Structural characterizations and up-conversion emission in Yb³⁺/Tm³⁺ co-doped ZnO nanocrystals by tri-doping with Ga³⁺ ions

Abstract

ZnO:Yb³⁺,Tm³⁺,Ga³⁺ up-conversion nanoparticles (UCNPs) with sizes of between 20 and 65 nm are synthesized by the hydrothermal method. The size of the nanocrystals is affected by different Ga³⁺ concentrations. The variation in size has influence on the particles' visible blue and red up-conversion (UC) emission. In this paper, X-ray powder diffraction results show different amounts of gallium substituting for Zn in the ZnO lattice. When the concentration of Ga³⁺ is 3 mol%, the resulting UCNPs exhibit strong UC emission, longer fluorescence lifetime and narrower band gap. Due to their adjustable size and UC fluorescence properties, ZnO:Yb³⁺,Tm³⁺,Ga³⁺ nanocrystals have broad prospects for future applications in the biomedical field.

---

1 Introduction

Materials doped with rare earth metals have attracted significant attention due to wide applications in science and technology. In particular, up-conversion nanoparticles (UCNPs) have recently gained interest for their optical properties,¹ which are used in optical transmission, medical diagnostics,^2,3 and biotechnology.⁴ Limited intensity, however, still restricts the practical application of UC phosphorescent nanoparticles. Among UCNPs, oxides have stronger chemical stability, higher mechanical strength and optical properties that make them potential candidates for biomedical and optical applications.^5–11 Realization of efficient UC in oxide nanoparticles will have a far-reaching influence on deploying their potential advantages.

Zinc oxide (ZnO) is an electrical insulator and n-type wide band gap semiconductor (E_g = 3.3 eV).^12–14 ZnO doped with rare earth metals has been established as a UC luminescence system. The Tm³⁺ ion is considered a good blue emitting lanthanide, and it plays an important role in data storage, information technology, laser printing, color display and health.¹⁵

There is a lot of evidence corroborating the notion that the UC luminescence intensity of Yb³⁺/Tm³⁺ doped ZnO is stronger than that of ZnO with Tm³⁺ ions alone, because Yb³⁺ as a sensitizer that can generate further UC luminescence.^16–19 However, despite the high UC luminescence of such systems, their UC fluorescence intensity remains insufficient for practical technological applications.

The UC luminescence was realized by taking advantage of the rich f–f level energy of rare-earth ions, but the main intra-4f electronic-dipole transitions are forbidden.²⁰ The forbidden transition can be destroyed by lattice vibration, magnetic dipole transition and the non-symmetry of the lattice system. The Ga³⁺ ion has a small lattice mismatch, low oxidation reactivity and a smaller cationic radius compared to the Zn²⁺ ion.^21–23 The small difference between the bond lengths of Ga–O and Zn–O makes Ga³⁺ better for doping in the ZnO host lattice than other ions.^24,25 These advantages make Ga³⁺ attractive for varying the nanosize of samples and decreasing the symmetry of the local electric field. The size of nanoparticles has a great influence on the intensity of UC radiation.

In this research, ZnO: Tm³⁺ (0.5 mol%), Yb³⁺ (7 mol%), and Ga³⁺ (0%, 1%, 2%, 3% or 4%) nanocrystals are successfully prepared via the hydrothermal method. Under the excitation of a 980 nm laser, UC emissions are observed at room temperature. The influence of the size of nanoparticles on the UC emission is investigated via different Ga³⁺ dopant concentrations. In addition, the UC enhancement is theoretically described by methods that aid in the observations of UC luminescence intensities, fluorescence lifetimes and energy gaps.

2 Experimental

2.1 Materials

All the rare earth nitric acid compounds RE(NO₃)₃·5H₂O (RE: Yb, Tm) and zinc nitrate (Zn(NO₃)₂·6H₂O) are analytically pure (99.99%). Gallium(III) nitrate (Ga(NO₃)₃·xH₂O, 99.99%) and trisodium citrate dehydrate (Na₃C₆H₅O₅·2H₂O, 99%) are purchased from Aladdin Industrial Corporation. The reagents are used in the experiment without further purification. For all experiments, deionized water is used.

2.2 Synthesis of Yb³⁺/Tm³⁺/Ga³⁺/ZnO

ZnO doped with 7 mol% of Yb³⁺, 0.5 mol% of Tm³⁺, and 0, 1, 2, 3, or 4 mol% of Ga³⁺ ions was synthesized by the hydrothermal method. The doping concentrations of Yb³⁺ and Tm³⁺ ions were determined by orthogonal experiments. A typical process was carried out as follows: the required molar amounts of Zn(NO₃)₂·6H₂O and Ga(NO₃)₃·xH₂O were completely dissolved in deionized water with magnetic stirring. Sodium citrate (0.5 mol L⁻¹) was added as a dispersant. Next, Tm(NO₃)₃·5H₂O and Yb(NO₃)₃·5H₂O were added to the abovementioned mixture under stirring for 2 h at room temperature. Finally, the resulting solution was transferred to the reaction kettle (30 mL) and reacted at 150 °C for 24 h. Nanoscale ZnO doped with different concentrations of Ga³⁺ was obtained.

2.3 Characterization

The power X-ray diffraction (XRD) patterns were obtained using a diffractometer with Cu Kα radiation (λ = 1.54 Å) that works at 40.0 kV and 30.0 mA. The 2θ angle of the XRD pattern was obtained at a scanning rate of 8° min⁻¹. Scanning electron microscopy (SEM SU8000) was used to observe the nanostructure of the samples. Double beam UV visible spectrophotometry (TU-1900) was employed to record the absorption band edge. The up-conversion emission spectra of the Yb³⁺/Tm³⁺/Ga³⁺/ZnO nanoparticles were obtained using 980 nm excitation.

3 Results

3.1 Structure and size of nanoparticles

Fig. 1 shows the XRD patterns of Yb³⁺/Tm³⁺/Ga³⁺/ZnO with different Ga³⁺ concentrations: 0 mol%, 1 mol%, 2 mol%, 3 mol%, and 4 mol%. The diffraction peaks of the samples are similar and are also in good agreement with the standard spectrum of hexagonal wurtzite structured ZnO. Analysis of the spectra reveals that there is no second phase associated with Ga³⁺ ions, which illustrates that their doping does not change the crystal structure. When the doping content is 3 mol% of Ga³⁺, the material shows sharper diffraction peaks.

Fig. 1 XRD patterns of Yb³⁺/Tm³⁺/ZnO with different Ga³⁺ concentrations.

Table 1 shows that the lattice constants remain almost unchanged with different Ga³⁺ doping levels. It reveals that Ga³⁺ ions enter into the ZnO host lattice without generating lattice strain. The average grain diameter of Yb³⁺/Tm³⁺/ZnO doped with 0, 1, 2, 3 or 4 mol% of Ga³⁺ ions was calculated by the Scherrer equation (eqn (1)) given as follows:

where D is the domain size; B is the half-width of the diffraction peak of interest; and θ is the angle of the corresponding diffraction peak. The results show that increasing the amount of Ga³⁺ from 0 mol% to 4.0 mol% gradually decreased the diameter size from about 74 nm to below 30 nm (see Table 1).

Table 1 Lattice constants and crystallite size of Yb³⁺/Tm³⁺/Ga³⁺/ZnO with different Ga³⁺ concentrations

Ga^3+ (mol%)	Δa = Δb (nm)	Δc (nm)	Crystallite size (nm)
0	0	0	74
1	0.0002	0.0002	44
2	0.0004	0.0003	39
3	0.0002	0.0002	31
4	0.0001	0.0001	30

---

In order to further investigate the effect of different Ga³⁺ dopant concentration on particle size and morphology, SEM images and gauss fitting of the nanoparticle sizes have been made in Fig. 2. According to the SEM results, the UCNPs are structured as nanorods. SEM (Fig. 2A) further showed that the diameter of the nanorods decreased with increasing of Ga³⁺ concentrations. The width of the nanorods is narrowest when the Ga³⁺ doping concentration is 4.0 mol%. The gauss fitting of nanorod diameter is shown in Fig. 2B. The size based on SEM is in reasonable agreement with those determined by XRD analysis, with the 0 mol%, 1.0 mol%, 2.0 mol%, 3.0 mol% and 4.0 mol% of Ga³⁺ doped particles having average diameters of 65.2 ± 2.9 nm, 50.5 ± 2.8 nm, 40.4 ± 5.5 nm, 30.3 ± 6.5 nm and 20.2 ± 6.2 nm, respectively. The diameter change is completely consistent with the variation of average grain diameter under different Ga³⁺ concentrations.

Fig. 2 Electron microscopy characterization of Yb³⁺/Tm³⁺/ZnO with different Ga³⁺ ions (A) SEM images of different Yb³⁺/Tm³⁺/ZnO with different Ga³⁺ ions (a–e) (B) histogram of size distribution evaluated from SEM images (a–e).

The size of the sample is controlled by the combination of dispersants (herein, sodium citrate) and the crystal growth rate. According to Fig. 2B, the particle size is reduced with increased concentration of Ga³⁺ ions. This phenomenon is mainly caused by the fact that the effective surface charge is affected by different Ga³⁺ concentrations, which is a key factor for the crystal growth rate. In the case of substitution of Zn²⁺ ions with smaller Ga³⁺ ions, the electron charge density of the crystal surface increases. This phenomenon and explanation can be supported by DFT calculations.²⁶

3.2 UC luminescence properties

Fig. 3 shows the UC emission of the ZnO nanoparticles doped with 0.5 mol% of Tm³⁺ and 7 mol% of Yb³⁺ ions; ZnO nanoparticles with 0 mol%, 1.0 mol%, 2.0 mol%, 3.0 mol%, 4.0 mol% of Ga³⁺, 0.5 mol% of Tm³⁺ and 7 mol% of Yb³⁺ ions, under 980 nm excitation of about 500 mW. The inset shows a blue UC emission picture of ZnO with 0.5 mol% of Tm³⁺, 7 mol% of Yb³⁺ and 3 mol% of Ga³⁺ ions. The blue UC emission is centered at 475 nm and corresponds to the ¹G₄ → ³H₆ transition of Tm³⁺ ions, whereas the red UC emission is centered at 652 nm and corresponds to the ¹G₄ → ³F₄ transition of Tm³⁺ ions. When the concentration of Ga³⁺ is 3.0 mol%, the blue and red luminescence reach maximum intensity; beyond this concentration, the blue emission at 475 nm decreases sharply. This phenomenon may be due to nanoparticle surface effects. Essentially, the higher concentration of Ga³⁺ makes the diameter of the grain smaller, as previously described. When the agglomeration behavior increases, the crystallinity of the samples also increases. The confusing interface of atom arrangement is increasing in the samples, which decreases the center of non-radiative transition at the surface of the samples. However, it can be easily seen that the blue and red UC luminescence of ZnO nanocrystals are enhanced with the Ga³⁺ ion dopant.

Fig. 3 Emission spectra of Yb³⁺/Tm³⁺/Ga³⁺/ZnO upon excitation at 980 nm for five different concentration of Ga³⁺ ions: 0 mol%, 1.0 mol%, 2.0 mol%, 3.0 mol%, and 4.0 mol%. The inset is a UC emission photograph of the sample with 3.0 mol% Ga³⁺ ions.

Fig. 4 shows the multiples of the blue and red UC emissions and the blue-to-red intensity ratios in the ZnO nanoparticles under 980 nm laser excitation of 500 mW. Fig. 4 presents the intensity of blue and red UC increases with Ga³⁺ ion doping. Under optimal experimental conditions, the maximum blue and red UC emissions reach about 6 times and 2 times, respectively, the concentration corresponding to 3 mol% of Ga³⁺ ions. The maximum blue-to-red intensity ratio reaches 3 times with 3 mol% Ga³⁺ concentration and the blue UC emission is far stronger than the red UC emission.

Fig. 4 Enhancement times of the blue and red UC and the blue-to-red intensity ratio in ZnO nanoparticles with different Ga³⁺ ion concentrations under laser excitation of 980 nm.

4 Discussion

4.1 The general theoretical description of UC blue and red luminescence mechanism

In order to understand the UC mechanism, log–log plots of the intensity versus input pump power of the blue, green and red UC emissions in Yb³⁺/Tm³⁺/Ga³⁺/ZnO sample are displayed in Fig. 5. In general, for an unsaturated UC process, the number of photons required to populate the upper emitting state can be obtained using the following equation:

I_f ∝ Pⁿ (2)

where I_f is the output fluorescence intensity; P is the pump power; and n is the photon number that is needed in the UC emission process. As illustrated in Fig. 5, the blue and red UC emissions yield n = 2.81 and n = 2.04, indicating that three 980 nm photons and two 980 nm photons are needed to produce the blue and red luminescence bands in Yb³⁺/Tm³⁺/Ga³⁺/ZnO.

Fig. 5 log–log plots of the intensity versus input pump power for the blue, green and red UC emissions in Yb³⁺/Tm³⁺/Ga³⁺/ZnO.

Fig. 6 illustrates the energy level diagrams of Tm³⁺ and Yb³⁺ ions, as well as the proposed UC mechanism under 980 nm excitation. As shown in Fig. 6, the ¹G₄ state of Tm³⁺ ion is populated via a three-photon energy transition process. As a sensitizer, Yb³⁺ ions receive a 980 nm photon in the ground-state (²F_7/2) and transfer it to the excited state (²F_5/2). This is followed by a strong photon energy transfer of Tm³⁺ ions in the ground state (³H₆). The Tm³⁺ ions absorb resonant photons in the ground state (³H₆) and promote them to ³H₅. The first process is ground state absorption (GAS). Subsequently, Tm³⁺ ions in the ³H₅ level decay to the ³F₄ level through non-radiative transition. The excited Yb³⁺ ions emit broad band resonant photons, which are absorbed by Tm³⁺ (³F₄). Then, particles are populated in the higher ³F_2,3 level and the non-radiative transition results in population of the ³H₄ level. This is called excited state absorption (ESA) process. The excited Yb³⁺ ions continue to offer a photon to Tm³⁺ ions (³H₄), which results in Tm³⁺ ions transfer to the ¹G₄ level. The excited ¹G₄ level relaxes to the ground state. The ¹G₄ → ³H₆ transition process results in intense blue (475 nm) UC emission. However, the red UC luminescence occurs via a cooperative sensitization UC mechanism, which is called two-photon absorption (TPA). The ¹G₄ state of Tm³⁺ ion simultaneously absorbs two photons to promote it to the ¹G₄, and then relaxes to the ³F₄ state. The ¹G₄ → ³H₄ transition process results in the intense red (652 nm) UC luminescence. This would explain why the blue UC emission is stronger than the red UC emission (see Fig. 4). Furthermore, suitable Ga³⁺ ions in the Yb³⁺/Tm³⁺/ZnO lattice participate in the energy transfer and favor the UC emission.

Fig. 6 The blue and red UC luminescence mechanism of Tm³⁺, Yb³⁺ and Ga³⁺ ions co-doped ZnO nanocrystals under 980 nm excitations.

4.2 Theoretical descriptions for UC enhancement

Fig. 7 shows the luminescence decay kinetics of samples with different Ga³⁺ ion concentration for blue (λ = 475 nm) UC emission. Hence, the double exponential fluorescence decay law is chosen for the data analysis (eqn (3)).

I(t) = I₀ + A_s e^−t/τ_s + A_f e^−t/τ_f (3)

where τ_s and τ_f are slow fluorescence lifetime and fast fluorescence lifetime, respectively; A_s and A_f are weighting factors of slow and fast fluorescence lifetime, respectively; and, I₀ is fluorescence intensity of the background. The UC fluorescence lifetimes were calculated with formula 4 (see Table 2).

Fig. 7 Fluorescence decay time of the ¹G₄ (Tm³⁺) → ³H₆ (Tm³⁺) transition in ZnO nanoparticles doped with 0, 1, 2, 3, 4 mol% of Ga³⁺ ions (a–e).

Table 2 Measured lifetimes (475 nm), experimental enhancements for the blue and red UC radiations and calculated blue UC luminescent population ratio (α_blue) in ZnO nanocrystals doped with 0.5 mol% of Tm³⁺, 7 mol% of Yb³⁺ ions and different amounts of Ga³⁺ ions

Ga^3+ (mol%)	τrad (μs)	αblue (error 4%)	Blue enhancement (times)	Red enhancement (times)
0	96	0.9	0	0
1	232	0.9	2	1
2	258	0.9	3	2
3	286	0.9	7	2.2
4	184	0.9	2	1

---

According to the mechanism of UC luminescence, the steady state rate equations of Yb³⁺–Tm³⁺ system is established:

R ∝ 1/τ (5.4)

α_blue = τ₃/τ_rad (5.5)

N = N₀ + N₁ + N₂ ≈ N₀ (5.6)

where N₀, N₁ (R₁, τ₁), N₂ (R₂, τ₂) and N₃ (R₃, τ₃) are the level densities (rate of decay, decay time) of the ³H₆, ³H₅, ³F_2,3 and ¹G₄ states of the Tm³⁺ ions, respectively; N_Yb and W are the level density of ³F_7/2 state of the Yb³⁺ ions and energy transfer rate between Yb³⁺ ions and Tm³⁺ ions, respectively; τ_rad is the radiation lifetime of the ¹G₄ state; and α_blue is the blue UC luminescent population ratio in the ¹G₄ state of Tm³⁺ ions. Since the blue UC emission is attributed to three photons processes, R₁ could not be ignored either. Based on eqn (5.1)–(5.6), the general theory about blue UC fluorescent radiations has been presented as eqn (6).

As shown in eqn (6) and (7), we put forward the claim that the intensity of UC fluorescence is related to fluorescence lifetime (τ_rad) and luminescence population ratio in the ¹G₄ state (α_blue). Supposing Ga³⁺ dopant concentration has an influence on UC fluorescence intensity, we can derive UC fluorescence enhancement in Yb³⁺/Tm³⁺ from eqn (8).

Enhancement = Bα_blue²(Ga)τ_rad²(Ga)/[α_blue²(0)τ_rad²(0)] (8)

Using eqn (7) and the blue-to-red ratios from Fig. 4, we can calculate luminescence population ratio (α_blue) with different Ga³⁺ ions (see Table 2). As illustrated in Table 2, UC blue luminescence population ratio (α_blue) is approximately equal in the different Ga³⁺ dopant concentrations and B is a constant. Accordingly, UC blue emission enhancement is proportional to the square of fluorescence lifetime. This theory is consistent with the experimental results of UC emission enhancement and provides confirmatory evidence that the Ga³⁺ ion doping is beneficial to UC fluorescence intensity. Transitions between the 4f states can be lifted by Zn²⁺ substitution with Ga³⁺ ions into the ZnO host lattice, partly because the symmetry of the local electric field decreases and the optically active lattice sites increase.

On the other hand, UCNPs are synthesized by a hydrothermal method, therefore water molecules exist in UCNPs. Water molecules collide with Tm³⁺ luminescence centers, resulting in a dynamic quenching procedure that can shorten the fluorescence lifetime. With the increase of Ga³⁺ ions, the diameter of the nanorod decreases. At the same time, the number of water molecules interacting with the surface of the sample is reduced, which results in an increase in fluorescence lifetime τ via weakening of the dynamic quenching procedure. When the concentration of Ga³⁺ ion is greater than 3 mol%, the fluorescence lifetime would be inhibited by the smaller size of the particles, which results in more evident agglomeration and poor dispersion. Large particles can increase the quenching ion, which will enhance the quenching effect and shorten the fluorescence lifetime.

4.3 Relationship between band gap and enhancement

The forbidden band gap of samples could be calculated according to UV visible absorption spectrum test. The formula for calculating the band gap of semiconductor is given in eqn (9) (Table 3):where A is the absorbance of UV spectrum; E_g is forbidden band gap; and K is constant. Fig. 8 shows the band gap of different samples. When the concentration of Ga³⁺ is 3 mol%, the forbidden band gap is the shortest (3.04 eV).

Table 3 UC luminescence intensity and forbidden band gap with different amounts of Ga³⁺ ions

Ga^3+ (mol%)	Eg (eV)	Blue emission (cps)	Red emission (cps)
0	3.24	580	80
1	3.19	660	121
2	3.14	1750	198
3	3.04	4100	230
4	3.17	1100	152

---

Fig. 8 The forbidden band gap of Yb³⁺/Tm³⁺/ZnO with different Ga³⁺ ions, the inset is 1 mol% of Ga³⁺ ions.

Up-conversion fluorescence and band gap are compared for different Ga³⁺ concentrations in Table 3. There is ample support to claim that the up-conversion luminescence intensity of blue emission is enhanced with the narrow bad gap; however, the red emission intensity hardly changes. According to Kubo's equation (eqn (10)):

where E_F is the Fermi level; N is the total of free electron content; and σ is the energy gap. When the band gap increases, the total of free electron content (N) decreases and the energy gap (σ) increases. The space of nanoparticles is expanded and the splitting energy level is easy to produced, causing fluorescence efficiency to drop. This is called quantum size effect.

5 Conclusion

Yb³⁺/Tm³⁺/Ga³⁺/ZnO nanocrystals have been successful prepared by hydrothermal method. The influence of different Ga³⁺ concentrations on UC emissions has been investigated. For UCNPs, the variation of Ga³⁺ dopant concentration makes a difference in the size of the nanoparticles, which in turn has an effect on their UC emissions. The Yb³⁺–Tm³⁺ distance also varies with size of UCNPs, which has a further influence on the UC emission.²⁷ X-ray powder diffraction results indicated that gallium substitutes for Zn in ZnO lattice. According to the Scherrer equation, the crystallite size of the samples decreases with increasing Ga³⁺ ion concentration in the reacting solution. By analyzing the micro topography, it can be concluded that the variation of the size of the nanoparticles is totally consistent with the change of crystallite size resulting from different amounts of Ga³⁺ ions. The data gathered in the decay lifetime test suggests that UC fluorescence lifetime of Yb³⁺/Tm³⁺/ZnO nanoparticles is enhanced with Ga³⁺ doping and analyzes the reasons by theoretical calculations. Doping Ga³⁺ ion are favorable for enhancing the UC luminescence intensity. In addition, changes to the band gap of Yb³⁺/Tm³⁺/Ga³⁺/ZnO explains the enhancement of UC luminescence intensity via the Kubo equation.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 11374080).

Notes and references

1. P. Rodriguezsevilla, H. Rodriguezrodriguez, M. Pedroni, J. García Solé and D. Jaque, Nano Lett., 2015, 15, 5068–5074.
2. F. Zhang, R. C. Haushalter, R. W. Haushalter, Y. Shi and Y. Zhang, Small, 2011, 7, 1972–1976.
3. W. Yan-Mei, C. Yao and H. Li-Jiao, Chem. Commun., 2014, 50(36), 4759–4762.
4. Z. Li, B. He, J. Huang, Z. Cheng and X. Xu, ACS Appl. Mater. Interfaces, 2014, 6, 7719–7727.
5. G. Chen, H. Liu, H. Liang, G. Someefalear and Z. Zhang, J. Phys. Chem. C, 2008, 112, 12030–12036.
6. D. J. Nelson and C. N. Brammer, Fluorine-related nanoscience with energy applications, ACS, 2011.
7. M. Challenor, P. Gong, D. Lorenser, M. Fitzgerald and S. Dunlop, ACS Appl. Mater. Interfaces, 2013, 5, 7875–7880.
8. C. Liang, W. Chao and L. Zhuang, Nanoscale, 2013, 5, 23–37.
9. F. Wang and X. Liu, J. Am. Chem. Soc., 2008, 130, 5642–5643.
10. J. Zhao, Y. Sun, X. Kong, L. Tian and Y. Wang, J. Phys. Chem. B, 2008, 112, 15666–15672.
11. M. Deng and L. Wang, Nano Res., 2014, 782–793.
12. L. Leonard, B. Joe and D. Steve, ACS Appl. Mater. Interfaces, 2015, 7, 152–157.
13. M. H. Huang, S. Mao, H. Feick, H. Yan and Y. Wu, Science, 2001, 292, 1897–1899.
14. A. Aloufy, J. Ebothé, N. Dumelié, I. Chaki and M. Abd-Lefdil, Phys. E, 2015, 73, 96–99.
15. G. V. Naik, J. Kim and A. Boltasseva, Opt. Mater. Express, 2011, 1, 1090–1099.
16. X. Ming, F. Lu, Z. Ji, M. Chen and J. Zhao, Nucl. Instrum. Methods Phys. Res., Sect. B, 2012, 274, 172–176.
17. I. Chaki, A. Belayachi and T. E. Bahraoui, Eur. Phys. J.: Appl. Phys., 2014, 68.
18. Y. Pu, F. Xu, Z. Jiang and Z. Ma, Appl. Phys. Lett., 2012, 101, 191903.
19. D. Wang, G. Xing, M. Gao, L. Yang and J. Yang, J. Phys. Chem. C, 2011, 115, 22729–22735.
20. G. Chen, H. Liu, H. Liang, G. Somesfalean and Z. Zhang, J. Phys. Chem. C, 2008, 112, 12030–12036.
21. J. T. Yan, C. H. Chen, S. F. Yen and C. T. Lee, IEEE Photonics Technol. Lett., 2010, 22, 146–148.
22. T. Y. Park, Y. S. Choi, J. W. Kang and J. H. Jeong, Appl. Phys. Lett., 2010, 96, 051124.
23. B. Y. Oh, M. C. Jeong, W. Lee and J. M. Myoung, J. Cryst. Growth, 2005, 274, 453–457.
24. H. Ray-Hua, S. Kun-Ching and Y. Chen-Yang, Opt. Express, 2013, 21, 14452–14457.
25. Z. Lichun, L. Qingshan and S. Liang, Opt. Express, 2013, 21, 16578–16583.
26. D. T. Klier and M. U. Kumke, J. Phys. Chem. C, 2015, 119, 3363–3373.
27. X. Teng, Y. Zhu and W. Wei, J. Am. Chem. Soc., 2012, 134, 8340–8343.

---