3
CdSe is in need of an explanation. The first and the most obvious
concentration for deposition of stoichiometric
References
2+
[1] E. Becquerel, C.R. Acad. Sci. 9 (1839) 561.
explanation would be the lower transport rate of Cd ions in com-
parison with the SeO
[
[
2] W.M. Brattain, C.G.B. Garrett, Bell Syst. Tech. J. 34 (1955) 129.
3] J.F. Dewald, in: B. Hannay (Ed.), Semiconductors, Reinhold, New York, 1957, p.
725.
2−
transport rate. The diffusion coefficient for
3
2+
−6
2
−1
Cd is reported to be in the range of 5.6–8.0 × 10 cm s [32,45],
2−
−6
2
−1
[4] M. Green, in: J.O’M. Bockris (Ed.), Modern Aspects of Electrochemistry, vol. 2,
Butterworths, London, 1959, p. 343.
while for SeO
it is ca. 5.6 × 10 cm s [32]. It is obvious that
3
such a small difference in the ionic transport cannot be the reason
for the ca. 5:1 excess of cadmium ions over selenite ions.
The possible explanation might be based on the reaction mech-
anism of the CdSe depositon. In the literature the proposition of
different deposition reaction mechanisms might be found (for some
review please check Ref. [46]). These mechanisms involve different
number of cadmium cations.
[
5] H. Gerischer, in: P. Delahay (Ed.), Advances in Electrochemistry and Electro-
chemical Engineering, vol. 1, Interscience, New York, 1961, p. 139.
[6] V.A. Myamlin, Y.V. Pleskov, Electrochemistry of Semiconductors, Nauka,
Moscow, 1965.
[
7] J. McHardy, F. Ludwig (Eds.), Electrochemistry of Semiconductors and Electron-
ics; Processes and Devices, Noyes Publication, Park Ridge, 1992.
[8] R. Williams, J. Chem. Phys. 32 (1960) 1505.
9] A. Fujishima, K. Honda, S. Kikuchi, J. Chem. Soc. Jpn. 72 (1969) 108.
[
[
[
10] A. Fujishima, K. Honda, Nature 37 (1972) 238.
11] Y. Nakato, S. Tonomura, H. Tsubomura, Ber. Bunsenges. Phys. Chem. 80 (1976)
1002.
The most often quoted mechanism [46] involves equal number
of cadmium and selenium based ions;
[
12] K. Ohashi, J. McCann, J.O’M. Bockris, Energy Res. Abstr. 1 (1977) 259.
2
−
+ 6H+ + 6e− → H Se + 3H O
H SeO
2
[13] W. Kautek, J. Gobrecht, H. Gerischer, Ber. Bunsenges. Phys. Chem. 80 (1980)
034.
14] A. Heller, R.G. Vadimsky, Phys. Rev. Lett. 46 (1981) 1153.
[15] M.A. Butler, D.S. Ginley, 163rd Electrochem. Soc. Meet., San Francisco, 1983
ext. abst. 773).
[16] M. Szklarczyk, J.O’M. Bockris, Appl. Phys. Lett. 42 (1983) 1035.
3
2
2
1
[
Cd2 + H Se → CdSe + 2H+
+
2
(
A reaction requiring the highest number of cadmium cations
was proposed by Tomkiewicz et al. [47] and it involves an initial
reduction of cadmium ions followed by chemical oxidation of Cd
by selenites;
[
17] E. Smotkin, A.J. Bard, A. Campion, M.A. Fox, T. Mallouk, S.E. Webber, J.M. White,
J. Phys. Chem. 90 (1986) 4604.
[
18] L.E. Brus, J. Chem. Phys. 80 (1984) 4403.
[19] L.E. Brus, J. Chem. Phys. 79 (1983) 5566.
20] M.L. Steigerwald, L.E. Brus, Acc. Chem. Res. 23 (1990) 183.
[21] K. Rajeshwar, N.R. de Tacconi, C.R. Chenthamarakshan, Chem. Mater. 13 (2001)
765.
[
+
3
Cd2 + 6e− → 3Cd
2
[
22] S.J. Rosenthal, J. McBride, S.J. Pennycook, L.C. Feldman, Surf. Sci. Rep. 62 (2007)
Cd + SeO32 + 6H+ → CdSe + 3H O + 2Cd
−
2+
3
2
111.
[
23] N.C. Greenham, X. Peng, A.P. Alivisatos, Phys. Rev. B 54 (1996) 17628.
The three cadmium cations requirement for one selenite anion
[24] L.G. Wang, S.J. Pennycook, S.T. Pantelides, Phys. Rev. Lett. 89 (2002) 075506.
[25] F.A. Kröger, J. Electrochem. Soc. 125 (1978) 2028.
[26] M.S. Kazacos, B.J. Miller, Electrochem. Soc. 127 (1980) 869.
[
[
in the above cited reactions is close to the concentration ratio
detected in this work. Furthermore, some excess of Cd2+ cations
in the solution is required to prevent CdSe dissolution [37]. The
proposed deposition mechanism confirms Kröger’s classification
of the CdSe deposition as class I of the deposition processes of
semiconductor film [25].
27] R. Tenne, G. Hodes, Ber. Bunsenges. Phys. Chem. 89 (1985) 74.
28] S.J. Lade, M.D. Uplane, C.D. Lokhande, Mater. Chem. Phys. 68 (2001) 36.
[29] C.M. Shen, X.G. Zhang, H.L. Li, Appl. Surf. Sci. 240 (2005) 34.
[
[
[
[
[
[
30] F. Loglio, M. Innocenti, F. D’Acapito, R. Felici, G. Pezzatini, E. Salvetti, M.L. Foresti,
J. Electroanal. Chem. 575 (2005) 161.
31] Q. Li, M.A. Brown, J.C. Hemminger, R.M. Penner, Chem. Mater. 18 (2006)
3432.
32] A. Lukinsas, I. Savickaja, V. Jasulaitien e˙ , J. Juodkazyt e˙ , T. Juodiene, P. Kalinauskas,
4
. Conclusion
Chemija 13 (2002) 89.
33] L.P. Colleti, B.H. Flowers, J.L. Stickney, J. Electrochem. Soc. 145 (1998)
1442.
34] M. Bouroushian, Z. Loizos, N. Spyrellis, G. Maurin, Appl. Surf. Sci. 115 (1997)
The bath composition is not the only parameter affecting the
stoichiometry of electro-deposited films, but it is a fundamental
one, which is pointed out in this work.
103.
35] T. Edamura, J. Muto, Thin Solid Films 226 (1993) 135.
It has been shown here that the ratio Cd2+:SeO3 = 5:1 is the
most satisfactory to obtain the deposit of the right elemental
stoichiometry (Cd:Se = 1:1) and photoreacting as an n-type semi-
conductor. The detected ratio of reagents can be explained on the
base of reaction mechanism involving three cadmium ions and nec-
essary excess of cadmium ions preventing CdSe deposit dissolution.
The presented procedure of CdSe electrosynthesis leads to the
deposition of direct semiconductor in the hexagonal structure. The
necessity for cadmium cations excess results from the proposed
mixed electrochemical/chemical deposition mechanism.
2−
[36] D. Lincot, Thin Solid Films 487 (2005) 40.
[
[
[
37] St. Kutzmutz, G. Láng, K.E. Heusler, Electrochim. Acta 47 (2001) 955.
38] R. Modolo, M. Traore, O. Vittori, Electrochim. Acta 31 (1986) 859.
39] P. Carbonnelle, L. Lamberts, Electrochim. Acta 340 (1992) 53.
[40] D. Sˇ imk u¯ nait e˙ , E. Iva sˇ kevi cˇ , V. Jasulaitien e˙ , V. Steponavi cˇ ius, Chemija 13 (2002)
219.
[
[
41] A.M. Kressin, V.V. Doan, J.D. Klein, M. Sailor, J. Chem. Mater. 3 (1991) 1015.
42] R.J. Bandaranayake, G.W. Wen, J.Y. Lin, H.X. Jiang, C.M. Sorensen, Appl. Phys.
Lett. 67 (1995) 831.
[43] B.O. Dabbousi, J. Rodriguez-Viejo, F.V. Mikulec, J.R. Heine, H. Mattousii, R. Ober,
K.F. Jensen, M.G. Bawendi, J. Phys. Chem. B 101 (1997) 9463.
44] D.J. Pe n˜ a, J.K.N. Mbindyo, A.J. Carado, T.E. Mallouk, C.D. Keating, B. Razawi, T.S.
Mayer, J. Phys. Chem. B 106 (2002) 7458.
[
[
[
45] S. Kariuki, H.D. Dewald, Electroanalysis 8 (1996) 307.
46] R.K. Pandey, Handbook of Semiconductor Electrodeposition, first ed., CRC Press,
Acknowledgement
1996.
[
47] M.T. Tomkiewicz, I. Ling, W.S. Parsons, J. Electrochem. Soc. 129 (1982)
2016.
We are grateful to the Ministry of Education and Science for the
support of this paper through grant program 3T08A 04128. We are