Indian J Phys DOI 10.1007/s12648-014-0597-0
ORIGINAL PAPER
Effect of zinc induced compressive stresses on different properties of copper oxide thin films H Faiz1, K Siraj1*, M S Rafique1, S Naseem2 and A W Anwar1 1
Department of Physics, Laser and Optronics Center, University of Engineering and Technology Lahore, Lahore, Pakistan 2
Center of Excellence in Solid State Physics, University of Punjab, Lahore, Pakistan Received: 09 May 2014 / Accepted: 27 August 2014
Abstract: Thin films of pure and zinc doped copper oxide have been grown on Si (1 0 0) substrate using pulsed laser deposition at varying concentrations of zinc. The effects of the doping concentration on structural, surface and optical properties of all thin films have been investigated. X-ray diffraction shows the presence of monoclinic CuO phase and compressive stresses in all thin films. The Raman spectra show an up-shift in the Raman peak position for doped thin films. Micrographs show smooth surface morphology, free from micron sized laser generated particulates. Spectroscopic ellipsometry has been used to study the optical constants (w, D, n, k, e1, e2). Optical bandgap energies have been calculated and found to be dependent on stresses in the films. Consequently, Zn doping induced stresses have strong effects on microstructure and optical properties of thin films. Keywords:
Pulsed laser deposition; CuO; Compressive stresses; Optical properties
PACS No.:
78.20.-e
1. Introduction There are many well studied transparent conducting oxide (TCO) materials like In2O3, ZnO and SnO2 used in solar cells, touch panels and in flat-panel display devices as transparent electrodes. These materials are n-type TCOs. TCOs exhibiting p-type conductivity, such as copper oxide (CuO) and copper aluminum oxide (CuAlO2) as well as thin films of CuGaO2 and SrCu2O2, have been developed in recent years [1]. The development of p-type TCO materials is one of the key technologies for pn-junction-based oxide devices, such as diodes, transistors and light-emitting diodes [1]. Cupric oxide (CuO) and cuprous oxide (Cu2O) are two important oxides of copper. Both these oxides exhibit p-type conductivity and have proved to be promising TCO materials for the fabrication of wide range of optoelectronic devices. Cu2O has a cubic structure with an optical bandgap of 2.1–2.6 eV while CuO has a monoclinic crystal structure with a bandgap of 1.9–2.1 eV. Non-toxic,
economic, abundant availability and relatively simple formation of oxide make copper oxide as an interesting material [2]. Apart from their semiconductor applications, these materials have been employed as heterogeneous catalysts, solid-state gas sensors, hetero-contacts and as electrode materials for lithium batteries and microwave dielectric materials [3]. CuO is a potential semiconductor for solar cell fabrication due to its suitable optical properties [4]. Different deposition techniques have been employed to deposit thin films of copper oxide. These techniques include chemical vapor deposition, pulsed laser deposition (PLD), molecular beam epitaxy, ion beam sputtering, reactive magnetron sputtering, thermal oxidation, ultrasonic spray pyrolysis and thermal evaporation [5– 11]. However, PLD technique has proved to be an excellent technique with the flexibility to control different deposition parameters to get high quality thin films. In the present work, thin films of pure and Zn doped copper oxide have been deposited by PLD technique and effect of stresses induced due to varying concentration of zinc doping on the structural, vibrational, surface and optical, properties of CuO thin films have been investigated.
*Corresponding author, E-mail:
[email protected]
Ó 2014 IACS
H Faiz et al.
2. Experimental details
3. Results and discussion 3.1. XRD analysis XRD patterns of all the films are shown in Fig. 1, which reveal that all the films have polycrystalline structure with (0 0 2) and (2 0 0) orientations of monoclinic CuO phase
Macrostrains
CuO (2 0 0)
SiO2
CuO (0 0 2)
-1.2x10-2
-1.0x10-2
-8.0x10-3
-6.0x10-3
-4.0x10-3
Intensity (arb. units)
The experiment was carried out in two steps. In the first step, circular targets of pure and Zn doped (1, 3, 4 and 5 atomic wt%) copper oxide, were prepared by solid state reaction method. Required amount of 99.99 % pure powders of CuO and ZnO were weighed using a micro balance and were finely mixed and ground in an agate mortar. These powders were then calcinated at 750 °C for 9 h in an electrical furnace. The powders were reground and compressed to form a pellet of 3 mm thickness and 6 mm diameter using hydraulic press applying a pressure of 8 metric tons for 30 min. All the pellets were finally sintered at 900 °C for 10 h in an electrical furnace. Total five pellets were prepared having zinc doping concentrations of 0, 1, 3, 4 and 5 atomic wt%. After sintering these pellets were ready to be used as targets. In second step, KrF excimer laser (Ex50, GAM LASER INC, USA having k = 248 nm, sl = 20 ns and /l = 1 J cm-2) beam was focused through a UV lens of focal length 40 cm at an angle of incidence 45° onto the target placed in the vacuum chamber evacuated at *10-5 Torr using a rotary and turbomolecular vacuum pumps. All the films were deposited on ultrasonically cleaned, single crystalline, one side polished p-Si (1 0 0) substrates of 10 9 10 9 1 mm3 size at temperature 200 °C. The target to substrate distance was 18 mm and the on-axis PLD geometry was employed for deposition. During deposition, the targets were rotated for homogeneous surface ablation and 10,000 number of laser pulses were used for each deposition. Prior to characterization, all the deposited films were annealed at 800 °C for 4 h. The thickness of the films was around 250 ± 5 nm. The deposited films were pure CuO, 1 % Zn–CuO, 3 % Zn–CuO, 4 % Zn–CuO and 5 % Zn–CuO. The crystalline structure of thin films was examined using PHILIPS PANalytical X-ray diffractometer with Cu as X-ray ˚ , CuKa). Vibrational modes were source (k = 1.5406 A studied by Resonance Raman Spectroscopy (RRS) using Lab RAM HR HORIBA JOBIN–YVON Raman spectrometer with spectral resolution of 1 cm-1. Surface morphology was analyzed using Hitachi S-3700 N Scanning Electron Microscope. For the analysis of optical properties, a Rotating Analyser Ellipsometer (J. A. Woolam Co. Inc. Various Angle Spectroscopic Ellipsometer) was used at an angle of incidence 60° and the scan data in terms of amplitude ratio (W) and phase difference (D) was very well fitted to Cauchy’s Model.
-1.4x10-2
5% Zn-CuO
-2.0x10-3 0
1
2
3
4
5
Zn doping concentration (at. wt. %)
4% Zn-CuO 3% Zn- CuO 1% Zn-CuO
Pure CuO
40
50
60
2θ (degree)
Fig. 1 XRD patterns of pure and Zn doped CuO thin films. Inset shows variation of macrostrain with increasing zinc concentration
only and no traces of Cu2O phase are observed. This is attributed to the post-deposition oxygen annealing at high temperature, because the monoclinic CuO phase (tenorite) is more stable at high temperatures than the cubic Cu2O phase (cuprite). Due to high annealing temperature, the grains acquire enough energy to orient in proper equilibrium sites resulting in an increase in the crystallinity of the thin films. No peak is observed for Zn or ZnO in any of the films, due to the substitution of Zn atoms on Cu lattice sites. SiO2 peak is observed due to unavoidable surface oxidation of silicon substrate. The preferred orientation of these thin films is found using following relation [12, 13] " #1 n I ðhi ki li Þ 1 X I ðhi ki li Þ Pð h i k i l i Þ ¼ ð1Þ Io ðhi ki li Þ n i¼1 Io ðhi ki li Þ where P is the texture coefficient, Io the standard intensity (ICDD 00-045-0937) and I is the observed intensity. The value of texture coefficient P provides useful information about the preferred orientation direction of the thin films. If this value is &1, for all the (hi ki li) planes observed in the XRD patterns, the films have no preferred orientation or they are randomly oriented. If the value is [1, then more number of grains are oriented in a given (hi ki li) direction, which is called the preferential orientation direction of the thin films. Finally if the value of texture coefficient is \1, but greater than zero, it shows the lacking of grains in that particular (hi ki li) direction [12, 13]. Table 1 shows that the texture coefficient for all the films is higher for (0 0 2) direction than that for the (2 0 0) direction having values [2 except for 5 % Zn–CuO thin film for which, the value is 1.94, indicating that all the films are preferentially oriented in the (0 0 2) direction. The value of full width at
Effect of zinc induced compressive stresses Table 1 XRD peak position, miller indices (h k l), FWHM, crystallite size, dislocation line density, texture coefficient and macrostrains for pure and Zn doped CuO thin films Sample name
2h value (degree)
(h k l)
FWHM (degree)
Crystallite size (nm)
Dislocation line ˚) density (A
Texture coefficient
Pure CuO
35.65
(0 0 2)
0.2663
31.3
3.2 9 1015
2.28
-5.2
1 % Zn–CuO
35.69
(0 0 2)
0.1786
46.7
4.5 9 1014
2.42
-12.6
3 % Zn–CuO 4 % Zn–CuO
35.61 35.89
(0 0 2) (0 0 2)
0.1709 0.3332
48.8 25.1
4.2 9 1014 1.5 9 1015
2.28 2.05
-11.6 -4.7
5 % Zn–CuO
35.62
(0 0 2)
0.4077
20.4
2.3 9 1015
1.94
-4.3
half-maximum (FWHM) of the (0 0 2) peak first decreases with Zn incorporation with lowest value for 3 % Zn–CuO thin film and then increases with further increase in the Zndoping concentration up to 5 % Zn–CuO thin film (Table 1). The crystallite size (D), dislocation line density (d) and macrostrain \e[ have been calculated using the following relations [13], D¼
0:9k b cos h
1 D2 d d0 hei ¼ d0
d¼
ð2Þ ð3Þ ð4Þ
where k, b, h, d and do represent wavelength of the X-rays, FWHM of the (0 0 2) peak, X-ray diffraction angle, the observed d-spacing and standard d-spacing of the pure CuO (bulk) respectively. Surface tensions, strains between the silicon substrate and polycrystalline material and inner strains as residual stresses in the films have been neglected. The peak position, miller indices (h k l), FWHM, crystallite size, texture coefficient, dislocation line density and macrostrain for all the films are listed in Table 1, which shows that the macrostrain has negative value for all thin films indicating the presence of compressive stresses in thin films as depicted by Eq. (4). The 1 % Zn–CuO thin film has maximum compressive stress as compared to all other thin films. On further doping of Zn, macrostrains are reduced and so a reduction in compressive stress is observed. The variation of macrostrains with increasing Zn concentration is plotted in the inset of Fig. 1. The value of compressive stress first increases with the incorporation of Zn, having maximum value for 1 % Zn–CuO thin film and then decreases, having lowest value for 5 % Zn–CuO thin film as shown in the inset of Fig. 1. XRD data shows a decrease in d-spacing for all the films as compared to standard data, which also indicates the presence of compressive stress in all thin films. As a result of this decrease in d-spacing and consequently due to compressive stress in the films, the 2h is shifted to a slightly higher values. This stress is induced due to zinc incorporation into the copper oxide thin films.
Macrostrain (910-3)
3.2. Raman spectroscopy Raman spectra of bulk monoclinic CuO, copper oxide nanorods, nanowhiskers, nanoflakes, copper oxide single crystal and pure thin films have been studied by many researchers [14–17]. The present study has been made to explore the effect of varying zinc doping concentration on the Raman peak shift of copper oxide thin films. CuO has monoclinic crystal structure and belongs to space group C62h (C 2/c) having four atoms per primitive cell. The lattice ˚ , b = 3.4230 A ˚, parameters of CuO are a = 4.6850 A ˚ c = 5.1320 A with a = 90°, b = 99°, c = 90° and it has twofold symmetry with b-axis as principal axis. According to group theory predictions, copper oxide has a total of nine optical vibrations among which, three (Ag ? 2Bg) are Raman active with even symmetry whereas the remaining six (3Au ? 3Bu) are IR active having odd symmetry [15, 18, 19]. The three Raman modes of bulk copper oxide are Ag (297 cm-1), Bg1 (341 cm-1) and Bg2 (631 cm-1). Raman tensors for the space group C2h as listed by Loudon [14] for the Ag and Bg mode of CuO are given as a Ag: 0 d
0 d b 0 0 c
0 Bg: e 0
e 0 f
0 f 0
Only oxygen atoms vibrate for the Raman active modes in CuO. The direction of vibration of atoms for Ag mode is in the y-direction (b-axis) and perpendicular to y-axis for the Bg modes as illustrated in Fig. 2(a). Figure 2(b) shows the Raman spectra of pure and zinc doped CuO thin films. The three Raman modes (Ag, Bg1 and Bg2) are observed for all the samples and are consistent with those reported earlier [15–17, 20, 21]. The high intensity and sharpness of the Raman peaks confirm the presence of monoclinic CuO phase only as no traces of peaks for Cu2O are observed. The Raman frequency peaks for all the observed modes in all the films show their better crystallinity as observed from XRD analysis. There is a slight up-shift in Raman peak position of Ag mode with increasing zinc incorporation as compared to the Raman spectrum of the pure CuO thin film. This shift in frequency is related to the change in crystallite size as well as to the stress/strain present in the
H Faiz et al.
(a)
Ag
Bg1
Bg2
y y
z
Oxygen, Copper
x
x
(b)
Ag
Si
Intensity (arb.units)
Bg1
Bg2
E2 (ZnO)
5% Zn- CuO 4% Zn- CuO 3% Zn- CuO 1% Zn-CuO Pure CuO 300
400
500
600
700
-1
Wave number (cm ) Fig. 2 (a) Raman modes of CuO. (b) Raman spectra of pure and Zn doped CuO thin films
thin films. This up-shift in the Raman peak position confirms the presence of compressive stress in Zn–CuO thin films. The maximum shift is observed for 1 % Zn–CuO thin film showing the presence of highest compressive stress and confirming the XRD results. For 3 % Zn–CuO thin film, the peak shift reduces a little as compared to 1 % Zn–CuO thin film showing a little reduction of compressive stresses. If the Raman frequency shift is positive, then the compressive strain is present in the film and vice versa [21]. At highest doping concentration of zinc (5 %), the intensity of Raman active modes of CuO is reduced and the FWHM has increased due to the appearance of E2 Raman active mode of ZnO at about 437 cm-1. CuO has monoclinic crystal structure and only oxygen atoms vibrate for the Raman active modes; whereas ZnO has hexagonal structure and the E2 mode of ZnO has the vibration of oxygen atoms only. So with the appearance of E2 mode of ZnO, the vibration of oxygen atoms in the crystal of CuO is distorted and as a result an up-shift as well as the broadening of the peak of Ag mode of CuO is observed. 3.3. Surface analysis Figure 3 shows the SEM micrographs of phase pure and Zn doped CuO thin films. These micrographs show that all the films are of good quality and have fine structure. Figure 3(a) shows that the phase pure CuO thin film has
smooth surface morphology and narrow size distribution of particles. The surface morphology of doped thin films, in Fig. 3(b)–3(e) qualitatively show the higher surface roughness as compared to phase pure CuO thin film. The surface morphology of 1 % Zn–CuO thin film, as in Fig. 3(b), appears to be most rough among all thin films. This is due to the stress present in the film because of zinc incorporation, as depicted by XRD analysis. Figure 3(c) shows the surface morphology of 3 % Zn–CuO thin film. It is obvious that the particle distribution on the surface of thin film becomes narrow and the average particle size is *0.255 lm. Figure 3(d) shows the SEM micrograph of 4 % Zn–CuO thin film and it is observed that the shape of the particles is clearly spherical for this thin film. The 5 % Zn–CuO thin film has best smoothness among all the doped thin films, as shown in Fig. 3(e). The size distribution appears to be narrow for this thin film also with an average particle size of *0.423 lm, which is larger than the grain size calculated from XRD data. It is observed that the particle size has increased with increasing doping concentration. This fact is attributed to grain agglomeration due to post deposition annealing at high temperatures. At high temperatures, the atoms attain sufficient energy to diffuse onto the surface and align along a particular direction, thus improving the quality of the thin film as reported earlier in [22, 23]. The same effect has been observed due to increasing substrate temperature also rather than annealing [10]. The particle distribution on the surface is random and the formation of particles with different sizes and shapes is observed. No grain boundaries are visible for these thin films. So the deposited thin films have smooth surface profile, which is free from micron sized clusters, fragments, molten globules and other big particulates, which are inherently linked with pulsed laser ablation process. The used optimized deposition parameters ensured the smooth surface profile.
3.4. Optical analysis The variation of optical constants such as refractive index (n), extinction coefficient (k), real dielectric constant (e1) and imaginary dielectric constant (e2) with wavelength as shown in Fig. 4(a)–4(d). Figure 4(a) shows that there is an increase in the refractive index value in the UV range (k & 300–400 nm), for all the thin films. Then it attains almost a constant value (1.3–1.5) in the visible wavelength range (k & 400–700 nm), indicating that all the thin films have good transparency in this range. Finally, the refractive index increases in the wavelength range k & 700– 1,000 nm (IR region) showing a maximum at k & 900 nm. The minimum value of the refractive index (n = 0.85) has been found for 1 % Zn–CuO thin film in the
Effect of zinc induced compressive stresses Fig. 3 SEM micrographs of pure and doped CuO thin films. (a) pure CuO, (b) 1 % Zn–CuO, (c) 3 % Zn–CuO, (d) 4 % Zn– CuO and (e) 5 % Zn–CuO at 35 K
UV region, which becomes almost 1.5 in the whole visible region. Then it remains at 1.5 up to 755 nm for this thin film in contrary to other thin films, which might be due to highest compressive stresses in the thin film. The highest refractive index (n = 3.1) is observed for 3 % Zn–CuO thin film in the IR region (k * 920 nm). Figure 4(b) shows the variation of the extinction coefficient with wavelength. The extinction coefficient first increases from 300 to 367 nm, then decreases gradually up to k & 460 nm; then broad peak is observed in the wavelength range of 400–1,000 nm showing a maximum around 775 nm except for 1 % Zn–CuO film, where the maximum is found near k & 900 nm, due to high compressive stress present in this film. The maximum value of the extinction coefficient is found for 3 % Zn–CuO thin film. The extinction coefficient is related to the absorption coefficient (a) according to the relation [24],
a ¼ 4pk=k ð5Þ Figure 4(c) shows a plot of real part of dielectric constant (e1) versus wavelength. The real part of dielectric constant (e1) is defined as [24] e 1 ¼ n2 k 2
ð6Þ
The real part of dielectric constant e1 represents the magnitude of polarization of a material. If the frequency of the incoming light resonates with the oscillatory frequency of the atoms/electrons of the medium, resonant oscillations occur and light is absorbed by the medium. As a result the real part of dielectric constant (e1) shows a peak value for that particular frequency/wavelength. The variation of e1 depends upon the frequency of the incoming light. In the IR region, where the angular frequency is very low, e1 is generally represented by es (static dielectric constant), which contains the contribution of atomic as well as of
H Faiz et al. 2.0
3.5
(b)
(a) 3.0 1.6
n
2.0
k
Pure CuO 1% Zn-CuO 3% Zn-CuO 4% Zn-CuO 5% Zn-CuO
2.5
1.5
Pure CuO 1% Zn-CuO 3% Zn-CuO 4% Zn-CuO 5% Zn-CuO
1.2
0.8
1.0 0.4
0.5 300
400
500
600
700
800
900
300
1000
400
500
600
700
800
900
800
900
1000
Wavelength (nm)
Wavelength (nm) 10
12
(c)
(d)
10
8
Pure CuO 1% Zn-CuO 3% Zn-CuO 4% Zn-CuO 5% Zn-CuO
8
6 6
4
ε2
ε1
Pure CuO 1% Zn-CuO 3% Zn-CuO 4% Zn-CuO 5% Zn-CuO
4
2
2
0
0 300
400
500
600
700
800
900
1000
300
400
500
600
700
Wavelength (nm)
Wavelength (nm)
Fig. 4 Variation of optical constants with wavelength for all thin films: (a) n, (b) k, (c) e1 and (d) e2
electric polarization. Therefore, the value of e1 is high in this region as can be seen in Fig. 4(c). For angular frequency higher than the IR region, the oscillating frequency of the atoms/electrons cannot follow the oscillation of the incoming light so atomic polarization does not appear and the value of e1 starts decreasing, as shown in Fig. 4(c). At further higher frequencies of the incoming light, the atomic polarization cannot even follow the oscillation of incoming light so that e1 attains value equal to the value in vacuum (e1 = 1) [24]. Figure 4(c) shows that e1 has a broad peak from 300 to 700 nm along with a maximum at k * 460 nm. It shows another broad peak in IR region from 700 to 1,000 nm, showing a maximum at k * 940 nm except for 1 % Zn–CuO thin film. The imaginary part of dielectric constant (e2) is the overview of optical properties of a material and is related to refractive index (n) and extinction coefficient (k) [24].
and stagnates up to 600 nm. Again it increases up to *820 nm and then it decreases again. The maximum of e2 occurs around 820 nm revealing maximum absorption is in the IR region for all the samples except 1 % Zn–CuO thin film. Also the peak value of e2 for this thin film is smaller than that of the other thin films due to higher compressive stress present in this film as the imaginary part e2 decreases with increasing stress. This is due to the distortion caused by the increase in stress, which enhances the phonon scattering and weakens the optical response of the material, thus resulting in a reduction of the e2 value [25]. In addition, in case of ionic crystals, atomic polarization is observed in the infrared region. The extinction coefficient also shows that maximum absorption is in the IR region and variation of e2 confirms this result. CuO is a direct bandgap semiconducting oxide and thus the bandgap holds the relation [24]
e2 ¼ 2nk
1=2 aE ¼ A E Eg
ð7Þ
Figure 4(d) represents the variation of imaginary dielectric constant (e2) with the wavelength. It shows that e2 increases from 300 to *400 nm, then decreases slightly
ð8Þ
where E is the photon energy, Eg is the optical bandgap energy of the material, a is the absorption coefficient and
Effect of zinc induced compressive stresses 1.5
2.0
2.5
3.0 3.5 E (eV)
4.0
1.25
4% Zn-CuO
( α hν) 2.1012 (eV/cm)2
(d)
2.0 1.5 1.0 0.5
1.5
1.0
0.5
0.0
1.5
4.5
2.0
2.5
3.0 E (eV)
3.5
4.0
1.5
4.5
2.0
2.5
3.0 E (eV)
3.5
4.0
2.8
5% Zn-CuO
(e)
Macrostrain (10-3)
1.5
( α hν) 2.1012 (eV/cm)2
0.5
0.0
0.0
2.5
1.0
3% Zn-CuO
(c)
1.00 0.75 0.50 0.25
-16
(f) 2.6
-12
2.4
-8
2.2
-4
2.0
0.00
0.0 1.5
2.0
2.5
3.0
3.5
4.0
1.8 1.5
4.5
Optical band gap energy (eV)
0.5
( α hν) 2.1012 (eV/cm)2
1.0
2.0
1% Zn-CuO
(b)
Pure CuO
(a)
( α hν) 2.1012 (eV/cm)2
( α hν) 2.1012 (eV/cm)2
1.5
2.0
2.5
3.0
3.5
4.0
4.5
E (eV)
E (eV)
0 2 4 Zn doping concentration (at. wt. %)
6
Fig. 5 The plot of (aht)2 versus photon energy for all thin films: (a) pure CuO, (b) 1 % Zn–CuO, (c) 3 % Zn–CuO, (d) 4 % Zn–CuO and (e) 5 % Zn–CuO, (f) shows variation of optical bandgap energy and macrostrains with doping concentration of Zn Table 2 Macrostrain, Raman shift and optical bandgap energy of pure and doped CuO thin films Sample name Pure CuO
Macrostrain (910-3) -5.2
Raman shift (cm-1)
Optical bandgap energy (eV)
294
2
1 % Zn–CuO
-12.6
297.8
2.4
3 % Zn–CuO
-11.6
297
2.5
4 % Zn–CuO
-4.7
294.1
2.3
5 % Zn–CuO
-4.3
299.2
1.95
A is a constant. The bandgap has been determined using Tauc’s plot as shown in Fig. 5(a)–5(e). The value of optical bandgap energy (Eg) has been obtained by extrapolating the linear portion of (aht)2 onto energy axis at a = 0. Figure 5(f) shows the variation of optical bandgap energy with the increasing doping concentration of zinc. It is evident from this figure that optical bandgap energy (Eg) increases from 2 to 2.5 eV as the doping concentration increases from zero up to 3 atomic wt%. Then Eg decreases for higher doping concentrations. The optical bandgap energy is plotted along with the corresponding macrostrains present in all thin films in Fig. 5(f) as a function of Zn concentration. It can be seen that the optical bandgap is broadened as the uniaxial compressive stress increases, consistent with the results reported in the literature [25]. The produced stresses create forbidden energy levels in the higher states of conduction band, leading to the widening
of the optical bandgap energy. The maximum shift in the value of optical bandgap energy calculated in the present work is 0.5 eV, which is 20 %. It is well known that the bandgap energy (Eg) increases with increasing compressive stresses but in this work, although the maximum stress is present in 1 % Zn–CuO thin film, which should show maximum Eg but its crystallite size is less than that of 3 % Zn–CuO thin film. So a little reduction in compressive stresses and an increase in crystallite size in 3 % Zn–CuO thin film as compared to 1 % Zn–CuO cause the highest bandgap energy (Eg) for 3 % Zn–CuO thin film. At higher Zn doping concentrations (4–5 atomic wt%), Zn?2 ions are not substituted by Cu?2 ions in CuO crystal network, as observed due to appearance of ZnO Raman peak in 5 % Zn–CuO thin film. So the 4–5 % Zn–CuO thin films show decreased stresses close to the pure CuO thin film, which cannot create the forbidden energy levels. Hence the optical bandgap energies also decrease for these thin films close to pure CuO thin film. The values of macrostrain, Raman shift and optical bandgap energies with increasing Zn doping concentration from 0 to 5 % are summarized in Table 2.
4. Conclusions Phase pure and Zn doped CuO thin films have been deposited by PLD technique at varying concentration of
H Faiz et al.
Zn. The effects of dopant induced stresses on microstructural and optical properties of these thin films have been investigated using XRD, Raman spectroscopy, SEM and spectroscopic ellipsometry. XRD analysis has revealed the polycrystalline structure with monoclinic CuO phase only for all the thin films. The presence of compressive stresses is confirmed in all the thin films and is highest for 1 % Zn– CuO thin film. The three Raman active modes of CuO, Ag (297 cm-1), Bg1 (341 cm-1) and Bg2 (631 cm-1) have been observed in the Raman spectra of all the thin films. An up-shift in the peak position of main Ag mode of Zn– CuO thin films, as compared to that of phase pure CuO thin film has been observed, as a consequence of the compressive stresses present in the thin films. SEM micrographs show smooth surface morphology for all the thin films. Analysis by spectroscopic ellipsometry shows that the optical properties (w, D, n, k, e1, e2 and optical bandgap energy) of CuO thin films are strongly affected by the stresses induced due to the Zn incorporation. Optical bandgap energies have been found to be highest for thin films having high compressive stresses. These thin films could have possible potential applications in optical and optoelectronic industries. Acknowledgment We acknowledge Mr. Irfan, Co-Director PITMAEM, PCSIR Laboratories Lahore, Pakistan for their cooperation in SEM analysis.
References [1] A N Banerjee, S Kundoo and K K Chattopadhyay Thin Solid Films 440 5 (2003)
[2] V Figueirdo et al. Appl. Surf. Sci. 254 3949 (2008) [3] M Engin, F Atay, S Kose, V Bilgin and I Akyuz J. Elect. Mater. 38 787 (2009) [4] S C Ray Sol. Energ. Mat. Sol. C 68 307 (2001) [5] M S Hegde Proc. Indian Acad. Sci. (Chem. Sci.) 113 445 (2001) [6] K P Muthe et al. Thin Solid Films 324 37 (1998) [7] K H Yoon, W J Choi and D H Kang Thin Solid Films 372 250 (2000) [8] J F Pierson, A T Keck and A Billard Appl. Surf. Sci. 210 359 (2003) [9] G Papadimitropoulos, N Vourdas, V Em Vamvakas and D Davazoglou J. Phys. Conf. Series 10 182 (2005) [10] S Kose, F Atay, V Bilgin and I Akyuz Mater. Chem. Phys. 111 351 (2008) [11] M F Al-Kuhaili Vacuum 82 623 (2008) [12] C S Barrett and T B Massalski Structure of Metals (Germany: Pargamon press Oxford) p 204 (1980) [13] V Bilgin, S Kose, F Atay and I Akyuz J. Mater. Sci. 40 1909 (2005) [14] R Loudon Adv. Phys. 13 423 (1964) [15] N Mukherjee et al. Mater. Lett. 65 3248 (2011) [16] A M Shanid, M A Khadar and V G Sathec J. Raman. Spectrosc. 42 1769 (2011) [17] T Yua, X Zhaoa, Z X Shena, Y H Wub and W H Sud J. Cryst. Growth 268 590 (2004) [18] H F Goldstein, D Kim, P Y Yu and L C Boumet Phys. Rev. B41 7192 (1990) [19] S Guha, D Peebles and T J Wieting Phys. Rev. B43 13092 (1991) [20] N D Hoa, N V Quy, H Jung, D Kim, H Kim and S Hong Sensors. Actuat. B 146 266 (2010) [21] I De Wolf Semicond. Sci. Technol. 11 139 (1996) [22] D Chuhan, V R Satsangi, S Dass and R Shrivastav Bull. Mater. Sci. 29 709 (2006) [23] K Siraj, M Khaleeq-ur-Rahman, S I Hussain, M S Rafique and S Anjum J. Alloys Comp. 509 6756 (2011) [24] H Fujiwara Principles of Optics in Spectroscopic Ellipsometry (England: Wiley) (Japanese edition) p 23 (2003) [25] Y Ping et al. Chin. Phys. B 21 016803 (2012)