Synthesis of Ni-doped ZnO nanostructures by low-temperature wet chemical method and their enhanced field emission properties

Abstract

In this study, we report an enhancement in the field emission (FE) properties of ZnO nanostructures obtained by doping with Ni at a base pressure of ∼1 × 10⁻⁸ mbar, which were grown by a simple wet chemical process. The ZnO nanostructures exhibited a single-crystalline wurtzite structure up to a Ni doping level of 10%. FESEM showed a change in the morphology of the nanostructures from thick nanoneedles to nanoflakes via thin nanorods with an increase in the Ni doping level in ZnO. The turn-on field required to generate a field emission (FE) current density of 1 μA cm⁻² was found to be 2.5, 2.3, 1.8 and 1.7 V μm⁻¹ for ZnO (Ni0%), ZnO (Ni5%), ZnO (Ni7.5%) and ZnO (Ni10%), respectively. A maximum current density of ∼872 μA cm⁻² was achievable, which was generated at an applied field of 3.1 V μm⁻¹ for a Ni doping level of 10% in ZnO. Long-term operational current stability was recorded at a preset value of 5 μA for a duration of 3 h and was found to be very high. The experimental results indicate that Ni-doped ZnO-based field emitters can open up many opportunities for their potential use as an electron source in flat panel displays, transmission electron microscopy, and the generation of X-rays. Thus, the simple low-temperature (∼80 °C) wet chemical synthesis approach and the robust nature of the ZnO nanostructure field emitter can provide prospects for the future development of cost-effective electron sources.

---

1. Introduction

Zinc oxide is one of the most striking semiconductors in the category of 1D and 2D nanomaterials, which is due to its remarkable and multifunctional properties such as a direct band gap of 3.37 eV and a high exciton binding energy of 60 meV.^1,2 ZnO also displays numerous morphologies such as nanocombs,^3,4 nanotubes,⁵ nanosprings,⁶ nanorods,⁷ nanorings,⁸ nanowires⁹ and nanoflowers.¹⁰ Owing to its significant physical and chemical properties, it is a potential candidate for various applications such as photoelectronic devices,^11,12 nanosensors,^13–17 nanogenerators,¹⁸ electronic devices,^19,20 dye-sensitized solar cells,²¹ spintronics,^22,23 detection of metal ions,²⁴ and photocatalysis,²⁵ etc. In the last few decades ZnO has been shown to be one of the most promising materials for FE devices, which is because of its high thermal stability and low electron affinity, as well as resistance to oxidation in harsh environments.^26,27 There are many reports on the enhancement of FE by ZnO; for example, Farid et al. observed enhanced FE properties in Cu-doped ZnO nanocomposite films synthesized by an electrochemical method²⁸ and Chang grew Ni-doped ZnO nanotower arrays on a silicon substrate using a thermal evaporation method (T = 1100 °C) and observed enhanced optical and field emission properties.²⁹ Similarly, Xing et al. reported the growth of ultrathin single-crystal ZnO nanobelts using an Ag-catalyzed vapor transport method and FE properties.³⁰ The key issues are the operating voltage range and the stability of the emission current, as the values of the turn-on and threshold fields are dependent on the morphology (shape and size) as well as the intrinsic physical properties such as the work function and electrical conductivity of the emitter. The approach towards improving the FE properties of semiconductors comprises either tailoring the geometry of the emitter or modifying its electronic properties. The first approach has limitations on the reduction in the size and shape of the nanostructures. Although the synthesis of an array of well-spaced anisotropic nanostructures that possess a very low apex radius (typically 20 nm) is feasible, the mechanical sturdiness of such an emitter is questionable. It is predictable that the mechanical stress induced by the presence of an intense electrostatic field may result in negative effects such as bending/fracture of the emitter. Similarly, for the second approach, the modified electronic properties obtained by doping/mixing with suitable elements are critical and sometimes the chemical reactivity of such structures may lead to a decline in their emission performance. Therefore, a suitable combination of both approaches is considered to be a promising method for improving FE performance, which has been successfully attempted by several researchers. Here, we report the synthesis of Ni-doped ZnO by low-temperature wet chemical methods with an enhancement in FE properties. We found that after doping with Ni the turn-on field decreased to ∼1.7 V μm⁻¹ and the emission current density increased to 872 μA cm⁻² at a field of 3.1 V μm⁻¹. This significant enhancement in FE properties was due to a change in morphology as well as an increase in the number of electrons in the conduction band.

2. Experimental details

Pure and Ni-doped ZnO nanostructures were synthesized by a low-temperature (synthesis temperature 80 °C) wet chemical method. Analytical grade zinc nitrate hexahydrate and nickel nitrate hexahydrate were used as the raw materials. A 100 mM solution of zinc nitrate was then prepared for doping with Ni and an appropriate amount of nickel nitrate was added to zinc nitrate, which were dissolved in 100 mL double-distilled water and then stirred. After constant stirring for 30 min, an aqueous ammonia solution was added dropwise continuously to the solution until the pH approached ∼11. Finally, a clear solution was obtained in the case of pure ZnO and sea-blue solutions in the case of Ni-doped solutions. Then, the solutions were kept at 80 °C for 2 h, and then the precipitates were separated and dried overnight at room temperature. Finally, the dried samples were annealed at 150 °C (for 2 h). The detailed synthesis is already reported elsewhere.³¹ Samples with different Ni doping levels of 0%, 5%, 7.5% and 10% were named as ZNi0, ZNi5, ZNi7.5 and ZNi10, respectively. These were the final compositions of the materials after analysis by EDX. The phase purity and morphology of ZnO nanostructures were investigated by X-ray diffraction (Bruker D8 Advance) and field emission scanning electron microscopy (Zeiss Supra 55). Field emission investigations were carried out in a planar diode configuration in all-metal ultrahigh vacuum (UHV) chambers, which were evacuated to a base pressure of ∼1 × 10⁻⁸ mbar. A typical diode configuration consisted of a phosphor-coated semitransparent indium tin oxide glass disc (a circular disc having a diameter of ∼50 mm), which acted as an anode. A Ni-doped ZnO nanostructure sprinkled onto a piece of UHV-compatible conducting carbon tape pasted on a copper rod holder (diameter ∼ 5 mm) served as the cathode. The emission current was measured by a Keithley electrometer (6514) by sweeping a DC voltage applied to the cathode in steps of 40 V (0–40 kV, Spellman, U.S.). The stability of the field emission current was investigated using a computer-controlled data acquisition system with a sampling interval of 10 seconds. Special care was taken to avoid any leakage current using shielded cables and proper grounding. Field emission images were recorded using a digital camera (Canon SX150 IS)

Ni–ZnO nanostructure growth mechanism

A possible growth mechanism of the nanomaterials can be described on the basis of chemical reactions and nucleation. A reasonable reaction mechanism is proposed for obtaining the doped metal oxides and is detailed below:

Zn(OH)₂ + Ni(OH)₂ → Ni–ZnO↓ + 3H₂O↑ (3)

Zn(OH)₂ ↔ Zn²⁺ + 2OH⁻ (4)

Zn²⁺ + 4NH₃ ↔ Zn(NH₃)₄²⁺ (5)

Zn²⁺ + 4OH⁻ ↔ Zn(OH)₄²⁻ (6)

Zn(OH)₄²⁻ ↔ ZnO + H₂O + 2OH⁻ (7)

Zn²⁺ + 2OH⁻ ↔ Zn(OH)₂ ↔ ZnO + H₂O (8)

Zn(NH₃)₄²⁺ + 2OH⁻ ↔ ZnO + 4NH₃ + 2H₂O (9)

These are possibly the reactions that are responsible for the growth of ZnO nanostructures.^16,32 The growth mechanism of ZnO nanostructures (nanorods, nanoflakes, nanocombs, nanosprings, nanobelts, doughnut-shaped particles, etc.) is controversial and still not very clear. The morphology of ZnO nanostructures is not only dependent on the preparation technique but also on other external conditions such as the reaction temperature, pH value, reaction time, solution concentration, doping, etc.^5,7 The most plausible growth mechanism known for ZnO nanostructures is either layer-by-layer along the axial direction or by screw dislocation; we have already discussed both mechanisms in our previous reports.^5,7,32 ZnO exhibits several crystal planes with different polarities (one basal polar oxygen face (0001̄), one polar zinc face (0001), six nonpolar (011̄1) faces and six symmetric nonpolar faces (011̄0) parallel to the c-axis). Li et al.³³ showed that different polar faces have different growth velocities, i.e., v(001) > v(011̄) > v(010) > v(011) > v(001̄). Therefore, growth along the (002) plane is faster than that along other polar faces and forms a hexagonal rod-like morphology. When Ni is doped into the ZnO matrix, it will slow down growth along the (002) direction and increase its thickness, so that the average diameter of nanorods is slightly increased. However, with a higher Ni doping level (10%) the rod-like morphology starts to change into flakes with a change in orientation along the (100) and (101) directions. Owing to the high concentration of Ni ions in the solution, nucleation in the ZnO structure becomes easier, which is due to the lower activation energy barrier of heterogeneous nucleation.²⁴ This helps ZnO to grow in a different direction as a result to form a flake-like morphology.

3. Results and discussion

Fig. 1(a) shows the XRD patterns of pure and doped samples. All the samples are phase-pure and exhibit a hexagonal phase of the wurtzite type of ZnO, with the preferred growth direction along the c-axis. No impurity peak is found that is related to a Ni phase, which indicates the successful incorporation of Ni in the ZnO structure; this is due to the comparable ionic radii of Ni²⁺ (0.69 Å) and Zn²⁺ (0.74 Å). The most intense peak for ZnO is due to (002) planes; this is due to the rod- or needle-like morphology of ZnO nanostructures.^5,29 Furthermore, with an increase in the Ni concentration there is a decrease in the intensity of the (002) peak and increases in the (100) and (101) peaks, which is due to a change in the orientation of the nanostructure or the morphology of doped ZnO, which is pointed out in previous studies.^1,4 Fig. 1(b) presents a shift in the (002) peak towards lower 2θ values up to a doping level of 7.5%, whereas a Ni doping level of 10% gives rise to an increase in the 2θ value. The reason for this may be differences in the doping concentration in dissimilar conditions, as well as many distortions in the ZnO host lattice. This creates lattice relaxation or compression in the host matrix because of vacancies or interstitial defects present in the host matrix. A similar trend of changes in 2θ values after doping with Ni was reported by Yu et al.³⁴ and Tong et al.³⁵ The interplanar spacing determined by XRD was used to estimate the lattice parameters. The XRD peak for (002) planes was used for the calculation of the grain size and the results are shown in Table 1.

Fig. 1 (a) XRD patterns of pure and Ni-doped ZnO nanostructures, (b) relative shift in the peak (002) planes with respect to the Ni doping level.

Table 1 Lattice parameters a and c of Ni-doped ZnO nanorods. The average grain sizes are calculated from the (002) reflections

Sample name	a (Å)	c (Å)	c/a ratio	Grain size (nm)
Pristine ZnO (ZNi0)	3.249	5.201	1.6008	67.46
ZnO with 5% Ni (ZNi5)	3.253	5.206	1.6003	75.34
ZnO with 7.5% Ni (ZNi7.5)	3.254	5.210	1.6011	76.03
ZnO with 10% Ni (ZNi10)	3.246	5.197	1.6010	62.29

---

Fig. 2 shows SEM images of pure and Ni-doped ZnO nanostructures (magnification 50 000×). In a wet chemical method, the structure and morphology of ZnO nanostructures depend upon external parameters such as the reaction temperature, reaction time, pH value, solution concentration, and dopant concentration, etc. These parameters have a crucial influence on the physical and chemical properties of ZnO, which we already discussed in a previous report.⁴ A typical SEM image of pure ZnO (Fig. 2(a)) (low-magnification (5000×) images are shown in Fig. S1†) shows the formation of nanoneedles in the form of a flower-like structure. The average diameter (at the middle) and length were estimated to be ∼200 nm and 1–2 μm, respectively. With Ni doping levels of 5% and 7.5%, the nanoneedles were converted into thick nanoneedles with a flat upper surface, with a slight increase in the average diameter from 280 to 294 nm, respectively. It was also observed that the conical tapered part of the nanoneedles cracked and started to develop into nanorods. Again, with a further increase in the Ni concentration, the nanoneedles were fully converted into nanorods and finally a Ni doping level of 10% gave rise to a decrease in the average diameter of the nanorods to less than ∼100 nm, with some being transformed into nanosheets between the thin nanorods. The energy-dispersive X-ray (EDX) spectra of all samples are shown in Fig. S2† and the percentages of Ni contained in the ZnO samples are shown in Table S1.†

Fig. 2 FESEM images of (a) ZNi0, (b) ZNi5, (c) ZNi7.5 and (d) ZNi10 (at 50 000×).

These 1D ZnO nanostructures were shown to be good field emitters because of their diverse nanostructures, which provide promising aspect ratios and appropriate work functions. FE properties can be altered by many factors such as the curvature, uniformity, size, and density of the emitter. The current density (J) versus applied electric field (E) characteristics of pure and doped samples are shown in Fig. 3(a). The values of the turn-on field and threshold field required to generate emission current densities of 1 and 10 μA cm⁻² were found to be 2.5 V μm⁻¹ and 2.8 V μm⁻¹ in the case of ZNi0 and 1.7 V μm⁻¹ and 2 V μm⁻¹ in the case of ZNi10, respectively. The emission current density was found to increase rapidly with an increase in the applied electric field, and emission current densities of 326 μA cm⁻² and 872 μA cm⁻² were obtained at an applied field of 3.8 V μm⁻¹ and 3.1 V μm⁻¹ for ZNi0 and ZNi10, respectively. The FE characteristics of pure and doped ZnO samples are shown in Table 2.

Fig. 3 Field emission of pure and Ni-doped ZnO. (a) Emission current density as a function of applied electrical field. (b) F–N plots.

Table 2 Comparison of field emission characteristics of pure and Ni-doped ZnO nanostructures

Sr. no.	Sample name	Turn-on field (V μm^−1) at 1 μA cm^−2	Threshold field (V μm^−1) at 10 μA cm^−2	Maximum current density (μA cm^−2) at corresponding applied field (V μm^−1)
1	ZNi0	2.5	2.8	∼326 at ∼3.8
2	ZNi5	2.3	2.6	∼156 at ∼3.4
3	ZNi7.5	1.8	2.2	∼528 at ∼3.4
4	ZNi10	1.7	2	∼872 at ∼3.1

---

The modified Fowler–Nordheim (F–N) equation in terms of the current density (J) and the applied electric field (E) is defined as:^36,37

where the applied electric field E = V/d, where V is the applied voltage and d is the separation between the anode and the cathode (∼2 mm). Furthermore, the emission current density J is estimated to be J = I/A, where I is the emission current and A is the total area of the emitter, a and b are constants, which are typically 1.54 × 10⁻¹⁰ A V⁻² eV and 6.83 × 10³ V eV^−3/2 μm⁻¹, respectively, ϕ is the work function of the emitter material, λ_M is a macroscopic pre-exponential correction factor, ν_F is the value of the principal Schottky–Nordheim barrier function (a correction factor), and β is the field enhancement factor. In the present study, the F–N plots were found to be nonlinear and such F–N plots have been reported for many semiconductor nanomaterials (Fig. 3(b)).

The enhancement in FE performance after doping with Ni in ZnO may be explained as follows: as we know, the turn-on and threshold fields depend upon the overall geometry of the emitter and also on its intrinsic electronic properties such as the charge carrier concentration, work function, etc. ZnO rods with Ni doping levels of 5% and 7.5% displayed an increase in diameter relative to that of pure ZnO, whereas ZnO rods with a Ni doping level of 10% exhibited a decrease. The overall morphology of ZNi5 and ZNi7.5 samples displayed a lower areal density with quite separate rods, which suggests a smaller screening effect in comparison to that of ZNi0. This decrease in screening resulted in a reduction in the turn-on voltage in both cases. In the ZNi10 sample, the decrease in diameter and presence of few nanosheets increased the number of emission centers; both these causes may result in a further decrease in the turn-on voltage. Another reason for a decrease in the turn-on voltage is an increase in the Ni doping level in ZnO, which increases the charge carrier concentration in the conduction band of ZnO, as well as causing a shift in energy levels, which has been proved by many experimental and theoretical studies.³⁸ In general, Ni is isoelectronic with Zn, so it should not behave as an acceptor or shallow donor into ZnO. Now the question is: where do the electrons come from? Katayama-Yoshida and Sato³⁹ show using ab initio electronic structure calculations that semimetallic behavior exists in transition metal-doped ZnO. In our samples, Ni²⁺ partially substituted Zn²⁺ in ZnO and, as we know, Zn²⁺ ions in the ZnO structure are located at the center of a tetrahedron surrounded by four oxygen atoms. Under the influence of the tetrahedral crystal field of ZnO, the d-states of Ni split into a higher triplet (t_2g) and a lower doublet (e_g) state, as shown in Fig. 4(a). The higher t_2g state hybridizes with the p-orbital of the valence band and further splits into t_bonding and t_antibonding states (Fig. 4(b)). The t_bonding states participate in Ni–O bonding and are localized. However, the antibonding states have higher energy and contain some mobile electrons and also the energy of the antibonding states lies very close to the conduction band. Hence, there is a high probability that electrons from these antibonding states will jump (acting as an impurity state) into the conduction band with a small increase in potential difference. Owing to increases in the Ni concentration in ZnO, increasing numbers of electrons are promoted into the conduction band, which results in an increase in FE preformation, as shown in schematic form in Fig. 4(c and d), in comparison to that of pure ZnO. In the present work, this effect has clearly been observed from the FE curve. Thus, an increase in the Ni concentration promotes a large number of electrons into the conduction band of ZnO, which ultimately results in a decrease in the turn-on voltage for ZNi5, ZNi7.5 and ZNi10 in comparison to that of ZNi0.

Fig. 4 (a) Electronic structure of a transition metal at a substitution site in a wurtzite structure. (b) Splitting of an impurity state under the influence of the crystal field of the ZnO host. Schematic energy band diagrams of (c) ZnO and (d) Ni-doped ZnO (E_g: energy gap, E_F: Fermi level, E_c: energy of conduction band, E_v: energy of valence band, ϕ: work function).

The current stability curves of pure and doped samples were also investigated at a preset value of 5 μA over a period of 3 h (Fig. 5). Successive current stability curves show no obvious decrease in current density. This is a very important feature, in particular for practical applications. The appearance of the “spike”-type fluctuations observed in the emission current may result from the adsorption/desorption and ion bombardment of residual gas molecules. The local work function of the emitter varies owing to the adsorption/desorption of gas molecules at the emitter surface and ion bombardment by residual gas molecules, which is due to the presence of a strong electrostatic field, which results in fluctuations. Typical FE images are shown in the insets of Fig. 5. The images show a number of tiny spots, which correspond to the emissions from the most protruding edges of the emitters. These results indicate the excellent emission stability of Ni-doped ZnO, which makes it highly valuable for practical applications as a field emitter. Furthermore, the field emission behavior of other pure and doped ZnO nanostructures was compared with data for Ni-doped ZnO, as shown in Table 3. Thus, the overall FE performances of Ni-doped ZnO nanostructures prepared by a low-temperature wet chemical method in the present report, such as the turn-on and threshold fields, are all better than those in previous reports on pure and doped ZnO nanostructures.

Fig. 5 Typical field emission current stability recorded at 5 μA indicating the stable field emission current of (a) pure ZnO, (b) ZnO doped with 5% Ni, (c) ZnO doped with 7.5% Ni, and (d) ZnO doped with 10% Ni (the insets show field emission patterns recorded during measurements of the long-term current stability of the emitters).

Table 3 Comparison of field emission characteristics of pure and doped ZnO nanostructures with those in the present work

Field emitter	Synthesis route	Turn-on field (V μm^−1)	Threshold field (V μm^−1)	Reference
ZnO nanoneedles	Vapor-phase growth	2.4	6.5	40
ZnO nanowires	Vapor deposition method	6.0	11.0	41
Ga-doped ZnO	Vapor-liquid-solid process	3.4	5.4	42
In-doped ZnO	Chemical vapor deposition	2.4	3.5	43
N-doped ZnO	Solvothermal synthesis	2.9	—	44
ZnO nanorods	Atomic layer deposition	2.85	—	45
ZnO nanowires/graphene	Hydrothermal method	2.0	—	46
Cu-doped ZnO quantum dots	Hydrothermal method	4.47	8.9	47
Ni-doped ZnO	Wet chemical method	1.7	2	This work

---

4. Conclusion

In conclusion, Ni-doped ZnO was synthesized by a simple low-temperature wet chemical method. XRD and SEM results confirm the successful incorporation of Ni ions in ZnO by variations in the lattice constants, as well as alterations in the nanostructure morphology of ZnO nanoneedles, respectively. Ni doping in ZnO decreases the turn-on field (from 2.5 to 1.7 V μm⁻¹) and the threshold field (from 2.8 to 2 V μm⁻¹), with a maximum current density of 872 μA cm⁻² at an applied field of 3.1 V μm⁻¹ for ZNi10. Prolonged current stability was also observed for a period of 3 h at a preset value of 5 μA. Therefore, these Ni-doped ZnO nanostructures are potential candidates for future applications in nanoelectronics, in particular in the front areas of flat panel displays and electron emitter devices.

Acknowledgements

This work was supported by the Department of Science and Technology, India by awarding a prestigious Ramanujan Fellowship (SR/S2/RJN-121/2012) to PMS. PMS acknowledges CSIR research grant no. 03(1349)/16/EMR-II. PMS is grateful to Prof. Pradeep Mathur, Director, IIT Indore, for encouraging the research work and providing the necessary facilities. The authors are thankful to SIC Indore for providing research facilities such as XRD and FESEM. P. K. Bankar acknowledges SPPU and DST for the financial support. Prof. M. A. More would like to thank the BCUD of Savitribai Phule Pune University for the financial support provided for the field emission work under CNQS-UPE-UGC program activity. The research work was also supported by the Department of Science and Technology (Government of India) under a Ramanujan Fellowship to Dr D. J. Late (grant no. SR/S2/RJN-130/2012), CSIR-NCL-MLP project grant 028626, DST-SERB Fast-track Young scientist project grant no. SB/FT/CS-116/2013, Board of Research in Nuclear Sciences (BRNS) (Government of India) grant no. 34/14/20/2015 and partial support by the INUP IITB project sponsored by DeitY, MCIT, Government of India.

References

1. R. F. Services, Science, 1997, 276, 895.
2. M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo and P. Yang, Science, 2001, 292, 1897.
3. Z. L. Wang, X. Y. Kong and J. M. Zuo, Phys. Rev. Lett., 2003, 91, 185502.
4. Y. H. Zhang, X. B. Song, J. Zheng, H. H. Liu, X. G. Li and L. P. You, Nanotechnology, 2006, 17, 1916.
5. A. K. Rana, et al., AIP Adv., 2015, 5, 097118.
6. X. Y. Kong and Z. L. Wang, Appl. Phys. Lett., 2004, 84, 975.
7. Y. Kumar, A. K. Rana, P. Bhojane, M. Pusty, V. Bagwe, S. Sen and P. M. Shirage, Mater. Res. Express, 2015, 2, 105017.
8. Z. L. Wang, X. Y. Kong, Y. Ding, P. X. Gao, W. L. Hughes, R. S. Yang and Y. Zhang, Adv. Funct. Mater., 2004, 14, 943.
9. Y. W. Heo, D. P. Norton, L. C. Tien, Y. Kwon, B. S. Kang, F. Ren, S. J. Pearton and J. R. LaRoche, Mater. Sci. Eng., R, 2004, 47, 1.
10. P. M. Shirage, Mater. Today, 2013, 16, 505.
11. M. T. Chen, M. P. Lu, Y. J. Wu, J. H. Song, C. Y. Lee, M. Y. Lu, Y. C. Chang, L. J. Chou, Z. L. Wang and L. J. Chen, Nano Lett., 2010, 10, 4387.
12. J. H. He, S. T. Ho, T. B. Wu, L. J. Chen and Z. L. Wang, Chem. Phys. Lett., 2007, 435, 119.
13. T. Y. Wei, P. H. Yeh, S. Y. Lu and Z. L. Wang, J. Am. Chem. Soc., 2009, 131, 17690.
14. P. H. Yeh, Z. Li and Z. L. Wang, Adv. Mater., 2009, 21, 4975.
15. D. Zhu, Q. He, Q. Chen, Y. Fu, C. He, L. Shi, X. Meng, C. M. Deng, H. M. Cao and J. G. Cheng, ACS Nano, 2011, 5, 4293.
16. A. K. Rana, et al., Appl. Surf. Sci., 2016, 379, 23.
17. P. M. Shirage, A. K. Rana, Y. Kumar, S. Sen, S. G. Leonardi and G. Neri, RSC Adv., 2016, 6, 82733.
18. Z. L. Wang, Adv. Funct. Mater., 2008, 18, 3553.
19. P. C. Chang and J. G. Lu, IEEE Trans. Electron Devices, 2008, 55, 2977.
20. J. H. He, P. H. Chang, C. Y. Chen and K. T. Tsai, Nanotechnology, 2009, 20, 135701.
21. M. Law, L. E. Greene, J. C. Johnson, R. Saykally and P. D. Yang, Nat. Mater., 2005, 4, 455.
22. R. Elilarassi and G. Chandrasekaran, J. Mater. Sci.: Mater. Electron., 2011, 22, 751.
23. S. A. Wolf, et al., Science, 2001, 294, 1488.
24. M. M. Rahman, et al., J. Ind. Eng. Chem., 2014, 20, 1071.
25. Q. Yin, et al., RSC Adv., 2016, 6, 38653.
26. C. A. Fancher, et al., J. Chem. Phys., 1998, 109, 8426.
27. M. Ahmad, H. Sun and J. Zhu, ACS Appl. Mater. Interfaces, 2011, 3, 1299.
28. F. J. Sheini, et al., Appl. Surf. Sci., 2010, 256, 2110.
29. Y. C. Chang, RSC Adv., 2014, 4, 56241.
30. G. Z. Xing, et al., Nanotechnology, 2010, 21, 25.
31. A. K. Rana, et al., Appl. Phys. Lett., 2015, 107, 231907.
32. R. Das, et al., RSC Adv., 2015, 5, 60365.
33. W. J. Li, E. W. Shi, W. Z. Zhong and Z. W. Yin, J. Cryst. Growth, 1999, 203, 186.
34. W. Yu, et al., J. Appl. Phys., 2008, 103, 093901.
35. L. N. Tong, et al., J. Appl. Phys., 2010, 108, 023906.
36. R. G. Forbes, J. Appl. Phys., 2008, 103, 114911.
37. R. G. Forbes, Appl. Phys. Lett., 2006, 89, 113122.
38. B. N. Pantha, A. Sedhain, J. Li, J. Y. Lin and H. X. Jiang, Appl. Phys. Lett., 2010, 96, 131906.
39. H. Katayama-Yoshida and K. Sato, Phys. B, 2003, 327, 337.
40. Q. Zhao, et al., Appl. Phys. Lett., 2005, 86, 203115.
41. C. J. Lee, et al., Appl. Phys. Lett., 2002, 81, 19.
42. L. W. Chang, et al., Appl. Surf. Sci., 2011, 257, 3145.
43. M. Ahmad, H. Sun and J. Zhu, ACS Appl. Mater. Interfaces, 2011, 3, 1299.
44. U. K. Gautam, et al., Adv. Funct. Mater., 2009, 19, 131.
45. Q. Ahsanulhaq, J. H. Kim and Y. B. Hahn, Nanotechnology, 2007, 18, 485307.
46. J. Ok Hwang, et al., J. Mater. Chem., 2011, 21, 3432.
47. L. Sun, Z. Lin, X. Zhou, Y. Zhang and T. Guo, J. Alloys Compd., 2016, 671, 473.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra21190a

---