One-step synthesis and upconversion luminescence properties of hierarchical In₂O₃:Yb³⁺,Er³⁺ nanorod flowers

Abstract

In₂O₃:Yb³⁺,Er³⁺ nanorod flowers (NRFs) are prepared by a simple hydrothermal method, where sucrose was used as a ligand. The obtained In₂O₃:Yb³⁺,Er³⁺ NRFs were carefully characterized by scanning electron microscopy (SEM), power X-ray diffraction (XRD), transmission electron microscopy (TEM) and steady/transient spectroscopy. The dependence of the upconversion luminescence (UCL) of In₂O₃:Yb³⁺,Er³⁺ NRFs on morphology, Yb³⁺ concentration and excitation power was carefully discussed. It is found that the luminescence intensity ratio of the red emission to green emission (R/G) depended on the morphology and Yb³⁺ concentration, and the green emissions are dominantly resulting from a three-photon populating process with higher Yb³⁺ doping concentration. More importantly, the concentration quenching in the In₂O₃:Yb³⁺,Er³⁺ NRFs was greatly suppressed due to boundary effects, and is beneficial for lighting and photon energy conversion devices.

---

1. Introduction

Since the concept of upconversion (UC) was proposed by Auzel in 1966, upconversion nanoparticles (UCNPs) have attracted considerable attention in many fields owing to their fascinating features, because of their ability of successive absorption of several low-frequency near infrared photons and generation of one ultraviolet or visible photon.^1–3 Furthermore, researchers found many superior advantages of UCNPs, such as multicolor emission, long lifetimes, low autofluorescence background and superior photo-stability, as well as high penetration depth, and they are widely applied in lasers, solar cells, displays, biosensors and bioimaging.^4–10 The inorganic host is the carrier of UC and it is very important for improving the efficiency of UCL, which is affected by the morphology of UCNPs. So far, many works have been performed to synthesize different shapes of UC nanocrystals (NCs).^11–16

Cubic sesquioxide In₂O₃ is recognized as one of the promising host materials for both up-/down-conversion luminescence through rare earth doping.^17–19 At present, the controllable preparation of high effective UCL of rare earth (RE) doped In₂O₃ NCs is still difficult. Several strategies have been proposed to solve these problems. For example, Chen's group synthesized Er³⁺ ions doped cubic In₂O₃ nanoparticles via a simple solvothermal method, and firstly demonstrated its UC and downconversion emissions under 808 and 980 nm excitation, respectively.²⁰ Xu et al. successfully fabricated In₂O₃:RE nanotubes by the electrospinning method for the first time, and the sensitivity of In₂O₃ nanotubes to H₂S were significantly improved due to the RE ions doping.²¹ Dutta's group also prepared Eu/Dy doped In₂O₃ nanoparticles by a sonochemical technique. Relatively weak Eu³⁺ emission could be detected in In₂O₃:Eu nanoparticles, which is attributed to the distorted environment around the Eu ions in the In₂O₃ lattice.²² So far, although a few works on In₂O₃:RE nanoparticles have been performed, the mechanism of UC processes of In₂O₃ NCs doped with Yb³⁺, Er³⁺ ions is not fully understood. Especially, the dependence of UCL on cross relaxation, population processes, excitation power density and RE doping concentration are not clear. Therefore, it is necessary to detailedly investigate the UCL property of In₂O₃:Yb³⁺,Er³⁺ NCs.

In this article, we prepared In₂O₃:Yb³⁺,Er³⁺ hierarchical NRFs by a facile hydrothermal method using sucrose as surfactant.²³ The formation, morphology and structure of the In₂O₃:Yb³⁺,Er³⁺ NRFs were examined by SEM, TEM, and XRD. And the influence of sucrose on the hydrothermally prepared In₂O₃:Yb³⁺,Er³⁺ NRFs is further elucidated. The investigation of UCL property of the In₂O₃:Yb³⁺,Er³⁺ NRF demonstrated that the UCL intensity and R/G ratios depended on particle size, excitation power and Yb³⁺ concentration. The as-obtained In₂O₃:Yb³⁺,Er³⁺ NRFs exhibited a high effective UC emission, which is a promising candidate for display and solar cells.

2. Experimental section

2.1 Synthesis In₂O₃ nanorods flowers

All the reagents (analytical-grade purity) were used without any further purification. In the preparation of (10 mol%)Yb³⁺, (1 mol%)Er³⁺ co-doped In₂O₃ NCs, InCl₃·4.5H₂O (266 mg), YbCl₃·6H₂O (33 mg), ErCl₃·6H₂O (3 mg), urea (15 mg) and moderate sucrose (342 mg) were dissolved in deionized water (36 mL) with adequately stirring for 30 min. In this reaction, sucrose acts as a template and linker to bind the nanoparticles, which leads to the formation of flower-like nanostructures. The presence of both sucrose and urea is a prerequisite for the formation of this porous hierarchical In₂O₃. Then, a Teflon-lined stainless steel autoclave was used for placed for the mixture solution and placed in 160 °C oven for 12 h. After the autoclave was cooled to room temperature naturally, the result product was obtained through centrifuging the result product using deionized water and ethanol for three times, and then dried at 80 °C for 1 day. In₂O₃:Yb³⁺,Er³⁺ NCs were obtained through calcining the precursor at 500 °C for 2 h. The same route is adopted for the preparation of (10 mol%)Yb³⁺ and xEr³⁺ (x = 2, 3, 5, 8 mol%) doped In₂O₃ NCs. The preparation process of In₂O₃:Yb³⁺,Er³⁺ NRFs is presented in Scheme 1.

Scheme 1 Schematic illustration for the synthesis of In₂O₃:Yb³⁺,Er³⁺ NRFs.

2.2 Characterization and measurements

The SEM images were obtained on a JEOL JSM-7500F microscope operating at 15 kV. The TEM, HR-TEM and selected-area electron diffraction (SAED) images were measured on a JEOL JEM-2100 microscope operated at 200 kV. The XRD pattern was recorded on a Rigaku D/max-rA power diffractometer. The UC emission spectra and dynamics of rare earth ions were measured on an Edinburgh fluorescence spectrometer (FLS980).

3. Results and discussion

3.1 Morphology and structure characterization of In₂O₃:Yb³⁺,Er³⁺ NRFs

Fig. 1 shows the XRD patterns of the calcining In₂O₃:Yb³⁺,Er³⁺ products and the corresponding reference samples (REF). It can be seen that the products are the pure cubic structure according to JCPDS card no. 06-416. With space group Ia3̄ (no. 206) and lattice parameters of a = 10.118 Å. No diffraction peaks from any other impurities were observed, indicating the high purity of the products. From the Debye–Scherer formula calculating, the average crystallite size of In₂O₃ is ∼15 nm. Fig. 2 shows the SEM images of In₂O₃:Yb³⁺,Er³⁺ NCs with varying amount of sucrose. In the absence of sucrose, the product is the mixture of irregular nanocubes and flower-like nanostructures (Fig. 2a), as the reference (REF) samples. When the moderate sucrose is added, the NRFs were appeared. Fig. 2b shows the low magnification SEM image of In₂O₃:Yb³⁺,Er³⁺ NRFs, demonstrated that the NRFs are high uniformity. The enlarged-magnification SEM images are presented in Fig. 2c and d, and we can see that each nanorod is uniform. Fig. 3 records the elaborate structural analysis of samples using TEM. In Fig. 3a, it can be seen that each nanorod of NRFs is homogenous, and the size of these nanorods is ∼100 nm in diameter and ∼1 μm in length (Fig. 3b). From the TEM images, we can deduce that the NRFs are self-assembled by many single nanorods. The inter-planar spacing is ∼0.292 nm through measuring the lattice fringes (Fig. 3c), matching with the distance of the (2 2 2) plane of the cubic In₂O₃ crystal. The SAED pattern of the In₂O₃:Yb³⁺,Er³⁺ NRFs (Fig. 3d) confirms that the obtained products are polycrystalline in structure. In order to further determine the structure and dopant concentrations of In₂O₃:Yb³⁺,Er³⁺ NRFs, the energy-dispersive X-ray (EDX) analysis of In₂O₃:(10 mol%)Yb³⁺,(2 mol%)Er³⁺ NRFs is measured in Fig. S1.† It shows the presence of In, Yb, Er and O element. Furthermore, the spectrum confirms that the mole ratio of Yb and Er in the NRFs is 8.47 and 1.7, respectively. According to this measuring result, we can deduce that the exact doping concentration of Er³⁺ ions is 0.85 mol%, 2.55 mol%, 4.24 mol%, 6.79 mol% in In₂O₃:(10 mol%)Yb³⁺,xEr³⁺ (x = 1, 3, 5, 8 mol%) NRFs.

Fig. 1 X-ray diffraction patterns of as-prepared In₂O₃:Yb³⁺,Er³⁺ products samples.

Fig. 2 (a) SEM image of the In₂O₃:Yb³⁺,Er³⁺ nanocrystals prepared without sucrose, SEM images of the as-synthesized In₂O₃:Yb³⁺,Er³⁺ NRFs (b) a panoramic and (c) an enlarged of a part of samples. (d) High-magnification SEM image of In₂O₃:Yb³⁺,Er³⁺ nanorod.

Fig. 3 Typical TEM images of In₂O₃:Yb³⁺,Er³⁺ NRFs (a) and single nanorod (b), HR-TEM image of In₂O₃:Yb³⁺,Er³⁺ NRFs (c), the corresponding SAED pattern (d).

3.2 Effect of morphology, Yb³⁺ concentration and excitation power on UCL

The typical UCL spectra of In₂O₃:Yb³⁺,(1 mol%)Er³⁺ REF and NRFs samples were investigated (Fig. 4). In the two samples, the ⁴F_3/2–⁴I_15/2 (blue), ²H_11/2, ⁴S_3/2–⁴I_15/2 (green) and ⁴F_9/2–⁴I_15/2 (red) transitions can be clearly identified. The intensity ratio of the red (⁴F_9/2–⁴I_15/2) emission to the green (²H_11/2, ⁴S_3/2–⁴I_15/2) emissions (R/G) in In₂O₃:Yb³⁺,Er³⁺ NRFs is higher than that in REF sample. The UCL property could be influenced by the following three factors: the nanocrystal size, crystallinity, and morphology. From the XRD and SEM analysis, we can deduce that the crystallinity and the size of In₂O₃:Yb³⁺,Er³⁺ NRFs and REF samples are both nearly same, indicating that the effect of the size and crystallinity might not be the dominant reason to the change of R/G ratio. The remarkable change in the R/G ratio between In₂O₃:Yb³⁺,Er³⁺ NRFs and REF samples could be attributed to the change of morphology. The In₂O₃:Yb³⁺,Er³⁺ NRFs constituted with dispersive nanorods has a large specific surface area, which introduced more surface ligands and surface defects. These groups and defects can bridge some non-radiative relaxation channels, such as ⁴I_13/2–⁴I_11/2 and ⁴S_3/2–⁴F_9/2, favoring the generation of the ⁴F_9/2–⁴I_15/2 red emission. In order to verify the conclusion, the In₂O₃:(10 mol%)Yb³⁺,(1 mol%)Er³⁺ power was prepared by the sol–gel method, the UCL spectra of In₂O₃:Yb³⁺,Er³⁺ power (the size > 10 μm) was measured under 980 nm excitation, shown in Fig. S2.† We can find that the R/G in the power is also dramatically lower than that in In₂O₃:Yb³⁺,Er³⁺ NRFs. The integral intensity of UCL for In₂O₃:Yb³⁺,Er³⁺ NRFs is lower than REF sample. Then the UCL intensity of In₂O₃:Yb³⁺,Er³⁺ NRFs exceeds REF samples with further increase the doping concentration of Er³⁺ (shown in the insert of Fig. 4). This phenomenon is known as concentration quenching and quite common in RE-doping NCs, which is due to an energy migration process among the same type of RE ions.²⁴ In In₂O₃:Yb³⁺,Er³⁺ NRFs, the nanorod is only ∼100 nm and dispersive each other, which can be clearly recognized by HR-TEM images (as shown in Fig. 3). The boundary effects in other traditional NCs would be unavoidably occurred. For example, in the Y₂O₃:Eu³⁺,Y₂SiO₅:Eu³⁺, and LaPO₄:Ce³⁺,Tb³⁺ inverse opal photonic crystals,^25–27 the quenching concentration of the RE emission increases more than relative to the corresponding bulk phosphors. In other words, the concentration quenching was suppressed considerably due to the boundary effects.

Fig. 4 Emission spectra of In₂O₃:Yb³⁺,Er³⁺ in REF and NRFs under 980 nm excitation. Insert: dependence of integral intensity on Er³⁺ ions concentration of In₂O₃:Yb³⁺,Er³⁺ REF and NRFs.

To further understand the UCL mechanism, the UCL dynamics of ⁴F_9/2–⁴I_15/2 transition of In₂O₃:Yb³⁺,(1 mol%)Er³⁺ REF and NRFs were measured (Fig. 5), and all the decay curves of Er³⁺ ions were fitted by the double-exponential function.²⁸ From the lifetime constants, we can concluded that: (1) in the lower doped concentration of Er³⁺ ions, the lifetimes in REF samples are longer than that in In₂O₃:Yb³⁺,Er³⁺ NRFs. As we known, the decay time constant is the reverse of the non-radiative and radiative decay rates. The non-radiative decay rate is significantly affected by the large phonon groups in the surface, crystal phase, lattice defect state and phonon energy. In In₂O₃:Yb³⁺,Er³⁺ NRFs, the non-radiative channels nearby Er³⁺ ions, such as lattice defect states/surface large phonon bands would be increased due to the increasing of the volume to surface ratio.²⁹ (2) In the In₂O₃:Yb³⁺,Er³⁺ NRFs, the lifetime constants decrease gradually with the increasing of Er³⁺ concentration compared to those of REF samples (as shown in the insert of Fig. 5), implying that the concentration quenching in the NRFs are suppressed. In the traditional In₂O₃:Yb³⁺,Er³⁺ phosphors, the long-range resonant energy transfer (ET) processes among Er³⁺ ions are very severe. Inevitably, luminescent quenching could occur because of the ET from Er³⁺ ions to defect states, which is dependent on doped concentrations. In the In₂O₃:Yb³⁺,Er³⁺ NRFs, these ET process should be restrained largely due to the structure of NRFs, because the size of each nanorod is ∼100 nm and the nanorods are dispersive. In this case, the ET among Er³⁺ ions can only happen within one nanorod, then, the photons will be scattered into the air instead of capturing by the defect states.

Fig. 5 The luminescent decay dynamics of the ⁴F_9/2–⁴I_15/2 transition at 660 nm in In₂O₃:Yb³⁺,Er³⁺ REF and NRFs. Insert: dependence of lifetimes on Er³⁺ concentration of In₂O₃:Yb³⁺,Er³⁺ REF and NRFs.

Fig. 6 shows the Yb³⁺ concentration dependent UCL spectra of the In₂O₃:Yb³⁺,Er³⁺ NRFs. From the spectra analysis of Fig. 6, the integral intensity of red emission gradually increased, whereas that of green emissions diminished gradually as the Yb³⁺ concentration increases. The emission spectra of the In₂O₃:Yb³⁺,Er³⁺ NRFs at different excitation powers were also measured (Fig. 7). We can found that the red emission is dominant in the low excitation power, and the R/G ratio is as high as ∼35 when the excitation power is 330 mW. The blue and green emissions increased gradually with increasing of excitation power. The R/G ratio versus excitation power is shown in the insert of Fig. 7, it is clear seen that the R/G ratios decreased with increasing of excitation power.

Fig. 6 UCL spectra in the In₂O₃:Yb³⁺,(1 mol%)Er³⁺ NRFs with different concentration of Yb³⁺.

Fig. 7 Power-dependent UC emission spectra of In₂O₃:(15 mol%)Yb³⁺,(1 mol%)Er³⁺ NRFs under 980 nm excitation. Insert: the R/G ratios of In₂O₃:Yb³⁺,Er³⁺ NRFs versus excitation power.

The ln–ln plots of the integral intensity versus excitation power for the red and green emissions of In₂O₃:(15 mol%)Yb³⁺,xEr³⁺ (x = 1, 3, 5, 8 mol%) NRFs are investigated (Fig. 8). As is well known, the visible output power intensity (I_V) will be proportional to power (n) of the infrared excitation (I_IR) power if the saturation effect can be neglected:³⁰ I_V ∝ I_IRⁿ, where n is the number of IR photons absorbed per visible photon emitted. The slopes n are determined to be 2.68, 2.62, 2.65, 2.4 for ²H_11/2–⁴I_15/2 transition, 2.87, 2.62, 2.73, 2.02 for ⁴S_3/2–⁴I_15/2 transition and 1.17, 1.71, 1.79, 1.86 for ⁴F_9/2–⁴I_15/2 transition in the In₂O₃:Yb³⁺,Er³⁺ NRFs at different Er³⁺ ions concentration of 1 mol%, 3 mol%, 5 mol% and 8 mol%, respectively. We can see that the slopes n are >2 for the green emissions and >1 for the red emission of Er³⁺. Thus the ²H_11/2/⁴S_3/2 levels are populating through three photons processes, and the ⁴F_9/2 level is populating through two-photons ET processes for all the In₂O₃:Yb³⁺,Er³⁺ NRFs. As far as we know, the three-photon populating phenomenon for the ²H_11/2/⁴S_3/2 levels is few reported in Er³⁺/Yb³⁺ codoped materials.

Fig. 8 Plot (ln–ln) of emission intensity versus excitation power in In₂O₃:(15 mol%)Yb³⁺,(1 mol%)Er³⁺ NRFs prepared at different concentration of Er³⁺ ions.

3.3 The mechanisms of UC population and luminescence processes in In₂O₃:Yb³⁺,Er³⁺ NRFs

The population mechanisms of ⁴S_3/2 and ⁴F_9/2 levels in In₂O₃:Yb³⁺,Er³⁺ NRFs are quite different to other UC nanocrystals from the above analysis. Fig. 9 presents UC population and emission diagram of the Er³⁺/Yb³⁺ doping system under 980 nm excitation. The ⁴I_11/2 level is populated by the electrons on the ground state ⁴I_15/2 via ET of neighboring Yb³⁺ ions. Subsequently, the ⁴I_13/2 level is populating through the non-radiative relaxation process of ⁴I_11/2–⁴I_13/2 via the assistance of large phonon groups, which is widely existed in NCs prepared via the hydrothermal method.³¹ In the second-step, the UC population of ⁴I_11/2–⁴F_7/2 and ⁴I_13/2–⁴F_9/2 occurred via ET and excited-state absorption, generating the red emission (⁴F_9/2–⁴I_15/2). The cross relaxation of ⁴F_7/2 + ⁴I_11/2 − ⁴F_9/2 + ⁴F_9/2 is significant for populating the ⁴F_9/2 level, when the excitation power or the Yb³⁺ concentration is sufficiently high. Then the Yb³⁺ concentration is lower in In₂O₃:Yb³⁺,Er³⁺ NRFs, the electrons in ⁴F_7/2 level is also populating ²H_11/2/⁴S_3/2 levels through non-radiative relaxation, generating ²H_11/2, ⁴S_3/2–⁴I_15/2 transitions. For populating ⁴S_3/2/²H_11/2 levels, the non-radiative relaxation of ⁴I_11/2–⁴I_13/2 is involved. The energy gap (ΔE) is ∼3700 cm⁻¹ between ⁴I_11/2 and ⁴I_13/2 level. For the bulk host, the phonon energy (hω) is ∼580 cm⁻¹,³² the matching of non-radiative relaxation of ⁴I_11/2–⁴I_13/2 requires at least seven host phonons, so this process is hardly happened. In In₂O₃:Yb³⁺,Er³⁺ NRFs prepared by the hydrothermal method, the large phonon groups are involved due to surfactant or surface defects, so the non-radiative relaxation of ⁴I_11/2–⁴I_13/2 can be bridged more effectively via the assistance of the large phonon group than the non-radiative relaxation of ⁴I_13/2–⁴I_15/2. Therefore, the linear decay is dominant for the ⁴I_11/2 level, because the effective non-radiative relaxation of ⁴I_11/2–⁴I_13/2 increases the linear decay processes, which leads the population of ⁴I_13/2–⁴F_9/2 to be more effective. This is a main reason for increasing the R/G ratio in In₂O₃:Yb³⁺,Er³⁺ NRFs.

Fig. 9 Energy level diagram of In₂O₃:Yb³⁺,Er³⁺ NRFs and UCL processes under 980 nm excitation.

The relation of R/G ratio and Yb³⁺ concentration can be explained below, the population of ⁴F_9/2 level is mainly originated from ET upconversion processes. The ET transfer from Yb³⁺ ions to Er³⁺ ions could be increased with the increase of Yb³⁺ concentration, thus the red emission (⁴F_9/2–⁴I_15/2) was accelerated because of an increase of the ET process. The quenching of green emission with the increasing Yb³⁺ concentration is due to the reversed ET from Er³⁺ to Yb³⁺ ions (⁴S_3/2(Er³⁺) + ²F_7/2(Yb³⁺) − ⁴I_13/2(Er³⁺) + ²F_5/2(Er³⁺)), which leads to the depopulation of the ⁴S_3/2 level. As we all known, the phonon energy of In₂O₃ is larger than that of Yb₂O₃, which may lead to the increase of nonradiative relaxation. However, all the emissions (including red emission and green emission) will be quenching due to the nonradiative relaxation. The quenching of the green emissions is mainly attributed to the reversed ET from Yb³⁺ ions to Er³⁺ ions, which can be confirmed by the fluorescence decay of ⁴S_3/2 level. Finally, the power-dependent UCL in In₂O₃:Yb³⁺,Er³⁺ NRFs can be well explained. The three-photon population processes ⁴F_9/2–⁴G_11/2 become more effective with the increasing of excitation power, leading the depopulation of the ⁴F_9/2 level and the population enhancement of ⁴G_11/2. The populations for ⁴F_3/2 and ⁴S_3/2/²H_11/2 levels also increase due to multi-phonon relaxation, so the R/G ratio decreases as the excitation power increases.

4. Conclusion

In₂O₃:Yb³⁺,Er³⁺ NRFs has been successfully prepared by a simple hydrothermal method combined with a succedent calcining process. It is also found that in In₂O₃:Yb³⁺,Er³⁺ NRFs, the concentration quenching is dramatically suppressed. The R/G ratio changed depending on the NP morphology, Yb³⁺ concentration and the excitation power. The green emissions are mainly originated from three-photon excitation in In₂O₃:Yb³⁺,Er³⁺ NRFs with higher Yb³⁺ concentration, which is different from other substrate materials. The In₂O₃:Yb³⁺,Er³⁺ NRFs are potential candidate for the application of lighting and solar cell.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work was supported by National Natural Science Foundation of China (Grant no. 11504188, 11504131, 51374132, 21571109), Natural Science Foundation of Henan Province (Grant no. U1504626).

References

1. F. Auzel, Chem. Rev., 2004, 104, 139–174.
2. P. P. Lei, R. An, S. Yao, Q. S. Wang, L. L. Dong, X. Xu, K. M. Du, J. Feng and H. J. Zhang, Adv. Mater., 2017, 29, 1700505.
3. M. Haase and H. Schafer, Angew. Chem., Int. Ed., 2011, 50, 5808–5829.
4. X. Chen, Y. S. Zhu, D. L. Zhou, W. Xu, J. Y. Zhu, G. C. Pan, Z. Yin, H. Wang, S. B. Cui and H. W. Song, J. Mater. Chem. C, 2017, 5, 2451–2458.
5. T. Hinamoto, H. Takashina, H. Sugimoto and M. Fujii, J. Phys. Chem. C, 2017, 121, 8077–8083.
6. I. Hyppänen, N. Höysniemi, R. Arppe, M. Schäferling and T. Soukka, J. Phys. Chem. C, 2017, 121, 6924–6929.
7. C. Yao, P. Y. Wang, R. Wang, L. Zhou, A. M. El-Toni, Y. Q. Lu, X. M. Li and F. Zhang, Anal. Chem., 2016, 88, 1930–1936.
8. Y. M. Li, Y. M. Li, R. Wang, Y. L. Xu and W. Zheng, New J. Chem., 2017, 41, 7116–7122.
9. S. Y. Han, A. Samanta, X. J. Xie, L. Huang, J. J. Peng, S. J. Park, D. B. L. Teh, Y. Choi, Y. T. Chang, A. H. All, Y. M. Yang, B. G. Xing and X. G. Liu, Adv. Mater., 2017, 29, 1700244.
10. Y. Liu, Q. Q. Su, M. Chen, Y. Dong, Y. B. Shi, W. Feng, Z. Y. Wu and F. Y. Li, Adv. Mater., 2016, 28, 6625–6630.
11. Y. S. Zhu, W. Xu, S. B. Cui, M. Liu, C. Lu, H. W. Song and D. H. Kim, J. Mater. Chem. C, 2016, 4, 331–339.
12. J. Wang, H. W. Song, W. Xu, B. Dong, S. Xu, B. T. Chen, W. Yu and S. Zhang, Nanoscale, 2013, 5, 3412–3420.
13. J. H. Kang, S. S. Lee, J. Guerrero, A. Fernandez-Nieves, S. H. Kim and E. Reichmanis, Adv. Mater., 2017, 29, 1606830.
14. Y. Luo, S. N. Du, W. Zhang, Z. F. Liao, F. Zuo and S. T. Yang, RSC Adv., 2017, 7, 37929–37937.
15. D. M. Yang, Y. L. Dai, P. G. Ma, X. J. Kang, Z. Y. Cheng, C. X. Li and J. Lin, Chem.–Eur. J., 2013, 19, 2685–2694.
16. S. W. Hao, G. Y. Chen, C. H. Yang, W. Shao, W. Wei, Y. Liu and P. N. Prasad, Nanoscale, 2017, 9, 10633–10638.
17. Y. G. Choi, S. M. Yu and W. J. Chung, Chem. Phys. Lett., 2008, 461, 290–293.
18. A. Podhorodecki, R. Kudrawiec, J. Misiewicz, N. V. Gaponenko and D. A. Tsyrkunov, Opt. Mater., 2006, 28, 685–687.
19. H. K. Kim, C. C. Li, G. Nykolak and P. C. Becker, J. Appl. Phys., 1994, 76, 8209–8211.
20. Q. B. Xiao, H. M. Zhu, D. T. Tu, E. Ma and X. Y. Chen, J. Phys. Chem. C, 2013, 117, 10834–10841.
21. L. Xu, B. Dong, Y. Wang, X. Bai, J. S. Chen, Q. Liu and H. W. Song, J. Phys. Chem. C, 2010, 114, 9089–9095.
22. D. P. Dutta, V. Sudarsan, P. Srinivasu, A. Vinu and A. K. Tyagi, J. Phys. Chem. C, 2008, 112, 6781–6785.
23. X. M. Xu, D. W. Wang, W. B. Wang, P. Sun, J. Ma, X. S. Liang, Y. F. Sun, Y. G. Ma and G. Y. Lu, Sens. Actuators, B, 2012, 171, 1066–1072.
24. Y. Ruan, Q. B. Xiao, W. Q. Luo, R. F. Li and X. Y. Chen, Nanotechnology, 2011, 22, 275701.
25. J. L. Ferrari, A. M. Pires and M. R. Davolos, Mater. Chem. Phys., 2009, 113, 587–590.
26. P. Ghosh, S. Sadhu and A. Patra, Phys. Chem. Chem. Phys., 2006, 8, 3342–3348.
27. Y. S. Zhu, Z. P. Sun, Z. Yin, H. W. Song, W. Xu, Y. F. Wang, L. G. Zhang and H. Z. Zhang, Dalton Trans., 2013, 42, 8049–8057.
28. M. Yu, J. Lin and J. Fang, Chem. Mater., 2005, 17, 1783–1791.
29. X. Bai, H. W. Song, G. H. Pan, Y. Q. Lei, T. Wang, X. G. Ren, S. Z. Lu, B. Dong, Q. L. Dai and L. B. Fan, J. Phys. Chem. C, 2007, 111, 13611–13617.
30. N. Niu, P. P. Yang, F. He, X. Zhang, S. L. Gai, C. X. Li and J. Lin, J. Mater. Chem., 2012, 22, 10889.
31. J. A. Capobianco, F. Vetrone, T. Alesio, G. Tessari and M. Bettinelli, Phys. Chem. Chem. Phys., 2000, 2, 3203–3207.
32. A. Raza, O. P. Agnihotri and B. K. Gupta, J. Phys. D: Appl. Phys., 1977, 10, 1871–1876.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7ra10582g

---