Materials Chemistry and Physics 147 (2014) 987e995

Contents lists available at ScienceDirect

Materials Chemistry and Physics

journal homepage: www.elsevier.com/locate/matchemphys

Influence of thermal annealing on microstructural, morphological,

optical properties and surface electronic structure of copper oxide thin
films
Funda Aksoy Akgul a, f, **, Guvenc Akgul b, f, *, Nurcan Yildirim c, d, Husnu Emrah Unalan d, f,
Rasit Turan e, f
a
Department of Physics, Nigde University, 51240 Nigde, Turkey
b
Bor Vocational School, Nigde University, 51700 Nigde, Turkey
c
Department of Physics Engineering, Ankara University, 06100 Ankara, Turkey
d
Department of Metallurgical and Materials Engineering, Middle East Technical University, 06800 Ankara, Turkey
e
Department of Physics, Middle East Technical University, 06800 Ankara, Turkey
f
Center for Solar Energy Research and Applications, Middle East Technical University, 06800 Ankara, Turkey

h i g h l i g h t s g r a p h i c a l a b s t r a c t

Effect of post-deposition annealing

on copper oxide thin films was
investigated.

Structural, optical, and electronic
properties of the thin films were
determined.

Oxidation states of copper oxide thin
films were confirmed by XPS
analysis.

Mixed phases of CuO and Cu2O were
found to coexist in copper oxide thin
films.

a r t i c l e i n f o a b s t r a c t

Article history: In this study, effect of the post-deposition thermal annealing on copper oxide thin films has been sys-
Received 7 March 2013 temically investigated. The copper oxide thin films were chemically deposited on glass substrates by
Received in revised form spin-coating. Samples were annealed in air at atmospheric pressure and at different temperatures
27 May 2014
ranging from 200 to 600 C. The microstructural, morphological, optical properties and surface electronic
Accepted 19 June 2014
Available online 8 July 2014
structure of the thin films have been studied by diagnostic techniques such as X-ray diffraction (XRD),
Raman spectroscopy, ultravioletevisible (UVeVIS) absorption spectroscopy, field emission scanning
electron microscopy (FESEM), atomic force microscopy (AFM), and X-ray photoelectron spectroscopy
Keywords:
A. Thin films
(XPS). The thickness of the films was about 520 nm. Crystallinity and grain size was found to improve
B. Sol-gel growth with annealing temperature. The optical bandgap of the samples was found to be in between 1.93 and
C. X-ray photo-emission spectroscopy (XPS) 2.08 eV. Cupric oxide (CuO), cuprous oxide (Cu2O) and copper hydroxide (Cu(OH)2) phases were
D. Electronic structure observed on the surface of as-deposited and 600  C annealed thin films and relative concentrations of
these three phases were found to depend on annealing temperature. A complete characterization

* Corresponding author. Bor Vocational School, Nigde University, 51700 Nigde,

Turkey. Tel.: þ90 (388) 311 45 27; fax: þ90 (388) 311 84 37.
** Corresponding author. Department of Physics, Nigde University, 51240 Nigde,
Turkey. Tel.: þ90 (388) 225 42 17; fax: þ90 (388) 225 01 80.
E-mail addresses: fundaaksoy01@gmail.com (F.A. Akgul), guvencakgul@gmail.
com (G. Akgul).

http://dx.doi.org/10.1016/j.matchemphys.2014.06.047
0254-0584/© 2014 Elsevier B.V. All rights reserved.
988 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
reported herein allowed us to better understand the surface properties of copper oxide thin films which
could then be used as active layers in optoelectronic devices such as solar cells and photodetectors.
1. Introduction
CuO (tenorite) and Cu2O (cuprite) are the two stable oxide forms
of copper. CuO has a monoclinic crystal structure and a bandgap of
1.2e1.9 eV [1,2]. This metal oxide has black color and high ab-
sorptivity and low thermal emittance [3]. On the other hand, Cu2O
with a brownish-red color has cubic crystal structure and a
bandgap of 2.0e2.2 eV [2]. Both of these oxides are semiconducting
in nature and exhibit p-type characteristic due to copper vacancies
in the structure. The advantages of copper oxides include their non-
toxic nature, abundance of their precursor materials, and cost-
effective synthesis routes. CuO attracted significant interest due
to its great potential for hydrogen (H2), carbon dioxide (CO2), car-
bon monoxide (CO) and nitrogen oxide (NO2) gas sensor applica-
tions [4e7]. Its high-temperature superconductivity properties
have also been investigated which is believed to result from the
specific coordination between Cu and O atoms [8]. In addition, this
compound has been employed for the fabrication of glucose sen-
sors [9,10], lithium-ion batteries [11,12], magnetic storage [13], and
catalysts [14]. Cu2O, on the other hand, is one of the first semi-
conductors to be discovered. Cu2O exhibits several attractive
characteristics for optoelectronic devices due to its high absorption
coefficient and reasonably good majority carrier mobility.
The copper oxides have also attracted interest due to their po-
tential applications in solar cell technology. High optical absorp-
tivity in the visible or near infrared region due to its direct bandgap
structure [15,16], non-toxicity and most particularly cost-efficient
production routes could make copper oxides alternative to silicon
which is predominantly used in conventional solar cells. Therefore,
CuO and Cu2O can potentially be used to fabricate thin film solar
cells and photodetectors. In particular, a heterojunction structure
with a complementing n-type semiconductor could lead to high
performance devices [17e19]. In such applications, there are many
factors that can affect the heterojunction formation, the efficiency
and performance of devices. Optical and microstructural properties
of copper oxide thin films are only two of them, which should be
carefully optimized for a decent device performance. It is then
crucial to determine these properties under different processing
conditions. Moreover, post-deposition treatments might be useful
to control and improve the microstructural, optical and electrical
properties of thin films. A detailed study of copper oxide thin films
is then of scientific and technological interest. Gaining a deeper
understanding of the properties of copper oxide films will allow
optimization of the processing conditions required to fabricate
highly efficient optoelectronic devices.
Copper oxide thin films can be prepared using a wide range of
methods including reactive evaporation [20], chemical deposition
[1,2], thermal oxidation [21e23], chemical vapor deposition (CVD)
[3], and solegel [4,24]. The deposition method and conditions
including the post-deposition treatments play an important role in
the determination of the physical and chemical properties of cop-
per oxide thin films. Among these methods, solegel is the most
widely used one due to its low cost, simplicity and controllability. It
is also possible to deposit films over large areas under atmospheric
conditions.
In this paper, we report on the variation of the microstructural,
morphological, optical properties and surface electronic structure
of the copper oxide thin films during the post-deposition heat
© 2014 Elsevier B.V. All rights reserved.
treatment, and showed that the copper oxide thin films with
required chemical compositions and physical properties can be
tailored using post-deposition annealing treatments.
2. Experimental
The solution for the deposition of thin films was prepared by
dissolving copper acetate [Cu(CH3COO)2$H2O, Merck, 99,98%] in
ethanol. Afterwards, lactic acid (C3H6O3) and triethylamine
(C6H15N) was added to the prepared solution. The pH of the ac-
quired solution was approximately 5.5. The thin films were
deposited by spin-coating technique on soda-lime silicate glass
substrates. Prior to deposition, the substrates were ultrasonically
cleaned in consecutive acetone, ethanol, and de-ionized water
(18.3 MU) baths for 15 min. Then, they were dried with nitrogen
gas. The spin-coating was applied at 2000 rpm. At the end of a
coating cycle, thin films were dried on a hot plate at 200  C for
2 min in air under ambient conditions. Relative humidity of air
environment was less than 30%. To avoid cracking in thin films and
inhomogeneous coverage, multiple coatings were made with dry-
ing in between each successive coating step. Fifteen coatings were
carried out and a film thickness of about 520 nm was obtained.
Finally, as-deposited films were annealed in a muffle furnace under
atmospheric pressure conditions in air at five different tempera-
tures (200, 300, 400, 500 and 600  C) for 30 min to investigate the
effect of annealing temperature on the microstructural, morpho-
logical, optical properties and surface electronic structure of the
thin films. First, the furnace temperature was elevated to post-
deposition annealing temperature with a heating rate of
20  C min1. Once the temperature reached the desired value, thin
film samples were placed inside the furnace. Although the
annealing temperature was set to the indicated value, due to our
furnace, temperature dropped to much lower levels (depending on
time of loading, sometimes to low 200  C's). It then takes some time
to get back to the indicated temperature. The samples annealed for
30 min after reaching the set temperature. Then the furnace was
turned off and thin film samples were allowed to cool down to
room temperature naturally to prevent film cracking by thermal
stress. An as-deposited sample was used as a reference.
In order to determine the crystal phase(s) of the deposited thin
films, a Rigaku miniflex XRD system equipped with Cu Ka radiation
(l ¼ 0.154 nm) was used. The XRD patterns were recorded in the
range of 10e80 with a scan speed of 2 /min. The surface
morphology and roughness of the thin films were characterized by
FESEM (Nova NanoSEM 430-operated at 10 kV) and AFM (Vecoo
MultiMode V), respectively. Optical properties of the samples have
been analyzed using an optical setup consisting of an 8-inch inte-
grating sphere with 1-inch ports (Newport, 70679NS), a mono-
chromator (Newport, 74100), a UV-enhanced Si photodiode
detector (Newport, 70356), and a collimated beam from a 100 W
halogen lamp. The samples were mounted at the front side of the
integrating sphere for transmittance measurements and at the back
side of the integrating sphere for reflectance measurements. The
wavelength range for both the reflectance and transmittance
measurements was varied between 400 and 1100 nm. The surface
electronic structure and chemical states of the atomic species in the
thin films were examined via XPS using a monochromatic Al Ka X-
ray source (1486.6 eV) (PHI 5000 VersaProbe). The XPS
 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
measurements were carried out in vacuum at a base pressure of
around 106 Torr. The binding energies were calibrated to the C 1s
line of residual carbon present on the surface of the thin films,
positioned at 284.5 eV (taken from NIST database) [25]. The XPS
data analysis was performed using the XPSPEAK41 software. In the
peak fitting, the Shirley background due to inelastically scattered
electrons was subtracted and then the Cu 2p3/2 peak with shake-up
satellite structure was fitted. Energy separation between different
chemical states of copper was kept constant during the fitting
procedure. All fitted peaks were shared same GaussianeLorentzian
mixing ratio. The integrated areas of the deconvoluted peaks were
used to determine the relative concentrations of copper species.
The microstructure of the copper oxide thin films was also inves-
tigated by Raman scattering using 532 nm line of Nd-YAG laser as
the excitation source. All measurements were carried out at room
temperature.
3. Results and discussion
Structural properties of the copper oxide thin films were char-
acterized by XRD. Fig. 1 illustrates XRD patterns of the as-deposited
and annealed thin films. From XRD, it is clear that all samples have
polycrystalline structure. The XRD traces exhibited two distinct
peaks of nearly equal intensities at diffraction angles of 35.6 and
38.7. These peaks were ascribed to the (111)/(002) and (111)/
(200) Bragg reflections indicating CuO was in the tenorite phase
(JCPDS card no. 45-0937) [2,26e29]. It can be seen that the
annealing temperature affects the XRD pattern of copper oxide
films i.e. the intensities of the observed peaks hence crystallinity of
the thin films increases with the annealing temperature. No
obvious peak shift, formation of Cu2O phase or any noticeable im-
purities were detected with an increase in the annealing temper-
ature. It was also clear that different Bragg reflections start to
appear at 400 C and evolve with annealing temperature. Therefore,
CuO thin films became notably crystalline with increasing anneal-
ing temperature. The formation of CuO phase in the as-deposited
thin film depends on the deposition techniques and conditions. It
has been reported that copper oxide films deposited by CVD
method onto a heated substrate (300  C) contain mixed phases of
CuO and Cu2O [30]. On the other hand, copper oxide films depos-
ited by ion beam sputtering on a substrate heated to 200  C were
found to constitute of only CuO phase [31]. Many other studies
Fig. 1. XRD patterns of copper oxide thin films annealed at different temperatures.
989
showed that the deposition of copper oxide thin films at room
temperature revealed films with the Cu2O phase; but, subsequent
annealing process transformed them to CuO phase at higher sub-
strate temperatures [15,32]. In the present study, formation of the
CuO phase in the as-deposited film can be explained in terms of
equilibrium phase diagram of the CueO system showing CuO as a
stable phase for temperatures below 300  C regardless of the par-
tial pressure of oxygen [15]. From thermodynamic considerations,
Gibbs free energy for the oxidation of Cu2O to CuO at a temperature
of 200  C is 3.73 kcal/mol. Therefore, the formation of CuO phase
in the as-deposited copper oxide thin film is reasonable and no
Cu2O phase is observed in this film as thermodynamically expected
[33]. Furthermore, annealing the films in air at higher temperatures
(300e600  C) did not change the phase composition. Once the CuO
phase is formed, it is insensitive to annealing. Thus, the post-
deposition annealing treatment only affected the crystal quality,
not the crystal phase in the copper oxide thin films deposited by
spin-coating.
Analysis of X-ray line profile is a simple and powerful method to
estimate crystallite size and internal lattice strain [34]. Among the
available methods, Williamson-Hall analysis is a simplified method
where both size and strain-induced broadening is deconvoluted by
considering the peak width as a function of 2q [35]. In the present
study, Williamson-Hall method was used for estimating average
crystallite size and effective lattice strain. Using the line profiles in
XRD spectra, crystallite sizes and the microstructural parameters
such as microstrain (ε) and dislocation density (d) can be deter-
mined from the full-width at half-maximum (FWHM) of the
diffraction peaks. The FWHM can be expressed as a linear combi-
nation of the contributions from the lattice strain and crystallite
size as [36]:
l
b cos q ¼ þ ε sin q (1)
D
where b is the FWHM of the diffraction peak measured in radians, q
is the diffraction angle, l is X-ray wavelength, D is crystallite size,
and ε is the effective lattice strain. In Fig. 2(a), bcosq/l is plotted
versus sinq/l for the as-deposited and annealed copper oxide thin
films. The crystallite size can be estimated from the linear fit to the
data shown in Fig. 2(a). The inverse of the y-intercept on the bcosq/l
axis gives a mean value of the crystallite size. The slope of the fitted
line indicates the presence of effective microstrain in the crystal
lattice of the copper oxide thin films.
The dislocation density (d) is defined as the length of dislocation
lines per unit volume of the crystal and it can be evaluated from the
crystallite size through [37]:
n
d¼ (2)
D2
where D is the crystallite size and n is a factor that equals unity
when the dislocation density is minimum.
The calculated microstructural parameters and average crys-
tallite sizes of the as-deposited and annealed thin films are listed in
Table 1. The variation of crystallite size and microstrain with
annealing temperature is shown in Fig. 2(b). As shown in Fig. 2(b),
calculated average crystallite size increased from 8.32 nm to
30.16 nm as annealing temperature increased. These relatively
small crystallite sizes may be due to the low growth rate of the CuO
phase [15]. The largest crystallite size was observed at an annealing
temperature of 600 C. This can be attributed to the thermal energy
produced by annealing which results in the formation of larger
crystallites. It is obvious that the crystallite size can be controlled
simply by varying the post-deposition annealing temperature. It
was also estimated that the values of the microstrain and
990 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995

Fig. 2. (a) WilliamsoneHall plots of copper oxide thin films annealed at different temperatures and (b) calculated crystallite size and microstrain of the films as a function of
annealing temperature. Lines are for visual aid.

dislocation density reduced from 0.78% to 0.34% and 1.45  1016 to measured Raman spectra from the thin films, three peaks were
0.11  1016 lines/m2 respectively with increasing annealing tem- observed, which were identified as the first order phonon scat-
perature. The lowest value of microstrain was obtained for the thin tering (298, 346 and 632 cm1). These strong peaks appeared at
film annealed at 600  C, as shown in Fig. 2(b). Such a decrease in high annealing temperatures of 500 C and 600 C. They are
microstrain could be owing to the decreased amount of point de- assigned to Ag and 2Bg CuO phonon peaks. The indicated Ag and 2Bg
fects and dislocations along grain boundaries with an increase in peaks shifted slightly to higher wave numbers and became stronger
crystallite size. Since the microstrain and dislocation density are and sharper with annealing temperature. FWHM value of the Ag
indication of dislocation network in thin films, the reduced amount peak changed significantly from 20 to 10 cm1 as shown in the inset
of microstrain and dislocation density shows the formation of high of Fig. 3. These results indicated that significant crystallization
quality CuO films at higher annealing temperatures. A comparative within the CuO thin film occurred at the annealing temperature of
look at microstructural parameters reported in Table 1 also indi- 600C, which was in accordance with obtained XRD results. A broad
cated that the dislocation density was decreased while the crys- hump centered around 570 cm1 was obtained for the as-deposited
tallite size was increased with annealing temperature. This increase and 200e400 C annealed films, which clearly indicated the pres-
of crystallite size can be attributed to a decrease in the amount of ence of Cu2O phase [39]. Thus, based on the Raman analysis, it can
lattice imperfections such as defects and grain boundaries in the be concluded that both CuO and Cu2O phases coexist in the pre-
crystalline structure. From XRD analysis it can be concluded that pared thin films. XRD analysis did not reveal the presence of Cu2O
the reduced microstrain and dislocation density with annealing phase in the as-deposited film and films annealed at 200e400  C.
temperature results in an increase in the crystallite size. This leads This is due to the fact that the Raman spectroscopy is very sensitive
to an improvement in the overall crystal structure. The thermal to both copper oxide phases. In addition, Raman spectroscopy is
energy produced by annealing enhanced diffusion of species and more surface sensitive than XRD and it gives more surface related
formed films with larger crystallites. information. It should also be pointed out that the increased crys-
Raman spectroscopy is a powerful method for the identification tallite size of CuO phase in thin films with annealing temperature
of different oxides through the change in polarizability. The room could have a remarkable effect on the Raman spectra. It is well
temperature Raman scattering experiments on the thin films have known that the intensity of the Raman scattering is directly pro-
been performed within the spectral region of 200e700 cm1. portional to number of scattering centers in the laser beam-
Respective Raman spectra are provided in Fig. 3. There are twelve illuminated volume of the investigated sample. This effect of
optical phonon branches theoretically expected in the Raman crystallite size on the Raman spectra was reported by Xu et al.
spectrum, which are described as 4Au þ 5Bu þ Ag þ 2Bg for CuO. where Raman peaks were shown to be stronger and sharper, and
Three modes represented by Ag þ 2Bg are Raman active [38]. In the shifted to higher wave numbers as grain size increased [38].
Fig. 4((a) and (b)) shows the surface FESEM images of the as-
deposited and 600  C annealed thin films, respectively. Energy
Table 1
dispersive X-ray (EDX) spectrum of the as-deposited and annealed
Microstructural properties of copper oxide thin films.
films can be seen on the right hand side of the SEM images, which
Annealing temperature Crystallite Microstrain Dislocation ascertain the presence of copper and oxygen. The atomic ratio of
( C) size (nm) (ε)  103 density (d)  1016
copper and oxygen obtained from the EDX spectrum was 43:57,
(lin2 m4) (lines/m2)
which could be assigned to CuO phase. It is evident that the surface
As-deposited 8.32 7.81 1.445
morphology is improved through post-deposition annealing.
200 8.96 7.76 1.245
300 9.71 7.71 1.06
Isotropic and highly uniform nano-spherical grains could be
400 13.71 7.35 0.532 observed for the thin films annealed at 600  C, as opposed to films
500 20.01 6.31 0.25 annealed at lower temperatures (not shown here). Besides, nano-
600 30.16 3.41 0.11 spherical grains were incorporated into bigger ones. The
 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995 991

Fig. 3. Raman spectra of copper oxide thin films annealed at different temperatures. Inset shows FWHM of Ag peak as a function of annealing temperature. Lines are for visual aid.

thickness of the copper oxide thin films were measured from cross-
sectional SEM images and a relatively uniform thickness of 520 nm
was found for all samples (Fig. 4(a) inset), independent of the
annealing temperature. Good adherence of the deposited thin films
was evident from the cross-sectional SEM images. The average
grain size for the 600  C annealed thin films was estimated to be
50 nm. It was rather difficult to determine the grain size of the as-
deposited thin film due to the quite fine and densely packed grains.
The surface morphology and roughness of the thin films were
examined by AFM. Fig. 5((a)e(c)) illustrates three dimensional AFM
images for the as-deposited, 400  C and 600  C annealed thin films,
respectively. Root mean square (RMS) roughness values of the thin
films were provided below the AFM images. The as-deposited thin
film showed a very smooth surface with a RMS roughness of 1.1 nm,
and it was composed of very fine grains. As the post-deposition
annealing temperature increased, the average grain size and

Fig. 4. FESEM image (top-view) and EDX spectrum of (a) as-deposited and (b) 600  C annealed copper oxide thin films. The inset shows the cross-sectional image of the as-
deposited film.
992 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
Fig. 5. 3D AFM images of (a) as-deposited thin film and films annealed at different temperatures of (b) 400  C and (c) 600  C.
roughness was found to increase. We attribute this to the poly-
crystalline nature and improved crystallinity of the films. A signif-
icant increase in the roughness suggests a decrease in the surface
smoothness. From the AFM analysis, a grain size in the range of
50e70 nm was estimated for the thin film annealed at 600  C. On
the other hand, SEM analysis revealed that most grains have an
average size of 50 nm. Different results in the average grain size of
the copper oxide thin films were derived from the XRD, SEM and
AFM measurements. The crystallite size obtained from XRD anal-
ysis is equivalent to the mean size of the crystallites that coherently
scatter X-rays [22,40]. The larger grain sizes observed in the SEM
and AFM images could be attributed to agglomeration of the
crystallites leading to larger grains. The surface roughness of the
thin films increased with annealing temperature based on AFM
results. It is clear that the increased crystallite size (evidenced by
XRD) increases the surface roughness of the thin films [41,42].
The optical properties of the thin films were investigated by
optical transmittance and reflectance measurements. Fig. 6(a)
shows the spectral dependence of the optical transmittance and
reflectance of the thin films. Films were found to be highly
absorbing in the visible region and transparent in near-infrared
region with an average transmittance of about 70%. The trans-
mittance of all samples sharply decayed for wavelengths below
700 nm. The reflectivity of the as-deposited and 200e400  C
annealed films were low and their spectra were relatively similar.
On the contrary, the peak of the reflectance curves increased at
higher annealing temperatures (500  C and 600  C). This could be
simply attributed to enhanced scattering with increased crystallite
size and surface roughness of the thin films. Experimentally
measured transmittance and reflectance spectra were used to
Fig. 6. (a) Transmittance and reflectance spectra of as-deposited and annealed thin films at various temperatures (b) Tauc plot for the as-deposited and 600  C annealed thin films.
calculate absorption coefficient and bandgap values of the samples.
The absorption coefficients were determined using the following
relation [43]:
2 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
!3
u
1 4ð1  RÞ2 ut ð1  RÞ4
a ¼ ln þ þR 5 2 (3)
d 2T 4T 2
where d is thickness, T and R are the transmittance and reflectance
values of the film in a strong absorption region at a particular
wavelength, respectively. The optical bandgap (Eg) of the films were
determined using the Tauc's relation [44]:
 n
ahn ¼ a0 hn  Eg (4)
where a is absorption coefficient, hn is the photon energy (eV), a0 is
a characteristic parameter that is independent of photon energy for
respective transitions and n ¼ 0.5, 1.5, 2 or 3 for allowed direct,
forbidden direct, allowed indirect and forbidden indirect electronic
transitions, respectively [45]. In the present case, the Eg was esti-
mated by assuming an allowed direct transition (n ¼ 0.5). The Eg
was obtained by extrapolating the linear portion of the plot as
shown in Fig. 6(b). Eg values for the samples were estimated to be
1.93e2.08 eV range, without a significant variation. These values
are consistent with the reported Eg values of copper oxide thin films
[22,46] and were in the expected range for CuO. Eg values decreased
slightly with annealing temperature. A few factors such as the
crystallinity, grain size and internal stress are known to influence
the Eg of the materials. In this study, the improved overall crystal
structure with the increasing annealing temperature was found
 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
from XRD results. This result implies that the crystallinity was one
of the factors to alter the direct bandgap energy. On the other hand,
the grain size could play an important role in the observed reduced
Eg values since the grain size of copper oxide films increased with
increasing the annealing temperature. It is also known that the
stress in thin films decreases with an increase in thin film thickness
[41]; however, our thin films had similar thicknesses. Improved
crystallinity and increased grain size could contribute to the Eg
shift. Similar observation of decrease in the optical bandgap with
the increase of annealing temperature has been reported in the
literature for chemically deposited and radio frequency magnetron
sputtered copper oxide thin films [2,41].
In order to further investigate the effect of the post-deposition
annealing on the surface chemical composition and electronic
structure of the copper oxide thin films, XPS measurements were
conducted. In the XPS study, probing depth of the X-ray with an
incident photon energy of 1486.6 eV (Al Ka X-ray source) corre-
sponds to ~2 nm, as can be estimated from the universal curve of
the inelastic mean free path versus initial electron energy in eV
[47]. During measurements, a slight ion bombardment with a
voltage of 500 V under 1  105 Torr of Argon gas was performed
in-situ for 45 s to reduce amount of carbon contamination to an
acceptable level without significantly perturbing the surface of the
thin films. Survey spectra of the as-deposited and 600  C annealed
copper oxide thin films (not shown here) consisted of well recog-
nized Cu 2s, Cu 2p, Cu 3s and Cu 3p core level peaks and carbon was
found to be the only contaminant observed on the surface of the
films.
Fig. 7(a) shows the high resolution Cu 2p XPS data recorded for
the as-deposited and 600  C annealed copper oxide thin films. It
can be seen that Cu 2p energy level of both samples was composed
of main characteristic doublet peaks corresponding to Cu 2p1/2 and
Cu 2p3/2 at ~952.3 eV and ~932.4 eV, respectively, and shake-up
satellite structures at about 8 eV and 10 eV binding energies
higher than that of the main Cu 2p peak. The peak separation in
Fig. 7. Cu 2p XPS core level spectra belonging to as-deposited and 600  C annealed copper oxide thin film samples (a) with decomposed components for Cu 2p3/2 (b and c).
993
other words the spin-orbit splitting between two main photo-
emission peaks (Cu 2p1/2 and Cu 2p3/2) was determined to be
around 19 eV, and was in good agreement with the values reported
in literature [48]. The satellite peaks on the high binding energy
side of the core level Cu 2p XPS data originate from multiple exci-
tations in copper oxides and they are known to be characteristics of
CuO phase [49,50]. Therefore, the presence of the intense shake-up
satellite structures observed in the Cu 2p XPS spectra of both
samples was an indication of the presence of CuO species in the
surface layer. The Cu 2p3/2 core level was employed to monitor the
surface oxidation through post-deposition annealing. The broad Cu
2p3/2 peak and shake-up satellite feature were deconvoluted into
five peaks, marked as peaks A1, A2, A3, S1 and S2 as shown in
Fig. 7((b) and (c)). The peak A1 and A2 were related to Cu2O (Cu1þ)
and CuO (Cu2þ) phases, respectively. There was an additional peak
at 935.65 eV (A3) in the XPS spectra of the samples, which was
consistent with the absorber copper hydroxide (Cu(OH)2) species
on the surface of thin films [51]. As evidenced in Fig. 7(b) and (c), a
significant difference in the intensity of the A1, A2, A3 features was
observed for the copper oxide thin film that was annealed at 600  C.
This result clearly indicates that the annealing affects the amount of
Cu1þ and Cu2þ species existing on the surface of the thin films. In
order to quantify the relative concentrations of Cu2O, CuO and
Cu(OH)2 species present on the thin film surfaces we used the
combination of the signals from the direct photoemission peaks
(A1, A2, A3) and the shake-up photoemission peaks (S1 and S2) in
the following equations,
A2 þ ðS1 þ S2Þ
%Cu2þ ¼ %Cu1þ
ðA1 þ A2 þ A3Þ þ ðS1 þ S2Þ
A1
¼ (5)
ðA1 þ A2 þ A3Þ þ ðS1 þ S2Þ
Curve fitting parameters and the calculated relative concentra-
tions are presented in Table 2.
994 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
Table 2
Curve fitting parameters and the % relative concentrations for as-deposited and
600  C annealed copper oxide thin films.
As-deposited Annealed at 600  C
Cu2O CuO Cu(OH)2 Cu2O CuO
Binding 932.44 933.64 935.65 932.44 933.64
energy (eV)
FWHM (eV) 1.54 1.54 1.54 2.24 2.24
Relative % 78 20 2 20 68
concentration
Quantitative analysis based on the area estimated from peak
fitting of the Cu 2p core level XPS peak indicated the relative con-
centration of CuO phase for the as-deposited thin film to be 20%
(Table 2). Following annealing at 600  C, CuO concentration was
increased to 68%, while Cu2O concentration decreased from 78 to
20% (Table 2). Upon annealing the thin films, oxygen atoms present
in the air ambient could be incorporated within the surface of thin
films leading to the formation of CuO from Cu2O. In addition, the
relative concentration of Cu(OH)2 calculated from Cu 2p XPS data
increased for the film annealed at 600  C. As a consequence, post-
deposition annealing treatment can noticeably affect the surface
phase composition of copper oxide thin films.
It is important to study the surface elemental composition,
chemical state and electronic structure as they influence the elec-
tronic and optical properties of thin films. In present study, the
extreme surface sensitivity of the XPS technique and its ability to
resolve small changes in the local binding configuration have been
utilized to investigate the surface chemical composition and elec-
tronic structure of the copper oxide thin films. The existence of
Cu1þ and Cu2þ was confirmed by XPS. For the as-deposited film,
relative concentration ratio of Cu1þ to Cu2þ phases was found to be
3.9. This result indicated that the surface of the as-deposited film is
rich in Cu2O phase. As mentioned previously, only crystalline CuO
phase was observed in the XRD data for the as-deposited copper
oxide thin films. Although XRD data did not provide an evidence of
Cu2O phase, XPS results revealed the presence of Cu1þ phase,
suggesting the formation of both Cu2O and CuO phases simulta-
neously on the surface of the as-deposited thin films. The observed
difference could be due to the fact that XPS technique, unlike XRD,
can only give information from the sample surface rather than the
bulk. Therefore, the surface composition (evaluated by XPS) of the
as-deposited copper oxide thin films strongly deviated from that of
the bulk (evaluated by XRD). The surface of the as-deposited film
was found to be rich in Cu2O phase. In addition, annealing at 600  C
modified the surface phase composition, where the relative con-
centration ratio (Cu1þ/Cu2þ) was determined to be 0.29.
The discrepancy in the obtained results from Raman, XRD and
XPS measurements stems from the different sampling depths of
these analytical characterization tools. The disagreements of
Raman, XRD and XPS results suggest that the crystal phase in the
surface region is different from that in the bulk during the post-
deposition annealing of the thin films. The difference between
XRD and XPS is that, XPS is a highly surface sensitive technique and
it can only get the compositional information from the first
monolayer at the thin film surface. The reason why Cu2O phase was
not detected by XRD is because XRD probes only crystalline phase
existing in the bulk structure of the thin films. Our XRD study did
not reveal the formation of Cu2O phase in the as-deposited and
600  C annealed thin films. However, the relatively high concen-
tration of this phase (78%) obtained from XPS measurements leads
us to conclude that Cu2O formed at the topmost layer of the as-
prepared thin films is not in crystalline form. On the other hand,
the absence of Cu2O phase in the 600  C annealed thin films can be
related to its trace amount, which is below the detection limit of
XRD since it cannot be in amorphous form at this temperature
[22,52]. This result is also confirmed by the Raman measurements.
Moreover, the amount of Cu2O as compared to the bulk could be
Cu(OH)2 negligible and/or the crystallite size of Cu2O could be too small to
be detected by XRD. Note that XPS is much more surface sensitive
935.65
compared to Raman spectroscopy. Hence, we believe that the Cu2O
2.24 phase obtained in the XPS measurements for the 600  C annealed
12 thin films only exists at the surface of these thin films.
4. Conclusions
In conclusion, copper oxide thin films were successfully
deposited on glass substrates using consecutive solegel and spin-
coating methods. The average thickness of the films was 520 nm.
A post-deposition annealing at various temperatures was applied
and the microstructural, morphological, optical properties and
surface electronic structure of the thin films were investigated. The
XRD results revealed the increase in crystallite size with annealing
temperature. The effect of the annealing temperature on the crys-
tallinity was also confirmed by Raman measurements. It has been
observed that the optical bandgap of the corresponding thin films
decreased slightly as the annealing temperature was increased. In
the as-deposited and 600  C annealed films, CuO and Cu2O phases
were found to co-exist, while CuO phase dominated the latter one.
High temperature annealing, through the incorporation of oxygen
atoms within the ambient converted Cu2O existing on the topmost
layer of the thin films to CuO. Finally, copper oxide is a promising
material for optoelectronic devices in both CuO and Cu2O phases
and the results reported herein will provide a roadmap for re-
searches who would be interested in the solution deposition of
copper oxide thin films as active layers in those devices.
Acknowledgments
G. A. and F. A. A. would like to give thanks to The Scientific &
Technical Research Council of Turkey (TUBITAK) 2218-National
Postdoctoral Research Fellowship Program for financial support. H.
E. U. acknowledges supports from the Distinguished Young Scien-
tists award of the Turkish Academy of Sciences (TUBA). METU
Central Laboratory facilities are also greatly acknowledged.
References
[1] M.R. Johan, M.S. Mohd Suan, N.L. Hawari, H.A. Ching, Int. J. Electrochem. Sci. 6
(2011) 6094.
[2] N. Serin, T. Serin, Ş. Horzum, Y. Çelik, Semicond. Sci. Technol. 20 (2005) 398.
[3] T. Maruyama, Sol. Energ. Mat. Sol. C 56 (1998) 85.
[4] S. Mridha, D. Basak, Semicond. Sci. Technol. 21 (2006) 928.
[5] M. Hübner, C.E. Simion, A. Tomescu-Sta
noiu, S. Pokhrel, N. Ba ^rsan, U. Weimar,
Sens. Actuators B Chem. 153 (2011) 347.
[6] M. Breedon, S. Zhuiykov, N. Miura, Mater. Lett. 82 (2012) 51.
[7] A. Chapelle, F. Oudrhiri-Hassani, L. Presmanes, A. Barnabe , P. Tailhades, Appl.
Surf. Sci. 256 (2010) 4715.
[8] A.Y. Oral, E. Menşur, M.H. Aslan, E. Başaran, Mater. Chem. Phys. 83 (2004) 140.
[9] W. Chen, L. Hong, A.L. Liu, J.Q. Liu, X.H. Lin, X.H. Xia, Talanta 99 (2012) 643.
[10] N.Q. Dung, D. Patil, H. Jung, D. Kim, Biosens. Bioelectron. 42C (2012) 280.
[11] E.A. Souza, R. Landers, L.P. Cardoso, T.G.S. Cruz, M.H. Tabacniks, A. Gorenstein,
J. Power Sources 155 (2006) 358.
[12] H. Pang, J. Deng, B. Yan, Y. Ma, G. Li, Y. Ai, et al., J. Electrochem. Sci. 7 (2012)
10735.
[13] R.V. Kumar, Y. Diamant, A. Gedanken, Chem. Mater. 12 (2000) 2301.
[14] R. Prasad, J. Therm. Anal. Cal. 85 (2006) 279.
[15] M.F. Al-Kuhaili, Vacuum 82 (2008) 623.
[16] Y.K. Jeong, G.M. Choi, J. Phys. Chem. Solids 57 (1996) 81.
[17] K. Akimoto, S. Ishizuka, M. Yanagita, Y. Nawa, G.K. Paul, T. Sakurai, Sol. Energy
80 (2006) 715.
[18] F. Gao, X.J. Liu, J.S. Zhang, M.Z. Song, N. Li, J. Appl. Phys. 111 (2012) 084507.
[19] B.D. Yuhas, P. Yang, J. Am. Chem. Soc. 131 (2009) 3756.
[20] B. Balamurugan, B.R. Mehta, Thin Solid Films 396 (2001) 90.
[21] A.H. Jayatissa, K. Guo, A.C. Jayasuriya, Appl. Surf. Sci. 255 (2009) 9474.
 F.A. Akgul et al. / Materials Chemistry and Physics 147 (2014) 987e995
[22] V. Figueiredo, E. Elangovan, G. Gonçalves, P. Barquinha, L. Pereira, N. Franco, et
al., Appl. Surf. Sci. 254 (2008) 3949.
[23] G. Papadimitropoulos, N. Vourdas, V.E. Vamvakas, D. Davazoglou, Thin Solid
Films 515 (2006) 2428.
[24] S.C. Ray, Sol. Energy Mater. Sol. Cells 68 (2001) 307.
[25] http://srdata.nist.gov/xps.
[26] D. Chauhan, V.R. Satsangi, S. Dass, R. Shrıvastav, Bull. Mater. Sci. 29 (2006) 709.
[27] C.Y. Huang, A. Chatterjee, S.B. Liu, S.Y. Wu, C.L. Cheng, Appl. Surf. Sci. 256
(2010) 3688.
[28] N.D. Hoa, S.Y. An, N.Q. Dung, N. Van Quy, D. Kim, Sens. Actuators B 146 (2010)
239.
[29] W.W. Wang, Y.J. Zhu, G.F. Cheng, Y.H. Huang, Mater. Lett. 60 (2006) 609.
[30] T. Maruyama, Sol. Energy Mater. Sol. Cells 56 (1998) 85.
[31] K.H. Yoon, W.J. Choi, D.H. Kang, Thin Solid Films 372 (2000) 250.
[32] M. Nair, L. Guerrero, O. Arenas, P. Nair, Appl. Surf. Sci. 150 (1999) 143.
[33] A.O. Musa, T. Akomolafe, M.J. Carter, Sol. Energy Mater. Sol. Cells 51 (1998)
305.
[34] B.D. Cullity, S.R. Stock, Elements of X-ray Diffraction, third ed., Prentice Hall
Publication, India, 2001.
[35] C. Suryanarayana, M. Grant Norton, X-ray Diffraction: A Practical Approach,
Springer, New York, 1998.
[36] G.K. Williamson, W.H. Hall, Acta Metall. 1 (1953) 22.
[37] G.B. Williamson, R.C. Smallman, Philos. Mag. 1 (1956) 34.
[38] J.F. Xu, W. Ji, Z.X. Shen, W.S. Li, S.H. Tang, X.R. Ye, et al., J. Raman Spectrosc. 30
(1999) 431.
995
[39] B. Balamurugan, B.R. Mehta, D.K. Avasthi, F. Singh, A.K. Arora, M. Rajalakshmi,
et al., J. Appl. Phys. 92 (2002) 3304.
[40] T. Ung ar, G. Tichy, J. Gubicza, R. Hellmig, J. Powder Diffr. 20 (2005) 366.
[41] J.Y. Park, T.H. Kwon, S.W. Koh, Y.C. Kang, Bull. Korean Chem. Soc. 32 (2011)
1331.
[42] S. Venkataraj, O. Kappertz, C. Liesch, R. Detemple, R. Jayavel, M. Wuttig,
Vacuum 75 (2004) 7.
[43] M. Kastner, Phys. Rev. Lett. 28 (1972) 355.
[44] J.C. Tauc, Optical Properties of Solids, North-Holland, Amsterdam, 1972.
[45] J.C. Tauc, A. Menth, J. Non-Cryst. Solids 8e10 (1972) 569.
[46] A.A. Ogwu, E. Bouquerel, O. Ademosu, S. Moh, E. Crossan, F. Placido, Acta
Mater. 53 (2005) 5151.
[47] C. Powell, A. Jablonski, NIST Inelastic Mean Free Path Database, National
Institute of Standards and Technology, Gaithersburg, MD, 2000.
[48] Z.H. Gan, G.Q. Yu, B.K. Tay, C.M. Tan, Z.W. Zhao, Y.Q. Fu, J. Phys. D Appl. Phys.
37 (2004) 81.
[49] S.K. Chawla, N. Sankarraman, J.H. Payer, J. Electron Spectrosc. Relat. Phenom.
61 (1992) 1.
[50] J. Ghijsen, L.H. Tjeng, J.V. Elp, H. Eskes, J. Westerink, G.A. Sawatzky, et al., Phys.
Rev. B 38 (1988) 11322.
[51] M.C. Biesinger, L.W.M. Lau, A.R. Gerson, R.S. Smart, Appl. Surf. Sci. 257 (2010)
887.
[52] G. Yang, A. Chen, M. Fu, H. Long, P. Lu, Appl. Phys. A 104 (2011) 171.