1. Introduction

As a direct band semiconductor with wide band-gap of 3.4 eV, ZnO has been explored as a doping host candidate for trivalent rare-earth (RE³⁺) ions, such as Eu³⁺, Er³⁺ and Tm³⁺ [1–3], due to their potential applications in flat panel displays and optoelectronic devices. As luminescent materials, the luminescence efficiency is one of the key factors that influence the practical applications. However, due to the large differences in atom radii and valence charge states between RE³⁺ and Zn²⁺, the trivalent RE ion is difficult to be doped into the ZnO lattice and thus most of the RE³⁺ ions were found either located on the surfaces or in the grain boundaries of ZnO crystal [4,5]. This leads to an inefficient energy transfer from ZnO to RE³⁺ and therefore a weak RE³⁺ emission. It is reported that co-doping RE³⁺ with Li⁺ in the ZnO host can greatly enhance the RE³⁺ emission intensities, although the role of Li⁺ ions is still unclear. Some considered that the co-doping Li⁺ ions might enhance the solubility of RE³⁺ ions in the ZnO and hence increase the number of the luminescent centers [4], or creates the oxygen vacancies that may act as the sensitizer for the energy transfer to the RE³⁺ ions. [6] While it was also suggested that the Li⁺ in ZnO host may distort the local symmetry and structure around RE³⁺ ions and thus enhance the energy transfer rate [7].

In this paper, we introduce a new type of ZnO-based luminescent materials, which mainly consist of ZnO with very small amount of LiYbO₂. The detailed microstructure analysis indicates that the LiYbO₂ phases are embedded in the ZnO substrate in the ZnO-LiYbO₂ hybrid phosphors and the diffusion layer were formed at the interfacial region at where the Li⁺, Yb³⁺ and Zn²⁺ ions diffused with each other. We then speculated that the high efficient energy transfer may benefit from the inter-diffusion of Li⁺, Yb³⁺ and Zn²⁺ in the interfacial region due to the strong interactions between ZnO and Yb³⁺. The spectral studies showed that the ZnO-LiYbO₂ hybrid phosphors can harvest the UV photons from a broad wavelength region and efficiently converted them into Yb³⁺ near-infrared emission in which wavelength the silicon solar cell shows the highest spectral response. As a result, this hybrid phosphor can be applied for the improvement of solar cells photovoltaic conversion efficiency via spectral modification. Comparing other spectral conversion luminescent materials with trivalent RE ions (Pr³⁺, Tm³⁺ and Tb³⁺) as energy absorption center, in which only a small portion of the solar spectrum is involved in the energy conversion process due to the nature of f-f transition [8–11], the ZnO-LiYbO₂ hybrid phosphor has the distinct advantages of broad band absorption in the near-UV region.

2. Experimental

The ZnO-LiYbO₂ hybrid phosphors were sintered by the solid-state reaction method in a weak reductive atmosphere by putting the crucible filled with raw materials into a bigger graphite crucible at 1050 °C for 2.5 hours. The starting materials were mainly analytical pure ZnO powders, mixing with 1mol% of high purity (99.99%) Li₂CO₃ and Yb₂O₃ each. The pure ZnO powder, the ZnO mixed with Yb₂O₃ or Li₂CO₃ were also sintered at the same condition for comparison. The X-ray diffraction (XRD) profiles were obtained on a Rigaku D/MAX-RA diffractometer using a Cu target. The scanning electron microscopy (SEM) and energy dispersive characteristic X-ray (EDX) observations were carried on using both Nova 200 NanoLab UHR FEG-SEM/FIB and FEI XL30 EFSEM. The excitation and emission spectra, the fluorescence decay curves and the time-resolved emission spectra in both visible and infrared regions were recorded by using a FLS920 fluorescence spectrophotometer.

3. Results and discussion

The structures of the ZnO-LiYbO₂ hybrid phosphors were studied by XRD, as shown in Fig. 1 . The XRD profiles acquired from the pure ZnO powder and the mixture of ZnO and Yb₂O₃ are also illustrated for comparison. In the pure ZnO sample, all the diffraction peaks can be indexed by the hexagonal Wurtzite ZnO, while in the mixture of ZnO and Yb₂O₃, the diffraction patterns are consisted of both Wurtzite ZnO and cubic Yb₂O₃ peaks. A closer comparison of ZnO peaks with and without Yb₂O₃ mixing shows no measurable change in the peak positions, indicating that Yb³⁺ ions cannot be doped into ZnO lattice by the solid-state reaction method, at least, the doping level cannot be detected by the XRD. In the ZnO-LiYbO₂ hybrid phosphors, new diffraction peaks other than those belonging to ZnO and Yb₂O₃ appear, which can be well indexed as LiYbO₂ phase [12].

 

Fig. 1 X-ray diffraction patterns of (a) pure ZnO, (b) 1mol% Yb₂O₃ mixed ZnO, (c) 1 mol% Yb₂O₃-1 mol% Li₂O mixed ZnO.

Download Full Size | PDF

The microstructure and composition of the ZnO-Yb₂O₃ mixtures and the ZnO-LiYbO₂ hybrid phosphors were further characterized by SEM and EDX, as shown in Fig. 2 .

 

Fig. 2 SEM images of (a): ZnO mixed with 1 mol% Yb₂O₃. (b): ZnO-LiYbO₂ hybrid phosphor (with starting concentration of 1 mol% Yb₂O₃ and Li₂O each). (c): enlarged image showing two Yb₂O₃ crystals adsorbed to ZnO and one LiYbO₂ crystal embedded on ZnO. (d): fracture surface between the ZnO substrate and a breakaway LiYbO₂.

Download Full Size | PDF

In the ZnO-Yb₂O₃ mixture, some large Yb₂O₃ aggregate as free-standing pieces, and some small Yb₂O₃ particles are adsorbed to the ZnO surfaces, as indicated with blue arrows and blue circles, respectively, in Fig. 2(a). However, in the ZnO-LiYbO₂ hybrid phosphors, besides the adsorbed Yb₂O₃ particles (indicated by blue circle), there are also some small particles embedded on the surface of ZnO, indicated by red circles in Fig. 2(b). A close look shows that the overall shape of these embedded particles is very different from that of the adsorbed Yb₂O₃ particles, as shown in Fig. 2(c). The adsorbed Yb₂O₃ particles have sharp edges and corners, while the embedded particles are generally in round shape with porous-like surfaces. EDX analyses show that both particles consist of Yb and O, while there is no detectable Yb in the ZnO substrate. According to XRD results, these embedded particles are most likely the products of solid-state reaction, LiYbO₂, while the adsorbed particles are the remaining Yb₂O₃. We also noticed that in the synthesis of LiYbO₂ compounds by the solid-state reaction method, the mixture of Yb₂O₃ and Li₂CO₃ were reacted in air at 900°C [12]. In our case, on contrast, the ZnO-LiYbO₂ hybrid phosphor were sintered in a reduced environment in order to obtain intense visible emission from ZnO. Therefore, we speculate that the ZnO surface may catalyze the reaction, providing O, and thus LiYbO₂ nuclear on the ZnO surface. The strong evidence for this speculation comes from the SEM images of the fracture surfaces of the breakaway LiYbO₂ nanoparticles, as shown in Fig. 2 (d). It is seen that the removal of a LiYbO₂ nanoparticle leaves a pit on the ZnO surface, and the EDX analysis shows that the remains in the fracture surface still contains a small amount of Yb, but no detectable Yb on the ZnO surface.

The spectral properties of the ZnO-LiYbO₂ hybrid phosphors, the pure ZnO and the Yb₂O₃ or Li₂O mixed ZnO were studied. In the mixture of ZnO with 1mol% Yb₂O₃, a strong broad green emission band peaking at 500 nm and a weak infrared emission around 980 nm were observed with 380 nm excitations, as shown in Fig. 3(a) . The green emission, known as intrinsic self-activated (SA) emission in ZnO materials sintered under oxygen-deficient condition, is due to the recombination of the excited electrons and deeply trapped holes [13,14]. The infrared emission is originated from the Yb^{3+ 2}F_5/2→²F_7/2 transition. It is also noticed that the excitation spectrum for the infrared emission overlaps very well with that for the SA emission. This indicates the occurrence of energy transfer from ZnO to Yb³⁺ ions, which subsequently leads to the infrared emission. However, the infrared emission was extremely weak due to the inefficient energy transfer since only a trace of Yb³⁺ ions might be incorporated into ZnO lattice, according to the XRD and SEM results.

 

Fig. 3 Excitation and emission spectra for (a): 1 mol% Yb₂O₃ mixed ZnO, λ_em=500 nm (black line), λ_em=980 nm (red line), and λ_ex=380 nm (blue line). (b) ZnO-LiYbO₂ hybrid phosphor (with starting concentration of 1 mol% Yb₂O₃ and Li₂O each), λ_em=540 nm (black line), λ_em = 980 nm (red line), and λ_ex=395 nm (blue lines). (c): pure ZnO (orange line) and 1 mol% Li₂O doped ZnO (green line), λ_em=500 nm, λ_ex=390 and 350 nm, respectively.

Download Full Size | PDF

Interestingly, in the ZnO-LiYbO₂ hybrid phosphors, with the excitation of ZnO absorption at 395 nm, the Yb³⁺ infrared emission intensity becomes three magnitudes higher than that in the ZnO-Yb₂O₃ mixture, as shown in Fig. 3(b). Moreover, the excitation bands for the SA emission and Yb³⁺ emission are significantly different from each other. For the SA emission, the narrow excitation band is almost quenched, on the contrary, which is largely enhanced for the Yb³⁺ emission.

In order to explain these phenomena, the excitation and emission spectra were also measured in the pure ZnO and Li₂O-doped ZnO, as shown in Fig. 3(c). It is seen that there are two excitation bands for the SA emission in pure ZnO, a broad one peaked at 350 nm and a narrow one located at 390 nm, corresponding to the band-band excitation and excitation absorption of ZnO, respectively [15]. In the Li₂O-doped ZnO, the narrow excitation band for the SA emission is remarkable decreased. This is consistent with the observations in the ZnO-LiYbO₂ hybrid phosphors, in which this narrow excitation band is also quenched. One possible explanation for this phenomenon is that in the Li₂O-doped ZnO, the Li⁺ ions may create defect energy levels, which provide a pathway for non-radiative recombination of electrons and holes, and thus lead to the decrease of the narrow excitation band for the SA emission. ¹³ On the other hand, the Li⁺ related defect energy levels may also act as donor levels to sensitize Yb³⁺ ions by favoring its emission intensity in the ZnO-LiYbO₂ hybrid phosphors. As the radius of Li⁺ is small, it is generally agreed that in the Li⁺ doped ZnO, Li⁺ ions can both occupy the interstitial sits or substitute Zn ions,^{5, 7} in this case, the Li related defect energy level is non-localized. However, the role of Li⁺ in the energy transfer from ZnO to Yb³⁺ need to be further investigated [16].

The greatly enhanced Yb³⁺ emission intensity in the ZnO-LiYbO₂ hybrid phosphors indicates that the energy transfer is more efficient than that in the ZnO-Yb₂O₃ mixture. The interpretation of this phenomenon is straightforward from the microstructure point of view. According to the SEM analysis, the Yb₂O₃ particles are only physically adsorbed to ZnO, in which case the energy transfer from ZnO to Yb³⁺ should be extremely weak. While the LiYbO₂ nanoparticles are grown in the ZnO surfaces with diffusion interfacial regions, therefore, the high efficient energy transfer from ZnO to Yb³⁺ can be expected at the interfacial region due to the strong interactions between ZnO and Yb³⁺ [17].

At last, it is worth to mention that in the ZnO-LiYbO₂ hybrid phosphors, the SA emission lifetime became longer under the band-band excitation at 350 nm, as shown in Fig. 4 , this is because the electrons were trapped by the defect energy level created by Yb³⁺ ions at the diffusion regions before recombine with the holes. According to the XRD and EDX measurement, the Yb³⁺ ions can’t be doped into ZnO lattice, in this case, the Yb³⁺ associated defect energy level is localized in the interfacial diffusion region only. This result also explained the red shift of the SA emission in ZnO-LiYbO₂ hybrid phosphors compared to that in the pure ZnO. The SA emission in the pure ZnO is due to the radiative recombination of electrons from a level close to the conduction band edge and deeply trapped holes [14,15], whereas in the ZnO-LiYbO₂ hybrid phosphors, the SA emission is due to the radiative recombination of the electrons trapped by the Yb³⁺ related defect energy levels that somewhat deeper into the forbidden band and the deeply trapped holes.

 

Fig. 4 Luminescence decay of the SA emission with 350 nm excitation in pure ZnO and ZnO-LiYbO₂ hybrid phosphor, respectively.

Download Full Size | PDF

4. Conclusion

In conclusion, the intense infrared emission associated with the Yb^{3+ 2}F_5/2→²F_7/2 transition was obtained due to the high efficient energy transfer from ZnO substrate in the ZnO-LiYbO₂ hybrid phosphors, in which the emission intensity is three magnitudes higher than that in the ZnO-Yb₂O₃ mixture where the Yb₂O₃ crystals are physically touched with ZnO. The greatly enhanced infrared emission in the ZnO-LiYbO₂ hybrid phosphors is benefited from the Yb³⁺ ions located at interfacial diffusion region due to the strong interaction with ZnO. The spectroscopy results clearly indicate that the ZnO-LiYbO₂ hybrid phosphors can harvest the energy from near-UV photons in a broad wavelength region and effectively convert them into the wavelength region that the silicon PV device exhibits the maximum spectral response. This work not only reveals the nature of the greatly enhanced Yb³⁺ infrared emission in the ZnO-LiYbO₂ hybrid phosphors, but also provides a new method to realize the high efficient broadband spectral modification by designing the phosphors with hybrid structures in micro scale.