Energy Transfer between Tb3+ and Eu3+ in LaPO4: Pulsed versus Switched‐off Continuous Wave Excitation

Abstract The energy transfer (ET) between Tb3+ and Eu3+ is investigated experimentally and with available theoretical models in the regime of high Tb3+ concentrations in ≈30 nm LaPO4 nanoparticles at room temperature. The ET efficiency approaches 100% even for lightly Eu3+‐doped materials. The major conclusion from the use of pulsed laser excitation and switched‐off continuous wave laser diode excitation is that the energy migration between Tb3+ ions, situated on La3+ sites with ≈4 Å separation, is not fast. The quenching of Tb3+ emission in singly doped LaPO4 only reduces the luminescence lifetime by ≈50% in heavily doped samples. Various theoretical models are applied to simulate the luminescence decays of Tb3+ and Tb3+, Eu3+‐doped LaPO4 samples of various concentrations and the transfer mechanism is identified as forced electric dipole at each ion.


DOI: 10.1002/advs.201900487
Dy-Tb, [5,6] Sm-Eu, [6][7][8] and Ho-Yb, [9,10] with the focus upon the applied optical properties of phosphors, such as the color tunability of luminescence and optical thermometry. However, Dutra et al. [11] have pointed out that in 2015 less than 3% of published studies concerning lanthanide ions made use of theoretical tools. In particular, the ET between Tb 3+ and Eu 3+ has attracted much attention (Table S4, Supporting Information). The 5 D 4 donor state emission from Tb 3+ is quenched and instead red emission from the Eu 3+ 5 D 0 state is observed when exciting into a Tb 3+ absorption band. The detailed energy transfer mechanism is not clear in many cases. For example, Moran et al. [12] studied the cubic system Cs 2 NaYCl 6 doped with Tb 3+ and Eu 3+ in a wide range of concentrations at 298 and 77 K. It was concluded that discrepancies between the calculated and experimental rate parameters are large, notably by four or five orders of magnitude for multipole-multipole mechanisms. Whereas it could be possible that different interaction mechanisms take part in different solid-state lattices, various studies have attributed the interaction mechanism between Tb 3+ and Eu 3+ to electric dipole-electric dipole (ED-ED), [13][14][15][16] or electric dipole-electric quadrupole (ED-EQ) [17] or exchange interaction. [18,19] The 5 D 4 → 7 F j (J = 6-4) transitions of Tb 3+ are forced ED allowed whereas the 5 D 0 ← 7 F 0 transition of Eu 3+ is dipole forbidden, [20] but at room temperature the 5 D 0 ← 7 F 1 forced ED pathway is also available. The term forced (or induced) ED refers to 4f-4f transitions being enabled via the wavefunction mixture with 5d (or higher) opposite-parity configurations through the crystal field symmetry or vibrations of the appropriate symmetry. Furthermore, the energy migration between the donor Tb 3+ ions has been stated as very fast [21] or slow. [22] For example, Blasse [23] envisaged that the migration rate between lanthanide ions such as Eu 3+ or Gd 3+ was up to 10 7 s -1 for a separation of 4 Å, whereas Dornauf and Heber [24] found negligible concentration quenching of Tb 3+ in LaP 5 O 14 . The process of migration between identical ions leads to an increase in lifetime, whereas termination of the process at a trap leads to a decrease.
The kinetics of an ET process as a function of concentration have been studied by several authors, for example, by Fong and Diestler [25] and Lupei. [26] A linear relation between the energy transfer rate and acceptor concentration has been taken to indicate a two-ion process, whereas a quadratic The energy transfer (ET) between Tb 3+ and Eu 3+ is investigated experimentally and with available theoretical models in the regime of high Tb 3+ concentrations in ≈30 nm LaPO 4 nanoparticles at room temperature. The ET efficiency approaches 100% even for lightly Eu 3+ -doped materials. The major conclusion from the use of pulsed laser excitation and switched-off continuous wave laser diode excitation is that the energy migration between Tb 3+ ions, situated on La 3+ sites with ≈4 Å separation, is not fast. The quenching of Tb 3+ emission in singly doped LaPO 4 only reduces the luminescence lifetime by ≈50% in heavily doped samples. Various theoretical models are applied to simulate the luminescence decays of Tb 3+ and Tb 3+ , Eu 3+ -doped LaPO 4 samples of various concentrations and the transfer mechanism is identified as forced electric dipole at each ion.

Introduction
Lanthanide ions present unique spectral properties, including long lifetime, multiple narrow and well-separated bands, and large effective shifts from excitation to emission wavelength. They can be model candidates as donor and acceptor in the study of energy transfer (ET). Hence, there have been a plethora of ET research studies concerning lanthanide ions, such as between the tripositive ions Ce-Eu, [1] Ce-Tb, [2] Tb-Eu, [3] Pr-Yb, [4] relation is indicative of a three-ion process. Lupei [26] studied the quenching of Er 3+4 S 3/2 emission in YAG: Er 3+ at high concentrations of Er 3+ and explained the variation of the energy transfer rate in terms of a three-body process where two acceptor ions are promoted from the electronic ground state to 4 I 9/2 and 4 I 13/2 J-multiplets. Hence, we were very surprised to observe the same type of interaction in two literature studies of Tb 3+ -Eu 3+ ET [27,28] (Figure 1a,b) where Eu 3+ concentration is varied when Tb 3+ concentration is high: for Cs 2 NaTb 1-x Eu x Cl 6 at 293 K and TlY 2.25-x Tb 0.75 Eu x F 10 at 295 K. However, at higher Eu 3+ concentrations in Cs 2 NaY 0.995-x Tb 0.005 Eu x Cl 6 at 295 K and Cs 2 NaTb 1-x Eu x Cl 6 at 80 K (Figure 1c,d), the variation of energy transfer rate with Eu 3+ concentration is clearly linear.
These unusual findings prompted us to reinvestigate the rate dependence of Tb 3+ -Eu 3+ ET as a function of concentration and demonstrate that in fact it is a two-body process and not a three-body process as proposed elsewhere. [20,27] We selected the lanthanide orthophosphate host because of its high insolubility, non-hygroscopicity, reliable stability against high temperature, low toxicity, and high quantum yield. [14] Naturally, it is a prerequisite to determine the photophysics of the singly doped host prior to an ET study involving two lanthanide ions, and we have performed this for LaPO 4 doped with Tb 3+ . Whereas most previous studies have utilized micrometer-sized samples synthesized at high temperature (Table S4, Supporting Information), we prepared nanomaterials doped with Tb 3+ and Eu 3+ because the occurrence of trap species, if important, would then be magnified in our study. We take the previous study of Chen and co-workers [29] on similar size nanoparticles of Eu 3+ -doped hexagonal hydrated TbPO 4 as a reference for comparison of our results obtained from nanomaterials prepared by a modified method of Haase and co-workers. [30,31] The relevant energy level schemes of Tb 3+ and Eu 3+ are displayed in Figure 2a.
The major aims of our study were to investigate (i) the importance of energy migration between terbium ions in the LaPO 4 host, where the donor-donor separation is about 4 Å; and (ii) to give a quantitative description of Tb 3+ -Eu 3+ ET in this host lattice, where complications such as charge inequality and multisite occupation of the ions are absent. With the use of different excitation techniques and various theoretical models, these aims have been fully addressed.

Theoretical Section
Various models have been employed to simulate the donor decay after pulsed excitation when energy is transferred from a donor to acceptor ions in order to ascertain the interaction mechanism. We largely focus upon ED-ED energy transfer because the relevant lanthanide ion transitions are of forced electric dipole character. The donor emission intensity at time t, relative to the initial intensity, when in the presence of an acceptor, is given by [32] where τ D is the lifetime of donor in the absence of the acceptor and k ET is the energy transfer rate (=P DA c A , where P DA is a parameter describing the donor-acceptor interaction and c A is the acceptor concentration). This equation holds for an isolated donor-acceptor pair, or for an acceptor in an average donor environment due to rapid energy migration between donors. In the absence of donor-donor migration, the decay is not monoexponential because the acceptors at sites q are situated at different distances [32,33] Adv. Sci. 2019, 6,1900487 where the sum over X 0q represents the total donor-acceptor transfer rate. In the continuum approximation (which predicts excessive energy transfer rates at very small donor-acceptor separations), the Inokuti-Hirayama model [34] for multipole interactions assumes a random arrangement of acceptors and the intensity of emission at time t is related to that initially by where s = 6, 8, or 10 according to the electric multipolar mechanism ED-ED, ED-EQ, or EQ-EQ, respectively; Γ is the gamma function; [34] x is the stoichiometric mole fraction, Z is the number of acceptors in the unit cell, and V is its volume; c 0 is the critical (reduced) acceptor concentration where R 0 is the critical distance. Instead, if transfer is considered to shells of acceptor ions in a crystal [24] 0 exp exp 4 where l Z is the total number of acceptor sites present in shell l at distance l R from the donor; Z is the total number of sites in k shells; and x is the acceptor molar concentration. N is the total number of acceptors = Zx. The second exponential function represents a correction factor for acceptors outside the chosen shells. The other symbols are defined above. An alternative description of the shell model for small (≈3 nm) spherical nanoparticles, based upon the master Equation (2), and correcting for the proximity of the donor to the surface, has been given by Rabouw et al. [35,36] This equation is not necessary in the present case since our particles are rather larger.  In Equations (2), (3), and (5), no account is taken of donordonor migration or donor-acceptor back transfer. The presence of 3D energy migration may be examined by the analysis of the (monoexponential) donor decay lifetime, τ, at very long times after the excitation pulse, the asymptotic behavior (t → ∞) where, in this case, τ D represents the intrinsic donor lifetime and for weak diffusion (when the donor-donor migration rate is smaller than the donor-acceptor transfer rate), with the rate Here, D is a diffusion constant and C is a donor-acceptor interaction constant. Alternatively, for weak diffusion, the simplest equation for the inclusion of migration between donor ions, for the case of ED-ED interaction is [37] 0 exp 4 3 where D is a diffusion constant and DA (6) α is the ED-ED interaction parameter for donor-acceptor transfer ( / ) . Analogous equations have been given for donor-acceptor exchange interaction, and for the occurrence of migration for types of interaction other than ED-ED, and the reader is referred to the original publications for details. [24,34,37,38] In principle, the fitting of donor decay by these equations yields the interaction mechanism, the critical distance (i.e., when the rate of ET is equal to the donor decay rate), and the magnitude of the diffusion constant. When the donor-donor migration rate exceeds that of donor-acceptor transfer, a random hopping model is more appropriate. [39] The above equations refer to the donor emission profile after a short excitation pulse. The donor emission decay profile following switched-off continuous excitation has been discussed by Eisenthal and Siegel [40] and Siebold and Heber. [41] Pulsed excitation provides an instantaneous population of randomly excited donors at t = 0 (cf. Equation (3)) with the acceptor ions unexcited. By contrast, switched-off continuous excitation gives the steady-state population of excited donors. In this case, the excited donors are not randomly distributed-the probability of an excited donor being near an acceptor is smaller than that of an excited donor with an acceptor further away. The donor decay will therefore differ from the case of pulsed excitation unless very rapid migration of energy occurs between the donors. Assuming a random population of unexcited donors and acceptors, in the absence of donor-donor migration, the donor decay at time t, after switched-off continuous excitation at time t = 0, can be represented in the case of ED-ED transfer by [40] 0 exp 2 e xp 1 erf and Here, τ is the donor lifetime in the absence of acceptors; N A is the number of unexcited acceptors; R 0 is the critical transfer distance for which the probability of nonresonance deactivation and resonance transfer is equal for a donoracceptor pair; and R v is the vessel radius. The donor decay under the same conditions for pulsed excitation is given by Equation (3).

Structure and Morphology of the Crystals
The morphology of La 0.10 Tb 0.90 PO 4 (i) and La 0.02 Tb 0.90 Eu 0.08 PO 4 (ii) nanocrystals, together with the high-resolution transmission electron microscopy (HRTEM) image of the latter (iii) are presented in Figure 2b. The nanocrystals are elongated compared with the size of the LaPO 4 nanocrystals prepared by Haase and co-workers, [31] which is due to the decreasing amount of the complexing agent triethyl phosphate employed. [30] The HRTEM image displays lattice fringes with spacing 0.5 nm for most particles which indicates a highly crystalline material. The measured average size of the nanocrystals is 30 nm compared with that of 34 nm determined from the line broadening in the X-ray diffraction (XRD) pattern (Figure 2c). The XRD pattern is consistent with the standard monoclinic card of TbPO 4 with a shift of the peaks to lower angles since Eu 3+ is a slightly larger ion than Tb 3+ . No additional peaks of other phases are present. The structure of the monoclinic form TbPO 4 (space group P2 1 /n, Z = 4 [42] ) is shown in Figure 2d and is the same structure for the other lanthanide ions from La to Gd. The closest distance between the nine-coordinate Tb 3+ ions, with site symmetry C s , is 3.97 Å. There is another form of TbPO 4 belonging to the tetragonal space group I4 1 /amd [43] with Tb 3+ ions situated at sites of D 2d point group symmetry, with closest separation 3.79 Å.

The Decay of Tb 3+ Emission in LaPO 4 :Tb 3+
The luminescence decay of the Tb 3+ 5 D 4 state was monitored for LaPO 4 doped with various concentrations of Tb 3+ , by exciting at 485 nm and collecting the emission of the 5 D 4 → 7 F 5 transition at 543 nm. The decay curves for switched-off continuous wave (cw) and for pulsed excitation are shown on a log scale in Figure 3a for one sample of La 0.10 Tb 0.90 PO 4 and show a distinct difference in the initial decay. The time width of the excitation pulse, in the microsecond regime, is therefore smaller than the energy migration hopping time. If the migration rate were fast, then the curves would be superimposable. The decay following cw excitation can be fitted by a monoexponential function in the region from 0 to 15 ms with the lifetime 1.556 ± 0.001 ms ( adj 2 R = 0.99933) whereas that for pulsed excitation can be fitted by biexponential decay with the above fixed lifetime and another one of 0.258 ± 0.001 ms ( adj 2 R = 0.99908). The decay after cw excitation represents the steady state of the migration and energy transfer to traps, whereas pulsed excitation provides an equal excitation probability for all donors. Hence, the difference could be due to the fact that some donors are near traps, that is, energy acceptors of a different species. The nature of the traps is subsequently discussed. The decays were fitted by Equation (9a) (for switched-off cw excitation, Figure 3b Table S1 (Supporting Information) and demonstrate a monoexponential decrease in lifetime with increasing concentration of Tb 3+ in La 1-x Tb x PO 4 (Figure 4b). The lifetime of the La 0.10 Tb 0.90 PO 4 nanocrystals (≈1.5 ms) is considerably longer than that reported at room temperature as 0.55 ms for TbPO 4 nanowires (2 µm × 40 nm) [45] and 0.98 ms for micrometer-size TbPO 4 [29] so that surface quenching in these latter two examples is clearly very important. The lifetime of a La 0.98 Tb 0.02 PO 4 bulk sample [46] was reported to be 3.1 ms which is similar to the measured lifetime of 3.04 ms for La 0.999 Tb 0.001 PO 4 nanocrystals herein.
The use of Equation (1), where τ D in this case is the lifetime at infinite dilution and k ET is the ET rate to traps, gives a linear relation of k ET with concentration (Figure 4c), which can be interpreted as showing that a two-body process is responsible for quenching: Tb 3+ -trap. The ET rate is slow: being only ≈350 s -1 for the La 0.10 Tb 0.90 PO 4 sample. Recently, Johnson et al. have argued that migration to killer sites (i.e., nonradiative sinks) is more important than cross-relaxation processes. [47] This alternative quenching process: cross-relaxation involving  two or three [48] Tb 3+ ions is not possible when considering the Tb 3+ energy level scheme in Figure 2a since the energy gap below 5 D 4 is nearly 15 000 cm -1 . Fitting the long-term decay curves with Equation (6) enables the determination of 1/τ Df , which is also not fast. This parameter gives a linear plot against Tb 3+ concentration, as expected from Equation (7) (Figure S1, Supporting Information).
More detailed fits to the decay curves of La 1-x Tb x PO 4 were then carried out. Figure 5a,b shows the section of Tb 3+ emission decay in the range from 0 to 4 ms with monoexponential (green), biexponential (red), and Inokuti-Hirayama, Equation (3) (blue), fittings for x = 0.9, 0.7. The Inokuti-Hirayama equation refers to energy transfer from a donor to a random array of acceptor ions without donor-donor migration. This model gives a poorer fit to the data. The biexponential fit is superior to the linear fit described above and was repeated for other Tb 3+ concentrations with the results listed in Table S2 (Supporting Information). The biexponential fit is explained by the occurrence of two different environments of Tb 3+ ions. In their study, van Hest et al. [46] have associated two environments of Eu 3+ in 3.9 nm nanocrystals of LaPO 4 with surface (τ = 1.9 ms) and interior (τ = 4.7 ms) sites since the amplitudes if their fit were 44 and 56%, respectively. The occurrence of several sites for Eu 3+ in La 0.98 Eu 0.02 PO 4 has previously been reported from 77 K emission spectra by Dexpert-Ghys et al., [49] who pointed out that in EuPO 4 and LaPO 4 the average LnO distances are 2.474 and 2.574 Å, respectively, so that in La 0.98 Eu 0.02 PO 4 some distortion occurs. We find for the x = 0.05-0.70 samples of La 1-x Tb x PO 4 that the mean amplitudes of the fast and slow components in the biexponential fits are about 32 and 68%, which differ considerably from the population values of about 4 and 96% for surface and interior ions in our nanoparticles. The energy transfer rate (k ET ) for the ions decaying by the fast component is 3.6 times faster than for the slower-decaying ions (Figure 5c). Our time-resolved spectra following pulsed excitation do not show evidence from band positions or intensities, under the experimental resolution of several nm, for the assignment of spectral features to Tb 3+ ions located at different sites. The faster decay is then associated with energy transfer to OHor H 2 O killer trap molecules (with the latter observed to be present from the FTIR spectrum) or other defect sites, followed by nonradiative decay. The slower decay is due to 5 D 4 emission at the regular Tb 3+ C s site in TbPO 4 .
In summary, the cw and pulsed excitation regimes distinguish behavior between Tb 3+ ions at the microsecond level and this is rationalized in terms of slow energy migration and the different decay of Tb 3+ at two different crystal environments. The quenching of emission with increasing Tb 3+ concentration is only moderate and emission is strong even in the 90% Tb 3+doped sample.
Considering the crystal structure of TbPO 4 ( Figure S2, Supporting Information), the migration between Tb 3+ ions is expected to be 3D. Consideration of the selection rules for 7 F 6 -5 D 4 migration [20] shows that it is forbidden to first order by the exchange mechanism and although spin-forbidden to first order, it could be due to the EQ-EQ or ED-ED mechanisms, where ED represents forced electric dipole.

Electronic Spectra
We have investigated the scenario of high Tb 3+ donor concentrations in the ET from Tb 3+ to Eu 3+ . Figure 6a shows the excitation spectrum of La 0.92 Eu 0.08 PO 4 when monitoring the Eu 3+ 5 D 0 → 7 F 4 emission at 697 nm. All features but one correspond to transitions from the electronic ground state to 4f 6 excited states of Eu 3+ , with the strongest band at 393 nm corresponding to 7 F 0 → 5 L 6 . The broad feature with maximum at 265 nm is the Eu-O charge transfer band. Notice that there is no Eu 3+ absorption band near 480-500 nm so when exciting this sample with 485 nm radiation, as Figure 6b shows on an expanded ordinate scale, there are no emission bands present. The excitation spectrum of La 0.10 Tb 0.90 PO 4 , Figure 6c exhibits the 7 F 6 → 5 D 4 absorption band at 485 nm, together with transitions to higher excited 4f 8 states of Tb 3+ , and to 4f 7 5d 1 at 260 nm. Note the absence of the broad, strong band at 316 nm observed by Chen and co-workers. [29] Hence, when exciting La 0.10 Tb 0.90 PO 4 at 485 nm, Figure 6d, the emission corresponds to 5   the codoped sample La 0.02 Tb 0.90 Eu 0.08 PO 4 when monitoring the Eu 3+ emission from 5 D 0 exhibits features of both Eu 3+ and Tb 3+ , thereby demonstrating ET from Tb 3+ to Eu 3+ . Now, when exciting into the Tb 3+ absorption band at 485 nm in La 0.02 Tb 0.90 Eu 0.08 PO 4 (Figure 6f), the emission from Tb 3+ is mainly quenched and all bands correspond to Eu 3+ 5 D 0 → 7 F J transitions. The time-resolved spectra of La 0.05 Tb 0.90 Eu 0.05 PO 4 show that the quenching of Eu 3+ emission (due to efficient energy transfer from Tb 3+ to Eu 3+ ) is more complete compared with La 0.75 Tb 0.20 Eu 0.05 PO 4 ( Figure S3, Supporting Information).
Adv. Sci. 2019, 6,1900487   There is no obvious emission from Tb 3+ in La 0.05 Tb 0.90 Eu 0.05 PO 4 even when the delay time is only 100 ns.

Efficiency of Energy Transfer
The efficiency of ET from Tb 3+ to Eu 3+ in La 0.10-x Tb 0.90 Eu x PO 4 was calculated by two methods. Figure 7a shows the decay curves of Tb 3+ emission in La 0.10-x Tb 0.90 Eu x PO 4 nanocrystals and the fitted monoexponential lifetime decreases from 1.6 ms (x = 0) with increasing of Eu 3+ mole ratio to 0.01 ms (x = 0.1) (  Figure 7c. The linear relation demonstrates that the ET involves one Tb 3+ ion and one Eu 3+ ion and is not a three-body process. The efficiency of ET from Tb 3+ to Eu 3+ achieves 99% for x = 0.05 (Table 1, column 6) and the ET rate is quite fast, 4.9 × 10 4 s -1 , accounting for the near-monoexponential decay in Figure 7a.    samples. The curves have been fitted by biexponential decay functions representing a donor-acceptor process in Figure S4 (Supporting Information). The donor lifetime was set as that in column 2 of Table 1. At low Eu 3+ concentrations, the build-up process from population by Tb 3+ 5 D 4 is clearly visible (refer to the inset in Figure 7b for x = 0.01). However, with increasing x the Tb 3+ decay is faster and the build-up is too fast to be observed in the figures ( Figure S4, Supporting Information, x = 0.05-0.10). The fitted 5 D 0 Eu 3+ lifetime does not change greatly-from 1.61 ms for x = 0.01 to 1.28 ms for x = 0.10-but the variation can be fitted by a monoexponential function ( Figure S4f, Supporting Information). The depopulation of the Eu 3+ 5 D 0 state is also due to a two-body process ( Figure S4g, Supporting Information).

Migration and Mechanism of Energy Transfer between Tb 3+ and Eu 3+
As discussed above for La 0.10 Tb 0.90 PO 4 , the rate of excitation migration between Tb 3+ ions is not fast. The 5 D 4 Tb 3+ emission decays after pulsed and switched-off continuous wave excitation in the Tb 3+ ,Eu 3+ doped system are displayed on a logarithmic ordinate scale in Figure 8a and show that the migration rate between Tb 3+ ions is slower than the microsecond time scale. The fits to the curves using Equations (9a) and (3) are shown on a linear scale in Figure 8b and from the determined ratio of c A /c 0 , both give the critical distance R 0 as 17 Å using Equation (4). Attempts to calculate the diffusion constant using the Yakota-Taminoto Equation (8), requiring one fixed and four variable parameters, were unsuccessful because there are too many parameters and the fits are not unique. Two models were employed to investigate the ET mechanism for the entire series of La 0.10-x Tb 0.90 Eu x PO 4 luminescence decays of Tb 3+ . The Inokuti-Hirayama equation (Equation (3)) gave the superior fits for s = 6 ( Figure 8c, Table 2). This ET model is based on the assumption that there is a continuous distribution of the acceptor sites around the donors. Dornauf and Heber [24] derived the model, Equation (5)  by the structure of the crystal lattice. Table S3 (Supporting Information) gives the radial distances from a donor ion to the acceptor sites in the first 80 shells in the LaPO 4 host lattice and the number of acceptor sites in each shell. For this system, the number of discrete lattice sites around the donor was taken to be Z = 115. This corresponds to a sphere with radius 12.4863 R k = Å. The curves can be well-fitted with Equation (5) only when s = 6 which means that the multipolar mechanism is ED-ED transfer. (Figure 8d) The latter fits used two adjustable parameters, I D (0) and R 0 , and the values of R 0 are listed in Table 3. The critical transfer distances values vary over a narrow range and are slightly smaller than the values calculated in Table 2 using the model of Inokuti-Hirayama. Figure 8e shows the sensitivity of the fits using Equation (5) to the number of shells employed. The fits improve when more shells are employed.

Conclusions
Luminescence quenching by migration of the excitation energy throughout the lattice which terminates at trap sites often limits the concentration of doped lanthanide ions in a solidstate material to less than a few mol%. The energy migration in highly doped materials has been stated as very fast or slow but no related experiment has proven the suggestion. We have chosen a system with a high mole ratio of terbium in order to investigate the migration process. We focused upon the luminescence performance of Tb 3+ doped into the LaPO 4 host lattice by employing pulsed and switched-off continuous wave excitation. The experiments gave different Tb 3+ luminescence decay profiles so that a fast migration regime does not occur following excitation. In fact, the luminescence lifetime is only reduced by ≈50% on going from a very dilute sample to La 0.10 Tb 0.90 PO 4 . The quenching has been associated with the presence of interstitial water. The process of migration itself is an ET process between like ions so that it depends upon the oscillator strengths of the donor and acceptor transitions and the spectral overlap integral. The slow migration between Tb 3+ ions results from (i) the weak oscillator strength of the 5 D 4 -7 F 6 transitions (taken as 3 × 10 -7 [50,51] ) and (ii) the diminished spectral overlap between absorption and emission because the spread of the 7 F 6 J-multiplet is larger than kT at room temperature. Figure S1 (Supporting Information) shows that the migration rate, k Df , is similar to the radiative rate. The value herein is similar to Adv. Sci. 2019, 6,1900487