ZnO QDs Deposited on Si by Sol-Gel Method: Role of Annealing Temperature on Structural and Optical Properties

Zinc acetate, propanol, and ethanolamine are used to synthesize ZnO quantum dots (QDs) via the sol–gel method. Spin coating at 220°C is then employed to prepare ZnO thin films consist of QDs. The effects of thermal annealing temperature on the structural and optical properties of the samples are investigated. The prepared QDs are characterized using X-ray diffraction (XRD), field emission scanning electron microscopy (FESEM), and photoluminescence spectroscopy (PL). XRD patterns reveal the crystalline nature of samples existing in the hexagonal wurtzite phase. Increasing annealing temperatures bring about changes to the crystal orientation through the c-axis and increase the size of the QDs from ~10 nm to ~22 nm; this increment can be explained by a coarsening mechanism and the Ostwald ripening process. The observed broadening of X-ray peaks confirms the evolution of crystalline phases in the ZnO QDs. Quantitative analysis of size-dependent strain effects is performed through the Williamson-Hall model. The QD mean size estimated from FESEM and XRD displays high consistency. Room-temperature PL peaks of 3.37ev are attributed to the radiative recombination of electrons and holes from the ZnO QDs/Si interface.


Introduction of the Problem
Zinc oxide (ZnO),an n-type metal oxide semiconductor with a wide band-gap(3.36 eV), has received considerable research attention because of its strong exciton binding energy of 60 meV at room temperature (Liu et al., 2013), large saturation velocity of 3.2×10 7 cm s −1 (Tekmen et al., 2013), and non-toxicity (Thomas et al., 2014). In general, ZnO exists as a hexagonal structure with c/a = 1.65. This form, however, may vary according to the growth conditions applied. Several synthesis methods, including sol-gel, hydrothermal, solvothermal, sonochemical, combustion, chemical vapor deposition (CVD), vapor-liquid solid (VLS), and pulsed laser deposition, have been developed over the years to control the growth of ZnO nanostructures (Kamalianfar et al., 2014;Zak et al., 2013;Razali et al., 2011;Yousefi et al., 2012;Yousefi et al., 2010;Zhanget al., 2004).Annealing is considered a promising route to tune the properties of ZnO nanostructure (Ashrafiet al., 2004;Radet al.,2013;Zamiri et al., 2013). The crystallite size and lattice parameters of a crystal system are greatly affected by the annealing temperature. The lattice strain and crystallite size are the two main factors leading to broadening of diffraction peaks. Lattice strain originates from lattice imperfections or dopants and alters peak intensities and positions (Ramgir et al., 2006). Uniform strains change peak positions, and non-uniform strains promote peak broadening.
In this study, we fabricate ZnO QDs on silicon substrates using a simple sol-gel method following by spin coating. The effects of annealing temperature on the crystalline structure and surface morphology of the samples are investigated in detail by X-ray diffraction (XRD), energy dispersive X-ray spectroscopy (EDX), and field emission scanning electron microscopy (FESEM). The results of this work may further establish the belief that by carefully controlling their process parameters, ZnO QDs may be utilized in spintronic applications. ZnO is used in several applications due to its high transparent and good electrical conductivity. Some of these applications are gas sensor and photovoltaic devices, water waste treatment, photo catalysis, chemical remediation, photo-initiazation of polymerization reactions, quantum dot devices, and solar energy conversion.

Experimental Details
The ZnO films are deposited by a sol-gel process using a spin coating method onto p-Si (100) substrates.Firstly, the NH 4 OH+H 2 O 2 +6 deionized (DI) should be boiled for 10 min with (18 MΩ-DI) water. Then before forming a ZnO layer on p-Si substrate, the native oxide on the polish surface of the substrate is removed from HF: H 2 O (1:10) solution. Finally, the wafer is rinsed in DI water at last dried by N 2 gas.
The zinc precursor solution is prepared by dissolving zinc acetate dihydrate (ZnAc; the starting material) in 2-propanol (the solvent) to obtain a final concentration of 0.1 M. Ethanolamine (EA; the stabilizer) is added to this mixture to obtain a stable and transparent solution. The molar ratio of ZnAc to EA is maintained at 1:1. The sol is stirred at 60 °C until a transparent and homogeneous solution is obtained.
The sol is allowed to drip onto the center of the substrate, which is immediately spun at 3000 rpm for 20 s. After each coating, the films are dried at 220°C for 10 min. The coating-drying cycles are repeated 10times, and the films are annealed at 425, 450, 475, or 500°C for 2h in ambient atmosphere of the samples are characterized by XRD (Bruker D8 Advance diffractometer) with Cu-Kα 1 radiation(1.5406 A°) at 40 kV and 100 mA over the 2 range of 25°-75°. A scanning speed of ~ 1.2°/min and a resolution of 0.011°are also employed. The elemental compositions of the samples are analyzed by an EDX diffractometer attached to a FESEM instrument (ZEISS SUPRA). Figure 1 illustrates the XRD patterns ofZnO QDs annealed at 425-500 °C. The peaks detected are indexed to the hexagonal structure of ZnO and conform to those described in PDF No. 00-036-1451. The intensity of all diffraction peaks increases with increasing annealing temperature. This increment is ascribed to the impacts of disorders or defects removed by the annealing process; removal of these defects decreases the strain and improves add atom mobility, thereby enhancing the crystallization quality of ZnO by providing an abundant supply of thermal energy (Mandal et al., 2008).

Results and Discussion
The lattice parameters are calculated by (Gondal et al., 2009): The planer distance is determined by Bragg equation as follows:

= 2 sin
Where d is the distance between two planes with lattice geometry The hexagonal structure volume is calculated by (Cullity,1956): Equations 1 and 2 are used to calculate the lattice constant values (a &c) and volume which are given in table1. The Zn-O bond length L in the wurtzite structure can be estimated from the expression (3) (Yakuphanoglu F et al., 2007(Yakuphanoglu F et al., , p.2180 which are given bellow: The texture coefficient (TC) reflects the texture of a particular plane and indicates the preferred growth orientation of a crystal; this coefficient is defined as (Barret &Massalski, 1980): Where TC (hkl) is the texture coefficient of the (hkl) plane, n is the number of reflections, and I (hkl) and I° (hkl) are the measured and relative intensities of the corresponding plane given in JCPDS Card No. 36-1451, respectively. Variations in TC among the peaks of the wurzite lattice are shown in Table 1. The highest TC occurs in the (101) plane. The dislocation density (Table 2), which represents the number of defects in the film, is determined from the following equation. Dislocation densities exhibit a marked reduction with increasing annealing temperature, indicating a lower concentration of lattice imperfections (Ergin et al., 2009).
Where n is a factor that equals unity at the minimum dislocation density and D is the grain size. The values of strain along the c-axis (ɛ zz ) are determined by using the following formula (Ghosh et al., 2004): Where c 0 is equal to 5.20661 A° which is the lattice parameter of the strained ZnO films obtained from unstrained lattice parameter of ZnO that extracted from X-ray diffraction data (JCPDS 36-1451). The strain value can be negative (compressive) or positive (tensile). As can be seen from Table 1 the heat-treated process gives smaller c values than the ZnO powder. Zhang et al., have also reported slightly smaller c values for nanocrystalline ZnO thin films prepared on p-type Si(1 0 0) substrates by a sol-gel method followed by heat-treatment (Zhang et al., 2006). ZnO QDs have a number of defects such as oxygen vacancies, lattice disorders, etc. As a result of annealing, these defects are removed and the lattice contracted. Also lattice relaxation due to dangling bonds should be considered. The dangling bonds on ZnO surface interact with oxygen ions from the atmosphere and due to electrostatic attraction, lattice is slightly contracted.
For hexagonal lattice, the stress (σ film ) parallel to film surface can be obtained from equation (7) (Shahzad et. al., 2013): The elastic stiffness constants C ij for single crystal wurtzite ZnO are C 11 = 208.8 GPa, C 12 = 119.7 GPa, C 13 = 104.2 GPa, and C 33 = 213.8 GPa. The relevant ɛ zz and σ film are listed in Table 2. Strain obviously decreases with increasing annealing temperature, because the small amount of intrinsic stress is presented which normally arise from expansion of film volume during increasing of annealing temperatures. The negative sign of ɛ zz belongs to thermal stress.
The XRD peak broadening analysis is used to evaluate the nanocrystallite size. The sizes are estimated from Scherrer equation= cos .
where D is the crystallite size, λ is the wavelength of the radiation (1.5406 Å for Cu-K α radiation), K is a constant with a value of about 0.94, β is the maximum full width at half maximum (FWHM), and θ is the peak position. The diffraction peak at (101) is considered with the maximum intensity. Table 2 compares the estimated sizes obtained from the three methods described above. Note that the Scherrer equation is limited in making accurate estimations of the crystallite size because it does not account for the strain present in the lattice because of imperfections or defects. Therefore, other methods are necessary to estimate the QD size accurately; one of these methods involves the use of the Williamson-Hall (W-H) plot. Here, the strain is assumed to be uniform in all crystallographic directions; this assumption considers the isotropic nature of the crystals wherein all of the material properties are independent of the direction along which they are measured.
FESEM images of the samples are clearly shown inFigure2, which reveals the presence of high-density ZnO QDs. The particle sizes (from ~10 nm to ~22 nm) estimated from the FESEM images conform to those www.ccsenet.org/mas Vol. 10, No. 4;2016 determined from the XRD results. The integrated volume and sizes of the QDs increase during annealing as some of the ZnO atoms move from the wetting layer into the QDs. The dot distributions coarsen during annealing over a portion of the temperature range. Thus, the observed coarsening represents the size increment toward the lowest-energy configuration of the system consistent with increasing temperature. The Ostwald ripening process is responsible for the coarsening mechanism. Consequently, the free energy of the dots decreases monotonically with increasing QD size. The term β cosθ is plotted with respect to 4·sinθ to determine the preferred orientation peaks of ZnO QDs with respect to the wurtzite hexagonal phase. The slope and y-intersect of the fitted line respectively represent the strain and particle size of the QDs. The plots display a positive strain for the ZnO QDs, which could be attributed to the lattice expansion observed during calculation of the lattice parameters. The results of W-H analysis of the ZnO QDs are shown in Figure 3. It is clearly seen from the table that by increasing the annealing temperature, the strain is reduced due to inter-diffusion of Si substrate into ZnO and formation of ZnO-Si alloy. The strain may be extracted from the slope and the crystalline size may be extracted from the y-intercept of the line obtained by fitting the data.
The EDX spectra of the samples are illustrated in Figure 4. All of characteristic peaks match the unique elements of Zn and O well. The peak of O decreases with increasing annealing temperature because of Zn diffusion into the Si substrate, which creates more oxygen vacancies.  The room-temperature PL spectra of the QDs are measured to investigate their optical properties as well as that of the resultant film. Figure 5 show that the PL spectra of the ZnO QDs show some dependence on the annealing temperature over the range of 425-500 °C. The peak at 368 nm (3.37eV) is characteristic of neutral-donor-bound-exciton (D0X) emission (Zhanget al., 2004;Ashrafi et al., 2004).The appearance of a weak shoulder at ~2.89 eV (violet region) is attributed to radiative recombination from intrinsic defects in the ZnO dots. Quantum confinement effects are not observed in the PL spectra because the size of the QDs is considerably larger than the exciton Bohr radius of ZnO (2.34 nm) (Wang et al., 2011).

Conclusion
Nearly spherical shaped high density ZnO QDs /Si are synthesized via facile yet accurate sol-gel method and spin coating. The influence of annealing temperature on ZnO QDs is found to alter the crystalline structures via www.ccsenet.org/mas Modern Applied Science Vol. 10, No. 4;2016 defect mediated strain interactions and broadens the spectral width. This in turn modifies the lattice parameters and elastic constants of the single crystalline QDs phase. The QDs size estimation through different methods reveals consistent measurements. The XRD measurement exhibits the wurtzite structure. The observed broadening is analyzed using the Scherrer formula and the W-H plot. The crystallite size is increased with increasing of annealing temperature. The FESEM image manifested presence of ZnO QDs with mean dot size between ~10 to 22 nm. EDX spectra confirm the inter-diffusion of ZnO into substrate layer, resulted to reduce the strain. The appearance of UV and blue-violet band in the room temperature PL spectra are ascribed to the electron-hole recombination from excited state to the ground-state and intrinsic defects belonging to ZnO QDs respectively. Quantum confinement effect is not observed as the QDs sizes are significantly larger than exciton Bohr radius for ZnO.