Optical and Structural Properties of Nickel Doped Zinc Oxide Grown by Metal Organic Chemical Vapor Deposition (MOCVD) at Different Reaction Chamber Conditions

Therefore, it is desirable to tune the bandgap of ZnO with the purpose of enhancing the solar cell’s performance. Such tunability in the band gap and consequently the optical absorption shift of ZnO are achievable by doping with transition metals (TMs) such as Cu, Fe, Ni, and Mn.^[5] In solar cell and photoelectrochemical (PEC) applications, it is necessary to achieve ZnO based materials with a lower bandgap, and this depends on the dopant type and the synthesis technique. Among TM sources for doping ZnO, Ni doping in ZnO could shift the optical band gap towards the red wavelengths.^[6] The chemical stability of Ni²⁺ while occupying Zn²⁺ sites makes it unique and make it identified as one of the most efficient doping elements as it enhances ZnO with optical and electrical properties with respect to unpaired electrons in the d-orbitals of Ni, which can interact with carriers in ZnO.^[7] The effects of doping Ni in ZnO are dramatically dependent on the growth technique used. For instance, zinc nickel oxide (Ni-doped ZnO) grown by spray pyrolysis deposition on quartz substrates showed a reduction in bandgap from 3.43 to 2.87 eV with increasing the Ni content from 2% to 15%.^[8] Meanwhile, The bandgap reduction from 3.2 to 1.4 eV is observed in the Ni-doped ZnO grown by DC/RF magnetron sputtering with the Ni doping up to 7%.^[9] The Ni-doped ZnO nanoparticles synthesized by the sol-gel technique had an increase in the bandgap from 3.29 to 3.32 eV with the Ni doping up to 6%.^[10]

Metal organic chemical vapor deposition (MOCVD) is an established deposition technique for controlling the properties of ZnO growth. To this end, numerous reports have been focused on the parameters (i.e. reactor gas delivery system and the reaction chamber) for the growth of II-VI semiconductors, specifically ZnO, by MOCVD technique.^[11,12] The structural and optical properties of ZnO could be systematically tuned by the growth conditions such as substrate temperature, chamber pressure, and disk rotation. However, there have been limited investigations for the growth of Ni-doped ZnO with regards to the optimum growth temperature and pressure for the decomposition of Ni precursor for replacing Zn⁺² with Ni⁺².^[13] In this work, the structural and optical properties of Ni-doped ZnO grown by MOCVD were studied as a function of the growth system parameters. The Ni incorporation in ZnO structure and the resulting structural and optical properties of Ni-doped ZnO were strongly dependent on the substrate temperature and growth system pressure.

2. Experimental section

Ni-doped ZnO thin films were grown on the c-plane (0001) sapphire substrates in a vertical injection MOCVD system.^[14] Diethylzinc (DEZn), nickelocene (Cp₂Ni), and oxygen (O₂) were used as precursors for Zn, Ni, and O, respectively, and nitrogen (N₂) was used as a carrier gas. The bubbler temperatures of DEZn and Cp₂Ni sources were maintained at 5 °C and 90-95 °C respectively to adequately vaporize the sources.^[13] DEZn begins to decompose at 357 °C but high crystal quality is achieved by increasing the growth temperature (450 °C and above) at pressure ≥ 30 Torr.^[14-17] The structural analysis of grown ZnO by the XRD shows an increased crystallinity for the growth temperature above 400 ℃.^[14] Also, the strong photoluminescence peak, a sign of high crystal quality, was observed for the ZnO grown samples by MOCVD, when the growth temperature was 550 ℃.^[17] On the other side, Cp₂Ni decomposition and NiO formation with (111) orientation occur at temperatures above 275 ℃.^[18-19] The Cp₂Ni deposition rate increases with rising the growth temperature but for temperature higher that 400 ℃, the deposition rate is saturated when the chamber pressure is 100 Torr.^[19] Hence, the growth conditions should be optimized in the purpose of decomposing both Cp₂Ni and DEZn and the oxidation of Ni and Zn concurrently. Two sets of samples were grown in this work, i.e., set A at 450 °C and set B at 550 °C with the chamber pressures from 22 to 100 Torr, at a disk rotation of 600 rpm. The Reynolds number for these conditions ranges from 700 to 2200, and results in a stable and uniform laminar gas flow for the adsorption of reactants on the sapphire substrates.^[15] DEZn and CP₂Ni had flow rates of 1 and 2.7 standard cubic centimeters per minute (sccm) respectively, considering a possibly low decomposition/ adsorption rate of the dopant. The oxygen flow rate was maintained at 300 sccm for both sets (VI/II ratio of 300). Suitable N₂ carrier gas flow rate was chosen to achieve an overall stable gas flow through the growth system. The growth conditions of the samples are summarized in Table 1.

Table 1. MOCVD growth details of Ni-doped ZnO.

The crystal structure of MOCVD-grown Ni-doped ZnO was investigated using a Bruker D-8 Focus X-ray diffractometer with a Cu X-ray source of 1.54 Å. The absorption measurements were acquired using an Agilent Cary 300 UV-Vis spectrophotometer. The room temperature (RT) photoluminescence (PL) spectra were measured using a Mini PL spectrometer by Photon Systems. Elemental composition of the samples was studied using an energy dispersive X-ray spectroscopy (EDX) Quanta 650.

3. Results and discussion

3.1 Structural characterization of Ni-doped ZnO

X-ray diffraction (XRD) diffraction patterns of the grown Ni-doped ZnO samples are shown in Fig. 1. The crystal orientation in the (002) (2θ of ~34.06°) direction is observed, along with ZnO (100) at 2θ=30.6°, ZnO (101) at 2θ=35.88° and NiO-related (111) phase at 2θ=37.11°. The peak positions and full width half maximum (FWHM) of the ZnNiO (002) peak are listed in Table 2. As seen in Fig. 1, the peak intensity at ZnO (002) decreases dramatically from the 110% of the sapphire peak intensity to 65% of the sapphire peak intensity when the chamber pressure increases and reaches to 100 Torr, which can be signs of defects in A-3 and B-3 samples. Moreover, the FWHM of the peaks increased with an increase in the pressure, which shows a deterioration in the crystal quality. The shoulder peaks in some samples show the signs of crystal oscillations or defects that could be introduced by Ni-doping.

Fig. 1 XRD patterns of MOCVD-grown Ni-doped ZnO with varying growth conditions.

Table 2. XRD measurement results of the Ni-doped ZnO
at (002) peak.

3.2 Optical characterization of Ni-doped ZnO

Room temperature (RT) optical absorption spectra of the Ni-doped ZnO thin films were investigated by a UV-VIS spectroscopy in the range of 200 - 800 nm. The results are shown in Fig. 2(a). A red-shift in sample B-3 is likely caused by the lattice defect, as observed in the XRD results. However, with regards to maintaining the crystalline nature and the phase purity of the ZnO in samples A-1 and A-2 (as XRD results shown), it may be concluded that the observed red shift in these samples is due to the formed impurity in the bandgap of the ZnO. The step at ~3.5 eV is due to the switching of lamps in the measurement equipment.

The Tauc plots are also used to determine the energy band gap of Ni-doped ZnO. The absorption coefficient (α) is depicted in Fig. 2(b) in terms of the wavelength. As seen, samples A-1 and A-2 have a direct bandgap equal to 3.27 and 3.25 eV respectively, corresponding to the extrapolated linear lines (dashed lines in Fig. 2(b)) on the x-axis. Meanwhile, both samples A-3 and B-3 with a 100 Torr growth pressure have indirect bandgaps with the band edge shift toward 3.1 eV and less.

Next, RT photoluminescence (PL) spectra of Ni-doped ZnO samples are shown in Fig. 3 to study the luminescence and possible defects in the grown Ni-doped ZnO crystal structures. The PL excitation wavelength of around 6 eV was used. As seen, only samples A-1 and A-2 show strong

  (a)                                                (b)

Fig. 2 (a) Absorbance spectra of MOCVD-grown Ni-doped ZnO and (b) Tauc plots of MOCVD-grown Ni-doped ZnO.

UV emission peaks centered around 3.3 eV, which corresponds to the ZnO crystal PL spectra and confirms a good crystal quality at lower pressures (≤ 30 Torr).[1-3] The strongest PL peak is observed in the Ni-doped ZnO grown at 450 ℃/30 Torr. The PL results of other samples, specifically A-3 and B-3 (grown at 100 Torr pressure), show a wide peak referring to a deep-level emission, which can be result of the recombination of electrons deeply trapped in oxygen or zinc vacancies. The emissions in the range between 2-2.6 eV, are associated with singly- and doubly- ionized oxygen vacancies.^[20] The emission intensity in this range for the sample A-1, which is grown at a lower temperature and pressure, is weaker than other grown samples Moreover, a peak appears at 4.13 eV in the PL spectra of the sample corresponding to the NiO phase, which is also observed in the XRD patterns.^[21] ZnO has a stable PL peak at 3.3 eV while NiO has a PL in the range of 3.5-4.2 eV.^[18,21] Considering the constituent elements of the Ni-doped ZnO and presence of a NiO-like phase in the X-ray diffraction measurement, the peak at 4.13 eV could be due to the NiO related phase.

                                     (a)                                           (b)

Fig. 3 (a) RT PL spectra of the Ni-doped ZnO samples with a growth temperature of 450 °C and varying pressure from 20 to 100 Torr. (b) RT PL spectra of Ni-doped ZnO samples with a growth temperature of 550 °C and varying pressures from 20 to 100 Torr.

To understand the effect of growth conditions on the Ni incorporation in ZnO structure, the chemical compositions of these grown Ni-doped ZnO are measured by the energy dispersive X-ray (EDX) spectroscopy analysis. The EDX results of samples A-1, A-2 and B-3, which have larger band edge shifts compared to others, are shown in Fig. 4 and the weight ratios of incorporated Zn, Ni and O in these samples are demonstrated in Table 3. It can be deduced that the Ni incorporation in the Ni-doped ZnO affects both zinc and oxygen contents, which can be a result of defect states introduced by the Ni-doping. In samples A-1 and A-2, O (wt%)/Zn (wt%) is 0.64 and 0.35 respectively, where their thickness is around 200 nm. However, for the sample B-3 with a thickness < 10 nm, O (wt%)/Zn (wt%) is 9.36 where Ni does not incorporate in the grown ZnO structure and the high oxygen to zinc ratio is likely because the oxygen is most likely from the sapphire instead of from any ZnO film with regards to a minimal film growth in the sample B-3.

     Fig. 4 Energy dispersive X-ray (EDX) spectroscopy for samples (a) A-1, (b) A-2 and (c) B-3.