Photoelectrochemistry of pure and core/sheath nanowire arrays of Cu 2S directly grown on copper electrodes

Article abstract of DOI:10.1149/1.1859991

Photoelectrochemistry (PEC) of straight and isolated Cu₂S nanowires arrayed on a copper foil is studied. A cathodic photocurrent is observed, which increases with the increasing negative bias of the film electrode, conforming to the p-type semiconducting nature of Cu₂S. PEC studies on the core/sheath nanowires of Cu₂S (Cu₂S nanowires coaled with other nanoparticles such as CdS, polypyrrole, and Au) have revealed distinctly different characteristics. Cu₂S/CdS nanowires exhibit a slightly higher photocurrent response due to the charge transfer between the p-type Cu₂S core and n-type CdS sheath, whereas the photocurrent response obtained for Cu₂S/PPy nanowires is generally an order of magnitude smaller than that for pure Cu₂S nanowires under the same bias voltage. For Cu₂S/Au nanowires, however. Au nanoparticles can readily accept the photogenerated electrons from the conduction band of Cu₂S; this redirects the electron transfer pathway and results in the anodic photocurrent response. On the basis of these findings, different photocurrent generation mechanisms are proposed.

Full text of DOI:10.1149/1.1859991

G220
Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
0013-4651/2005/152͑3͒/G220/7/$7.00 © The Electrochemical Society, Inc.
Photoelectrochemistry of Pure and CoreÕSheath Nanowire
Arrays of Cu2S Directly Grown on Copper Electrodes
Shangfeng Yang,a Xiaogang Wen, Weixin Zhang, and Shihe Yangz
Department of Chemistry, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon,
Hong Kong
Photoelectrochemistry ͑PEC͒ of straight and isolated Cu2S nanowires arrayed on a copper foil is studied. A cathodic photocurrent
is observed, which increases with the increasing negative bias of the film electrode, conforming to the p-type semiconducting
nature of Cu2S. PEC studies on the core/sheath nanowires of Cu2S (Cu2S nanowires coated with other nanoparticles such as CdS,
polypyrrole, and Au͒ have revealed distinctly different characteristics. Cu2S/CdS nanowires exhibit a slightly higher photocurrent
response due to the charge transfer between the p-type Cu2S core and n-type CdS sheath, whereas the photocurrent response
obtained for Cu2S/PPy nanowires is generally an order of magnitude smaller than that for pure Cu2S nanowires under the same
bias voltage. For Cu2S/Au nanowires, however, Au nanoparticles can readily accept the photogenerated electrons from the
conduction band of Cu2S; this redirects the electron transfer pathway and results in the anodic photocurrent response. On the basis
of these findings, different photocurrent generation mechanisms are proposed.
© 2005 The Electrochemical Society. ͓DOI: 10.1149/1.1859991͔ All rights reserved.
Manuscript submitted May 17, 2004; revised manuscript received September 8, 2004. Available electronically January 31, 2005.
The favorable material characteristics of Cu2S makes it suitable
for large-scale photovoltaic ͑PV͒ applications, particularly in com-
bination with CdS or ZnxCd1ϪxS.1 p-n Heterojunctions based on
Cu2S/CdS thin film solar cell show peculiarities, including the
crossover of the I-V characteristic in the dark and under illumina-
tion, and the light bias effect.1-3 Theoretical analyses have revealed
that the Cu2S/CdS and Cu2S/ZnCdS thin film PV systems are
capable of achieving practical conversion efficiencies of 11 and 14-
15%, respectively.1 A thin film solar cell based on a single-
component Cu1.8S was developed by constructing nanoporous het-
erojunctions of TiO2 /Cu1.8S, showing an open-circuit voltage
͑OCV͒ of about 200 mV and a short-circuit current of 30 ␮A/cm2 at
2.8 kW/m2.4 For dye-sensitized solid-state PV cells based on
TiO2/dye/CuI thin films, a power conversion efficiency of 2.8% was
recently reported.5 On the other hand, photoelectrochemical ͑PEC͒
cells based on nanocrystalline materials have emerged in recent
years as a new generation of PV devices and offer the prospect of
enhanced efficiency and cheap fabrication together with other attrac-
tive features such as flexibility.6-9 PEC properties of nanoparticles
or quantum dots of SnO2 ,10 TiO2 ,11-12 ZnO,13 CuO͑Cu2O),14
CdS,15-17 ZnS,18 CdSe19 have been reported.
Nanowires and nanotubes may become important building
blocks for nanoscale optoelectronics because they can function as
miniaturized devices as well as electrical interconnects.20 Nanode-
vices such as field-effect transistors,21 single-electron transistors,22
metal-semiconductor juctions,23 and interwire crossed junctions24
have been demonstrated. However, less attention has been paid to
the PEC properties of nanowires despite the exciting prospects. PEC
of oriented nanorod thin films of hematite (␣-Fe2O3) including dye
sensitization was reported by Hagfeldt et al. revealing the consider-
able enhancement of the photoconversion efficiency compared to
that based on the hematite nanoparticles.25 A more recent study
based on the photoresponses toward the water-splitting reaction of
the thin film and nanowire electrodes of n-type TiO2 indicated that a
more than twofold increase in maximum photoconversion efficiency
was observed when a single-layer thin film of n-TiO2 was replaced
by nanowires.26
reaction under ambient conditions. By using a nanostructured thin
film based on perpendicularly oriented Cu2S nanowires, one hopes
to reduce the recombination losses at grain boundaries between the
nanoparticles and thus enhance the power conversion efficiency.25,26
Several core/sheath nanowire systems have also been studied and
compared. By examining the effects of the nanowire coating, bias
voltage, and electrolyte solution, we have proposed plausible photo-
current generation mechanisms, which are also presented.
Experimental
Materials.—The preparation of straight and isolated Cu2S
nanowires arrayed on copper foils by using a gas-solid reaction un-
der ambient conditions was reported previously.27 Briefly, cleaned
copper foils (0.5 ϫ 0.5 cm͒ were placed in a homemade reactor.
The reactor was first pumped to vacuum, then a gas flow was di-
rected to the reactor which consists of a mixture of oxygen ͑99.8%͒
and hydrogen sulfide ͑99.8%, Aldrich͒ at a given molar ratio and the
total pressure in the reactor was kept at ϳ1.05-1.08 atm. After 10 h
reaction, the copper surface became black and fluffy, indicating the
formation of dense Cu2S nanowire arrays. Nanoparticles ͑NP͒ were
coated onto the Cu2S nanowires to form Cu2S/NP core/sheath
nanowires (NP ϭ CdS,b polypyrrole ͑PPy͒,c,28 Aud,29͒. KCl ͑RDH
AG͒, Na2SO3 ͑sodium sulfite, RDH AG͒, and Na2C2H4O6•2H2O
͑sodium tartrate dibasic dihydrate, Fisher͒ were used as received.
PEC measurements.—The PEC measurements were carried out
with a model 600 electrochemical analyzer from CH Instruments,
Inc., U.S. A 500 W mercury-xenon lamp ͑Hamamatsu, Japan͒ was
used as the light source. The beam intensity of the light was cali-
brated with a power meter ͑Newport Co., U.S.͒, but no corrections
were made for the reflections. The collimated white light beam
passed through an FS-3 filter (od ϭ 0.05, Newport Co., U.S.͒,
Ͻ
which was used to eliminate light with ␭
330 nm. The irradiation
passed through a homemade water filter before entering the PEC cell
to minimize thermal effects by removing IR lights. The electro-
chemical cell was under a flowing nitrogen atmosphere ͑but without
bubbling͒ to eliminate the influence of oxygen. Before measure-
ments, the electrolyte solution was bubbled with N2 (purity
Ͼ 99.9%) for ϳ30 min. A three-electrode configuration was used
for the quartz cell. The Cu2S (Cu2S/NP) nanowires on Cu foil were
used as the working electrode. Typically, The thickness and length
of the Cu2S nanowires are 60-100 nm and several micrometers,
respectively, and the outer sheath layers are around 20-50 nm thick
for the core/sheath nanowires, as described above and else-
Here, we report the PEC study of straight and isolated Cu2S
nanowires arrayed on copper foils, prepared by using a gas-solid
a Present address: Leibniz-Institute for Solid State and Materials Research Dresden,
01069 Dresden, Germany.
z E-mail: chsyang@ust.hk
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
G221
Figure 2. Representative photocurrent traces vs. time obtained from a Cu2S
nanowire film on Cu electrode ͑A͒ and pure Cu foil ͑B͒. ͓Ϫ0.50 V bias
voltage; 0.1 M KCl electrolyte solution ͑pH 7͒; light intensity ϭ 38.0 mW
cmϪ2͔. The asterisks indicate the light on-off cycles.
tates the efficient charge carrier transport among the nanowires with
the abated charge recombination losses in the film. Furthermore, by
uniformly coating of NP as the outer sheaths, charge transfer is
expected to occur between the core of Cu2S and the sheath of
NP, which may subsequently tune the PEC performance of the
nanowires.
Figure 1. Top: schematic illustration showing the preparation of the
Cu2S/NP core/sheath nanowires. ͑A͒ Representative TEM image of Cu2S
nanowires; ͑B͒ TEM image of a Cu2S/Au core/sheath nanowire; ͑C͒ magni-
fied TEM image of a Cu2S/Au core/sheath nanowire.
where.27-29,b,c,d The counter electrode was a Pt wire, and the refer-
ence electrode was Ag/AgCl. The electrolytes were aqueous solu-
tions of KCl ͑pH 6.5͒, Na2SO3 ͑pH 9͒, and Na2C2H4O6
•2H2O ͑sodium tartrate dibasic dehydrate, pH 7͒ with concentra-
tions of 0.1, 1.0, and 0.5 mol dmϪ3 ͑M͒, respectively. Due to the
opaqueness of the substrate ͑Cu foil͒, the nanowire working elec-
trode was always illuminated through the electrolyte/electrode inter-
face ͑EE͒.30 More than six nanowire samples were measured under
each experimental condition to ensure the reproducibility of the pho-
tocurrent data.
Photoelectric response of Cu2S nanowires on Cu elec-
trodes.—Shown in Fig. 2 is a typical photocurrent response curve
for Cu2S nanowires ͑A͒ on Cu electrode at a bias voltage of Ϫ0.50
V and the white light beam irradiation intensity of 38.0 mW cmϪ2
and that for pure Cu substrate ͑B͒ as a reference.e Clearly, the ob-
served cathodic photocurrent response, which is absent for the pure
Cu substrate, is from the photoactive Cu2S nanowires. Under the
relatively low irradiation intensity condition we used, the net steady-
state cathodic photocurrent was calculated to be ϳ0.11 mA/cm2
with little attenuation for the two successive cycles ͑curve A͒. Al-
though the PEC data for Cu2S nanoparticles is not available pres-
ently, this photocurrent response is estimated to be comparable to
Results and Discussion
Nanowire structures.—Figure 1 illustrates the typical transmis-
sion electron microscopy ͑TEM͒ images of Cu2S and Cu2S/NP core/
sheath nanowires ͑here NP ϭ Au is shown as an example͒. The
latter was converted from the former as shown in the figure. It has
been established that the straight and isolated Cu2S nanowires are
uniformly arrayed on copper foils,27-29,b,c,d we expect that this facili-
those of TiO2 ,11 CdS nanoparticles,15-17 and TiO2 /Cu1.8
S
heterojunctions4 by considering the different irradiation conditions.
Such a high photocurrent response is understandable, because the
uniformly arrayed nanowire nature of Cu2S, as demonstrated
elsewhere,27-29,b,c,d may facilitate the efficient charge-carrier trans-
port among the nanowires with the abated charge recombination
losses in the film.25,26 However, the dark current due mainly to
electrochemical reactions prior to the light illumination has the same
magnitude as the photocurrent in this case. First, this is because the
Cu substrate is actually electroactive in an aqueous solution ͓e.g.,
2H2O ϩ Cu → H2↑ ϩ 2OHϪ ϩ Cu2ϩ → H2↑ ϩ Cu͑OH)2↓] ac-
cording to the cyclic voltammetry ͑CV͒ analysis ͑not shown͒. This
can also be confirmed by noticing that the dark current of the pure
Cu foil ͑curve B͒ is much higher than that with other inert substrates
such as indium tin oxide ͑ITO͒ which was used in our other work.30
Moreover, the dark current for the Cu2S nanowire film ͑see curve A͒
appears to be much higher than that of pure Cu electrode, suggesting
the partial reduction of Cu2S under the dark condition. Because the
absorption spectrum of Cu2S nanowires has not be determined yet
due to the opaqueness of Cu substrate, the incident photon-to-
current conversion efficiency ͑IPCE͒ could not be calculated at
b The pretreated Cu2S nanowires were fixed on a glass capillary and immersed into
40 mL CdCl2 aqueous solution with a concentration of 1 ϫ 10Ϫ2 M. After 6 h, 10 mL
hydrogen sulfide gas was injected into the solution slowly using a syringe. After coating
for 2 h, the samples were collected and washed three times with deionized water and
dried in air.
c 0.1 mL pyrrole was first dissolved in 40 mL chloroform. Then the Cu2S nanowire
arrays were immersed in the solution and stirred for 2 h. Afterward, the Cu2S nanowires
was raised and positioned right on the surface of the chloroform solution without mag-
netic stirring. 20 mL of a (NH4)2S2O8 aqueous solution ͑0.1 M͒ was then added slowly
to the chloroform solution. A thin white film was formed at the interface within half an
hour, which turned to gray/black within another half an hour. By then the film nearly
covered the whole interface. The film became increasingly black with increasing poly-
merization time.28
d First the Au-coating solution was prepared by dissolving 0.395
g
HAuCl4•3H2O in 10 mL deionzed water. The solution was then diluted to different
concentrations and an appropriate amount of a NaOH solution ͑1 M͒ was added to
adjust the pH value of the solution. The pretreated Cu2S nanowires were fixed on a
glass capillary and immersed into the coating solution. After coating for 1 h, the nanow-
ires were collected and washed three times with deionized water following with abso-
lute alcohol.29
e Prior to the photoelectrochemical measurement, the pure Cu foil was first treated
by 6.0 M HCl for 10 min to remove the oxide layer, and then cleaned for ϳ5 min in an
ultrasonic bath of absolute ethanol.
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G222
Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
Figure 3. ͑a͒ Photocurrent (Iph) as a function of the bias voltage for a Cu2S
nanowire film on Cu electrode ͑0.1 M KCl electrolyte solution ͑pH 7͒; light
intensity ϭ 38.0 mW cmϪ2. Inset: representative photocurrent traces under
different bias voltage. ͑b͒ I2ph as a function of the bias voltage derived from a.
Figure 4. Representative photocurrent trace vs. time obtained with a Cu2S
nanowire film on Cu electrode in different electrolyte solutions of Na2SO3
͑1.0 M; pH 12͒ ͑a͒ and Na2C2H4O6 ͑0.5 M; pH 14͒ ͑b͒. ͑Ϫ0.50 V bias
voltage; light intensity ϭ 38.0 mW cmϪ2͒. The arrows indicate the light
on-off cycles.
present. Nevertheless, the present study demonstrates the potential
use of Cu2S nanowires for facilitating the photoelectric conversion.
An interesting feature of the observed cathodic photocurrent
trace shown as A of Fig. 2 is the photocurrent overshooting upon
light illumination. This becomes more evident when the bias volt-
ages are more positive ͑see also the inset of Fig. 3͒. One would
expect the overshooting to be suppressed completely by applying
further higher negative bias. However, due to the higher dark current
under the very high negative one, we are unable to detect photocur-
rents with bias more negative than Ϫ0.70 V. As reported by Hickey
et al. based on the PEC studies of CdS nanoparticle modified tin
oxide electrodes in an aqueous 1.0 M Na2SO3 electrolyte,16 a simi-
lar phenomena was observed, in which an anodic photocurrent tran-
sient was obtained with different shapes under different applied po-
tentials. This was explained by the back electron transfer process
competing with hole scavenging by the solution species at relatively
negative potentials.16 Bakkers et al. also reported the photocurrent
transients of different shapes obtained in an aqueous 1.0 M KCl
electrolyte.17 We believe that a similar photocurrent generation
mechanism is operative in the case of Cu2S nanowires, which is
discussed in detail below.
the Cu electrode through the nanowire film to the electrolyte, which
is consistent with the p-type semiconducting nature of Cu2S.1 Fur-
thermore, shown in Fig. 3b is the dependence of I2ph on the bias
voltage for the Cu2S nanowire film electrode, from which the flat-
band potential (EFB) is determined ͑Ϫ0.295 V͒ by extrapolating the
linear region of the curve of I2ph-V to zero bias voltage.30-32 Here
EFB is the upper limit of a blocking potential region for p-type
semiconductor where the electrode response is dominated by the
space-charge region layer.
Effect of electrolyte solution.—The effect of the electrolyte solu-
tion on the photocurrent response for the better known n-type CdS
has been revealed based on the PEC studies of the nanoparticles or
quantum dots.15-17 For example, aqueous Na2SO3 solution could
suppress the photodissolution of CdS,15,33 while tartrate and trietha-
nolamine are well-known stabilizing agents for CdS.17,34 In both
cases, the electrolyte plays the role as the hole scavenger.15-17,33,34
An immediate interest thus arises to investigate whether such a
modulation effect of the electrolyte solution would exist in the
present Cu2S nanowire system especially for the transient-state pho-
tocurrent overshooting stated above. We then replaced the electro-
Effect of bias voltage.—The steady-state cathodic photocurrent
(Iph) as a function of bias voltage (Vbias) from Ϫ0.50 to Ϫ0.15 V
͑vs. Ag/AgCl͒ of the Cu2S nanowires is shown in Fig. 3a. The ca-
thodic photocurrent increases with the increasing negative bias of
the nanowire film electrode, exhibiting a sharp increase for the por-
tion from Ϫ0.275 to Ϫ0.35 V, then starts to level off at Ϫ0.40 V.
The fact that the cathodic photocurrent increases with the increasing
negative bias of the film electrode indicates the electrons flow from
f
lyte with sodium sulfite (Na2SO3 , 1.0 M, pH 12͒ and sodium tar-
f
trate (Na2C2H4O6 , 0.5 M, pH 14͒ for comparison. When Na2SO3
is used as the electrolyte solution ͑Fig. 4a͒, the photocurrent over-
f The pH of Na2SO3 and Na2C2H4O6 electrolyte solutions were adjusted by using
1.0 M NaOH.
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
G223
Table I. Characteristic of the PEC cells based on Cu2S and Cu2SÕNP nanowire films on Cu electrodes under different conditions.
d
Photocurrent
Nanowire
film
EFB
Correlation
Electrolyte
solution
a
Iph ͑mA/cm2͒
͑V vs. Ag/AgCl͒
coefficient Rd,e
0.11b
0.14b
Ϫ0.295
Ϫ0.258
Ϫ0.363
Ϫ0.278
0.981
0.993
0.976
0.962
0.1 M KCl
͑pH 7͒
0.1 M KCl
͑pH 7͒
Cu2S
Cu2S/CdS
Cu2S/PPy
Cu2S/Au
0.0033b
Ϫ0.091c
0.5 M Na2C2H4O6
͑pH 14͒
0.5 M Na2C2H4O6
͑pH 14͒
a Under white light illumination (intensity ϭ 38.0 mW cmϪ2͒; bias voltage ϭ Ϫ0.50 V. Iph is the mean value based on three reproducible measurements.
b Cathodic photocurrent.
c Negative value denotes anodic photocurrent.
d Obtained by linear fit analysis which is based on the corresponding data points in the linear region.
e Parameter for linear fit analysis.
shooting observed for the case of KCl being the electrolyte generally
appears to be diminished and a more stable cathodic photocurrent is
obtained. However, the dark current of the same magnitudes as for
the former case is unstable over time, indicating that some addi-
tional electrochemical reactions occur under dark condition. In this
case, the cathodic photocurrent obtained under the same conditions
is smaller than that when KCl was used ͑see Fig. 2͒. This result
indicates that Na2SO3 is not as effective for promoting the photo-
current response of Cu2S as for CdS although the transient-state
photocurrent overshooting is suppressed. On the other hand, as
shown in Fig. 4b, the use of sodium tartrate (Na2C2H4O6) as the
electrolyte solution seems to diminish the transient-state photocur-
rent overshooting too, confirming its similar stabilizing effect on the
photocurrent response of Cu2S.17,34 Nevertheless, the dark current is
an order of magnitude higher, indicating again the increasing elec-
trochemical reactions in the dark. The photocurrent response ob-
tained in this case is comparable to the case when Na2SO3 was used
but smaller than that when KCl was used. Moreover, the observed
photocurrent response appears to be slower upon applying/removing
light illumination in this case. Normally it takes 3-4 s for the pho-
tocurrent to reach the saturation value, suggesting that the electron
transfer from the Cu electrode to the nanowires and then to the
electrolyte is slower. These results reveal that the additions of
Na2SO3 and Na2C2H4O6 diminish the cathodic photocurrent re-
sponse instead, presumably because such hole scavengers could sup-
press the charge carrier separation process as hole is the major car-
rier for p-type Cu2S.
Na2C2H4O6 ͑pH 14͒ together with the cathodic photocurrent trace
͑inset͒. The measured photocurrent under the same bias voltage is
generally an order of magnitude smaller than that for pure Cu2S
nanowire film ͑see Fig. 4b͒, presumably due to the misalignment of
the band levels and/or the poor adhesion of the PPy sheath layer.
Similar to the case for pure Cu2S film, no obvious photocurrent
overshooting is detected. However, the slow response for the pho-
tocurrent response for pure Cu2S by using Na2C2H4O6 as electrolyte
͑see Fig. 4b͒ is not observed here ͑see inset͒, indicating that electron
transfer is faster in this case. In addition, contrary to the case of
Cu2S/CdS nanowire, EFB experiences a negative shift of ϳ70 mV
͑see Table I͒, indicating that the charge distribution between Cu2S
and PPy is quite different from the former case. In fact, the obtained
EFB is also more positive than that determined for pure PPy film
͑Ϫ0.474 V vs. Ag/AgCl͒,36 suggesting the possible charge transfer
from PPy to Cu2S.
Recently, Kamat et al. showed that metal nanoparticles or islands
deposited on semiconductor surfaces play an important role in cata-
lyzing interfacial charge-transfer.11,35,37 Particularly, PEC perfor-
mance of nanostructured TiO2 films was improved by adsorption of
Au nanoparticles based on the charge distribution between photoex-
cited n-type TiO2 and Au nanoparticles.37 Moreover, PEC of Au
nanoparticle arrays are also reported by Lahav et al.38 It thus ap-
pears interesting to investigate whether the PEC properties of Cu2S
core could be modulated by the Au sheath for Cu2S/Au nano-
wires.29,d Upon coating Au nanoparticle layer, as shown in Fig. 7a,
the photocurrent becomes anodic instead of cathodic which is gen-
erally observed for the above cases, suggesting a significant change
for the electron-transfer process in this case. The anodic photocur-
rent is stable and no obvious photocurrent overshooting is detected,
indicating that the photocurrent originates from the electron transfer
from the electrolyte to the Cu electrode through the nanowire film,
which is mediated by Au nanoparticles layer. Furthermore, the pho-
tocurrent decreases with the decreasing negative bias voltage of the
film electrode in the range of Ϫ0.65 to Ϫ0.20 V ͑Fig. 7b͒. This
unusual dependence is quite surprising because it is the characteris-
tic of a cathodic photocurrent instead and contradictory to the anodic
nature of the photocurrent observed here. For example, the anodic
photocurrent for n-type semiconductors such as TiO2 and CdS nor-
mally exhibit an increase of anodic photocurrent upon decreasing
negative bias voltage.11,12,15-17,37 One plausible interpretation is that
although the electron transfer from the Au to the Cu electrode be-
comes hindered upon applying more negative bias voltage, the com-
peting back electron transfer from Au to the surface states ͑SS͒ of
Cu2S ͑cathodic͒ would be even more abated ͑see also Scheme Ib͒,
and the net result then is the increase of the anodic photocurrent. If
the bias voltage is too negative ͑say, ϽϪ0.65 V͒, however, the elec-
tron transfer from Au to the Cu electrode is much hindered and thus
Photoelectric response of core/sheath nanowires.—The suc-
cess in preparing Cu2S/NP ͑NP ϭ CdS, PPy, Au͒ core/sheath nano-
wiresb-d,28,29 enables us to extend the PEC investigation to these
nanocomposites with the possibility of charge transfer between the
nanowire core and the outer nanoparticle sheath. Upon coating with
the CdS sheath,b no difference on the shape of cathodic photocurrent
trace under the same measurement condition ͑0.1 M KCl͒ was de-
tected, while the photocurrent response appears to be slightly higher
as evidenced by the Iph-V curve shown in Fig. 5. This is understand-
able because the charge transfer between the p-type Cu2S core and
n-type CdS sheath is expected to occur in the Cu2S/CdS p-n hetero-
junction nanowire, resulting in enhanced energy conversion
efficiency.1-3 Moreover, EFB appears to shift positively by ϳ40 mV
͑see inset͒, confirming the charge transfer from Cu2S to CdS ͑see
Table I͒.11,35 By optimizing the relative thickness of the core and
sheath layers, a higher enhancement of the efficiency is expected.
A conducting polymer, polypyrrole ͑PPy͒, is also coated as the
sheath layer onto the Cu2S nanowire core.28,c PEC studies of PPy in
both aqueous and nonaqueous solutions had revealed its p-type
semiconductor properties.36 Figure 6 demonstrates the Ip2h-V curve
for the Cu2S/PPy core/sheath nanowire film measured in 0.5 M
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G224
Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
Figure 5. Photocurrent (Iph
)
as a function of the bias voltage for a
Cu2S/CdS ͑A͒ and pure Cu2S ͑B͒ nanowire films on Cu electrodes. ͑0.1 M
KCl electrolyte solution ͑pH 7͒; light intensity ϭ 38.0 mW cmϪ2͒. Note that
curve B is copied from Fig. 3a for comparison. Inset: I2ph as a function of the
bias voltage derived from curve A.
the anodic photocurrent begins to decrease. In fact, as determined
from the I2ph-V curve ͑inset of Fig. 7b͒, the charge transfer from
Cu2S to Au is deduced by the positive shift of EFB ͑see Table I͒.
Photocurrent generation mechanism.—The results presented
above show that different photocurrent generation mechanisms may
operate in the different nanowire films. According to the PEC theo-
ries, the photoexcitation of semiconductor upon light illumination
generates charge carriers, which are separated at the electrode/
electrolyte interface and then transferred between the surface of the
electrode and the underlying substrate.11,15-17,37 Proposed in Scheme
I is the possible mechanism for the electron-transfer processes oc-
curring for different nanowire films. As shown in Scheme Ia, pho-
Figure 7. ͑a͒ Representative photocurrent traces for a Cu2S/Au nanowire
films on Cu electrode. ͓Ϫ0.50 V bias voltage; 0.5 M Na2C2H4O6 electrolyte
solution ͑pH 14͒; light intensity ϭ 38.0 mW cmϪ2.͔ ͑b͒ Anodic photocurrent
(Iph) as a function of the bias voltage. ͓0.5 M Na2C2H4O6 electrolyte solu-
tion ͑pH 14͒; light intensity ϭ 38.0 mW cmϪ2͔. Inset: Ip2h as a function of the
bias voltage derived from b.
*
toexcitation of Cu S (Cu S) results in an electron in the conduc-
͓
͔
2
2
tion band ͑CB͒ transferring to the electron acceptor H2O in the
electrolyte. Subsequently, the Cu2S molecule is replenished by an
electron transferred from the Cu electrode, completing the external
circuit with the generation of cathodic photocurrent. The plausible
half reactions are proposed as follows
*
Oxidation of (Cu2S)
Cu2ϩ ϩ CuS ϩ 2e → Cu2S
͓1͔
͓2͔
Reduction of H2O
2H2O ϩ 2e → H2↑ ϩ 2 OHϪ
The overall reaction can then be described as
Cu S͒ ϩ 2H2O → Cu2ϩ ϩ CuS↑ ϩ H2↓ ϩ 2 OHϪ
*
͑
2
→ Cu͑OH)2↑ ϩ CuS↑ ϩ H2↓
͓3͔
To account for the photocurrent overshooting observed in the case
when KCl was used as the electrolyte, one expects to consider the
existence of the SS, which had been proposed to interpret the anodic
photocurrent transient for CdS nanoparticles.15,16 In this case, the
e͑back͒ may compete with the electron transfer from CB of Cu2S to
the electrolyte, resulting in the occurrence of the transient photocur-
rent ͑see Scheme Ia͒.
Figure 6. I2ph as a function of the bias voltage for a Cu2S/PPy nanowire films
on Cu electrode. ͓0.5 M Na2C2H4O6 electrolyte solution ͑pH 14͒; light
intensity ϭ 38.0 mW cmϪ2.͔ Inset: representative photocurrent traces ob-
tained at Ϫ0.50 V bias voltage.
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
Conclusions
G225
We have demonstrated the PEC of straight and isolated Cu2S
nanowires arrayed on copper foils. The observed cathodic photocur-
rent increases with the increasing negative bias of the film electrode,
indicating that the electrons flow from the Cu electrode through the
nanowire film to the electrolyte, which is consistent with the p-type
semiconducting nature of Cu2S. An interesting photocurrent over-
shooting upon light illumination on and off is detected when KCl
was used as the electrolyte, which is believed to originate from back
electron transfer competing with the electron transfer from CB of
*
(Cu2S) to the electrolyte. This feature is not observed when
Na2SO3 or Na2C2H4O6 are used as electrolytes. PEC studies on the
core/sheath nanowires of Cu2S/NP ͑NP ϭ CdS, PPy, Au͒ indicate
different charge-transfer pathways between the Cu2S core and the
nanoparticle sheath. Cu2S/CdS nanowires exhibit a slightly higher
photocurrent response due to charge transfer between the p-type
Cu2S core and n-type CdS sheath in the p-n heterojunction. The
photocurrent response obtained for Cu2S/PPy nanowires is generally
an order of magnitude smaller than that for pure Cu2S nanowires
under the same bias voltage, presumably due to the misalignment of
the energy bands and/or the poor adhesion of the PPy sheath layer.
For Cu2S/Au nanowires, the photocurrent response is completely
reversed to anodic because Au nanoparticles can readily accept the
photogenerated electrons from the CB of Cu2S and thus redirect the
electron transfer pathway. By transferring the nanowire films onto
transparent conducting substrates such as tin oxide or indium tin
oxide, the absorption spectrum of Cu2S nanowires could be obtained
and the IPCE may be determined. Further optimization of the
present PEC cell may be accomplished by using different substrates,
optimizing the relative thickness of the sheath layer and the Cu2S
core, and utilizing dye sensitization.
Schematic representation of the electron transfer processes for the photocur-
rent generation. ͑a͒ Cu2S ͑solid lines͒ or Cu2S/CdS core/sheath ͑dotted lines͒
nanowires. The dashed line denotes e͑back͒, which exists only when KCl is
used as electrolyte. ͑b͒ Cu2S/Au core/sheath nanowires. The dashed line
denotes e͑back͒, which competes with the anodic photocurrent generation
process. (EF : Fermi level; CB: conduction band; VB: valence band; SS:
surface states.͒
Acknowledgments
This work was supported by an RGC grant administered by the
UGC of Hong Kong.
References
In the existence of the CdS sheath layer, because the bandgap of
CdS is much larger than that of Cu2S by neglecting the slight de-
viations at nanosize2 and the thickness of CdS is smaller in the
Cu2S/CdS nanowires, it is expected that only Cu2S is primarily ex-
cited. After photoexcitation of Cu2S, the electron transfer process
from Cu2S to the electrolyte may be mediated via the conduction
band of CdS owing to the characteristic of p-n heterojunction ͑the
dotted lines in Scheme Ia͒, contributing to the enhanced photocur-
rent response for Cu2S/CdS nanowires. On the other hand, for the
Cu2S/PPy nanowires, PPy may just play the role of an electron
transport layer instead of mediator like CdS due to the possible
reasons stated above.
1. K. L. Chopra and S. R. Das, Thin Film Solar Cells, Plenum, New York ͑1983͒.
2. F. Pfisterer, Thin Solid Films, 431, 470 ͑2003͒.
3. A. Goetzberger, C. Hebling, and H.-W. Schock, Mater. Sci. Eng., R., 40, 1 ͑2003͒.
4. L. Reijner, B. Meester, A. Goossens, and J. Schoonman, Chem. Vap. Deposition, 9,
15 ͑2003͒.
5. M. Rusop, T. Shirata, P. M. Sirimanne, T. Soga, and T. Jimbo, Jpn. J. Appl. Phys.,
Part 1, 42, 4966 ͑2003͒.
6. M. Gratzel, Nature (London), 414, 338 ͑2001͒.
¨
7. A. Aruchamy, Photoelectrochemistry and Photovoltaics of Layered Semiconduc-
tors, Kluwer Academic, Dordrecht ͑1992͒.
8. A. Hagfeldt and M. Gratzel, Acc. Chem. Res., 33, 269 ͑2000͒.
¨
9. D. A. Tryk, A. Fujishima, and K. Honda, Electrochim. Acta, 45, 2363 ͑2000͒.
10. I. Bedja, S. Hotchandani, and P. V. Kamat, J. Phys. Chem., 98, 4133 ͑1994͒.
11. N. Chandrasekharan and P. V. Kamat, J. Phys. Chem. B, 104, 10851 ͑2000͒.
12. F. Cao, G. Oskam, G. J. Meyer, and P. C. Searson, J. Phys. Chem., 100, 17027
͑1996͒.
13. S. Hotchandani and P. V. Kamat, J. Electrochem. Soc., 139, 1630 ͑1992͒.
14. K. H. Yoon, W. J. Choi, and D. H. Kang, Thin Solid Films, 372, 250 ͑2000͒.
15. S. G. Hickey and D. Riley, J. Phys. Chem. B, 103, 4599 ͑1999͒.
16. S. G. Hickey, D. J. Riley, and E. J. Tull, J. Phys. Chem. B, 104, 7623 ͑2000͒.
17. E. P. A. M. Bakkers, E. Reitsma, J. J. Kelly, and D. Vanmaekelbergh, J. Phys.
Chem. B, 103, 2781 ͑1999͒.
After Au nanoparticles were deposited on the Cu2S nanowires,
however, such an interfacial charge-transfer process may experience
a distinct change.11,35,37 Here Au nanoparticles can readily accept
the photogenerated electron from the CB of Cu2S, and thus redirect
the electron transfer pathway, resulting in the anodic photocurrent
response. In fact, besides the coating on the top of the Cu2S core
layer, Au sheath layers may also coat the surfaces of the individual
Cu2S nanowire ͑see Fig. 1c͒, this would accomplish the electron
transfer from Au to the Cu electrode ͑see Scheme Ib͒. The possible
reactions involved could be stated as below
18. T. Torimoto, A. Obayashi, S. Kuwabata, H. Yasuda, H. Mori, and H. Yoneyama,
Langmuir, 16, 5820 ͑2000͒.
19. B. Alperson, H. Demange, I. Rubinstein, and G. Hodes, J. Phys. Chem. B, 103,
4943 ͑1999͒.
20. C. Dekker, Phys. Today, 52, 22 ͑1999͒.
21. S. J. Tans, R. M. Vershueren, and C. Dekker, Nature (London), 393, 49 ͑1998͒.
22. M. Bockrath, D. H. Cobden, P. L. McEuen, N. G. Chopra, A. Zettl, A. Thess, and
R. E. Smalley, Science, 275, 1922 ͑1997͒.
23. J. Hu, M. Ouyang, P. Yang, and C. M. Lieber, Nature (London), 399, 48 ͑1999͒.
24. X. Duan, Y. Huang, Y. Cui, J. Wang, and C. M. Lieber, Nature (London), 409, 66
͑2001͒.
25. N. Beermann, L. Vayssieres, S.-E. Lindquist, and A. Hagfeldt, J. Electrochem. Soc.,
147, 2456 ͑2000͒.
26. S. U. M. Khan and T. S. Sultana, Sol. Energy Mater. Sol. Cells, 76, 211 ͑2003͒.
27. S. H. Wang and S. H. Yang, Chem. Mater., 13, 4794 ͑2001͒.
28. W. X. Zhang, X. G. Wen, and S. H. Yang, Langmuir, 19, 4420 ͑2003͒.
Oxidation of OHϪ
O2 ϩ 2H2O ϩ 4e → 4 OHϪ
͓4͔
*
Reduction of (Cu2S)
2Ϫ
*
Cu S͒ ϩ 2e → 2Cu↑ ϩ S
͓5͔
͑
2
2 Cu S͒ ϩ 4 OH → 4Cu↑ ϩ 2S2Ϫ ϩ O2↓ ϩ 2H2O ͓6͔
Ϫ
*
͑
2
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G226
Journal of The Electrochemical Society, 152 ͑3͒ G220-G226 ͑2005͒
29. X. G. Wen and S. H. Yang, Nano. Lett., 2, 451 ͑2002͒.
30. S. F. Yang, L. Z. Fan, and S. H. Yang, J. Phys. Chem. B, 108, 4394 ͑2004͒.
31. H. O. Finklea, Semiconductor Electrodes, Elsevier, New York ͑1988͒.
32. S. Licht, Semiconductor Electrodes and Photoelectrochemistry, Wiley-VCH, Wein-
heim ͑2002͒.
34. S. Ogawa, F. F. Fan, and A. J. Bard, J. Phys. Chem., 99, 11182 ͑1995͒.
35. M. Jakob, H. Levanon, and P. V. Kamat, Nano. Lett., 3, 353 ͑2003͒.
36. C. Zhao, H. Wang, and Z. Jiang, Appl. Surf. Sci., 207, 6 ͑2003͒.
37. P. V. Kamat, J. Phys. Chem. B, 106, 7729 ͑2002͒.
¨
38. M. Lahav, V. Heleg-Shabtai, J. Wasserman, E. Katz, I. Willner, H. Durr, Y.-Z. Hu,
33. T. Inoue, T. Watanable, A. Fujishima, K. Honda, and K. Kohayakawa, J. Electro-
chem. Soc., 124, 719 ͑1977͒.
and S. Bossmann, J. Am. Chem. Soc., 122, 11480 ͑2000͒.
Downloaded on 2015-01-28 to IP 169.230.243.252 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).