Bi2+-doped MBPO5 (M = Ba2+, Sr2+, Ca2+), synthesized in air via solid state reaction, are considered as novel orange and red phosphors for white light emitting diodes with improved colour quality. Absorption of Bi2+ due to 2P1/2→2S1/2 and 2P1/2→2P3/2 could be observed and quantified. Excitation to 2P3/2 is accompanied by vibronic sidebands, and corresponding emission behaviour is found. The electron-phonon coupling strength increases in the order M = Ba2+→Sr2+→Ca2+. In the case of MBPO5:Bi2+, one-, two- and even three-phonon sidebands could clearly be observed. The crystal structure of all three compounds belongs to space group P3121. Bi2+ is incorporated on M2+ sites, and reduction of Bi3+ to Bi2+ occurs for reasons of charge compensation. In accordance with crystallographic data, fluorescence decay behaviour indicates that only one type of Bi2+-emission centers is present.
© 2009 OSA
Due to the large number of possible valence states, strong interaction with the surrounding lattice and, in some cases, cluster formation, as dopant in inorganic matrices, Bismuth exhibits a broad variety of optoelectronic properties and potential applications [1–9]. Contrary especially to Bi3+ and Bi0, however, research on Bi2+-doped materials is still rare [10–14]. Because of the relatively high oxidation tendency, only a handful of matrix materials are know in which efficient stabilization of the Bi2+ ion is possible. On the other side, its electronic band structure  clearly suggests application in alternative red- or orange emitting phosphors that can be excited by blue or UV light emitting diodes (LEDs). Consequently, Bi2+-doped strontium borates have been considered recently for this application .
In the late 1990s, MBPO5 (M = Ca2+, Sr2+ and Ba2+) was proposed by Srivastrava as an alternative host for Bi2+ . Spectroscopic examination at low temperature revealed the known excitation bands of 2P1/2→2S1/2 (UV), 2P1/2→2P3/2 (2) (blue) and 2P1/2→2P3/2 (1) (orange) and emission in the red for BaBPO5:Bi2+, whereas 2P1/2→2S1/2 could not be seen in CaBPO5: Bi2+ and SrBPO5: Bi2+ .
According to Srivastava’s report, heat treatment at reducing atmosphere was necessary to obtain red emission from Bi2+ centers in MBPO5. On the other hand, more recent observations from Liang and Su suggest differently [15,16]: they have shown what was later described as an abnormal reduction process of Sm3+ and Eu3+ to Sm2+ and Eu2+, respectively, to occur under oxidizing atmosphere when either of the two was incorporated in an MBPO5 matrix. Following this idea and considering the fact that at elevated temperature, Bi2O3 readily decomposes into BiO [2,9], the aim of the present study is to evaluate whether MBPO5:Bi2+ can be synthesized in the same way.
Lattice constants of MBPO5 were originally reported by Bauer ([17,18], Table 1). Kniep et al.  as well as Rulmont and Tarte , however, suggested that the structure of MBPO5 is in fact close to trigonal stillwellite, (Ce,La,Ca)BSiO5, with only half the c value as compared to Bauer’s first results. CaBPO5 and SrBPO5  were consequently assigned to space group P3121 with a = 6.6799Å and c = 6.6121Å, and a = 6.8488Å and c = 6.8159Å (Table 1 , ) respectively. On the other hand, single-phase synthesis was not achieved in these early works. Space-group P3121 was confirmed by Shi et al. who reported lattice constants a = 7.1094Å and c = 6.9895Å for BaBPO5 that was prepared via a standard solid state reaction (Table 1 , ). Contrary, in a more recent study, Pushcharovsky et al. assigned the structure of BaBPO5 to P322 . Since knowledge on the crystal structure is of high importance for understanding and assigning spectroscopic properties and, particularly, electron-phonon interactions, these reported discrepancies deserve further consideration. In the present study, in situ high temperature X-ray diffraction (HT-XRD) was therefore employed to first determine optimum synthesis conditions for the MO-B2O3-P2O5 system, and to then synthesize single phase MBPO5:Bi2+ in air. Rietveld refinement was used to confirm that all three compounds belong to the trigonal space group P3121. Excitation and emission properties were quantified at room temperature by photoluminescence spectroscopy, and the appearance of multi phonon side bands of 2P1/2→2P3/2 could be related to the crystal structure.
Doped and un-doped samples of MBPO5 (M = Ca2+, Sr2+ and Ba2+) were prepared via solid state reaction. For that, analytical grade reagents CaCO3, SrCO3, BaCO3, H3BO3, NH4H2PO4 and Bi2O3 were used as raw materials. Individual batches of 20 g were weighed according to nominal compositions M1-xBPO5: xBi (x = 0, 0.005) and mixed thoroughly. Boric acid and ammonium-dihydrogen-phosphate were added in excess of 3% to compensate for volatilization losses. Batches were pre-calcined at 500°C for 5 h and ground for a second time. Subsequently, a small sample was taken from each batch for HT-XRD examination. For that, about 200 mg of the powder were suspended in isopropanol, and about two drops of this suspension were brought on a platinum substrate which acted as both sample holder and heating element. Finally, the major part of the batches of Ca1-xBPO5: xBi (x = 0, 0.005) and Ba1-xBPO5: xBi was heat-treated at 900°C for 10 h, Sr1-xBPO5: xBi at 1000°C for 10 h. In all heating steps, heating rates of 1.4 K/min were applied.
Crystal structures of the obtained samples were examined by room-temperature XRD. Dynamic and static optical emission and excitation spectra were recorded with a high-resolution photoluminescence spectrometer (Horiba Jobin Yvon Fluorolog-3), using a static Xe lamp (450 W) and a Xe flashlamp (75 W) as excitation sources. Fourier transform infrared (FTIR) spectra were recorded on a Nicolet Impact 420 spectrometer on samples embedded in KBr pellets at room temperature.
3. Results and discussion
3.1 Crystal structure
HT-XRD patterns for pre-calcined Ba0.995BPO5: 0.005Bi are shown in Fig. 1 . At temperatures below 800 °C, the spectra are dominated by the two precursor phases β-Ba(PO3)2 (JCPDS Card No. 43-518) and Ba4B2O7 (JCPDS Card No. 24-84). Starting at around 850 °C, diffraction peaks due to planes (200) (2θ = 28.88), (003) (2θ = 38.48), and (121) (2θ = 40.80) of BaBPO5 begin to appear. These grow continuously with further increasing temperature up to > 900 °C, where they are joined by additional reflexes at 2θ = 29.30 (012) and 2θ = 38.82 (202). Approaching 1000°C, most BaBPO5 reflexes start to disappear as a result of structural breakdown. At 1050°C, planes (012) and (003) can no longer be indexed. This suggests a preferred reaction temperature in the range of 900-950 °C. On the other hand, it explains why synthesis according to Srivastava’s procedure (at 1000°C , ) does not result in a single-phase material. Similar results are found for the synthesis CaBPO5 (onset of growth at ~800 °C, optimum at 900-1100 °C, breakdown at 1150 °C). SrBPO5 exhibits a significantly more narrow synthesis window, i.e. 950-1000°C and breakdown occurs already at 1050°C due to the competing phase Sr2P2O7 (JCPDS Card No. 75-1490).
Based on above results, Ca0.995BPO5: 0.005Bi, Sr0.995BPO5: 0.005Bi and Ba0.995BPO5: 0.005Bi were synthesized as single-phase materials as already described. Their room-temperature XRD patterns are shown in Fig. 2 . Rietveld refinement (FullProf [24,25] software package) was performed on the basis of space group P3121. Calculated diffraction patterns are compared to experimental data in Fig. 2, and corresponding lattice parameters are given in Tab.1. High consistency between calculated and experimental data clearly confirms that all three compounds are isostructural, confirming conclusions from Refs [19,22]. As a result of increasing d spacing and, hence, increasing lattice parameters, in the series M = Ca2+→Sr2+→Ba2+, all diffraction peaks gradually shift towards lower θ. This is a direct consequence of increasing cation radius, i.e. from 1.00 Å for Ca2+ to 1.18 Å for Sr2+ and 1.35 Å for Ba2+ (for sixfold coordination), and agrees well with calculated data (column 2 and 3 in Table 1).
FTIR spectroscopy reveals a corresponding picture. The network of MBPO5 consists essentially of corner-sharing BO4 tetrahedra in infinite loop-branched chains, whereby along the c-axis, free corners of BO4 are connected by PO4 tetrahedra to form a 3-dimensional network (, Fig. 3(b)). FTIR spectra of all three compounds are therefore dominated by analogous absorption peaks that can be assigned to those BO4 and PO4 groups (Fig. 3(a) ). Peaks at ~1200, ~1100, ~1050, and ~970 cm−1 originate from asymmetric stretching ν 3 modes of BO4 and PO4; ~900 cm−1 is due to ν 3 (BO4) + ν 1(PO4); ~850/~800 and ~750cm−1 are asymmetric and symmetric B-O-P stretching vibrations, respectively; ~650/~600 cm−1 are due to symmetric vibration/bending modes of BO4; ~560/520cm−1 is due to the bending mode ν 4 of PO4; ~470 cm−1 is the deformation mode ν 2 of PO4 and BO4 [20,22,26,27].
3.2 Optoelectronic properties
Room-temperature excitation and emission spectra of MBPO5:Bi2+ (prepared in air) are shown in Fig. 4 . Positions of excitation and emission peaks are given in Tab. 2 and, where applicable, compared to Srivastrava’s original data (prepared in N2/H2, spectra taken at 15 K). Consistent with the expectations from Srivastrava’s data, red emission was observed in all three compounds. This clearly evidences that Bi2+ can be stabilized in MBPO5 even when synthesized in air. Noteworthy, UV excitation peaks (2P1/2→2S1/2) could be detected also in CaBPO5:Bi2+ and SrBPO5: Bi2+, at 231 and 234 nm, respectively, for the first time (Fig. 4 and Table 2 ).
Formation and stabilization of Bi2+, even under oxidizing atmosphere, may be explained on basis of a charge compensation model [16,17,28]. When Bi3+ is doped into the lattice of MBPO5, due to the good match of ionic radii, it will be incorporated on M2+-sites. Such a [Bi3+ M2+]· defect requires negative charge compensation, what can be achieved only via the creation of a M2+ vacancy [VM2+]”, or an interstitial oxygen defect [Oi]”. At sufficiently high temperature, these negative defects will be thermally activated and, eventually, captured by [Bi3+ M2+]·, thus reducing it to [Bi2+ M2+]x In this way, the charge imbalance due to the aliovalent substitution of Bi3+ for M2+ will be removed. As discussed in Refs [16,17], the shielding environment of PO4-BO4 loops prevents future oxidation and, hence, stabilizes the [Bi2+ M2+] x defect.
Detailed examination of the emission spectra reveals shoulder peaks towards lower energy, at 676nm, 688nm and 686nm for CaBPO5:Bi2+, SrBPO5:Bi2+ and BaBPO5:Bi2+, 1080, 1066 and 1024cm−1 apart from the dominating peak, respectively. This frequency shift corresponds to the asymmetric stretching modes of BO4 and PO4 anion groups (Fig. 3). The shoulder peaks should therefore be assigned to the asymmetric stretching vibration of the respective anion groups. Corresponding phonon side bands occur for the excitation transitions from 2P1/2 in all three compounds (insets of Fig. 4). Noteworthy, single-phonon side bands were recently observed in the excitation and emission spectra of Bi2+-doped SrB4O7 and SrB6O10 .
While in the excitation spectra of CaBPO5: Bi2+ and SrBPO5: Bi2+, the phonon side bands are strongly overlapped by the main peaks of electronic excitation from 2P1/2, a more complex situation can be identified for BaBPO5:Bi2+. In the uncorrected excitation spectrum taken at 15K, Srivastava found a vibronic side band centered at 587nm, 959cm−1 higher than the lowest energy band at 622nm, but he did not notice the side band of the transition at 432 nm band. However, the present data clearly reveals the side bands of 2P1/2→2P3/2 (1) and 2P1/2→2P3/2 (2) at 562 nm, 1639 cm−1 above the main peak, and at 372 nm, 3626cm−1 above the corresponding main peak, respectively.
Crystal refinement reveals that in BaBPO5, only one Ba2+-site is present. As shown in the inset of Fig. 5 , it is surrounded by three types of oxygen, four O1, four O2 and two O3. The average bond length is 2.8687 Å. Assuming that Bi2+ is indeed incorporated into the MBPO5 lattice only on Ba2+-sites, there should be only a single type of Bi2+ emission centers present in doped material. This is confirmed by time-resolved fluorescence spectroscopy (Fig. 5 and Fig. 6 ): for BaBPO5: Bi2+, the fluorescence decay curve follows a first order exponential decay equation (Fig. 5). Excitation at wavelengths 619, 562, 432 and 372 nm always leads to the emission at 641 nm. Emission lifetimes for at all these different excitation wavelengths fall within an error range of ± 1μs around an absolute value of 22 µs, what corresponds to the experimental error of the employed set-up. Similar behaviour is found for CaBPO5: Bi2+ and SrBPO5: Bi2+ (column 5 of Tab. 2).
Figure 6 shows emission spectra of Ba0.995BPO5: 0.005Bi2+ after excitation at 430 nm as taken after different delay times (room temperature). With increasing delay time, emission intensity monotonically decreases while the line shape remains unchanged. Adding to the data shown in Fig. 5, this observation represents further evidence that only a single emission center is present in Ba0.995BPO5: 0.005Bi2+. Excitation bands at 619 nm, 430 nm, 562 nm and 372 nm all belong to the same active center, and represent the transition of 2P1/2→2P3/2 (Table 2) and corresponding vibronic sidebands.
The vibronic transition probability is determined by two contributions: so-called M and Δ processes [29,30]. The M process describes a vibrationally induced forced eletric dipole transition: coupling with an infrared active vibrational mode leads to opposite parity wavefunctions. The Δ process represents the Frank-Condon principle. It arises from a change in the equilibrium position of the excited state against the ground state. Vibronic transitions that involve multiple phonons are expected only for the Δ process, especially at low temperature. For instance, in Na5La(MoO4)4:Pr3+, for the transition of 3H4→3P2 (Pr3+), two vibronic sidebands exist 800 cm−1 and 1620 cm−1 above the zero phonon line (at 4.2 K). These can be ascribed to one- and two-phonon replicas due phonon-electron coupling of the molybdate stretching vibration at ~800cm−1) .
The highest-energy phonon of BaBPO5 is the v 3-mode of the BO4 group at ~1194 cm−1 (Fig. 3). This is much lower than the energy distance of the vibronic sidebands from the respective main peak (1639cm−1 and 3626cm−1). The sidebands are therefore assigned to multi-phonon interactions with the 2P1/2→2P3/2 transition.
Vibronic sidebands are much less pronounced in CaBPO5:Bi2+ and SrBPO5:Bi2+, what is a clear sign for different electron-phonon coupling strength . Qualitatively, this can be confirmed further by considering the Huang-Rhys factor S . S represents a measure of electron-phonon coupling strength and can be estimated from (2S-1)·ħω = Δυ ,where ħω is the maximum phonon energy and Δυ the Stokes shift. Values of ħω can be taken from FTIR data, 1194 cm−1 for BaBPO5, 1207 cm−1 for SrBPO5 and 1211 cm−1 for CaBPO5 (Fig. 3), Δυ -data are listed in column 4 of Tab. 2. Using these data, values of 0.73, 0.76, and 1.04 are derived for S of BaBPO5, SrBPO5 and CaBPO5, respectively. This means that the strength of electron-phonon coupling indeed increases from BaBPO5 to SrBPO5 to CaBPO5.
In summary, Bi2+-doped MBPO5 (M = Ba2+, Sr2+, Ca2+) were synthesized in air via a solid-state reaction at 900-1000°C. The reduction process of Bi3+ to Bi2+ occurs for reasons of charge compensation, and Bi2+-ions are stabilized on M2+-sites as a result of the surrounding crystal lattice. All three compounds are isostructural. Their crystal structure belongs to space group P3121. Fluorescence decay behaviour indicates that in accordance with crystallographic data, only one type of Bi2+-emission center is present in MBPO5. In all compounds, the transition of 2P1/2→2S1/2 was directly observed. For 2P1/2→2P3/2, vibronic sidebands were observed in both excitation and emission spectra. In the case of BaBPO5:Bi2+, one-, two- and three- phonon processes could clearly be identified. Strong absorption in the blue spectral range suggests high interest for application in WLED devices, for example in combination with a blue LED chip and/or a yellow-emitting phosphor.
Financial support by the Deutsche Forschungsgemeinschaft under grant no. WO 1220/2-1 is gratefully acknowledged.
References and links
1. K. Murata, Y. Fujimoto, T. Kanabe, H. Fujita, and M. Nakatsuka, “Bi-doped SiO2 as a new laser material for an intense laser,” Fusion Eng. Des. 44(1-4), 437–439 ( 1999). [CrossRef]
2. M. Peng, J. Qiu, D. Chen, X. Meng, L. Yang, X. Jiang, and C. Zhu, “Bismuth- and aluminum-codoped germanium oxide glasses for super-broadband optical amplification,” Opt. Lett. 29(17), 1998–2000 ( 2004). [CrossRef] [PubMed]
3. V. V. Dvoyrin, V. M. Mashinsky, L. I. Bulatov, I. A. Bufetov, A. V. Shubin, M. A. Melkumov, E. F. Kustov, E. M. Dianov, A. A. Umnikov, V. F. Khopin, M. V. Yashkov, and A. N. Guryanov, “Bismuth-doped-glass optical fibers - a new active medium for lasers and amplifiers,” Opt. Lett. 31(20), 2966–2968 ( 2006). [CrossRef] [PubMed]
4. M. Peng, D. Chen, J. Qiu, X. Jiang, and C. Zhu, “Bismuth-doped zinc aluminosilicate glasses and glass-ceramics with ultra-broadband infrared luminescence,” Opt. Mater. 29(5), 556–561 ( 2007). [CrossRef]
5. M. Peng and L. Wondraczek, “Bismuth-doped oxide glasses as potential solar spectral converters and concentrators,” J. Mater. Chem. 19(5), 627 ( 2009). [CrossRef]
6. S. Zhou, N. Jiang, B. Zhu, H. Yang, S. Ye, G. Lakshminarayana, J. Hao, and J. Qiu, “Multifunctional Bismuth-Doped Nanoporous Silica Glass: From Blue-Green, Orange, Red, and White Light Sources to Ultra-Broadband Infrared Amplifiers,” Adv. Funct. Mater. 18(9), 1407–1413 ( 2008). [CrossRef]
7. M. Peng, Q. Zhao, J. Qiu, and L. Wondraczek, “Generation of emission centers for broadband NIR luminescence in bismuthate glass by femtosecond laser irradiation,” J. Am. Ceram. Soc. 92(2), 542–544 ( 2009). [CrossRef]
8. S. Zhou, W. Lei, N. Jiang, J. Hao, E. Wu, H. Zeng, and J. Qiu, “Space-selective control of luminescence inside the Bi-doped mesoporous silica glass by a femtosecond laser,” J. Mater. Chem. 19(26), 4603–4608 ( 2009). [CrossRef]
9. M. Peng, C. Zollfrank and L. Wondraczek, “Origin of broad NIR photoluminescence in bismuthate glass and Bi-doped glasses at room temperature” J. Phys.: Condens. Matter, 21, 285106–1-6, (2009).
10. G. Blasse, A. Meijerink, M. Nomes, and J. Zuidema, “Unusual bismuth luminescnece in strontium tetraborate (SrB4O7:Bi),” J. Phys. Chem. Solids 55(2), 171–174 ( 1994). [CrossRef]
11. M. Hamstra, H. Folkerts, and G. Blasse, “Materials chemistry communications. Red bismuth emission in alkaline-earth-metal sulfates,” J. Mater. Chem. 4(8), 1349–1350 ( 1994). [CrossRef]
12. A. Srivastava, “Luminescence of divalent bismuth in M2+BPO5 (M2+=Ba2+, Sr2+ and Ca2+),” J. Lumin. 78(4), 239–243 ( 1998). [CrossRef]
14. L. Su, P. Zhou, J. Yu, H. Li, L. Zheng, F. Wu, Y. Yang, Q. Yang, and J. Xu, “Spectroscopic properties and near-infrared broadband luminescence of Bi-doped SrB4O7 glasses and crystalline materials,” Opt. Express 17(16), 13554–13560 ( 2009). [CrossRef] [PubMed]
15. H. Liang, Y. Tao, W. Chen, X. Ju, S. Wang, and Q. Su, “The luminescent properties of lanthanide ions activated BaBPO5 in VUV-Vis range,” J. Phys. Chem. Solids 65(6), 1071–1076 ( 2004). [CrossRef]
16. Q. Su, H. Liang, T. Hu, Y. Tao, and T. Liu, “Preparation of divalent rare earth ions in air by aliovalent substitution and spectroscopic properties of Ln2+,” J. Alloy. Comp. 344(1-2), 132–136 ( 2002). [CrossRef]
17. H. Bauer, “Über eine Reihe isotyper Erdalkaliboratphosphate und –arsenate vom Typus 2MeO.X2O5.B2O3,” Z. Anorg. Allg. Chem. 337(3-4), 183–190 ( 1965). [CrossRef]
18. H. Bauer, “Die Verbindungen 2BaO.P2O5.B2O3 und 2BaO.As2O5.B2O3,” Z. Anorg. Allg. Chem. 345(5-6), 225–229 ( 1966). [CrossRef]
19. R. Kniep, G. Gözel, B. Eisenmann, C. Röhr, M. Asbrand, and M. Kizilyalli, “„Borophosphate - Eine stiefmütterlich behandelte Verbindungsklasse: Die Kristallstrukturen von MII[BPO5] (MII = Ca, Sr) und Ba3[BP3O12],” Angew. Chem. Int. Ed. Engl. 33, 749–750 ( 1994). [CrossRef]
20. A. Rulmont and P. Tarte, “Lanthanide borogermanate LnBGeO5: synthesis and structural study by X-Ray diffractometry and vibrational spectroscopy,” J. Solid State Chem. 75(2), 244–250 ( 1988). [CrossRef]
21. S. Pan, Y. Wu, P. Fu, G. Zhang, Z. Li, C. Du, and C. Chen, “Growth, Structure, and Properties of Single Crystals of SrBPO5,” Chem. Mater. 15(11), 2218–2221 ( 2003). [CrossRef]
22. Y. Shi, J. Liang, H. Zhang, Q. Liu, X. Chen, J. Yang, W. Zhuang, and G. Rao, “Crystal Structure and Thermal Decomposition Studies of Barium Borophosphate, BaBPO5,” J. Solid State Chem. 135(1), 43–51 ( 1998). [CrossRef]
23. D. Pushcharovsky, E. Gobetchia, M. Pasero, S. Merlino, and O. Dimitrova, “Hydrothermal synthesis and crystal structures of Li,Ba-nanoborate, LiBaB9O15, and Ba-borophosphate, BaBPO5,” J. Alloy. Comp. 339(1-2), 70–75 ( 2002). [CrossRef]
24. http://www.ill.eu/sites/fullprof/ ( 2009)
25. J. Rodriguez-Carvajal, “Recent advances in magnetic structure determination by neutron powder diffraction,” Physica B 192(1-2), 55–69 ( 1993). [CrossRef]
26. A. Baykal, M. Kizilyalli, G. Gözel, and R. Kniep, “Synthesis of strontium borophosphate SrBPO5 by solid state and hydrothermal methods and characterisation,” Cryst. Res. Technol. 35(3), 247–254 ( 2000). [CrossRef]
27. C. Lin, Y. Luo, H. You, Z. Quan, J. Zhang, J. Fang, and J. Lin, “Sol-gel-derived BPO4/Ba2+ as a new efficient and environmentally-friendly bluish white luminescent material,” Chem. Mater. 18(2), 458–464 ( 2006). [CrossRef]
28. M. Peng, Z. Pei, G. Hong, and Q. Su, “Study on the reduction of Eu3+→Eu2+ in Sr4Al14O25: Eu prepared in air atmosphere,” Chem. Phys. Lett. 371(1-2), 1–6 ( 2003). [CrossRef]
29. G. Blasse, “Vibronic transitions in rare earth spectroscopy,” Int. Rev. Phys. Chem. 11(1), 71–100 ( 1992). [CrossRef]
30. C. de Mello Donegá, M. Crombag, A. Meijerink, and G. Blasse, “Vibronic transitions in the luminescence spectra of Pr3+ in Na5La(MoO4)4,” J. Lumin. 60–61, 74–77 ( 1994). [CrossRef]
31. K. Huang and A. Rhys, “Theory of Light Absorption and Non-Radiative Transitions in F-Centres,” Proc. R. Soc. Lond. A Math. Phys. Sci. 204(1078), 406–423 ( 1950). [CrossRef]
32. M. Nazarov, B. Tsukerblat, and D. Noh, “Vibronic coupling parameters and Stokes shift in thiogallate phosphors,” J. Phys. Chem. Solids 69(10), 2605–2612 ( 2008). [CrossRef]