Luminescence Thermal Quenching of M2SiO4:Eu2+ (M = Sr, Ba) Phosphors

Shirun Yan
DOI: https://doi.org/10.1016/j.jre.2019.03.020
IF: 4.632
2020-01-01
Journal of Rare Earths
Abstract:In this paper, luminescence thermal quenching of M2SiO4:Eu2+ (M = Sr, Ba) orthosilicate phosphors and mechanisms for thermal quenching proposed by different authors are briefly reviewed. Depending on preparation conditions and/or Eu2+-doping concentrations, the quenching temperature (T1/2) and activation energy for thermal quenching of the same orthosilicate phosphor reported by different authors are inconsistent. The common conclusion is that T1/2 of the intermediate compound (Ba1−xSrx)2SiO4:Eu2+ (x ≈ 0.5) is higher than that of either Sr2SiO4:Eu2+ or Ba2SiO4:Eu2+ end-member. Moreover, T1/2 of the best-performing SrBaSiO4:Eu2+ is evidently lower than that of YAG:Ce3+ and some Eu2+-doped nitride phosphors. Regarding the quenching mechanism, most of the investigators attributed thermal quenching to a thermally assisted 4f–5d cross-over in the configuration coordinate diagram. Only a few authors ascribed thermal quenching to a thermally assisted photoionization of 5d electron to conduction band of the host. Nonetheless, a close inspection of T1/2 and Stokes shift derived from the vibrational spectra of the intermediate compound and end-member phosphors indicates that the 5d electron photoionization model instead of the 4f–5d crossing decay model should be the genuine mechanism for the thermal quenching of M2SiO4:Eu2+ (M = Sr, Ba) phosphors. Since the relationship between T1/2 and Stokes shift of the phosphors does not support the 4f–5d crossing decay model. The ionization probability of the 5d electron depends on the energy gap (EdC) between 5d1 level of the Eu2+ and conduction band minimum (CBM) of the host at higher temperatures. Lattice thermal expansion would result in an elevating 5d1 level of the Eu2+ along with a diminishing CBM of the host and as a consequence a reduction in EdC and an enhanced photoionization probability at elevated temperatures. A less rigid lattice and hence a larger coefficient of thermal expansion of M2SiO4 hosts should be the physical origin of poorer thermal quenching properties of the orthosilicate phosphors.
What problem does this paper attempt to address?