Luminescence Behavior of GdVO 4 : Tb Nanocrystals in Silica Glass-Ceramics

: Glass ceramics with GdVO 4 : Tb nanocrystals impregnated in the highly transparent silica glass were prepared by the porous glass and sintering process and conﬁrmed by XRD, Raman spectrum, and TEM. Spectral analysis shows that there are multifarious energy transfer processes in GdVO 4 : Tb nanocrystals, such as VO 43 − → Tb 3 + , Gd 3 + → VO 43 − , Gd 3 + → VO 43 − → Tb 3 + , and Gd 3 + → Tb 3 + , and the main one is VO 43 − → Tb 3 + . In this process, 3 T 1,2 → 1 A 1 transition of VO 43 − transfers the energy to Tb 3 + and generates 5 D 3 → 7 F 6,5,4,3,2 and 5 D 4 → 7 F 5 transitions of Tb 3 + . The energy transfer e ﬃ ciency is 29.5%, and the excitation wavelength range of Tb 3 + ions can be shifted from 230–260 nm to 280–365 nm. This shows that GdVO 4 can e ﬀ ectively change the excitation wavelength of Tb 3 + , which is conducive to the application of Tb 3 + ions excited by LED light sources.


Introduction
In recent years, rare-earth ion doped phosphors have been widely studied due to their excellent luminescent properties. Among many rare earth elements, the 5 D 4 → 7 F 5 transition of Tb 3+ leads to a very strong green photoluminescence (PL) at around 545 nm. This kind of green fluorescence has strong potential applications in blue-green laser, undersea communication, and medical equipment [1,2]. However, due to the strictly forbidden 4f → 4f transitions, the optimal excitation wavelength of Tb 3+ is near 240 nm, which is too short for a commercial Light-Emitting Diode (LED) light source, thus directly affecting the application of Tb 3+ ions [3,4]. An effective way to change the excitation wavelength of Tb 3+ is to introduce a suitable sensitizer, which can transfer the energy absorbed from LED light source to Tb 3+ . A large number of reports have been focused on this aspect, mainly through Ce 3+ or Eu 2+ to sensitize Tb 3+ [1,5,6].
Rare earth doped gadolinium vanadate crystal (GdVO 4 : RE) is considered as an excellent luminescent material because of its excellent chemical stability, high thermal conductivity, strong energy absorption, and high energy transfer efficiency [7,8]. GdVO 4 : Tb can effectively improve the PL intensity and transfer excitation wavelength of Tb 3+ due to its efficient absorption of ultraviolet light and energy transfer (ET) from Gd 3+ and VO 4 3− to Tb 3+ , so as to expand the application of Tb 3+ luminescent materials [9,10]. There are many papers about the GdVO 4 : Tb, but most of them were nanocrystalline phosphors. If the nanocrystals can be synthesized into transparent glass, the application field of the luminescent nanocrystals will be further improved. Due to their composition having more than 97% SiO 2 and distributed interconnected nanopores, nano-porous silica glass (NPSG) has been researched. Lots of works have reported that NPSG can be used in many fields after doping and sintering due to its ability to effectively disperse rare-earth ions [11,12]. It is not difficult to imagine if luminescence nanocrystals can be grown in the connected nanopores, the size of the nanocrystals will be limited, and a transparent luminescence material could be obtained. In addition, luminescence nanocrystals can be effectively dispersed by interconnected nanopores in the glass, avoiding the clusters of luminescence ions and improving their luminescence properties. Simultaneously, as the protective shell of nanocrystals, silica glass can effectively improve the chemical, mechanical, and high temperature stability of luminescent materials.
The transparent glass ceramics doped with luminescent ions can effectively combine crystal and glass with many excellent properties, which makes it have many applications in the fields of luminescent, laser and nonlinear optics, and has been favored by many scholars in recent years [13][14][15][16]. Therefore, if we combine the excellent properties of GdVO 4 : Tb and silica glass together to synthesize a transparent glass ceramics, then this transparent luminescence glass-ceramics not only can be applicable to phosphor [17,18], but also to laser, such as laser medium materials [12,19], fiber core of microstructure fiber, and high-power fiber [20][21][22].
In general, the purpose of this research is to synthesize a transparent silica glass-ceramics (SGC) containing GdVO 4 : Tb nanocrystallites. The synthesized SGC contains the advantages of both silica glass and GdVO 4 : Tb crystal and is a green luminescent material. Moreover, we would comprehensively study the changes in the luminescent properties of Tb 3+ and the ET process of Gd 3+ and VO 4 3− to Tb 3+ in this SCG.

Synthesis of NPSG and SGC
The phase-separated alkali-borosilicate glass was made first. Then, the borate phase was removed by leaching with hot nitric acid solutions and NPSG was obtained. The obtained glass has a content of silica as high as 97% or more [23]. After that, NPSG were immersed in Gd 3+ , V 5+ , and Tb 3+ mixed-ion solutions (Gd(NO 3 ) 3 ·6H 2 O, VOSO 4 , and Tb(NO 3 ) 3 ·6H 2 O) for 1-2 h and dried at room temperature for 20-40 h. Subsequently, the porous glass embedded with active ions was sintered at 1050 • C and 1150 • C in an oxygen atmosphere [11]. Finally, SGC and uncrystallined glass were obtained after being ground and polished. Basic information of various samples with different ion concentration and crystalline statuses are listed in Table 1.

Instruments and Measurements
The X-ray diffraction (XRD) patterns of samples were carried out on a diffractometer with Cu Kα radiation (Smartlab9, Rigaku, Japan). Measurements were performed in the angular range 10 • ≤ 2θ ≤ 60 • . The Micro-Raman spectra were collected using a laser Raman microscope (Renishaw InVia, Wotton-under-Edge, UK) equipped with a 488-nm laser. The micromorphology of samples was characterized by transmission electron microscopy (TEM) (FEI Tecnai G2 F20, Thermo Fisher Scientific, Waltham, MA, USA). Transmittance spectra were recorded with a UV/visible/near-infrared spectrophotometer (Jasco V-570, Tokyo, Japan). The thickness of the samples is 1.5 mm. The photoluminescence excitation (PLE) spectra, photoluminescence (PL) spectra, and fluorescence lifetimes (FL) were measured at room temperature on a high-resolution spectrofluorometer (Photon Technology International (Canada) Inc. QM/TM/NIR, London, ON, Canada), utilizing a Xenon lamp as excitation source. Since the excitation and emission spectra are all arbitrary units, the experimental errors will not be discussed. The test errors of absorption spectra and lifetime are all within 5%. However, such errors do not affect the final research results. Figure 1a shows the XRD profiles of all samples. For the sample T, no diffraction peak was observed except for the typical amorphous broad scattering patterns, which can be ascribed to the silica glass substrate. In addition to the amorphous scattering patterns, the samples xTGV(x = 0.1-0.4) showed weak diffraction peaks at 2θ = 24.67 • , 33.23 • , and 49.15 • due to the (2 0 0), (1 1 2), and (3 1 2) reflections of tetragonal GdVO 4 . These peaks indicate that GdVO 4 nanocrystals were present in SGC.

Structure and Morphology
Crystals 2020, 10, x FOR PEER REVIEW 3 of 12 characterized by transmission electron microscopy (TEM) (FEI Tecnai G2 F20, Thermo Fisher Scientific, Waltham, MA, USA). Transmittance spectra were recorded with a UV/visible/near-infrared spectrophotometer (Jasco V-570, Tokyo, Japan). The thickness of the samples is 1.5 mm. The photoluminescence excitation (PLE) spectra, photoluminescence (PL) spectra, and fluorescence lifetimes (FL) were measured at room temperature on a high-resolution spectrofluorometer (Photon Technology International (Canada) Inc. QM/TM/NIR, London, ON, Canada), utilizing a Xenon lamp as excitation source. Since the excitation and emission spectra are all arbitrary units, the experimental errors will not be discussed. The test errors of absorption spectra and lifetime are all within 5%. However, such errors do not affect the final research results.

Structure and Morphology
Figure 1Error! Reference source not found.a shows the XRD profiles of all samples. For the sample T, no diffraction peak was observed except for the typical amorphous broad scattering patterns, which can be ascribed to the silica glass substrate. In addition to the amorphous scattering patterns, the samples xTGV(x = 0.1-0.4) showed weak diffraction peaks at 2θ = 24.67°, 33.23°, and 49.15° due to the (2 0 0), (1 1 2), and (3 1 2) reflections of tetragonal GdVO4. These peaks indicate that GdVO4 nanocrystals were present in SGC. Micro-Raman spectroscopes of samples and GdVO4 crystal were also tested and shown in Figure  1b. Several bands near 441, 495, 600, 800, and 1060 cm −1 can be seen in each sample, which can be assigned to the vibrational mode of SiO2 glass [24][25][26][27][28]. In addition, some vibrations were detected in the Raman spectra of xTGV(x = 0.1-0.4), which were located at 263, 380, 485, 820, and 880 cm −1 . These peaks due to the modes of (VO4) 3− in GdVO4 nanocrystals [29][30][31][32]. The observed peaks were detailed analysis in our early work [11]. Obviously, the existence of GdVO 4 : Tb was also confirmed by the micro-Raman spectra.
TEM and HR-TEM of sample 0.3TGV were investigated to exhibit the morphology of the grains GdVO4: Tb and its distribution in SGC. TEM micrograph in Figure 2a indicated that nanoparticles were uniformly distributed in SCG, and the diameters were less than 50 nm. Meanwhile, the HR-TEM micrograph in Figure2b,c shows that a more obvious lattice fringe existed in the sample. The lattice fringes spaced by 0.36 nm and 0.29 nm coincide with the (2 0 0) and (2 1 1) planes of tetragonal GdVO4, respectively [11,33]. The results of the TEM and HR-TEM pictures prove the existence of GdVO4: Tb nanocrystals in SGC further. Micro-Raman spectroscopes of samples and GdVO 4 crystal were also tested and shown in Figure 1b. Several bands near 441, 495, 600, 800, and 1060 cm −1 can be seen in each sample, which can be assigned to the vibrational mode of SiO 2 glass [24][25][26][27][28]. In addition, some vibrations were detected in the Raman spectra of xTGV(x = 0.1-0.4), which were located at 263, 380, 485, 820, and 880 cm −1 . These peaks due to the modes of (VO 4 ) 3− in GdVO 4 nanocrystals [29][30][31][32]. The observed peaks were detailed analysis in our early work [11]. Obviously, the existence of GdVO 4 : Tb was also confirmed by the micro-Raman spectra.
TEM and HR-TEM of sample 0.3TGV were investigated to exhibit the morphology of the grains GdVO 4 : Tb and its distribution in SGC. TEM micrograph in Figure 2a indicated that nanoparticles were uniformly distributed in SCG, and the diameters were less than 50 nm. Meanwhile, the HR-TEM micrograph in Figure 2b,c shows that a more obvious lattice fringe existed in the sample. The lattice fringes spaced by 0.36 nm and 0.29 nm coincide with the (2 0 0) and (2 1 1) planes of tetragonal GdVO 4 , respectively [11,33]. The results of the TEM and HR-TEM pictures prove the existence of GdVO 4 : Tb nanocrystals in SGC further.  Figure 3 shows the transmittance spectra of xTGV (x = 0.1-0.4). It can be seen that the transmittance of the samples gradually decreased with the increase of the Tb 3+ content, and the absorption edges shifted to long wavelengths, which implies that the increase of Tb 3+ promotes the growth of microcrystalline in high silica glass. The results of some optical scattering and decreasing transmission is owing to the different refractive index between the particles and matrix glass. Another possibility of the red shifting is from nephelauxetic effect. Since the atomic number of Tb 3+ is greater than Gd 3+ , and the outer layer of Tb 3+ has a higher electron density, it causes a decrease in the energy of upper level VO4 3− through the electron cloud expansion. Therefore, the higher the concentration of Tb 3+ , the higher the electron density in the whole glass, which increased the covalency around Tb 3+ ions and led to the red shift of the absorption edge [34,35].   Figure 3 shows the transmittance spectra of xTGV (x = 0.1-0.4). It can be seen that the transmittance of the samples gradually decreased with the increase of the Tb 3+ content, and the absorption edges shifted to long wavelengths, which implies that the increase of Tb 3+ promotes the growth of microcrystalline in high silica glass. The results of some optical scattering and decreasing transmission is owing to the different refractive index between the particles and matrix glass. Another possibility of the red shifting is from nephelauxetic effect. Since the atomic number of Tb 3+ is greater than Gd 3+ , and the outer layer of Tb 3+ has a higher electron density, it causes a decrease in the energy of upper level VO 4 3− through the electron cloud expansion. Therefore, the higher the concentration of Tb 3+ , the higher the electron density in the whole glass, which increased the covalency around Tb 3+ ions and led to the red shift of the absorption edge [34,35].  Figure 3 shows the transmittance spectra of xTGV (x = 0.1-0.4). It can be seen that the transmittance of the samples gradually decreased with the increase of the Tb 3+ content, and the absorption edges shifted to long wavelengths, which implies that the increase of Tb 3+ promotes the growth of microcrystalline in high silica glass. The results of some optical scattering and decreasing transmission is owing to the different refractive index between the particles and matrix glass. Another possibility of the red shifting is from nephelauxetic effect. Since the atomic number of Tb 3+ is greater than Gd 3+ , and the outer layer of Tb 3+ has a higher electron density, it causes a decrease in the energy of upper level VO4 3− through the electron cloud expansion. Therefore, the higher the concentration of Tb 3+ , the higher the electron density in the whole glass, which increased the covalency around Tb 3+ ions and led to the red shift of the absorption edge [34,35].

PLE/PL Properties
With a fixed concentration of Gd 3+ and V 5+ as 1 × 10 20 ion/cm 3 , samples with different contents of Tb 3+ were fabricated, the PLE spectra and PL spectra were measured and shown in Figure 4. It can be seen that the peak shapes of the PLE and PL spectra for these samples were similar. The main difference was changes in PL intensity. The PL intensity first increases, then starts to decrease with the increased contents of Tb 3+ , and reaches the strongest when the doping concentration is 0.3 × 10 20 ion/cm 3 , which is due to the fluorescence quenching due to the high concentration of Tb 3+ . Therefore, the concentration of Tb 3+ was fixed at 0.3 × 10 20 ion/cm 3 in the above and subsequent studies. With a fixed concentration of Gd 3+ and V 5+ as 1 × 10 20 ion/cm 3 , samples with different contents of Tb 3+ were fabricated, the PLE spectra and PL spectra were measured and shown in Figure 4. It can be seen that the peak shapes of the PLE and PL spectra for these samples were similar. The main difference was changes in PL intensity. The PL intensity first increases, then starts to decrease with the increased contents of Tb 3+ , and reaches the strongest when the doping concentration is 0.3 × 10 20 ion/cm 3 , which is due to the fluorescence quenching due to the high concentration of Tb 3+ . Therefore, the concentration of Tb 3+ was fixed at 0.3 × 10 20 ion/cm 3 in the above and subsequent studies. The PLE spectra of the 5 D4 → 7 F5 transition of Tb 3+ monitored at 542 nm are shown in Figure 4a. It can be seen that the main excitation peak of all samples was located near 330 nm, which reflects an energy transfer process between VO4 3− and Tb 3+ because the excitation peak is mainly caused by the absorption band of VO4 3− [36][37][38] in contrast to the PLE spectrum of single Tb 3+ doped sample T, which will be discussed in the following sections. In addition to the absorption band at 330 nm, a weak absorption peak near 375 nm also can be seen, which can be ascribed to the absorption of the Tb 3+ ion 7 F6 → 5 D3 transition [39,40]. Another change is slight red-shifting of the excitation peak with the concentration increase of Tb 3+ . The reason for this may be the electron cloud density increased with the increase content of Tb 3+ , which led to a decrease in the energy of the upper level of VO4 3− group [36]. The PL spectra of the 5 D4 → 7 F5 transition of Tb 3+ excited at 330 nm are shown in Figure 4b. The emission of Gd 3+ 6 P7/2-8 S7/2 transitions occurred at 310 nm under 274 nm excited, and therefore does not appear in Figure 4b. Similar to the PLE spectra, the intensity of the emission peak also increased first and then decreased with the increase of Tb 3+ concentration, which is also caused by the fluorescence quenching. The broadband emission at 400-650 nm was caused by the 3 T1,2 → 1 A1 transition of VO4 3− in each emission spectrum [11]. In addition to this emission band, the PL peaks at 425, 450, 467, and 542 nm were caused by the 5 D3 → 7 F4,3,2 and 5 D4 → 7 F5 transitions of Tb 3+ ions [39,41]. This further illustrated that ET existed in these samples, so the ET process will be discussed in the subsequent chapter.

Energy Transfer Mechanism
As mentioned, ET between VO4 3− and Tb 3+ was observed in previous results. In order to clarify the ET processes between various active ions in GdVO4: Tb nanocrystals embedded in the synthesized SGC, the samples GV and T shown in Table 1 were fabricated, and PLE and PL spectra were measured, as displayed in Figure 5. The PLE spectra of the 5 D 4 → 7 F 5 transition of Tb 3+ monitored at 542 nm are shown in Figure 4a. It can be seen that the main excitation peak of all samples was located near 330 nm, which reflects an energy transfer process between VO 4 3− and Tb 3+ because the excitation peak is mainly caused by the absorption band of VO 4 3− [36][37][38] in contrast to the PLE spectrum of single Tb 3+ doped sample T, which will be discussed in the following sections. In addition to the absorption band at 330 nm, a weak absorption peak near 375 nm also can be seen, which can be ascribed to the absorption of the Tb 3+ ion 7 F 6 → 5 D 3 transition [39,40]. Another change is slight red-shifting of the excitation peak with the concentration increase of Tb 3+ . The reason for this may be the electron cloud density increased with the increase content of Tb 3+ , which led to a decrease in the energy of the upper level of VO 4 3− group [36]. The PL spectra of the 5 D 4 → 7 F 5 transition of Tb 3+ excited at 330 nm are shown in Figure 4b. The emission of Gd 3+6 P 7/2 -8 S 7/2 transitions occurred at 310 nm under 274 nm excited, and therefore does not appear in Figure 4b. Similar to the PLE spectra, the intensity of the emission peak also increased first and then decreased with the increase of Tb 3+ concentration, which is also caused by the fluorescence quenching. The broadband emission at 400-650 nm was caused by the 3 T 1,2 → 1 A 1 transition of VO 4 3− in each emission spectrum [11]. In addition to this emission band, the PL peaks at 425, 450, 467, and 542 nm were caused by the 5 D 3 → 7 F 4,3,2 and 5 D 4 → 7 F 5 transitions of Tb 3+ ions [39,41]. This further illustrated that ET existed in these samples, so the ET process will be discussed in the subsequent chapter.

Energy Transfer Mechanism
As mentioned, ET between VO 4 3− and Tb 3+ was observed in previous results. In order to clarify the ET processes between various active ions in GdVO 4 : Tb nanocrystals embedded in the synthesized SGC, the samples GV and T shown in Table 1 were fabricated, and PLE and PL spectra were measured, as displayed in Figure 5. Firstly, the PLE spectra of samples T and 0.3TGV of the 5 D4 → 7 F5 transition of Tb 3+ monitored at 542 nm were measured. For comparing, the PLE spectrum of sample GV monitored at 520 nm was also shown in Figure 5a. It can be seen that sample T also have some f →f transition absorptions in the range of 275-385 nm, including 7 F6 → 5 H7， 7 F6 → 5 D2, and 7 F6 → 5 D3 besides the main absorption band of 4f 8 → 4f 7 5d 1 transition of Tb 3+ ion near 244 nm [42]. However, they are very weak. On the other hand, sample GV also shows strong absorption peaks of 1 A1 → 1 T1,2 transitions of VO4 3− in the range of 225-375 nm [36][37][38]. Moreover, besides the absorption peaks of VO4 3− , a weak peak of 7 F6 → 5 D3 transition of Tb 3+ ion near 380 nm could also be seen in sample 0.3TGV. It is easy to see by comparing the samples T and 0.3TGV that the most obvious change was the FWHM range of excitation wavelength of Tb 3+ shifting from about 230-260 nm to 280-365 nm, and the strongest excitation peak shifting from around 244 to 330 nm. At the same time, the PL intensity of Tb 3+ ions also increased. These results proved the existed ET process of VO4 3− → Tb 3+ in SGC. Comparing the sample GV and 0.3TGV, it can be seen that there were two changes in the PLE spectra. One such change was a weak new absorption peak, which is resulting from 7 F6 → 5 D2 transition of Tb 3+ ions, while the other was the position red-shifting of the optimal excitation wavelength, which may be due to the absorption red shifting of matrix after microcrystallization. In addition, it can be observed that the emission from 3 T1,2 → 1 A1 transition of VO4 3− was strong in Figures 4b and 5b, implying that there was low energy transfer efficiency from VO4 3− to Tb 3+ . Thus, the disappearance of 4f 8 → 4f 7 5d 1 transition of Tb 3+ ion near 244 nm for 0.3TGV may be due to the 5d 1 energy level causing a strong split after the Tb 3+ entered the Gd 3+ lattice, which resulted in a large red-shift of the 4f 8 → 4f 7 5d 1 transition and overlapped with the 1 A1 → 1 T1,2 transitions of VO4 3− . Figure 5b shows the PL spectra of samples GV, T, and 0.3TGV, and their respective characteristic excitation wavelengths were marked in the Figure 5b. The different between the PL spectra of GV and 0.3GVT were mainly two points. One was that the luminescence peak resulted from the 3 T1,2 → 1 A1 transition of VO4 3− at 400-650 nm was widened after adding Tb 3+ ions, while the other was the emergence of some new emission peaks near 380, 425, 450, 467, and 542 nm, which can be ascribed to the 5 D3 → 7 F6,5,4,3,2 and 5 D4 → 7 F5 transitions of Tb 3+ ions [39,41]. This shows again that partial energy of the 3 T1,2 → 1 A1 transition transferred to the Tb 3+ ion. In order to illustrate the above process, a schematic of the ET process in SGC is suggested in Figure 6, and the process of VO4 3− → Tb 3+ is ET2. Firstly, the PLE spectra of samples T and 0.3TGV of the 5 D 4 → 7 F 5 transition of Tb 3+ monitored at 542 nm were measured. For comparing, the PLE spectrum of sample GV monitored at 520 nm was also shown in Figure 5a. It can be seen that sample T also have some f →f transition absorptions in the range of 275-385 nm, including 7 F 6 → 5 H 7 , 7 F 6 → 5 D 2 , and 7 F 6 → 5 D 3 besides the main absorption band of 4f 8 → 4f 7 5d 1 transition of Tb 3+ ion near 244 nm [42]. However, they are very weak. On the other hand, sample GV also shows strong absorption peaks of 1 A 1 → 1 T 1,2 transitions of VO 4 3− in the range of 225-375 nm [36][37][38]. Moreover, besides the absorption peaks of VO 4 3− , a weak peak of 7 F 6 → 5 D 3 transition of Tb 3+ ion near 380 nm could also be seen in sample 0.3TGV. It is easy to see by comparing the samples T and 0.3TGV that the most obvious change was the FWHM range of excitation wavelength of Tb 3+ shifting from about 230-260 nm to 280-365 nm, and the strongest excitation peak shifting from around 244 to 330 nm. At the same time, the PL intensity of Tb 3+ ions also increased. These results proved the existed ET process of VO 4 3− → Tb 3+ in SGC. Comparing the sample GV and 0.3TGV, it can be seen that there were two changes in the PLE spectra. One such change was a weak new absorption peak, which is resulting from 7 F 6 → 5 D 2 transition of Tb 3+ ions, while the other was the position red-shifting of the optimal excitation wavelength, which may be due to the absorption red shifting of matrix after microcrystallization. In addition, it can be observed that the emission from 3 T 1,2 → 1 A 1 transition of VO 4 3− was strong in Figures 4b and 5b, implying that there was low energy transfer efficiency from VO 4 3− to Tb 3+ . Thus, the disappearance of 4f 8 → 4f 7 5d 1 transition of Tb 3+ ion near 244 nm for 0.3TGV may be due to the 5d 1 energy level causing a strong split after the Tb 3+ entered the Gd 3+ lattice, which resulted in a large red-shift of the 4f 8 → 4f 7 5d 1 transition and overlapped with the 1 A 1 → 1 T 1,2 transitions of VO 4 3− . Figure 5b shows the PL spectra of samples GV, T, and 0.3TGV, and their respective characteristic excitation wavelengths were marked in the Figure 5b. The different between the PL spectra of GV and 0.3GVT were mainly two points. One was that the luminescence peak resulted from the 3 T 1,2 → 1 A 1 transition of VO 4 3− at 400-650 nm was widened after adding Tb 3+ ions, while the other was the emergence of some new emission peaks near 380, 425, 450, 467, and 542 nm, which can be ascribed to the 5 D 3 → 7 F 6,5,4,3,2 and 5 D 4 → 7 F 5 transitions of Tb 3+ ions [39,41]. This shows again that partial energy of the 3 T 1,2 → 1 A 1 transition transferred to the Tb 3+ ion. In order to illustrate the above process, a schematic of the ET process in SGC is suggested in Figure 6, and the process of VO 4 3− → Tb 3+ is ET 2 .
Crystals 2020, 10, 396 7 of 12 Crystals 2020, 10, x FOR PEER REVIEW 7 of 12 In addition to the emission of Tb 3+ , there were also emission of Gd 3+ and VO4 3− in samples; and there were some overlap between the position of the emission peak of Tb 3+ and the excitation wavelength of VO4 3− as shown in Figure 5. Therefore, in order to verify whether there was a reverse ET from Tb 3+ to VO4 3− in the SGC, the sample 0.3TV was made and the PLE spectra of the samples GV, 0.3TV, and 0.3TGV monitored at 520 nm were measured and shown in Figure 7a. It can be seen that there were weak absorption peaks at 230-270 nm besides peaks of 1 A1 → 1 T1,2 transitions of VO4 3− in both samples of 0.3TV and 0.3TGV. This may be evidence of the reverse ET process of Tb 3+ → VO4 3− in the SGC, but the transfer efficiency is low. This process is shown as ET3 in Figure 6. In addition, it can be clearly seen in the Figure 7a that the excitation peaks of both samples of 0.3TV and 0.3TGV shifted to long wavelength compared to GV. As mentioned earlier, this may be due to the changes of energy levels of VO4 3− group, which caused by the increased electron cloud density. The emission of Tb 3+ in the sample T has been mentioned above, and thus is not repeated here. As can be seen, all samples behaved weak emission bands of 3 T1,2 → 1 A1 transition of VO4 3− except for sample T. There are two possibilities for the phenomenon. One may be that VO4 3− has some absorptive capacity for the ultraviolet at about 244 nm, which results in the emission of VO4 3− .The other may be resulting from the reverse ET process of Tb 3+ → VO4 3− . Namely, Tb 3+ ions transferred the In addition to the emission of Tb 3+ , there were also emission of Gd 3+ and VO 4 3− in samples; and there were some overlap between the position of the emission peak of Tb 3+ and the excitation wavelength of VO 4 3− as shown in Figure 5. Therefore, in order to verify whether there was a reverse ET from Tb 3+ to VO 4 3− in the SGC, the sample 0.3TV was made and the PLE spectra of the samples GV, 0.3TV, and 0.3TGV monitored at 520 nm were measured and shown in Figure 7a. It can be seen that there were weak absorption peaks at 230-270 nm besides peaks of 1 A 1 → 1 T 1,2 transitions of VO 4 3− in both samples of 0.3TV and 0.3TGV. This may be evidence of the reverse ET process of Tb 3+ → VO 4 3− in the SGC, but the transfer efficiency is low. This process is shown as ET 3 in Figure 6. In addition, it can be clearly seen in the Figure 7a that the excitation peaks of both samples of 0.3TV and 0.3TGV shifted to long wavelength compared to GV. As mentioned earlier, this may be due to the changes of energy levels of VO 4 3− group, which caused by the increased electron cloud density. In addition to the emission of Tb 3+ , there were also emission of Gd 3+ and VO4 3− in samples; and there were some overlap between the position of the emission peak of Tb 3+ and the excitation wavelength of VO4 3− as shown in Figure 5. Therefore, in order to verify whether there was a reverse ET from Tb 3+ to VO4 3− in the SGC, the sample 0.3TV was made and the PLE spectra of the samples GV, 0.3TV, and 0.3TGV monitored at 520 nm were measured and shown in Figure 7a. It can be seen that there were weak absorption peaks at 230-270 nm besides peaks of 1 A1 → 1 T1,2 transitions of VO4 3− in both samples of 0.3TV and 0.3TGV. This may be evidence of the reverse ET process of Tb 3+ → VO4 3− in the SGC, but the transfer efficiency is low. This process is shown as ET3 in Figure 6. In addition, it can be clearly seen in the Figure 7a that the excitation peaks of both samples of 0.3TV and 0.3TGV shifted to long wavelength compared to GV. As mentioned earlier, this may be due to the changes of energy levels of VO4 3− group, which caused by the increased electron cloud density. The emission of Tb 3+ in the sample T has been mentioned above, and thus is not repeated here. As can be seen, all samples behaved weak emission bands of 3 T1,2 → 1 A1 transition of VO4 3− except for sample T. There are two possibilities for the phenomenon. One may be that VO4 3− has some absorptive capacity for the ultraviolet at about 244 nm, which results in the emission of VO4 3− .The other may be resulting from the reverse ET process of Tb 3+ → VO4 3− . Namely, Tb 3+ ions transferred the  Figure 7b. The emission of Tb 3+ in the sample T has been mentioned above, and thus is not repeated here. As can be seen, all samples behaved weak emission bands of 3 T 1,2 → 1 A 1 transition of VO 4 3− except for sample T. There are two possibilities for the phenomenon. One may be that VO 4 3− has some absorptive capacity for the ultraviolet at about 244 nm, which results in the emission of VO 4 3− .The other may be resulting from the reverse ET process transferred back to Tb 3+ . In these processes, a part of the energy generated 5 D 4 → 7 F 5 transition of Tb 3+ ions, and another part of the energy was lost in the form of heat, which in turn reduces the PL intensity of VO 4 3− in the sample 0.3TV and 0.3TGV. On the other hand, as mentioned above, the of 4f 8 → 4f 7 5d 1 transition would be a red shift in a strong crystal field, so it resulted in a weak emission of Tb 3+ under the excitation wavelength of 244 nm as shown in Figure 7b. In our previous report [11], it was found that there were various energy transfer processes in GdVO 4 : Eu nanocrystalline silica glass, including Gd 3+ → Eu 3+ , VO 4 3− → Eu 3+ , Gd 3+ → VO 4 3− , and Gd 3+ → VO 4 3− → Eu 3+ . In this article, the Tb 3+ is also in the GdVO 4 matrix. Therefore, based on the above discussion and previous report, it can be considered that there are also multiple ET processes of VO 4 3− → Tb 3+ , Gd 3+ → VO 4 3− , and Gd 3+ → VO 4 3− → Tb 3+ in the SGC, that is, the ET 2 and ET 1 processes in Figure 6. Is there an energy transfer process of Gd 3+ → Tb 3+ in the SGC? In order to verify the speculation, the PLE spectra of samples T and 0.3TGV were measured with a monitored wavelength of 542 nm, as shown in Figure 8a. It can be seen that there is a weak shoulder near 260-280 nm of sample 0.3TGV, which may be due to the 8 S 7/2 → 6 I J transition of Gd 3+ [11] and it may make the intensity of this region enhanced [37,43]. However, the 4f 8 → 4f 7 5d 1 transition of Tb 3+ is also absorbed nearby as shown in Figure 8a, and the shoulder may also be caused by this absorption.
Crystals 2020, 10, x FOR PEER REVIEW 8 of 12 absorbing energy to VO4 3− , and then VO4 3− transferred back to Tb 3+ . In these processes, a part of the energy generated 5 D4 → 7 F5 transition of Tb 3+ ions, and another part of the energy was lost in the form of heat, which in turn reduces the PL intensity of VO4 3− in the sample 0.3TV and 0.3TGV. On the other hand, as mentioned above, the of 4f 8 → 4f 7 5d 1 transition would be a red shift in a strong crystal field, so it resulted in a weak emission of Tb 3+ under the excitation wavelength of 244 nm as shown in Figure 7b.
Is there an energy transfer process of Gd 3+ → Tb 3+ in the SGC? In order to verify the speculation, the PLE spectra of samples T and 0.3TGV were measured with a monitored wavelength of 542 nm, as shown in Figure 8a. It can be seen that there is a weak shoulder near 260-280 nm of sample 0.3TGV, which may be due to the 8 S7/2 → 6 IJ transition of Gd 3+ [11] and it may make the intensity of this region enhanced [37,43]. However, the 4f 8 → 4f 7 5d 1 transition of Tb 3+ is also absorbed nearby as shown in Figure 8a, and the shoulder may also be caused by this absorption. In order to clarify the role of Gd 3+ , sample G single doped with Gd 3+ was prepared, and the PL spectra of G, GV, and 0.3TGV were tested with the excitation wavelength of 274 nm. The results are shown in Figure 8b. The peak at about 310 nm was reduced to 1/5 of the original intensity in sample G due to the strongly actual intensity (Iactual) of the 6 PJ → 8 S7/2 transition of Gd 3+ . Compared with sample G, the 6 PJ → 8 S7/2 transition emission intensity of Gd 3+ near 310 nm in GV and 0.3TGV decreased sharply, which was replaced by the 3 T1,2 → 1 A1 transition emission peak of VO4 3− . This indicates that Gd 3+ mainly transfers energy to VO4 3− . However, a weak luminescence peak of the 5 D4 → 7 F5 transition of Tb 3+ ions can be seen near 542 nm in the sample 0.3 TGV, which also indicates that Gd 3+ may transfer some energy to Tb 3+ and includes the processes of Gd 3+ → VO4 3− → Tb 3+ . In general, the both energy transfer processes of Gd 3+ → VO4 3− → Tb 3+ and Gd 3+ → Tb 3+ can be seen in the SGC, which are shown as ET1, ET2, and ET4 in Figure 6. However, there is no way to obtain each weight of both processes yet.
In the above analysis and discussion, the ET processes of VO4 3− → Tb 3+ ，Gd 3+ → VO4 3− ，Gd 3+ → VO4 3− → Tb 3+ and Gd 3+ → Tb 3+ have been proven, and VO4 3− → Tb 3+ was still the main process. The decay time (τ) of Gd 3+ in samples GV and (0.1-0.4) TGV were measured，they are basically the same, about 600 μs. The decay time of Tb 3+ in samples (0.1-0.4) TGV were also measured，they are basically the same, about 290 μs. Therefore, only the decay time of VO4 3− and transfer efficiency (η) of VO4 3− → Tb 3+ were measured and discussed. Figure 9 depicts the decay curves of the 3 T1,2 → 1 A1 transition of VO4 3− in GV and 0.3TGV. In order to clarify the role of Gd 3+ , sample G single doped with Gd 3+ was prepared, and the PL spectra of G, GV, and 0.3TGV were tested with the excitation wavelength of 274 nm. The results are shown in Figure 8b. The peak at about 310 nm was reduced to 1/5 of the original intensity in sample G due to the strongly actual intensity (I actual ) of the 6 P J → 8 S 7/2 transition of Gd 3+ . Compared with sample G, the 6 P J → 8 S 7/2 transition emission intensity of Gd 3+ near 310 nm in GV and 0.3TGV decreased sharply, which was replaced by the 3 T 1,2 → 1 A 1 transition emission peak of VO 4 3− . This indicates that Gd 3+ mainly transfers energy to VO 4 3− . However, a weak luminescence peak of the 5 D 4 → 7 F 5 transition of Tb 3+ ions can be seen near 542 nm in the sample 0.3 TGV, which also indicates that Gd 3+ may transfer some energy to Tb 3+ and includes the processes of Gd 3+ → VO 4 3− → Tb 3+ . In general, the both energy transfer processes of Gd 3+ → VO 4 3− → Tb 3+ and Gd 3+ → Tb 3+ can be seen in the SGC, which are shown as ET 1 , ET 2 , and ET 4 in Figure 6. However, there is no way to obtain each weight of both processes yet.
In the above analysis and discussion, the ET processes of VO 4   Both of the decay curves followed a double-exponential decay. The equation , can be used to calculated the average decay times, where τ1 is the short decay component and τ2 is the long-decay component, respectively, while parameters A1 and A2 are the fitting constants [33]. The equation η 1 τ/τ can be used to estimate the efficiency η of the ET, where τ0 is the fluorescence lifetime before doping and τ is the fluorescence lifetime after doping. As can be seen, the average lifetime of the VO4 3− → Tb 3+ transition decreases from 154.6 μs to 108.9 μs with the addition of Tb 3+ , which indicates that energy of the 3 T1,2 → 1 A1 transitions transfers from VO4 3− to Tb 3+ ions, and the transfer efficiency was 29.6%. The transfer efficiency was not high, and it was consistent with the results in Figures 4b and 5b.

Conclusions
The glass-ceramics with GdVO4: Tb nanocrystals embedded in the highly transparent silica glass were prepared by the porous glass and sintering process. The existence of nanocrystals was confirmed by XRD, Raman spectrum, and TEM. The fluorescence performance of Tb 3+ in the transparent SGCs with different concentrations Tb 3+ were studied by PLE and PL spectra, and the best performance of 0.3TGV was obtained. In order to clarify the energy transfer mechanism, various samples with single, double, and triple active ions in SGCs were prepared and their spectra were analyzed. The results showed that there was an ET process of VO4 3− → Tb 3+ in the SGC after the formation of nanocrystals by the comparison of the spectra of GV, T, and 0.3TGV. Namely, partial energy of the 3 T1,2 → 1 A1 transition of VO4 3− transferred to the Tb 3+ ion and generated the emission of 5 D3→ 7 F6,5,4,3,2 and 5 D4→ 7 F5. Besides the VO4 3− → Tb 3+ , processes of Gd 3+ → VO4 3− , Gd 3+ → VO4 3− → Tb 3+ , and Gd 3+ → Tb 3+ have been confirmed in the SGC through spectral comparison of samples, but the main process is VO4 3− → Tb 3+ . Finally, the transfer efficiency of the ET of VO4 3− → Tb 3+ was obtained by measuring and calculating the decay time of 3 T1,2 → 1 A1 transition of VO4 3− ion before and after Tb 3+ doping in the SGC, which was 29.6%.
To sum up, these ET processes in GdVO4: Tb nanocrystals enhanced the luminescence intensity of Tb 3+ in the SGCs and shifted the FWHM range of excitation wavelength of Tb 3+ from 230-260 nm to 280-365 nm, which is conducive to the application of Tb 3+ ions excited by LED light sources. Consequently, this new SGC can be used in the fields of LED excitation luminescence, displays, and lasers fiber that require high transparency, high power, and high chemical stability.  Both of the decay curves followed a double-exponential decay. The equation τ = A 1 τ 1 2 + A 2 τ 2 2 /(A 1 τ 1 + A 2 τ 2 ), can be used to calculated the average decay times, where τ 1 is the short decay component and τ 2 is the long-decay component, respectively, while parameters A 1 and A 2 are the fitting constants [33]. The equation η = 1 − τ/τ 0 can be used to estimate the efficiency η of the ET, where τ 0 is the fluorescence lifetime before doping and τ is the fluorescence lifetime after doping. As can be seen, the average lifetime of the VO 4 3− → Tb 3+ transition decreases from 154.6 µs to 108.9 µs with the addition of Tb 3+ , which indicates that energy of the 3 T 1,2 → 1 A 1 transitions transfers from VO 4 3− to Tb 3+ ions, and the transfer efficiency was 29.6%. The transfer efficiency was not high, and it was consistent with the results in Figures 4b and 5b.

Conclusions
The glass-ceramics with GdVO 4 : Tb nanocrystals embedded in the highly transparent silica glass were prepared by the porous glass and sintering process. The existence of nanocrystals was confirmed by XRD, Raman spectrum, and TEM. The fluorescence performance of Tb 3+ in the transparent SGCs with different concentrations Tb 3+ were studied by PLE and PL spectra, and the best performance of 0.3TGV was obtained. In order to clarify the energy transfer mechanism, various samples with single, double, and triple active ions in SGCs were prepared and their spectra were analyzed. The results showed that there was an ET process of VO 4 3− → Tb 3+ in the SGC after the formation of nanocrystals by the comparison of the spectra of GV, T, and 0.3TGV. Namely, partial energy of the 3 T 1,2 → 1 A 1 transition of VO 4 3− transferred to the Tb 3+ ion and generated the emission of 5 D 3 → 7 F 6,5,4,3,2 and 5 D 4 → 7 F 5 . Besides the VO 4 3− → Tb 3+ , processes of Gd 3+ → VO 4 3− , Gd 3+ → VO 4 3− →Tb 3+ , and Gd 3+ → Tb 3+ have been confirmed in the SGC through spectral comparison of samples, but the main process is VO 4 3− → Tb 3+ . Finally, the transfer efficiency of the ET of VO 4 3− → Tb 3+ was obtained by measuring and calculating the decay time of 3 T 1,2 → 1 A 1 transition of VO 4 3− ion before and after Tb 3+ doping in the SGC, which was 29.6%. To sum up, these ET processes in GdVO 4 : Tb nanocrystals enhanced the luminescence intensity of Tb 3+ in the SGCs and shifted the FWHM range of excitation wavelength of Tb 3+ from 230-260 nm to 280-365 nm, which is conducive to the application of Tb 3+ ions excited by LED light sources.