Synthesis, characterization and photoelectrochemical properties of poly(3,4-dioctyloxythiophene)-CdS hybrid electrodes

Article abstract of DOI:10.1016/j.electacta.2007.11.009

CdS-poly(3,4-dioctyloxythiophene) (CdS-PDOT) hybrid electrode has been prepared by electrosynthesis of PDOT on Au substrate followed by electrodeposition of Cd and its chemical transformation into CdS. The polymer and semiconductor obtained by this method form hemispherical structures dispersed on the substrate. The synthesized composites were characterized by UV-vis absorption spectra and energy dispersive X-ray spectra (EDS). The AFM images of the electrodes covered with different amounts of each component were correlated with photoactivity of the hybrid electrodes. Photoresponses of Au/PDOT-CdS electrodes under illumination in aqueous solution of Na₂S were also compared with those of CdS without polymer. Enhancement of the photocurrent achieved for some polymer-to-semiconductor ratio is discussed in terms of the hybrid electrode morphology and hole-mediating properties of PDOT. The power conversion efficiency of the device based on CdS-PDOT hybrid electrode was determined from photocurrent-potential behavior of two electrode system, Au/CdS-PDOT/0.1 M Na₂S/Pt with a variable resistance in series in the external circuit.

Full text of DOI:10.1016/j.electacta.2007.11.009

Available online at www.sciencedirect.com
Electrochimica Acta 53 (2008) 2984–2993
Synthesis, characterization and photoelectrochemical properties of
poly(3,4-dioctyloxythiophene)–CdS hybrid electrodes
∗
´
Monika Refczynska, Jozef Mieczkowski, Magdalena Skompska
Department of Chemistry, Warsaw University, ul. Pasteura 1, 02-093 Warsaw, Poland
Received 10 August 2007; received in revised form 2 November 2007; accepted 2 November 2007
Available online 13 November 2007
Abstract
CdS–poly(3,4-dioctyloxythiophene) (CdS–PDOT) hybrid electrode has been prepared by electrosynthesis of PDOT on Au substrate followed
by electrodeposition of Cd and its chemical transformation into CdS. The polymer and semiconductor obtained by this method form hemispherical
structures dispersed on the substrate. The synthesized composites were characterized by UV–vis absorption spectra and energy dispersive X-ray
spectra (EDS). The AFM images of the electrodes covered with different amounts of each component were correlated with photoactivity of the
hybrid electrodes. Photoresponses of Au/PDOT–CdS electrodes under illumination in aqueous solution of Na2S were also compared with those
of CdS without polymer. Enhancement of the photocurrent achieved for some polymer-to-semiconductor ratio is discussed in terms of the hybrid
electrode morphology and hole-mediating properties of PDOT. The power conversion efficiency of the device based on CdS–PDOT hybrid electrode
was determined from photocurrent–potential behavior of two electrode system, Au/CdS–PDOT/0.1 M Na2S/Pt with a variable resistance in series
in the external circuit.
© 2007 Elsevier Ltd. All rights reserved.
Keywords: CdS; Poly(3,4-dioctyloxythiophene); Polymer–semiconductor hybrid electrode; AFM; Photovoltaic properties
1. Introduction
polymer [2,5–7]. The latter method is advantageous because of
the increase of p–n junction interface and in effect the increase of
photovoltaic efficiency. However, the semiconductor nanocrys-
tals are not easily dispersed in conjugated polymer. A high
surface energy of nanocrystals leads to their aggregation. The
most efficient method of avoiding this effect is stabilization of
nanocrystals by surrounding surfactant. According to the lit-
erature, the hybrid device consisting of a polymer layer and
monodispersed CdS capped with thioglycerol, spin coated on
the polymer film possesses a dual properties of photocurrent
generation and electroluminescence [8]. However, the surfactant
usuallytendsto isolatethesemiconductorfromconductingpoly-
mer phase leading to enhanced electron–hole recombination and
decrease of quantum efficiency of the light energy conversion.
Another way of fabrication of semiconductor nanoclus-
ter–polymer composites is co-deposition of the two components
[9]. The semiconductor nanoparticles may be also covalently
linked to the monolayer assembled on Au electrode [10–12].
Preparation of the polymer–semiconductor blends is rel-
atively easy but control of amount of two composites is
problematical. In the present work we demonstrate a simple
method of preparation of poly(3,4-dioctyloxythiophene)–CdS
Conducting polymer–semiconductor hybrids represent a
novel class of materials for low-cost photovoltaic devices.
Combining of p-type conducting polymers and n-type semicon-
ductors is advantageous for separation of the charges generated
under illumination due to a high electron affinity of inorganic
semiconductor and relatively low ionization potential of the
polymer [1,2]. This allows for transport of the charge carriers in
the separate materials with a low probability of recombination.
The polymer–semiconductor composites are usually pre-
pared in the form of bilayers of the type ITO (or Au)/conducting
polymer/semiconductor/metal. The ITO or Au substrates may
be covered with the conducting polymer by electropolymer-
ization or spin-casting a solution of the polymer soluble in
chloroform. Semiconductor can be synthesized as a film by
chemical bath deposition [3], self-assembling [4] or deposited
in the form of dispersion of nanoparticles in soluble conducting
∗
Corresponding author. Tel.: +48 22 822 02 11; fax: +48 22 822 59 96.
E-mail address: mskomps@chem.uw.edu.pl (M. Skompska).
0013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2007.11.009

M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
2985
composite for photovoltaic cells. Polymer is obtained by
electrodeposition, whereas CdS is formed by electrochemi-
cal/chemical synthesis. Electrochemical methods offer a wide
range of control of the polymer and CdS amounts. In this work
we optimize the conditions to obtain the hybrid electrode of
enhanced photoactivity in comparison with that of pure CdS.
light intensity was measured with IL1700 Research Radiometer,
(International Light, USA).
The efficiency of conversion of light energy to electric energy
(power efficiency) was measured in a two electrode system,
Au/CdS–PDOT/0.1 M Na2S/Pt with a variable resistance in
series in the external circuit. The solution was deoxygenated
with argon prior the experiments.
2. Experimental
3. Results and discussion
All electrochemical measurements reported in this paper
were performed by means of Autolab (EcoChemie, The Nether-
lands) in a three electrode cell with a platinum gauze counter
electrode, Ag/AgCl, Cl− (std., aq.) reference electrode and Au
working electrode. The Au substrate was polished to a mirror
finish with alumina slurry and washed in an ultrasonic bath.
Procedure of synthesis and characterisation 3,4-dioctyloxy-
thiophene (DOT) have been described elsewhere [13]. All other
chemicals, LiClO4 (Aldrich), Cd(ClO4)2 hydrate (Aldrich),
Na2S·9H2O (Aldrich) were used as received. Electrosynthe-
sis of the polymer was performed from the solution of 10 mM
DOT in acetonitrile (AN) (Aldrich, HPLC grade) containing
0.1 M LiClO4 supporting electrolyte. Amount of the polymer
(in mol cm−2) deposited on the electrode was estimated from
the equation Γ = Qredox/zF, where Qredox is the redox charge of
the polymer during cycling between 0.3 and 0.9 V in the solution
of 0.1 M LiClO4/AN, z is a number of electrons involved in the
redox reaction of one monomer unit (z = 0.3) and F is a Faraday
constant. The estimations were not based on the polymerization
charge because of some charge loses due to formation of soluble
oligomers [13].
3.1. Electrochemical behavior of Au/PDOT electrodes in
aqueous solutions of 0.1 M LiClO4 and in 0.1 M Na2S
Poly(3,4-dioctyloxythiophene) (PDOT) was deposited on Au
electrode from acetonitrile solution of 10 mM monomer + 0.1 M
LiClO4 by cycling within the potential range from 0 to 1.34 V
at the scan rate 100 mV s−1. Exemplary electropolymerization
curves are presented in Fig. 1.
As visible in the inset in Fig. 1, the PDOT is electroactive
not only in acetonitrile (as most of polythiophenes) but also in
aqueous solution, owing to the presence of oxygen in the side
groups [13]. This feature is promising in respect of application
of this class of the polymers in the polymer/CdS composites.
Since the photodissolution of CdS is quenched in aqueous
solution of Na2S [14], the stability of PDOT in this medium is
also required. To check it and establish the most suitable poten-
tial range for further electrochemical studies, the Au/PDOT
electrode was studied in deoxygenated solutions of 0.1 M Na2S
in two different potential ranges: from −1.0 to −0.3 V (Fig. 2a)
and from −1.0 to 0.9 V (Fig. 2b). For comparison, the simi-
lar experiments were performed in 0.1 M Na2S on a bare Au
electrode. After 10 scans in each potential range the Au/PDOT
electrodewastransferredbackinto0.1 MLiClO4/ANandcycled
Electrodeposition of Cd was carried out from acetonitrile
containing 0.1 M Cd(ClO4)2 + 0.1 M LiClO4. All solutions were
thoroughly deaerated by purging with dry argon for 15 min prior
the experiments.
The UV–vis spectra were done using a double-beam UV/vis
spectrometer (Lambda 12, Perkin-Elmer). The PDOT, CdS
and PDOT/CdS for spectrophotometric measurements were
deposited on ITO (indium–tin oxide)/glass electrodes (Balzers,
Lichtenstein) with an area of about 2.5 cm2. The spectra were
obtained for dry electrodes.
A nanoscope III (Digital Instruments, USA) was used for
imaging the surfaces of the samples studied. The samples were
prepared on Au substrate with an area of 0.86 cm2. AFM images
in tapping mode were obtained with standard phosphorous n-
doped Si cantilevers of 10 nm tip radius.
Energy dispersive X-ray spectra (EDS) were obtained by
means of scanning electron microscope (LEO 435VP) equipped
with a Roentec EDX analyser (model M1). Quoted elemental
ratios are the mean of six values acquired by probing the film in
different locations.
Photoelectrochemical experiments were carried out in the
three electrode cell with a Pt counter electrode, Ag/AgCl, Cl−
(std.) reference electrode and Au/CdS or Au/PDOT–CdS work-
ing electrode in deaerated aqueous solution of 0.1 M Na2S,
prepared from deionised water (>18 mꢀ). The working elec-
trodes were illuminated by 450 W arc xenon lamp (Oriel). The
incident power flux was passing through a 18 cm water filter. The
Fig. 1. Cyclic voltammograms for electropolymerization of DOT on Au elec-
trode from the solution of 0.01 M DOT + 0.1 M LiClO4/AN. Inset: stabilized
cyclic voltammograms of PDOT in the monomer-free solutions 0.1 M LiClO4
in AN (curve 1) and H2O (curve 2); curve 3 corresponds to CV obtained again
in AN after studies in aqueous solution.

2986
M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
Fig. 2. (a) Cyclic voltammograms of Au (curve 1) and Au/PDOT (curve 2) electrodes in aqueous solution of 0.1 M Na2S in the potential range from −1.0 to −0.3 V.
Inset: comparison of CVs at Au in two different potential ranges: from −1.0 to −0.3 V (curve A) and −1.0 to 0.9 V (curve B); (b) cyclic voltammograms of Au
(curve 1) and Au/PDOT (curve 2) electrodes in 0.1 M Na2S in the potential range from −1.0 to 0.9 V; inset: subsequent CVs at Au/PDOT electrode; (c) comparison
of the cyclic voltammograms in 0.1 M LiClO4/AN for the fresh Au/PDOT electrode (curve 1) and after 10 cycles in 0.1 M Na2S in the potential range from −1.0 to
−0.3 V (curve 2) and in the range from −1.0 to 0.9 V (curve 3).
Deposition of 4 nmol cm−2 PDOT on Au electrode does not
eliminate these peaks but only leads to some decrease of their
intensity (curve 2 in Fig. 2a). This suggests that the polymer did
not cover the whole substrate and SH- ions are still adsorbed
on Au. As visible in Fig. 2c, the cycling of the polymer in
Na2S between −1.0 and −0.3 V did not change the polymer
electroactivity in acetonitrile solution (cf. curves 1 and 2).
Extension of the polarization range of Au electrode to 0.9 V
leads to the appearance of an anodic peak at −0.24 V (inset in
Fig. 2a, peak B), related to oxidation of sulfides to sulfur and
in the same range as before the treatment in the sulfides (Fig. 2c)
to monitor the changes in the polymer electroactivity.
A pair of oxidation and reduction peaks, located respec-
tively at −0.86 and −0.91 V, obtained during polarization of
Au electrode in the potential range from −1.0 to −0.3 V (curve
1 in Fig. 2a) results from redox reaction of adsorbed SH− ions
[15,16]:
Au–(SH−)ads ↔ Au–Sads + H+ + 2e
(1)
where the subscript ads refers to the adsorbed species.

M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
2987
polysulfides [15–18]:
(Au–S) + xHS− + xOH− → (Au–S)Sx + xH2O + 2xe
(2)
(Au–S)Sx + SH− + OH− → (Au–S)Sx−n + Sn+12− + H2O
(3)
where and x ≥ n.
During the reverse scan the reduction of sulfur species occurs
in two steps:
2−
(Au–S)Sx + 2e → (Au–S)Sx
(4)
(5)
(Au–S)Sx2− + 2(x − 1)e → Au–S + xS2−
giving rise to two cathodic peaks (Bꢀ and Bꢀꢀ) at −0.65 and
−0.94 V.
Fig. 3. Cyclic voltammograms of Au electrodes covered with different amounts
PDOT, in the solution of 5 mM Cd(ClO4)2 + 0.1 M LiClO4/AN in the poten-
As visible in Fig. 2b, two anodic peaks corresponding to
oxidation of sulfides are observed at the Au/PDOT electrode.
The first one at −0.24 V (peak B) corresponds to the reaction at
uncovered Au substrate, whereas the second one at 0.67 V (peak
C) to oxidation of sulfides at the polymer/solution interface, after
polymer transformation from insulating to conducting form.
Enormous height of the second peak in comparison with the first
one is likely due to developed surface area of the deposited poly-
mer. Decrease of the peak intensity in the subsequent scans (see
inset) suggests that the process is irreversible and the deposited
sulfur blocks the polymer surface leading also to decrease of
its electroactivity. This is evidenced by the change of the shape
of voltammogram of Au/PDOT after transferring back into AN
solution (see curve 3 in Fig. 2c).
tial range 0.1 V → −1.0 V → 0.9 V → 0.1 V at the scan rate 100 mV s−1
.
The amount −o2f the polymer deposited on Au was: (1) 0 nmol cm−2, (2)
17.5 nmol cm , (3) 48 nmol cm−2, (4) 76 nmol cm−2, (5) 122 nmol cm−2 and
(6) 161 nmol cm−2
.
comparison to that on the bare Au electrode. However, for the
polymer coverage below 17.5 nmol cm−2 the shift is relatively
small (20 mV) and the deposition charge of cadmium is close
to that obtained on the bare electrode. This is not surprising
because, as it was concluded above, some parts of the sub-
strate remained uncovered by the polymer and then Cd is mainly
deposited on the bare Au.
For the higher amounts of the polymer the reduction peak
shifts more strongly with complementary decrease of the depo-
sition charge, as expected for the deposition on a low-conducting
substrate. There is also a strong unbalance between reduction
and reoxidation charges under the peaks A and Aꢀ. At the same
time, an additional wave (Aꢀꢀ) appears at the positive potentials
at the onset of the polymer oxidation (Bꢀ). It corresponds to
dissolution Cd deposited on the polymer surface—the position
of peak (Aꢀꢀ) shifts towards more positive potentials with the
increase of the polymer coverage (due to increase of the poly-
mer film resistance) and the charge under this peak is close to
the “missing” charge for cadmium reoxidation (Table 1). Some
charge unbalance still observed for polymer-covered electrode
may be due to hindrances in oxidation of Cd deposited on the
polymer surface and/or in the pores.
The results presented above lead to the conclusion that the
anodic potential range in the studies of Au/PDOT–CdS hybrid
electrode in Na2S should limited to −0.3 V.
3.2. Deposition of Cd on polymer-modified electrode
The reduction and oxidation peaks of Cd/Cd2+ couple
at the bare Au electrode in acetonitrile solution of 5 mM
Cd(ClO4)2 + 0.1 M LiClO4 occur at −0.61 and −0.23 V, respec-
tively (Fig. 3, curve 1). One may expect that process of Cd
deposition/dissolution on the PDOT film should be hindered
because of low conductivity of the polymer in the range of neg-
ative potentials, in the range between p- and n-doping [19], and
dependent on the amount of the deposited polymer.
Polymerization of DOT was carried out according to
the standard procedure described above. Then, the polymer-
modified electrode was placed in acetonitrile solution of
5 mM Cd(ClO4)2 + 0.1 M LiClO4 and cycled within the poten-
tial range: 0.1 V → −1.0 V → 0.9 V → 0.1 V to investigate the
electrochemical behavior not only in the range of cadmium
deposition/dissolution but also in the zone of the polymer
electroactivity. The cyclic voltammograms obtained for the elec-
trodes covered with increasing amount of PDOT are presented in
Fig. 3. The data presented in the graph and in the Table 1 indicate
that the position of the reduction peak of Cd2+ at PDOT-modified
electrode (peak A) shifts towards more negative potentials in
3.3. Formation of CdS and PDOT–CdS composites on the
electrodes
Formation of CdS was performed according to the modi-
fied electrochemical/chemical procedure developed by Penner
et al. for synthesis of semiconductor quantum dots [20]. Firstly
Cd was deposited on the bare electrode (Au, Pt or ITO) or
on the electrode covered with PDOT (deposited by the stan-
dard procedure) by a potential step from 0 to −0.74 V versus
Ag/AgCl/Cl− in the solution of 5 mM Cd(ClO4)2 in AN. The
amount of deposited cadmium was controlled by the reduction

2988
M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
Table 1
The charge densities for deposition/dissolution of Cd on the bare Au electrode and Au covered with different amounts of PDOT
Γ PDOT (nmol cm−2
)
0
2.7
17.5
48
76
122
161
For polarization from 0.1 to −1.0 and back to 0.1 V
Epc (V)
−0.610
−7.7
7.3
−0.622
−7.8
6.5
−0.632
−7.88
6.1
−0.630
7.6
5.5
−0.660
7.5
3.7
−0.678
7.4
2.8
−0.766
7.0
1.6
Qdep(Cd) (mC cm−2) (A)
Qdiss(Cd) (mC cm−2) (Aꢀ)
For polarization: 0 V → −1.0 V → 0.9 V → 0 V and subtraction of oxidation and reduction charges of the polymer
Qdep(Cd) (mC cm−2
Qdiss(Cd) (mC cm−2
)
7.62
7.21
7.57
7.17
7.67
7.2
7.66
6.85
7.6
6.8
7.4
6.2
)
charge. Then, Cd was oxidized to Cd(OH)2 by immersion of the
electrode in aqueous solution of 0.1 M NaOH for several min-
utes. Finally, Cd(OH)2 was transformed into CdS by dipping
the electrode for several minutes in aqueous solution of 0.1 M
Na2S.
The transformation of Cd deposited on ITO or ITO/PDOT
into CdS was monitored by absorption optical spectra. After
each stage of transformation the ITO-modified substrate was
taken out from the solution and the spectrum was done in the air
within the wavelength from 1100 to 330 nm.
TheEDSspectraofCdSpreparedbychemicaltransformation
of Cd, presented in Fig. 5, indicated that the CdS crystals are
cadmium rich with Cd/S ratio between 2.5 and 3. The peaks
at 3.13, 3.32, 3.52 and 3.72 keV with the height ratio 8:4:2:1
correspond to Cd whereas these at 2.31 and 2.46 eV (3:1) are
due to the presence of S.
This Cd:S ratio is similar to that obtained from XPS spectra
for CdS nanocrystals formed on HOPG by Penner et al. [22].
3.4. Photoeffects of PDOT/CdS composites in aqueous
solution of Na2S
As visible in Fig. 4a, the spectrum after immersion of
ITO/Cd in NaOH remained nearly unchanged, whereas dipping
ofITO/Cd(OH)2 inNa2Sledtotheappearanceofwelldeveloped
absorption peak at about 400 nm.
The electrodes for photoelectrochemical studies were pre-
pared on a gold substrate according to the standard procedures
described above, i.e. polymerization of DOT by cyclic voltam-
metry, deposition of Cd by a potential step to −0.74 V and
chemical transformation of Cd into CdS. Au substrate is more
suitable than Pt because of better quality of ohmic contacts with
semiconducting CdS film due to formation of Au–Cd alloy [23].
In order to study the influence of the amount of each compo-
nent on the photoeffect, three electrodes with different amounts
of PDOT and CdS were prepared, as indicated in Table 2. For
comparison, the electrodes covered with only one component,
PDOT or CdS were also formed.
The spectrum of ITO covered with PDOT (Fig. 4b, curve 4),
with characteristic vibronic structure, is typical of the neutral
poly(3,4-dialkoxythiophenes) [13]. The data were collected in
air but no increase of the absorption in the NIR range due to oxi-
dation of the polymer was observed. After deposition of Cd on
ITO/PDOT electrode and its transformation into CdS, an addi-
tional peak appeared (Fig. 4b, curve 3), similar to that visible in
the spectrum of ITO/CdS. Although it is difficult to determine
the absorption peak onset to estimate the value of band gap, it is
evidently above 510 nm. Thus, it means that the obtained CdS
particles are rather of submicron size with the bandgap close to
that of the bulk CdS, i.e. 2.4 eV [21].
Photoelectrochemical measurements were carried out in
deoxygenated aqueous solution of 0.1 M Na2S. In the presence
Fig. 4. Visible spectra obtained for: (a) ITO/Cd and (b) ITO/PDOT/Cd before (curves 1) and after (curves 2) treatment in 0.1 M NaOH. Curve 3 denotes the spectra
after transformation of Cd(OH)2 into CdS by dipping in Na2S; curve 4 in (b) corresponds to the spectrum of the neutral PDOT on ITO. All spectra were done for dry
samples in air.

M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
2989
Fig. 5. EDS pattern of CdS deposited by electrochemical/chemical method on Pt.
of sulfides the holes generated in the semiconductor under illu-
mination participate in the process of oxidation of S2− to S
quenching photodissolution of CdS:
ered with the same amount of CdS but without polymer. As
visible in Fig. 6a, the photoeffect obtained for the electrode
covered with 3.75 nmol cm−2 PDOT and 7.7 nmol cm−2 CdS
(PDOT4CdS8) is the same as that registered for the electrode
decorated with the same amount of CdS but without polymer.
The situation is quite different at the higher surface coverage of
CdS (20 nmol cm−2) (Fig. 6b). The photocurrent is significantly
higher for the hybrid electrode PDOT4CdS20 than that for CdS
without polymer (CdS20) (cf. curves 1 and 2 in Fig. 6b).
On the other hand, the increase of the surface coverage of the
polymer to 20 nmol cm−2 at the same amount of CdS leads to
more complicated behavior: at E < −0.8 V the photoactivity of
both electrodes is the same, whereas at the higher potentials the
photocurrent is higher for CdS without polymer (Fig. 6c). Thus,
it is evident that the ratio Γ PDOT/Γ CdS is a crucial parameter for
photoelectrochemical properties of hybrid electrodes.
S
2− + 2h+ → S
(6)
However, the sulfur does not accumulate on the electrode sur-
face but it is involved in the reaction with S2− to formation of
polysulfides [24]:
2−
S
2− + S → S2
(7)
(8)
Sx2− + S → Sx+1
2−
The electrodes were firstly cycled in dark and then under illu-
mination. In all cases several cycles were done to stabilize the
dark and the photocurrent. The experiments performed on the
electrodes covered only with PDOT (PDOT4 and PDOT20, see
Table 2) showed that no photocurrent was generated in polymer
film cycled in aqueous solution of 0.1 M Na2S in the potential
range from −1.0 to −0.4 V. The differences between photocur-
rent and dark current for the electrodes covered with about
4 nmol cm−2 PDOT and two different amounts of CdS, 8 and
20 nmol cm−2 are presented in Fig. 6a and b, respectively. The
same experiments were also carried out on the electrodes cov-
In order to explain this behavior the morphology of the elec-
trodes covered with different amounts of CdS and PDOT was
examined by means of AFM in tapping mode.
3.5. AFM images of Pt electrodes decorated with PDOT
and CdS
The AFM images were performed both for the Au substrate
covered with one of the components (PDOT or CdS) as well as
with polymer–semiconductor composites, prepared as described
above. The images were done for three different polymer surface
coverages (Γ ): 3, 10 and 17 nmol cm−2 and two amounts of Cd:
Table 2
The surface coverages of PDOT and Cd in the electrodes prepared for the
photoelectro-chemical studies
Electrode
PDOT (nmol cm−2
)
Cd (nmol cm−2
)
3 and 10 nmol cm−2
.
PDOT4CdS8
PDOT4CdS20
PDOT20CdS20
PDOT4
CdS8
PDOT20
3.75
3.75
20
4
–
20
–
7.7
20
20
–
7.7
–
According to the literature, the size and monodispersity of
CdS structures obtained by E/C method is the same as metal
crystallites deposited in the electrochemical step due to particle-
by-particle conversion from metal to semiconductor [25].
As visible in Fig. 7a and b, at low coverages (between
2 and 3 nmol cm−2) the substrate is decorated with spheri-
CdS20
20

2990
M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
Fig. 6. Comparison of the photoeffects (difference between photocurrent and a dark current) obtained for the electrodes: (a) PDOT4CdS8 (curve 1) and CdS8 (curve
2) (see description in Table 2), (b) PDOT4CdS20 (curve 1) CdS20 (curve 2) and (c) PDOT20CdS20 (curve 1) CdS20 (curve 2) during voltammetric cycles at the
scan rate 100 mV s−1 in the solution of 0.1 M Na2S. Curves 3 and 4 correspond, respectively, to the voltammograms obtained at the bare Au and Au covered with
corresponding amount of PDOT.
cal structures randomly distributed on the surface. The size
of PDOT particles is 150–240 nm, being markedly greater
than the size of CdS grains, 40–50 nm. A broad size distri-
bution of the polymer particles results from the method of
electrodeposition. During each voltammetric cycle the new
“nucleation” centers are formed, whereas the old ones are grow-
ing up. A general image of the substrate covered with CdS
obtained at low deposition charge is similar to that presented
by Gorer et al. [25] but a distribution of the particles is less
uniform and their size is about 10 times larger. This is due
to different electrodeposition conditions i.e. different substrate,
non-aqueous solvent, longer potential pulses and smaller over-
potential applied. However, the goal of our experiments was also
different and deposition of nano-sized CdS particles was not our
intention.
As visible in Fig. 7c, at the higher PDOT coverages the poly-
mer grains tend to coalescence forming the islands of dense,
three-dimensional aggregates of independent particles, which
finally start to cover the whole substrate. However, even at
Γ = 17 nmol cm−2 some parts of the substrate remained still
uncovered (Fig. 7e), which is consistent with the conclusions
derived from the cyclic voltammograms obtained for deposi-
tion of Cd and oxidation of sulfides on the PDOT-modified
electrodes.
At the CdS coverage of 10 nmol cm−2, the surface is covered
with a uniform “layer” of independent particles and the size
distribution of CdS grains is still relatively narrow (60–80 nm)
(Fig. 7d). Thus, it seems that deposition of Cd occurs via instan-
taneous nucleation, as has been suggested by Gorer et al. [25].
According to the literature, the nearest neighbor distance of

M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
2991
Fig. 7. AFM images for Au substrate covered with: (a) 2.8 nmol cm−2 of PDOT, (b) 2 nmol cm−2 of CdS, (c) 10 nmol cm−2 of PDOT, (d) 10 nmol cm−2 of CdS and
(e) 17 nmol cm−2 of PDOT.
0.289 nm for hexagonal packed layer Cd gives a coverage of
2.16 nmol cm−2. The studies of underpotential Cd deposition
by means of EQCM on the polycrystalline Au from aqueous
solution of HClO4 gave even lower value of 1.4 nmol cm−2
[26]. Thus, at Γ = 10 nmol cm−2 the surface should be cov-
ered with amount corresponding to about five monolayers
of Cd.
Now, having a picture of the PDOT and CdS deposits, one
can explain the differences in the photoeffects of hybrid elec-
trodes at various polymer and cadmium coverages. It should
be stressed however, that the images presented in Fig. 7 do
not reflect exactly the distribution of the particles on the
whole substrate of an area 0.86 cm2—some parts were cov-
ered by more and other by less densely distributed particles.

2992
M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
Fig. 8. Surface three-dimensional topography of the bare Au substrate (a); PDOT (4 nmol cm−2) (b); CdS (10 nmol cm−2) (c); PDOT–CdS hybrid electrode
(4 nmol cm−2 PDOT and 10 nmol cm−2 Cd) (d).
Nevertheless, one can expect that at low amounts of both com-
ponents, the particles of PDOT and CdS probably exist on
the substrate independently and the polymer does not influ-
ence the photoeffect of CdS. At the higher amount of CdS
(PDOT4CdS20) both type of the species (polymer and semi-
conductor) are likely located in close proximity to one another.
Since the polymer and CdS have different electron affinity the
charge carriers generated in the semiconductor under illumi-
nation may be easily separated – the electrons are transported
to the metal substrate, whereas the holes are transferred to
the polymer and then may be consumed in the reaction with
S2- from the solution. Accommodation of the holes by the
polymer results in significant inhibition of the electron–hole
recombination in the semiconductor and in the increase of the
photoeffect in comparison to that for CdS without polymer
(CdS20).
The arrangements of PDOT, CdS and both components in the
hybrid electrode are clearly illustrated in the three-dimensional
images presented in Fig. 8.
At the higher polymer coverages (PDOT20CdS20), the area
of uncovered Au is limited and Cd deposition takes place also on
the surface of PDOT. However, the relatively high resistivity of
the polymer in the range of negative potentials results in slowing
down the transport of the electrons from CdS to the metal contact
and therefore, the photocurrent is lower than that observed for
CdS without polymer.
3.6. Photovoltaic properties of the electrochemical cell
based on CdS–PDOT hybrid electrode
To determine the conversion efficiency of light energy to
electrical energy (power efficiency), η, the photoelectrochem-
ical cell was connected in series with a variable load resistance.
The current–potential curve was obtained for resistances rang-
ing from 20 Mꢀ to 100 ꢀ. The measurements were performed
in dark and under illumination and the results were used to
determine the power characteristic of the cell.
The power efficiency, in percent, was calculated using the
equation:
Pmax
η (%) =
× 100
(9)
Pin
where Pmax is a maximum power of the cell, whereas Pin denotes
the power input (in our case 200 mW cm−2). The quality of the
cell was also determined by a fill factor, FF, defined as
I
maxUmax
FF =
(10)
IscUoc
where Imax and Umax are the photocurrent and photopotential
corresponding to the maximum power, Isc is a short-circuit cur-
rent, i.e. maximum photocurrent observed when the resistance
of the circuit between the cell terminals is zero (at zero photopo-
tential) whereas Uoc is the open-circuit potential corresponding

M. Refczynska et al. / Electrochimica Acta 53 (2008) 2984–2993
2993
Table 3
energy conversion efficiency of 0.1%, open circuit voltage of
0.52 V, short circuit photocurrent density of 0.58 mA cm−2 and
fill factor of 0.51. The improvement of these parameters can be
likely achieved by annealing of the hybrid electrode.
Photovoltaic properties of hybrid CdS–PDOT/0.1 M Na2S electrochemical cell
Isc (mA cm−2
Uoc (V)
FF
Pmax(out) (mW)
η (%)
)
0.58
0.52
0.51
0.03
0.1
Acknowledgements
Support for this work by the Polish Ministry of Education
and Science under grants 3T08A04128 and N20409131/2142 is
gratefully acknowledged.
to zero photocurrent (when the external resistance is very large
(R → ∞)).
The photovoltaic properties of the cell based on CdS–PDOT
hybrid electrode composed of 4 nmol cm−2 PDOT and
20 nmol cm−2 CdS are listed in Table 3.
References
[1] N.C. Greenham, P. Xiaogang, A.P. Alivisatos, Phys. Rev. B 54 (1996)
17628.
[2] D.S. Gringer, N.C. Greenham, Phys. Rev. B 59 (1999) 10622.
[3] P. Chartier, H. Nguyen Cong, C. Sene, Solar Energy Mater. 52 (1998) 413.
[4] T. Cassagneau, T.E. Mallouk, J.H. Fendler, J. Am. Chem. Soc. 120 (1998)
7848.
[5] W.U. Huynh, J.J. Dittmer, A.P. Alivisatos, Science 295 (2002) 2425.
[6] E. Arici, H. Hoppe, F. Scha¨ffler, D. Meissner, M.A. Malik, N.S. Sariciftci,
Thin Solid Films 451/452 (2004) 612.
[7] W.U. Huynh, J.J. Dittmer, W.C. Libby, G.L. Whiting, A.P. Alivisatos, Adv.
Funct. Mater. 13 (2003) 73.
[8] K.S. Narayan, A.G. Manoy, J. Nanda, D.D. Sarma, Appl. Phys. Lett. 74
(1999) 871.
[9] S. Pethkar, R.C. Patil, J.A. Kher, K. Vijayamohanan, Thin Solid Films 349
(1999) 105.
[10] T. Nakanishi, B. Ohtani, K. Uosaki, J. Phys. Chem. B 102 (1998) 1571.
[11] L. Sheeney-Haj-Ichia, J. Wasserman, I. Willner, Adv. Mater. 14 (2002)
1323.
[12] E. Granot, F. Patolsky, I. Willner, J. Phys. Chem. B 108 (2004) 5875.
[13] A. Szkurlat, B. Palys, J. Mieczkowski, M. Skompska, Electrochim. Acta
48 (2003) 3665.
The obtained photovoltaic parameters are lower than those
reported recently by Kang and Kim [27] (η = 0.6%, FF = 0.496,
Voc = 0.858 V and Isc = 1.39 mA cm−2) for the system composed
of vertically aligned CdS nanorods in poly[2-methoxy-5-
(2ꢀethylhexoxy)-1,4-phenylene-vinylene (MEH-PPV). How-
ever, the electron transport in the aligned nanorods is remarkably
faster than that in the spherical structures obtained by electrode-
position. According to the literature [28,29], the enhancement of
solar energy conversion efficiency may be achieved by thermal
treatment of the hybrid system. This effect was attributed to reor-
ganization of the CdS nanocrystals/polymer interface leading to
an increase of exciton dissociation efficiency [29]. This raises
expectation that annealing should also improve the properties of
the CdS–PDOT composite and our work will be continue in this
direction.
4. Conclusions
[14] A.B. Ellis, S.W. Kaiser, J.M. Bolts, M.S. Wrighton, J. Am. Chem. Soc. 99
(1977) 2839.
[15] R.O. Lezna, N.R. De Tacconi, A.J. Arvia, J. Electroanal. Chem. 283 (1990)
319.
[16] X. Gao, Y. Zhang, M.J. Weaver, Langmuir 8 (1992) 668.
[17] A.N. Buckley, I.C. Hamilton, R. Woods, J. Electroanal. Chem. 216 (1987)
213.
[18] N. Myung, S. Ham, B. Choi, N.R. de Tacconi, K. Rajeshwar, J. Electroanal.
Chem. 574 (2005) 367.
[19] M. Skompska, J. Mieczkowski, R. Holze, J. Heinze, J. Electroanal. Chem.
57 (2005) 9.
[20] M.A. Anderson, S. Gorer, R.M. Penner, J. Phys. Chem. B 101 (1997) 5895.
[21] A. Henglein, Chem. Rev. 89 (1989) 1861.
[22] S. Gorer, J.A. Ganske, J.C. Hemminger, R.M. Penner, J. Am. Chem. Soc.
120 (1998) 9584.
[23] J.J. Kelly, J.M.G. Rikken, J.W.M. Jacobs, A. Valster, J. Vac. Sci. Technol.
B 6 (1988) 48.
[24] B. Miller, A. Heller, Nature 262 (1976) 680.
[25] S. Gorer, G.S. Hsiao, M.G. Anderson, R.M. Stiger, J. Lee, R.M. Penner,
Electrochim. Acta 43 (1998) 2799.
Au/poly(3,4-dioctyloxythiophene)–CdS hybrid electrodes
have been prepared by means of four-step electrochemi-
cal/chemical method. Au was firstly decorated with the polymer
by cyclic voltammetry and then, CdS was formed by potentio-
static deposition of Cd followed by its chemical transformation
intoCd(OH)2 inNaOHandfinallytoCdSbyimmersioninNa2S.
The CdS crystals obtained by this procedure are cadmium rich
with Cd/S ratio from 2.5 to 3.
CdS and PDOT are deposited in the form of hemispherical
structures dispersed on the substrate. Their size and density may
be easily controlled by the electrodeposition charge.
The amount of both components and the CdS:PDOT ratio
are crucial for photoelectroactivity of the hybrid electrodes in
aqueous solution of 0.1 M Na2S. In the most suitable conditions
the polymer structures adjacent to the semiconductor particles
play role of acceptors of the holes generated under illumination,
whereas the semiconductor transports the electrons to the metal
substrate. The charge separation results in significant inhibition
of the electron–hole recombination and enhancement of the pho-
toeffect. The photovoltaic device based on the spherical CdS
structures and electrochemically synthesized PDOT provided
[26] M.R. Deakin, O. Melroy, J. Electroanal. Chem. 239 (1988) 321.
[27] Y. Kang, D. Kim, Solar Energy Mater. Solar Cells 90 (2006) 166.
[28] W.J.E. Beek, M.M. Wienk, R.A.J. Janssen, Adv. Funct. Mater. 16 (2006)
1112.
[29] L. Wang, Y.S. Liu, X. Jiang, D.H. Quin, Y. Cao, J. Phys. Chem. C 111
(2007) 9538.