6408 J. Phys. Chem.

C 2010, 114, 6408–6412

A Simple Two-Step Electrodeposition of Cu2O/ZnO Nanopillar Solar Cells

Jingbiao Cui*
Department of Physics and Astronomy, UniVersity of Arkansas at Little Rock, Little Rock, Arkansas 72204

Ursula J. Gibson
Thayer School of Engineering, Dartmouth College, HanoVer, New Hampshire 03755
ReceiVed: January 15, 2010; ReVised Manuscript ReceiVed: February 23, 2010

Nanopillar heterojunctions composed of n-type ZnO nanowire arrays embedded in p-type Cu2O thin films
were fabricated by using a simple two-step electrodeposition method. Structural, optical, and electrical properties
of the nanopillar junctions were investigated. Nanopillar radial junction arrays have the potential for improved
performance in solar cells due to increased junction area and improved charge carrier collection. Improved
efficiency in the Cu2O/ZnO nanopillar junctions was experimentally observed compared to planar thin film
junctions prepared under similar conditions. This study demonstrates that electrodeposition, which is easily
adapted to other chemical systems, is a promising technique for large-scale fabrication of low-cost nanopillar
solar cells.

1. Introduction Although Cu2O/ZnO thin film solar cells have been exten-
sively studied in recent years,14-18 to our knowledge, true radial
Solar energy is considered the cleanest and least limited nanopillar cells have not been reported so far in this system.
energy source despite the dominance of fossil fuels through the Hsueh et al. reported on deposition of Cu2O onto ZnO nanowires
last century. Silicon-based solar cells currently dominate the for solar cells.19 Their method resulted in axial rather than radial
solar energy market due to well-developed fabrication techniques junctions because only the tops of the ZnO nanowires were in
and relatively high energy conversion efficiencies.1,2 To further contact with the Cu2O thin films, which were deposited by
improve the energy conversion efficiency and lower the fabrica- sputtering. In this work, we report on an economic approach to
tion cost, solar cells based on other inorganic and organic fabricate ZnO nanopillars embedded in Cu2O thin films using
materials have been extensively investigated in the last 20 simple two-step electrodeposition. This new approach is a low-
years.3-6 The development of nanowire arrays in the past decade temperature, low-cost, and template-free process with potential
has opened the door for fabricating highly efficient nanopillar for large-scale production, and these materials have reduced
solar cells. Several device structures have been proposed, environmental hazards compared to Cd-based cells. The Cu2O
including nanopillar arrays with axial and radial junctions (core/ thin films with an optimal band gap of 2.1 eV and theoretical
shell structures)7-9 and nanopillar collectors embedded in energy conversion efficiency of 20% (ref 20) absorb the light
absorbing thin films.10 All-inorganic nanopillar arrays with and the ZnO nanopillars act as charge collectors. Planar
various semiconductors such as Si, GaN, CdS, and GaAs have junctions comprised of Cu2O and ZnO thin films were prepared
been experimentally demonstrated. An overall efficiency of under similar conditions for comparison. The radial nanopillar
about 3.4% has been reported in Si nanopillar radial junctions.11 junctions showed improved energy conversion efficiency which
These novel device structures are potentially useful for advanced was attributed to the increased junction area. While much effort
solar cell applications. is needed to optimize device overall performance of Cu2O/ZnO
Use of inorganic nanopillars as charge collectors and organic nanopillar junctions, we demonstrate that electrodeposition is
materials as absorbers has been widely studied in dye-sensitized suitable for fabricating low-cost embedded nanopillar solar cells.
solar cells.5,12,13 However, there are very few reports on
embedded nanowires in inorganic thin films for solar cells, likely 2. Experimental Details
due to the complicated fabrication process. Very recently, 2.1. Growth of ZnO. Both ZnO planar thin films and
nanopillar solar cells with efficiencies up to 6% were demon- nanopillar arrays were prepared by using a low-temperature
strated with CdS nanowire arrays embedded in CdTe thin electrodeposition process.21,22 Commercially available indium
films.10 Fan and colleagues used CdS nanowires with exposed tin oxide (ITO) coated glass substrates with sheet resistance of
length comparable to the distance between nanowires, making 10 Ω were purchased from Structure Probe, Inc. (SPI Supplies).
it possible to deposit CdTe into the gaps by vapor deposition.10 The substrates were sonicated in acetone, rinsed in alcohol and
However, longer nanowires with small separations are desirable deionized water, and blown dry with nitrogen gas. ZnO thin
for fabricating embedded nanopillar solar cells due to their larger films were directly electrodeposited on the substrates, using
junction area and high efficiency. Electrodepositon represents 0.025 M zinc nitrate aqueous solution at a temperature of
a promising technique for fabricating embedded nanopillar solar 70 °C. An electric potential of -2.5 V was applied to the
cells. substrates relative to a gold counterelectrode placed in the
solution, which resulted in a growth rate of 2.5 µm per hour.
* To whom correspondence should be addressed. Phone: (001) 501-569 The detailed growth conditions for ZnO nanopillars can be
8962. Fax: (001) 501-569 3314. E-mail: [email protected]. found in the literature.22 Briefly, ZnO nanopillar arrays were
10.1021/jp1004314  2010 American Chemical Society
Published on Web 03/10/2010
Electrodeposition of Cu2O/ZnO Nanopillar Solar Cells J. Phys. Chem. C, Vol. 114, No. 14, 2010 6409

Figure 1. Experimental procedures for the fabrication of Cu2O/ZnO

nanopillar heterostructures.

electrodeposited in an equimolar aqueous solution of 0.0063

M zinc nitrate and hexamine. The growth was performed at
95 °C with an applied potential of -2.5 V. Nanopillars with
lengths from a few hundred nanometers to several micrometers Figure 2. SEM images of the surface of a ZnO thin film (a), a Cu2O
were obtained, depending on the growth time. Growth of ZnO thin film (b), and the cross section of a Cu2O/ZnO planar junction (c).
nanopillars directly on ITO glass showed a relatively low
nucleation density. To enhance the nucleation, a 50 to 150 nm
planar layer of ZnO was deposited on top of the ITO glass by
e-beam evaporation and annealed at 450 °C in air for 60 min
prior to the growth of ZnO nanopillars.
2.2. Growth of Cu2O. polycrystalline Cu2O thin films were
grown at 60 °C by electrodeposition. The source materials were
copper sulfate (4 g) and lactic acid (6 g) dissolved in 300 mL
of deionized water. Concentrated sodium hydroxide solution was
slowly added to adjust the pH value to 11-12 while stirring. A
potential between -0.7 and -0.9 V was applied to the ZnO/
ITO/glass substrates resulting in a current density of 0.2 mA/
cm2. Pure Cu2O phase with a growth rate of about 30-40 nm/h
was obtained under these growth conditions. Figure 3. Absorption spectra of ITO coated glass, a ZnO thin film on
Figure 1 shows the procedure for growing Cu2O/ZnO ITO/glass, and Cu2O thin films of different thicknesses on ZnO/ITO/
glass.
nanopillar heterojunctions. A ZnO nucleation film is deposited
on ITO glass (Figure 1a) and then ZnO nanopillars are grown
by electrodeposition (Figure 1b). To avoid the etching of ZnO that Cu2O/ZnO planar heterojunctions can be fabricated by the
nanopillars in the Cu2O growth solution, a negative potential is two-step electrodeposition process, consistent with previous
applied to the ZnO nanopillar substrate prior to immersion in publications.15,16 Our materials had somewhat different electrical
the copper salt bath. It was found that the potential applied for properties, as described below.
Cu2O thin film growth ensures little etching of ZnO nanopillars Optical absorption measurements, to identify band edges,
even at times greater than 30 h. The Cu2O/ZnO nanopillar were performed on Cu2O/ZnO planar junctions rather than
heterojunctions are complete when the Cu2O thin film becomes nanopillar structures due to optical scattering from the nanopillar
continuous and embeds the nanopillars (Figure 1c). arrays. Figure 3 shows the absorption spectra of ITO glass, a
2.3. Material Characterization. The structure, morphology, ZnO thin film, and a Cu2O/ZnO bilayer. The ITO glass has little
and composition of the samples were characterized by scanning absorption from near-UV to the near-infrared region. The
electron microscopy (SEM), energy dispersive X-ray spectros- deposition of 500 nm ZnO onto ITO results in an obvious
copy (EDX), and X-ray diffraction (XRD). UV-vis absorption absorption below 380 nm, which is due to the band edge of
spectra of thin film samples from 300 to 900 nm were measured ZnO. No absorption features were seen in the visible region.
by an Ocean Optics USB4000 spectrometer. Circular Au contact However, the absorption spectrum was significantly altered after
pads with diameters from 1 to 5 mm were thermally evaporated subsequent deposition of Cu2O for 3 h. The appearance of an
onto the top of Cu2O/ZnO structures and a portion of the absorption band around 550 nm is attributed to the formation
exposed ITO substrate for electrical measurements. Electrical of nanocrystalline Cu2O. The short growth time resulted in the
properties and photovoltaic effects were measured with use of formation of isolated Cu2O nanocrystals as confirmed by SEM
a computer controlled Keithley 2400 source-measure unit. (data not shown). A continuous Cu2O film of 500 nm thickness
During photovoltaic characterization, a power meter was used was obtained after growth for 14 h, resulting in an absorption
to monitor the intensity of the white light source. edge around 600 nm. The band gap of Cu2O is 2.1 eV (590
nm).23 Therefore the strong absorption below 600 nm is
3. Results and Discussion attributed to band edge absorption of the Cu2O thin film.
3.1. Cu2O/ZnO Planar Heterojunctions. To establish a 3.2. Cu2O/ZnO Nanopillar Heterojunctions. Figure 4
baseline, Cu2O/ZnO planar heterojunctions were first deposited shows SEM images of ZnO nanopillar arrays and the electrode-
by using the electrodeposition process. Figure 2 shows the top posited Cu2O/ZnO nanopillar heterostructures. Top-down and
view of a ZnO thin film (Figure 2a) and a Cu2O thin film (Figure cross-section views shown in Figure 4a,b indicate that the
2b), and a cross section of Cu2O/ZnO bilayers (Figure 2c). Both average length and diameter of the ZnO nanopillars are
Cu2O and ZnO thin films have polycrystalline structures. respectively around 2 µm and 100 nm. Subsequent growth of
Individual layers of Cu2O, ZnO, and ITO on the glass substrate Cu2O for 14 h with use of the nanopillar arrays as a substrate
are visible in the cross-section (Figure 2c). The thicknesses of resulted in formation of isolated Cu2O crystals, even though
both Cu2O and ZnO thin films are around 500 nm. We confirm this growth time was sufficient to form a continuous film on a
6410 J. Phys. Chem. C, Vol. 114, No. 14, 2010 Cui and Gibson

Figure 4. SEM images of ZnO nanopillars and Cu2O/ZnO nanopillar Figure 5. EDX spectra of Cu2O/ZnO nanopillar heterostructures: (a)
heterostructures. (a) Top view of ZnO nanopillars; (b) cross section of ZnO nanopillars; (b) Cu2O/ZnO nanopillar heterostructures with Cu2O
ZnO nanopillars; (c, d) Cu2O/ZnO nanopillar heterostructures after growth for 14 h; and (c) Cu2O/ZnO nanopillar heterostructures with
growth of Cu2O for 14 (c) and 28 h (d). Circled areas highlight the Cu2O growth for 28 h.
partially embedded ZnO nanopillars in the Cu2O crystals. (e) Cross
section of the sample shown in panel d. Note that the spaces among
ZnO nanopillars are completely filled with Cu2O. The rectangular region
highlights imprints left when ZnO nanopillars were pulled out during
sample fracture.

planar ZnO film. Nucleation is inhibited on the nanopillar array

substrate, as evidenced by the larger crystallite size, and the
growth rate of Cu2O may be reduced inside the arrays due to
the diffusion-limited availability of source materials. Note that
the Cu2O crystals grew among the nanopillars as highlighted
in the circled areas of Figure 4c. ZnO nanopillars are encased
by the Cu2O crystals, forming nanowire radial junctions.
The surface morphology of the sample after growth of Cu2O
for 28 h is shown in Figure 4d. This top view image indicates Figure 6. X-ray diffraction patterns of ZnO nanopillars (a) and Cu2O/
that a continuous film of Cu2O is obtained. To determine ZnO nanopillar heterostructures (b). The stars in part b label the
whether the gaps among ZnO nanopillars were completely filled diffraction peaks of ZnO.
with Cu2O beneath the top layer, a cross-sectional image of the
sample was taken and is shown in Figure 4e. It can be seen
that Cu2O thin films completely fill the spaces among ZnO
nanopillars. Some of the nanopillars were pulled out of the Cu2O
crystals during sample preparation for SEM measurements,
which left nanopillar imprint marks as highlighted in the
rectangular region. However, some of the nanopillars survived
the fracture process and remained partially embedded inside the
Cu2O thin films. These observations indicate unambiguously
that the electrodeposited Cu2O fills the narrow spaces between
the ZnO nanopillars to form Cu2O/ZnO radial nanopillar
heterojunctions. This is an advantage of electrodeposition over
vapor-phase techniques which are typically incapable of filling
the deep interpillar voids. Figure 7. Dark current-voltage (I-V) curve of Cu2O/ZnO nanopillar
The composition of the electrodeposited Cu2O/ZnO nanopillar heterojunctions. The inset depicts the device structure for electrical
heterostructures was analyzed by EDX measurements. The characterization.
atomic ratio of Cu to O in our Cu2O thin films was measured
to be 2.0:1.0, consistent with Cu2O formation. The ratio of Zn patterns taken on both pure ZnO nanopillars and Cu2O/ZnO
to O was found to be 48:52 in the ZnO thin films. Figure 5 nanopillar heterostructures. A very strong ZnO (002) peak was
shows the EDX spectra taken on three typical samples of pure observed in the pure ZnO nanopillars, indicating that oriented
ZnO nanopillars, ZnO nanopillar with Cu2O grown for 14 h, nanopillars were obtained. After deposition of Cu2O for 28 h,
and that with Cu2O grown for 28 h. Only a narrow energy range various diffraction peaks from Cu2O were observed in addition
is shown here in order to see the details of Cu (0.95 keV) and to those from ZnO nanopillars. The lattice constant of the
Zn (0.99 keV) signals. As the growth time of Cu2O increases, electrodeposited Cu2O is consistent with the reported value in
the relative intensity of the Cu signal increases while that of the literature.24
Zn decreases. After growing Cu2O for 14 h, the Zn signal still The electrical properties and photovoltaic effects of the
dominates in the spectrum. After growing Cu2O for 28 h, electrodeposited Cu2O/ZnO nanopillar heterostructures were
however, the Cu signal is the dominant peak in the energy investigated at room temperature. Figure 7 shows a typical dark
window. current-voltage (I-V) curve of the heterostructures. The inset
The growth of Cu2O among ZnO nanopillars was also of Figure 7 shows the Cu2O/ZnO nanopillar device structure
investigated by XRD measurements. Figure 6 shows XRD for electrical characterization. The nonlinear I-V curve is typical
Electrodeposition of Cu2O/ZnO Nanopillar Solar Cells J. Phys. Chem. C, Vol. 114, No. 14, 2010 6411

an increase of nanopillar length implies a longer growth time

for Cu2O in order to fully fill the spaces among nanopillars. (3)
Improved homogeneity in length and separation of ZnO
nanopillars over a large area. This effort will help minimize
leakage current and increase fill factor. Leakage current can be
a serious problem for nanopillar junctions. Uniformity among
individual nanopillar junctions would increase the fill factor of
the whole device.
Although the overall performance of electrodeposited Cu2O/
ZnO nanopillar solar cells is not optimized yet, our study sheds
light on a simple and economic approach to fabricating these
novel nanopillar devices. The advantage of this approach is that
it is simple, low-cost, template-free, and has potential for large-
Figure 8. Current density versus voltage of Cu2O/ZnO nanopillar solar
cells. The light intensity was 100 mW/cm2.
scale production. Importantly, electrodeposition can effectively
fill the deep narrow gaps among nanowires, which is challenging
with use of vapor deposition. This unique feature makes
of p-n junctions. The p-n heterojunctions are formed at the electrodeposition better suited for fabricating nanopillar arrays
interfaces of n-type ZnO nanopillars and p-type Cu2O because embedded in thin films for solar cell applications.
ZnO nanopillars grown without intentional doping generally
show n-type semiconducting properties while Cu2O grows as a 4. Summary
p-type semiconductor.
Photovoltaic effects of Cu2O/ZnO nanopillar heterostructures Nanopillar hereojunctions are successfully fabricated by
were measured during illumination through the ITO glass embedding ZnO nanowires in Cu2O thin films with use of a
substrate. Data from Cu2O/ZnO planar thin film junctions are simple two-step electrodepositon. The Cu2O with a direct band
included in Figure 8 for comparison. The light intensity was gap of 2.1 eV acts as a solar absorber while ZnO nanopillars
maintained at 100 mW/cm2 for both measurements. Typical solar are charge carrier collectors. Compared with planar solar cell
cell behavior was observed in the I-V curves. The short circuit structure prepared under similar conditions, an improvement in
current density of the nanopillar junctions was 8.2 mA/cm2, energy conversion efficiency was observed in the nanopillar
which is almost two times larger than that obtained from the solar cells. This improved performance is attributed to the
planar junctions (4.3 mA/cm2). However, the open circuit increased p-n junction area in the latter case. Although much
voltage was 0.29 V, somewhat less than the 0.33 V observed effort is still needed to get a better understanding and to further
for the thin film junctions. Since the open circuit voltage depends improve the overall device performance, our study demonstrates
on the band alignment of p- and n-type semiconductors, the the promise of this process for fabricating nanopillar solar cells.
different values observed suggest a shift associated with the Electrodeposition can be easily adapted to other chemical
ZnO nanowire configuration, possibly due to the dominance of systems, and thus may have an important role in nanopillar solar
different exposed crystal faces in the nanowires. The increased cell development.
short circuit current of nanopillar junctions is likely due to the
increased p-n junction area in the nanopillar structures. The References and Notes
fill factor (maximum output power divided by open circuit
(1) Schropp, R. E.; Zeman, M. In Amorphous and microcrystalline
voltage and short circuit current) is similar for both devices, silicon solar cells: modeling, materials, and deVice technology; Kluwer
i.e., 0.36 for the nanopillar junctions and 0.38 for the planar Academic Publishers: Norwell, MA, 1998; p 107.
junctions. Due to the large short circuit current of the nanopillar (2) Emery, A. G. K.; Hishikawa, Y.; Warta, W. Prog. PhotoVoltaics:
Res. Appl. 2008, 16, 61–67.
cell, its energy conversion efficiency was found to be 0.88%, (3) Grätzel, M. J. Photochem. Photobiol. C 2003, 4, 145–153.
as compared to 0.55% for the planar cell. (4) Olson, D. C.; Piris, J.; Collins, R. T.; Shaheen, S. E.; Ginley, D. S.
Open circuit voltages up to 0.595 V have been reported for Thin Solid Films 2006, 496, 26–29.
Cu2O/ZnO planar junctions,18 significantly higher than observed (5) Greene, L. E.; Law, M.; Yuhas, B. D.; Yang, P. D. J. Phys. Chem.
B 2007, 111, 18451–18456.
in this study. Mittiga and colleagues used high temperature (6) Pasquier, A. D.; Mastrogiovanni, D. D. T.; Klein, L. A.; Wang,
grown Cu2O doped with Cl for conductivity enhancement.18 This T.; Garfunkel, E. Appl. Phys. Lett. 2007, 91, 183501.
intentional doping helped improve the electrical properties of (7) Kayes, B. M.; Atwater, H. A.; Lewis, N. S. J. Appl. Phys. 2005,
97, 114302.
Cu2O films, which result in a large open circuit voltage in the (8) Czaban, J. A.; Thompson, D. A.; LaPierre, R. R. Nano Lett. 2009,
corresponding solar cells. The relatively small open circuit 9, 148–154.
voltage in our nanopillar junctions may result from the low (9) Garnett, E. C.; Yang, P. D. J. Am. Chem. Soc. 2008, 130, 9224.
dopant concentration in Cu2O since no intentional doping was (10) Fan, Z. Y.; Razavi, H.; Do, J. W.; Moriwaki, A.; Ergen, O.; Chueh,
Y. L.; Leu, P. W.; Ho, J. C.; Takahashi, T.; Reichertz, L. A.; Neale, S.;
performed. Other reasons may include the quality of the junction Yu, K.; Wu, M.; Ager, J. W.; Javey, A. Nat. Mater. 2009, 8, 648–653.
interfaces, which have not yet been optimized. Improvement (11) Tian, B. Z.; Zheng, X. L.; Kempa, T. J.; Fang, Y.; Yu, N. F.; Yu,
of the open circuit voltage may be achieved by alterations in G. H.; Huang, J. L.; Lieber, C. M. Nature 2007, 885, 889.
(12) Law, M.; Greene, L. E.; Johnson, J. C.; Saykally, R.; Yang, P. D.
the growth conditions as well as surface treatment of ZnO prior Nat. Mater. 2005, 4, 455–459.
to Cu2O growth. (13) Diamant, Y.; Chappel, S.; Chen, S. G.; Melamed, O.; Zaban, A.
Much effort is still needed to increase the energy conversion Coord. Chem. ReV. 2004, 248, 1271–1276.
(14) Tanaka, H.; Shimakawa, T.; Miyata, T.; Sato, H.; Minami, T. Thin
efficiency of Cu2O/ZnO nanopillar solar cells. It may include Solid Films 2004, 469-470, 80–85.
the following: (1) Improvement of material quality. Both Cu2O (15) Jeong, S. S.; Mittiga, A.; Salza, E.; Masci, A.; Passerini, S.
and ZnO are naturally grown semiconductors; however, inten- Electrochim. Acta 2008, 53, 2226–2231.
tional doping will help improve their electrical properties for (16) Katayama, J.; Ito, K.; Matsuoka, M.; Tamaki, J. J. Appl. Electro-
chem. 2004, 34, 687–692.
charge collection. (2) Increases in p-n junction area by (17) Izaki, M.; Mizuno, K.; Shinagawa, T.; Inaba, M.; Tasaka, A. J.
increasing the density or length of ZnO nanopillars. Note that Electrochem. Soc. 2006, 153, C668–C672.
6412 J. Phys. Chem. C, Vol. 114, No. 14, 2010 Cui and Gibson

(18) Mittiga, A.; Salza, E.; Sarto, F.; Tucci, M.; Vasanthi, R. Appl. Phys. (22) Cui, J. B.; Gibson, U. J. J. Phys. Chem. B 2005, 109, 22074.
Lett. 2006, 88, 163502. (23) Jones, P. M.; May, J. A.; Reitz, J. B.; Solomon, E. I. J. Am. Chem.
(19) Hsueh, T. J.; Hsu, C. L.; Chang, S. J.; Guo, P. W.; Hsieh, J. H.; Soc. 1998, 120, 1506.
Chend, I. C. Scr. Mater. 2007, 57, 53–56. (24) Poizot, P.; Hung, C. J.; Nikiforov, M. P.; Bohannan, E. W.; Switzer,
(20) Tanaka, H.; Shimakawa, T.; Miyata, T.; Sato, H.; Minami, T. Appl. J. A. Electrochem. Solid-State Lett. 2003, 6, C21–C25.
Surf. Sci. 2005, 244, 568.
(21) Izaki, M.; Omi, T. J. Electrochem. Soc. 1997, 144, 1949–1952. JP1004314