Visible Light Photocatalytic Activity of NaYF 4 :(Yb,Er)-CuO/TiO 2 Composite

: TiO 2 photocatalysts are limited to ultraviolet light photo-activation, however, by coupling with NaYF 4 :(Yb,Er) they can utilize visible light activation, but with low efficiencies. In order to enhance visible light photo-activity, CuO nanoparticles were coupled with NaYF 4 :(Yb,Er) by mechanical ball-milling and coated with TiO 2 -sol. The coupling of CuO nanoparticles with NaYF 4 :(Yb,Er)/TiO 2 caused the formation of a NaYF 4 :(Yb,Er)-CuO/TiO 2 composite capable of visible absorption with a remarkably reduced band gap of ~2.5 eV. The NaYF 4 :(Yb,Er)-CuO/TiO 2 composite in H 2 O 2 showed the most efficient methylene blue (MB) degradation efficiency of more than 99.5% after exposure to visible light.

Interestingly, the coupling of TiO 2 with upconversion phosphors can remarkably extend utilization of the whole solar spectrum [17]. The major advantage of upconversion phosphor is the ability to convert even low near-infrared (NIR) photons to higher energy photons in the UV and visible light range [18,19]. In this way, the NaYF 4 :(Yb, Er) upconversion phosphor characteristic emerges from Yb 3+ ions which absorb the 980 nm wavelength spectrum and transfer energy to Er 3+ ions which then emit visible light photons. Er 3+ ions emit visible light at 450, 525, and 650 nm wavelengths due to transitions from its ground state 4 I 15/2 to 4 F 7/2 , 4 H 11/2 , 4 F 9/2 and 4 G 9/2 , respectively [20][21][22]. For this reason, Yb 3+ ions are termed sensitizers while Er 3+ ions are activators. The outstanding photo-stability of upconversion phosphors has propelled their extensive research for dye sensitized solar cells, bio-imaging and UV-vis-NIR photocatalysis applications [19,23].
In a NaYF 4 :(Yb,Er)/TiO 2 composite, NaYF 4 :(Yb,Er) phosphor can offer an efficient catalyst support and provide the interfacial energy band bending so that the TiO 2 absorption edge can be extended to the visible region [14,22]. However, the visible light photocatalysis of a NaYF 4 :(Yb,Er)/TiO 2 composite shows low efficiency. For example, complete photo-degradation of dye pollutants mixed with phosphor-TiO 2 photocatalyst occurs after 10 h of exposure to a visible light source [17,24].
Photocatalysis efficiency depends on various factors which include intensity of illuminating light, temperature, concentration of dye pollutants and the amount of doping elements in the photocatalyst  (302) respectively. The thermal treatment may have induced the phase transition from α to β-NaYF 4 . Furthermore, due to the co-existence of three phases (cubic NaYF 4 , hexagonal NaYF 4 and CuO), it is expected that intermetallic compounds are formed during thermal treatment. Although no other Cu phases were identified at different annealing conditions, a 450 • C annealing temperature is known for pure CuO phases with band gaps between 1.2 eV and 2 eV [41]. Therefore, NaYF 4 :(Yb,Er)-CuO composites annealed at 450 • C were utilized for TiO 2 coating. TiO 2 peaks could not be observed in the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite because its amount was beyond the XRD detection limit.   Figure 2b are dense spherical particles. In Figure 2c, NaYF4:(Yb,Er)-CuO composite powders consist of particles of sizes below 1 µm. It is evident that mechanical ball-milling might reduce the size of NaYF4:(Yb,Er) phosphor agglomerates. The NaYF4:(Yb,Er)-CuO/TiO2 composite in Figure 2d consists of agglomerates with dense particles of nanometer range, due to TiO2 coating. Figure 2e shows the particle size distribution for the NaYF4:(Yb,Er)-CuO/TiO2 composite. The particle sizes are ranging between 0.8 and 4 µm. The EDS spectra for the NaYF4:(Yb,Er)-CuO/TiO2 composite confirmed the presence of all elements, as shown in Figure 2f.   Figure 2d consists of agglomerates with dense particles of nanometer range, due to TiO 2 coating. Figure 2e shows the particle size distribution for the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite. The particle sizes are ranging between 0.8 and 4 µm. The EDS spectra for the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite confirmed the presence of all elements, as shown in Figure 2f.   Figure 2b are dense spherical particles. In Figure 2c, NaYF4:(Yb,Er)-CuO composite powders consist of particles of sizes below 1 µm. It is evident that mechanical ball-milling might reduce the size of NaYF4:(Yb,Er) phosphor agglomerates. The NaYF4:(Yb,Er)-CuO/TiO2 composite in Figure 2d consists of agglomerates with dense particles of nanometer range, due to TiO2 coating. Figure 2e shows the particle size distribution for the NaYF4:(Yb,Er)-CuO/TiO2 composite. The particle sizes are ranging between 0.8 and 4 µm. The EDS spectra for the NaYF4:(Yb,Er)-CuO/TiO2 composite confirmed the presence of all elements, as shown in Figure 2f.   Figure 3 shows N2 adsorption-desorption isotherms for the NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2 photocatalyst composites. The specific surface areas in the table insert in Figure 3 are 15.9 m 2 /g and 12.4 m 2 /g for NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2, respectively, while the pore size and adsorbed volume are 8.93 nm and 0.0355 cm 3 /g for NaYF4:(Yb,Er)/TiO2 and 13.3 nm and 0.0413 cm 3 /g for NaYF4:(Yb,Er)-CuO/TiO2, respectively. The specific surface area of pure NaYF4:(Yb,Er) in other previous research is 11.5 m 2 /g [22]. The differences in the specific surface area, pore volume and diameter are dependent on the synthesis variables and changes in crystalline composition of the photocatalysts. Therefore, the low surface area and high volume adsorbed/pore diameter observed in the NaYF4:(Yb,Er)-CuO/TiO2 composite confirms the change in crystallinity with CuO nanoparticles coupling.   Figure 3 are 15.9 m 2 /g and 12.4 m 2 /g for NaYF 4 :(Yb,Er)/TiO 2 and NaYF 4 :(Yb,Er)-CuO/TiO 2 , respectively, while the pore size and adsorbed volume are 8.93 nm and 0.0355 cm 3 /g for NaYF 4 :(Yb,Er)/TiO 2 and 13.3 nm and 0.0413 cm 3 /g for NaYF 4 :(Yb,Er)-CuO/TiO 2, respectively. The specific surface area of pure NaYF 4 :(Yb,Er) in other previous research is 11.5 m 2 /g [22]. The differences in the specific surface area, pore volume and diameter are dependent on the synthesis variables and changes in crystalline composition of the photocatalysts. Therefore, the low surface area and high volume adsorbed/pore diameter observed in the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite confirms the change in crystallinity with CuO nanoparticles coupling.  Figure 4a shows the UV-vis absorption spectra for NaYF4:(Yb,Er), CuO and NaYF4:(Yb,Er)-CuO/TiO2 samples vs. wavelength. The NaYF4:(Yb,Er)-CuO/TiO2 composite exhibited similar absorption characteristics to CuO, but with an extended UV absorption edge at 380 nm. This absorption edge is related to both TiO2 nanoparticles and NaYF4:(Yb,Er). The enlarged UV-vis absorption spectra in Figure 4b clearly shows absorption peaks at 380 nm, 522 nm and 655 nm which are characteristic absorption peaks for NaYF4:(Yb,Er) upconversion phosphors. These Er 3+ absorption peaks were originated from the ground states 4 I15/2 to 4 G9/2, 4 F7/2 and 4 S3/2, respectively [42]. However, in the TiO2 coated NaYF4:(Yb,Er) phosphor, the 380 nm peak is not observed because it corresponds to the TiO2 absorption range. The estimated band energies for CuO, NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2 are 1.5 eV, 3.18 eV and 2.5 eV, respectively, as shown in Figure 4c. The remarkable band gap reduction in NaYF4:(Yb,Er)-CuO/TiO2 typically signifies the composites can be photo-activated with both UV and visible light for photocatalysis reactions.  Figure 4a shows the UV-vis absorption spectra for NaYF 4 :(Yb,Er), CuO and NaYF 4 :(Yb,Er)-CuO/TiO 2 samples vs. wavelength. The NaYF 4 :(Yb,Er)-CuO/TiO 2 composite exhibited similar absorption characteristics to CuO, but with an extended UV absorption edge at 380 nm. This absorption edge is related to both TiO 2 nanoparticles and NaYF 4 :(Yb,Er). The enlarged UV-vis absorption spectra in Figure 4b clearly shows absorption peaks at 380 nm, 522 nm and 655 nm which are characteristic absorption peaks for NaYF 4 :(Yb,Er) upconversion phosphors. These Er 3+ absorption peaks were originated from the ground states 4 I 15/2 to 4 G 9/2 , 4 F 7/2 and 4 S 3/2 , respectively [42]. However, in the TiO 2 coated NaYF 4 :(Yb,Er) phosphor, the 380 nm peak is not observed because it corresponds to the TiO 2 absorption range. The estimated band energies for CuO, NaYF 4 :(Yb,Er)/TiO 2 and NaYF 4 :(Yb,Er)-CuO/TiO 2 are 1.5 eV, 3.18 eV and 2.5 eV, respectively, as shown in Figure 4c. The remarkable band gap reduction in NaYF 4 :(Yb,Er)-CuO/TiO 2 typically signifies the composites can be photo-activated with both UV and visible light for photocatalysis reactions.  Figure 5a shows the photoluminescence spectra of NaYF4:(Yb,Er), NaYF4:(Yb,Er)-CuO and NaYF4:(Yb,Er)-CuO/TiO2. The intensity of the photoluminescence (PL) peaks denotes the extent of energy transfer efficiency [18]. Pure NaYF4:(Yb,Er) phosphor has visible light emission peaks at 450 nm, 545 nm, 660 nm and a near infrared peak at 825 nm. However, the emission peaks diminish with coupling NaYF4:(Yb,Er) with CuO nanoparticles, TiO2 coating on NaYF4:(Yb,Er) phosphor and in NaYF4:(Yb,Er)-CuO composite. Although energy transfer efficiency is evaluated by the decrease in peak intensities, in TiO2 coated samples, the peak reduction is generally referenced to light emission hindrance by TiO2 nanoparticles [14]. However, the overall peak suppression can be explained by considering the band structure of the NaYF4:(Yb,Er)-CuO/TiO2 composite. Figure 5c shows the schematic band structure of the NaYF4:(Yb,Er)-CuO/TiO2 composite and the energy transfer processes labelled (A), (B), (C) and (D). Firstly, if process (A) is considered as excitation of NaYF4:(Yb,Er) phosphor with 980 nm illumination, because of its luminescent characteristics, energy photons in the visible and NIR are emitted. Thus, as seen in Figure 5a, emission peaks were observed at 450 nm, 545 nm, 660 nm and 825 nm. Secondly, if process (B) is considered as the energy transfer route for light photons emitted from NaYF4:(Yb,Er), the visible light peaks are absorbed by CuO. However, the emission peaks are suppressed at (C) in the NaYF4:(Yb,Er)-CuO composite. Finally, when considering light photons emitted by NaYF4:(Yb,Er) phosphor at (B) are transferred to TiO2 at (C) via CuO, energy losses are inevitable at (D). The reasons for peak suppression being that energy is lost through multi-phonon vibrations and the blockage of phosphor emission centers by TiO2 and CuO nanoparticles [17].  Figure 5a shows the photoluminescence spectra of NaYF 4 :(Yb,Er), NaYF 4 :(Yb,Er)-CuO and NaYF 4 :(Yb,Er)-CuO/TiO 2 . The intensity of the photoluminescence (PL) peaks denotes the extent of energy transfer efficiency [18]. Pure NaYF 4 :(Yb,Er) phosphor has visible light emission peaks at 450 nm, 545 nm, 660 nm and a near infrared peak at 825 nm. However, the emission peaks diminish with coupling NaYF 4 :(Yb,Er) with CuO nanoparticles, TiO 2 coating on NaYF 4 :(Yb,Er) phosphor and in NaYF 4 :(Yb,Er)-CuO composite. Although energy transfer efficiency is evaluated by the decrease in peak intensities, in TiO 2 coated samples, the peak reduction is generally referenced to light emission hindrance by TiO 2 nanoparticles [14]. However, the overall peak suppression can be explained by considering the band structure of the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite. Figure Figure 6a shows the degradation efficiency of the methylene blue (MB) solution with H2O2 only (no photocatalyst), without and with H2O2 for NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2 under visible light for 1 h. The NaYF4:(Yb,Er)-CuO/TiO2 composites exhibit the fastest degradation efficiency, whereby 60% of the MB solution is degraded within 1 h. The photocatalytic reaction of NaYF4:(Yb,Er)/TiO2 without H2O2 oxidant molecules is very slow, but with hydrogen peroxide the reaction is enhanced. This is attributed to the unique role of hydrogen peroxide decomposing in the presence of a photocatalyst and light to give hydroxyl and superoxide ions which attack and degrade MB-dye molecules [39]. The photocatalytic degradation of MB solution with NaYF4:(Yb,Er)-CuO/TiO2 follows first order reaction kinetics as shown in Figure 6b. The rate constant for NaYF4:(Yb,Er)-CuO/TiO2 is 0.008 min −1 . Whilst for NaYF4:(Yb,Er)/TiO2 without and with H2O2 is 0.0015 min −1 and 0.0043 min −1 , respectively. Accordingly from the rate constants, the degradation with NaYF4:(Yb,Er)-CuO/TiO2 is five times and almost two times faster than in NaYF4:(Yb,Er)/TiO2 without and with H2O2, respectively.  molecules is very slow, but with hydrogen peroxide the reaction is enhanced. This is attributed to the unique role of hydrogen peroxide decomposing in the presence of a photocatalyst and light to give hydroxyl and superoxide ions which attack and degrade MB-dye molecules [39]. The photocatalytic degradation of MB solution with NaYF 4 :(Yb,Er)-CuO/TiO 2 follows first order reaction kinetics as shown in Figure 6b. The rate constant for NaYF 4 :(Yb,Er)-CuO/TiO 2 is 0.008 min −1 . Whilst for NaYF 4 :(Yb,Er)/TiO 2 without and with H 2 O 2 is 0.0015 min −1 and 0.0043 min −1 , respectively. Accordingly from the rate constants, the degradation with NaYF 4 :(Yb,Er)-CuO/TiO 2 is five times and almost two times faster than in NaYF 4 :(Yb,Er)/TiO 2 without and with H 2 O 2 , respectively.  Figure 6a shows the degradation efficiency of the methylene blue (MB) solution with H2O2 only (no photocatalyst), without and with H2O2 for NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2 under visible light for 1 h. The NaYF4:(Yb,Er)-CuO/TiO2 composites exhibit the fastest degradation efficiency, whereby 60% of the MB solution is degraded within 1 h. The photocatalytic reaction of NaYF4:(Yb,Er)/TiO2 without H2O2 oxidant molecules is very slow, but with hydrogen peroxide the reaction is enhanced. This is attributed to the unique role of hydrogen peroxide decomposing in the presence of a photocatalyst and light to give hydroxyl and superoxide ions which attack and degrade MB-dye molecules [39]. The photocatalytic degradation of MB solution with NaYF4:(Yb,Er)-CuO/TiO2 follows first order reaction kinetics as shown in Figure 6b. The rate constant for NaYF4:(Yb,Er)-CuO/TiO2 is 0.008 min −1 . Whilst for NaYF4:(Yb,Er)/TiO2 without and with H2O2 is 0.0015 min −1 and 0.0043 min −1 , respectively. Accordingly from the rate constants, the degradation with NaYF4:(Yb,Er)-CuO/TiO2 is five times and almost two times faster than in NaYF4:(Yb,Er)/TiO2 without and with H2O2, respectively.   Figure 7a shows the photo-degradation efficiency of MB solution with and without H2O2 for three samples: NaYF4:(Yb,Er)/TiO2, NaYF4:(Yb,Er)-CuO and NaYF4:(Yb,Er)-CuO/TiO2 under visible light. Clearly, the NaYF4:(Yb,Er)-CuO/TiO2 composite in H2O2 has the most efficient photo-activity by decomposing the MB solution more than 99.5% after exposure to visible light. This photocatalytic activity has higher efficiency as compared to only TiO2-coated phosphor in our previous work [16] and other research [33,43]. The coupling of CuO nanoparticles with NaYF4:(Yb, Er)/TiO2 caused the formation of visible light-active nanocomposites. Thus, the NaYF4:(Yb, Er)-CuO/TiO2 nanocomposites and H2O2 oxidant molecules enhanced photo-activation and separation of electronhole pairs for improved photocatalytic activity. Figure 7b shows the recyclability of the NaYF4:(Yb,Er)-CuO/TiO2 composite under visible light for three cycles. After the first run, the photocatalytic degradation efficiency of the MB solution decreased from almost 100% to 80% and almost 60% in the second and third cycle, respectively. The noticeable difference in photocatalyst efficiency is due to several factors relating to surface-adsorption characteristics, such as Cu leaching [44] and the possible loss of TiO2 nanoparticles [43]. Specifically, the centrifugal process and several washing steps during material preparation may result in a loss of TiO2 nanoparticles.  has the most efficient photo-activity by decomposing the MB solution more than 99.5% after exposure to visible light. This photocatalytic activity has higher efficiency as compared to only TiO 2 -coated phosphor in our previous work [16] and other research [33,43]. The coupling of CuO nanoparticles with NaYF 4 :(Yb, Er)/TiO 2 caused the formation of visible light-active nanocomposites. Thus, the NaYF 4 :(Yb, Er)-CuO/TiO 2 nanocomposites and H 2 O 2 oxidant molecules enhanced photo-activation and separation of electron-hole pairs for improved photocatalytic activity. Figure 7b shows the recyclability of the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite under visible light for three cycles. After the first run, the photocatalytic degradation efficiency of the MB solution decreased from almost 100% to 80% and almost 60% in the second and third cycle, respectively. The noticeable difference in photocatalyst efficiency is due to several factors relating to surface-adsorption characteristics, such as Cu leaching [44] and the possible loss of TiO 2 nanoparticles [43]. Specifically, the centrifugal process and several washing steps during material preparation may result in a loss of TiO 2 nanoparticles.  Figure 7a shows the photo-degradation efficiency of MB solution with and without H2O2 for three samples: NaYF4:(Yb,Er)/TiO2, NaYF4:(Yb,Er)-CuO and NaYF4:(Yb,Er)-CuO/TiO2 under visible light. Clearly, the NaYF4:(Yb,Er)-CuO/TiO2 composite in H2O2 has the most efficient photo-activity by decomposing the MB solution more than 99.5% after exposure to visible light. This photocatalytic activity has higher efficiency as compared to only TiO2-coated phosphor in our previous work [16] and other research [33,43]. The coupling of CuO nanoparticles with NaYF4:(Yb, Er)/TiO2 caused the formation of visible light-active nanocomposites. Thus, the NaYF4:(Yb, Er)-CuO/TiO2 nanocomposites and H2O2 oxidant molecules enhanced photo-activation and separation of electronhole pairs for improved photocatalytic activity. Figure 7b shows the recyclability of the NaYF4:(Yb,Er)-CuO/TiO2 composite under visible light for three cycles. After the first run, the photocatalytic degradation efficiency of the MB solution decreased from almost 100% to 80% and almost 60% in the second and third cycle, respectively. The noticeable difference in photocatalyst efficiency is due to several factors relating to surface-adsorption characteristics, such as Cu leaching [44] and the possible loss of TiO2 nanoparticles [43]. Specifically, the centrifugal process and several washing steps during material preparation may result in a loss of TiO2 nanoparticles.  The overall photocatalytic activity can be influenced by the chemical composition, specific surface area, optical property of the photocatalyst and the presence of H2O2 oxidant molecules [39]. Firstly, the change in compositional crystallinity by coupling NaYF4:(Yb,Er)/TiO2 with CuO evidently improved the photocatalytic activity. Moreover, the co-existence of CuO, α-NaYF4 and β-NaYF4 phases as confirmed by XRD further increases the TiO2 absorption edge towards visible light. Secondly, it is well known that the high specific surface area is related to the enhancement of photocatalytic activity. Nevertheless, the case is unique in the NaYF4:(Yb,Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2 photocatalyst composites. Thus, NaYF4:(Yb,Er)/TiO2 showed higher specific area (15.9 m 2 /g), but exhibited lower photocatalytic activity than the NaYF4:(Yb,Er)-CuO/TiO2 (12.4 m 2 /g) [45]. The explanation for this discrepancy is regardless of the specific surface area; the dominating property is the existence of a visible light-active CuO compound. Therefore, with CuO nanoparticle coupling (as the third compound in the NaYF4:(Yb,Er)-CuO/TiO2 composite) visible light absorption by composites is improved and these composites have more efficient charge transfer properties than only TiO2 coated phosphor [33]. Thirdly, the photoactivity is low without H2O2 oxidant molecules in the photo-reaction mixture of TiO2 coated composites (NaYF4:(Yb, Er)/TiO2 and NaYF4:(Yb,Er)-CuO/TiO2) and NaYF4:(Yb,Er)-CuO as exhibited in Figure 7a. Although photo-degradation up to 40% is observed in the TiO2-coated composites, the NaYF4:(Yb,Er)-CuO degraded only 20% of MB solution. This is due to the higher oxidizing ability of TiO2 as compared to CuO [46]. In contrast, the photocatalyst mixture with H2O2 oxidant molecules showed an improvement in photocatalytic activity of the NaYF4:(Yb,Er)/TiO2, NaYF4:(Yb,Er)-CuO and NaYF4:(Yb,Er)-CuO/TiO2. Thus, the H2O2 oxidant was photo-degraded to produce hydroxyl and super oxide ions which might degrade the MB-dye solution. The superior oxidizing power of H2O2 molecules has been utilized in several research studies to enhance the photocatalytic activity of pure CuO and its composite materials [47]. Finally, the major advantage of coupling CuO and TiO2 nanoparticles with phosphor is the inevitable alteration of the NaYF4:(Yb,Er)-CuO/TiO2 composites band structure, as observed in both visible light absorption and band gap estimations in Figure 4. The NaYF4:(Yb,Er) can also facilitate the electronhole charge separation process which enhances photocatalytic efficiencies [17]. Additionally, the phosphor crystalline surface not only supports the nanoparticles, but it also facilitates energy transfer within the composite system as described in Figure 5c. Accordingly, the composites can also be utilized efficiently in solar harvesting systems. The overall photocatalytic activity can be influenced by the chemical composition, specific surface area, optical property of the photocatalyst and the presence of H 2 O 2 oxidant molecules [39]. Firstly, the change in compositional crystallinity by coupling NaYF 4 :(Yb,Er)/TiO 2 with CuO evidently improved the photocatalytic activity. Moreover, the co-existence of CuO, α-NaYF 4 and β-NaYF 4 phases as confirmed by XRD further increases the TiO 2 absorption edge towards visible light. Secondly, it is well known that the high specific surface area is related to the enhancement of photocatalytic activity. Nevertheless, the case is unique in the NaYF 4 :(Yb,Er)/TiO 2 and NaYF 4 :(Yb,Er)-CuO/TiO 2 photocatalyst composites. Thus, NaYF 4 :(Yb,Er)/TiO 2 showed higher specific area (15.9 m 2 /g), but exhibited lower photocatalytic activity than the NaYF 4 :(Yb,Er)-CuO/TiO 2 (12.4 m 2 /g) [45]. The explanation for this discrepancy is regardless of the specific surface area; the dominating property is the existence of a visible light-active CuO compound. Therefore, with CuO nanoparticle coupling (as the third compound in the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite) visible light absorption by composites is improved and these composites have more efficient charge transfer properties than only TiO 2 coated phosphor [33]. Thirdly, the photoactivity is low without H 2 O 2 oxidant molecules in the photo-reaction mixture of TiO 2 coated composites (NaYF 4 :(Yb, Er)/TiO 2 and NaYF 4 :(Yb,Er)-CuO/TiO 2 ) and NaYF 4 :(Yb,Er)-CuO as exhibited in Figure 7a. Although photo-degradation up to 40% is observed in the TiO 2 -coated composites, the NaYF 4 :(Yb,Er)-CuO degraded only 20% of MB solution. This is due to the higher oxidizing ability of TiO 2 as compared to CuO [46]. In contrast, the photocatalyst mixture with H 2 O 2 oxidant molecules showed an improvement in photocatalytic activity of the NaYF 4 :(Yb,Er)/TiO 2, NaYF 4 :(Yb,Er)-CuO and NaYF 4 :(Yb,Er)-CuO/TiO 2 . Thus, the H 2 O 2 oxidant was photo-degraded to produce hydroxyl and super oxide ions which might degrade the MB-dye solution. The superior oxidizing power of H 2 O 2 molecules has been utilized in several research studies to enhance the photocatalytic activity of pure CuO and its composite materials [47]. Finally, the major advantage of coupling CuO and TiO 2 nanoparticles with phosphor is the inevitable alteration of the NaYF 4 :(Yb,Er)-CuO/TiO 2 composites band structure, as observed in both visible light absorption and band gap estimations in Figure 4. The NaYF 4 :(Yb,Er) can also facilitate the electron-hole charge separation process which enhances photocatalytic efficiencies [17]. Additionally, the phosphor crystalline surface not only supports the nanoparticles, but it also facilitates energy transfer within the composite system as described in Figure 5c. Accordingly, the composites can also be utilized efficiently in solar harvesting systems.
The synthesized samples of NaYF 4 :(Yb,Er)-CuO/TiO 2 were characterized for crystallinity by a X-ray diffraction (XRD, D8 Discover, Bruker AXS GmbH, Karlsruhe, Germany) with Cu Kα radiation. The morphology of the powder samples was analyzed by scanning electron microscope (SEM, Hitachi S-4300, Hitachi Ltd., Tokyo, Japan). The chemical composition of the composite sample was examined by energy dispersive spectroscopy (EDS, JEOL, JP/JSM-6330F, Kyoto, Japan). The particle size analysis of the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite was obtained by a particle size analyzer (Microtrac, S-3500, Largo, FL, USA). The specific surface area of the photocatalyst was determined by the Brunauer-Emmett-Teller (BET) method on an N 2 based chemisorption-physisorption micropore analyzer (3flex Micromeritics, Norcross, GA, USA). The photoluminescence (PL) spectra were obtained by a fluorescence spectrophotometer (F-4200, Hitachi, Tokyo, Japan) with an excitation slit of 5 nm and 10 nm emission slit. UV-vis diffuse reflectance spectra were observed using an UV-vis spectrophotometer (UV-1601, Shimadzu, Kyoto, Japan).
The photocatalytic activity was examined by monitoring the UV-vis absorbance of 2 ppm methylene blue (MB, 95.0%, Duksan Pure Chemical Co., Ltd., Ansansi, Kyunggido, Korea) solution under 200 W (clear, Ilkwang Co., Ltd., Dalseogu, Daegu, Korea) visible light irradiation with a 410 nm filter on top of beaker (UV L41 (W) 82 mm, Kenko Zeta, Tokina Co., Ltd., Tokyo, Japan). Prior to irradiation, photocatalyst powders weighing 100 mg were dispersed in a 100 mL beaker of MB solution. Then, the photo-reaction mixture was magnetically stirred for 30 min in darkness to achieve adsorption-desorption equilibria. For comparisons, photo-degradation efficiency was examined with 0.2 mL and without H 2 O 2 (28.0%, Duksan Pure Chemical Co., Ltd., Ansansi, Kyunggido, Korea) oxidant molecules. The degradation rate of MB solution mixed with photocatalyst composites was studied by sampling a small amount of solution photo-degraded at 15 min intervals for 1 h. Then, the photocatalyst-mixed solution was further exposed to visible light for 2 h to achieve a 3 h degradation cycle. After the first 3 h cycle, the photocatalyst-mixed solution was centrifugated and washed with ethanol in water to remove adsorbed dye. Then, the photocatalyst powders were dried in an electric oven for 100 • C for 2 h. The recyclability test was carried out for 2 more cycles with the collected NaYF 4 :(Yb,Er)-CuO/TiO 2 powders using the described photoactivity measurement. The photo-reaction beaker was placed in an ice-water bath to avoid temperature rise during the experiment. Photocatalyst concentration was maintained at 1 mg/mL of MB solution with 0.2 mL H 2 O 2 oxidant. The photo-degraded MB solution was centrifugated and examined for absorption by an UV-vis-NIR spectrophotometer (UV-3150, Shimadzu, Kyoto, Japan).

Conclusions
A visible light photo-active NaYF 4 :(Yb,Er)-CuO/TiO 2 composite has been synthesized. The increase in annealing temperature enhanced the crystallinity of the NaYF 4 :(Yb,Er)-CuO/TiO 2 composite. The combination of NaYF 4 :(Yb,Er), CuO and TiO 2 greatly enhances UV-vis optical properties and photocatalytic degradation efficiency for MB solution under visible light irradiation. The NaYF 4 :(Yb,Er)-CuO/TiO 2 photocatalyst exhibited some recyclability of up to 60% in the third cycle, which could be beneficial for possible application in photocatalyst systems.