Home
Search
Collections
Journals
About
Contact us
My IOPscience
Luminescence properties of different Eu sites in LiMgPO4:Eu2+, Eu3+
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 J. Phys.: Condens. Matter 26 385401 (http://iopscience.iop.org/0953-8984/26/38/385401) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 130.92.9.56 This content was downloaded on 12/09/2014 at 16:07
Please note that terms and conditions apply.
Journal of Physics: Condensed Matter J. Phys.: Condens. Matter 26 (2014) 385401 (10pp)
doi:10.1088/0953-8984/26/38/385401
Luminescence properties of different Eu sites in LiMgPO4:Eu2+, Eu3+ A Baran1, S Mahlik1, M Grinberg1, P Cai2, S I Kim2 and H J Seo2 1
Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80–952 Gdansk, Poland Department of Physics and Center for Marine-Integrated Biomedical Technology, Pukyong National University, Busan 608–737, Republic of Korea 2
E–mail: [email protected] Received 3 June 2014, revised 23 July 2014 Accepted for publication 5 August 2014 Published 2 September 2014 Abstract
The effect of temperature on the luminescence properties of LiMgPO4 doped with Eu3+ and Eu2+ are presented. Depending on the excitation wavelength, luminescence spectra consist of two distinct broad emission bands peaking at 380 nm and 490 nm related to 4f65d1 → 4f7 (8S7/2) luminescence of Eu2+ and to europium-trapped exciton, respectively, and/or several sharp lines between the 580 nm and 710 nm region, ascribed to the 5D0 → 7FJ (J = 0, 1, 2, 3 and 4) transitions in Eu3+. To explain all the features of the Eu2+ and Eu3+ luminescence we discussed the existence of two different Eu sites substituting for Li+, with short and long distance compensation. The evident effect of increasing the intensity of the Eu2+ luminescence with increasing temperature was observed. It was considered that the charge compensation mechanism for Eu3+ and Li+ as well as Eu2+ replacing Li+ in the LiMgPO4 is a long distance compensation that allows for the existence of some of the europium ions either as Eu3+ at low temperature or as Eu2+ at high temperature. We concluded that Eu2+ in the Li+ site with long distance compensation yields only 4f65d1 → 4f7 luminescence, whereas Eu2+ in the Li+ site with short distance compensation yields 4f65d1 → 4f7 luminescence and europium-trapped exciton emission. Keywords: LiMgPO4:Eu2+, Eu3+ phosphor, luminescence properties, time-resolved spectroscopy, long distance compensation model, the fermi level (Some figures may appear in colour only in the online journal)
typically originates from the intra-configurational 4f → 4f transitions which are almost independent of ligand field strength and even the effect of the local symmetry. Mostly, the room temperature emission of Eu3+ originates from the 5D0 state and locates about 595 nm (5D0 → 7F1) and 610‒620 nm (5D0 → 7F2). As a result, Eu3+ ion provides an orange-red luminescence. The luminescence properties of the Eu2+ depend strongly on the matrix. The energy of the lowest state of the 4f65d1 electronic configuration of Eu2+ in lattice is diminished with respect to the free ion by quantity of the centroid shift and crystal field splitting, therefore the energy of the 4f65d1 state is usually lower than the energy of the first excited state of the ground 4f7 electronic configuration 6P7/2 [7]. In such a case, the luminescence spectrum of the Eu2+ ions consists of the broad band related to the parity allowed 4f65d1 → 4f7 (8S7/2) transition. In lattices where
1. Introduction LiMgPO4 belongs to a large family of phosphates with formula ABPO4 where A and B are mono and divalent cations, respectively, with structures dependent on the relative size of the ions [1]. If the size of A and B are small, such as Li1+ and Mg2+, the resulting compound adopts the olivine structure [2]. The mono-phosphates, e.g. LiMPO4 (M = Mg, Mn, Co, Ni) have attracted attention due to their magnetoelectric properties [3]. LiMgPO4 doped with Tb and B have been investigated due to their potential application as dosimetry phosphors [4–6]. Among the rare earth elements, europium is an exceptional dopant because it can exist in divalent and trivalent states. Both these ions show many different characteristics as emitting centers. In the case of Eu3+, photoluminescence 0953-8984/14/385401+10$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401 6
6
1
energy of the P7/2 state is lower than the 4f 5d state, the emission spectrum consists of the sharp lines related to the 6 P7/2 → 8S7/2 transition. Such emission has been observed in some silicates [8], fluorites [9–12] and sulphates [13]. Eu2+ can potentially replace Li+ and Mg2+ in the LiMgPO4 lattices. The large difference between ionic radiuses of sixfold coordinated Eu2+ and Mg2+/Li+, which are equal to 1.17 Å and 0.76 Å/ 0.72 Å, respectively [14], causes the large lattice distortion [15], or can move the Eu2+ ions to the interstitial position. On the other hand, Eu3+ ion has an ionic radius equal to 0.947 Å and therefore it fits better to the Mg2+ /Li+ sites than Eu2+. As one expects, LiMgPO4 can contain simultaneously the Eu2+ and Eu3+ ions and the relative amount depends on synthesis conditions. The co-existence of Eu2+ and Eu3+ can be obtained by annealed material containing Eu2+ in ambient atmosphere, when energy necessary for the creation of defects compensating the Eu3+ in divalent or monovalent metal sites is smaller than the difference between energies of the ground states of Eu2+ (4f7 electronic configuration) and Eu3+ (4f6 electronic configuration [16]). Luminescence of Eu2+ and Eu3+ in the same materials has been observed in β-Ca2SiO4 [16], oxifluoride glass ceramics [17], borates [18, 19] and fluorides [20, 21]. In both glassy and crystalline materials, Eu2+ is achieved from the reduction of Eu3+ to Eu2+ [19, 22]. It has been shown that Eu2+ and Eu3+ coexist in Ca3Y2(SiO4)2 [23–25] where Eu2+ and Eu3+ occupy Ca2+ and Y3+ sites, respectively. Generally, luminescence of Eu3+ provides the red component to the blue or green luminescence of Eu2+. Therefore, materials doped simultaneously with Eu2+ and Eu3+ can be good candidates for white phosphors for light-emitting diodes (LEDs). This prompts us to study LiMgPO4 doped with Eu2+ and Eu3+. In this contribution, spectroscopic properties and the effect of temperature on the luminescence of the LiMgPO4 doped with Eu3+ and Eu2+ (mol 3%) are presented. We discuss the fundamental aspects of the existence of Eu3+ and Eu2+ in the LiMgPO4 related to the compensation mechanism, which in our opinion, different from the case of β-Ca2SiO4 [16], is also the long distance compensation.
Figure 1. (a) XRD patterns of the LiMgPO4; (b) crystal structure of the LiMgPO4.
in the reducing atmosphere were mixed and placed in the covered corundum crucible and the crucible was then buried in carbon sticks. Subsequently, the crucible was heated at 800 °C for 20 h with a reduction atmosphere. The final product was obtained after cooling to ambient temperature naturally. More details on the synthesis procedure can be found in [15, 26]. Sintered samples were characterized by x-ray powder diffraction (XRD) patterns using a Philips Xpert/MPD diffraction system with Cu Kα (λ = 1.5405 Å). By comparison of the XRD patterns of the samples with Joint Committee on Powder Diffraction Standards (JCPDS no. 32–0574), no impurities were detected and all the reflections could be well indexed to a LiMgPO4 single-phase, olivine-type orthorhombic structure with the space group Pnma and lattice parameters of a = 10.147 Å, b = 5.909 Å, c = 4.692 Å and Z = 4 [27]. The XRD pattern is presented in figure 1(a). LiMgPO4 crystal tetrahedral PO4 and octahedral LiO6, and MgO6 groups constitute the structure forming a 3D network with perpendicular tunnels along the [0 1 0] and [0 0 1] directions occupied by Li ions. The average bond distances Li-O and Mg-O are 2.143 and 2.105 Å within polyhedra LiO6 and MgO6, respectively. The Eu2+/Eu3+ ions are prone to occupy
2. Experimental techniques 2.1. Synthesis and structure
The samples of the LiMgPO4:Eu2+, Eu3+ were synthesized by the conventional solid-state reaction. The raw materials were LiCO3 (99.9%), 4(MgCO3), Mg(OH)2, 5H2O (magnesium carbonate basic pentahydrate, 99.9%), NH4H2PO4 (99.9%) and Eu2O3 (99.9%). The starting materials with stoichiometric amounts were mixed in a small covered corundum crucible. The Eu2O3 was added in an amount to obtain 0.5, 1 and 3 mol% of Eu in the sample. For further spectroscopic analysis, the sample containing 3 mol % of Eu was selected. The mixture was first heated to 300 °C and kept at this temperature for 5 h in air. After a second homogenization in the mortar with acetone, the sample was heated to 600 °C and kept at this temperature for 10 h in air. The samples prepared 2
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401 +
the Li sites on the account of the longer Li-O distance in LiMgPO4 and the fact that Li+ occupy the structural tunnels in the 3D network [15]. The crystal structure of the LiMgPO4 is presented in figure 1(b). 2.2. Spectroscopy
To perform spectroscopic experiments, a powdered sample was placed in a quartz cuvette type 210.003-QS produced by Hellma Analytics GmbH & Co. KG. The dimensions of the cell are as follows: height – 24 mm, width – 24 mm, light path – 0,1 ± 0,005 mm, volume – 18 μl. Steady-state luminescence and luminescence excitation spectra at room temperature were obtained using a Horiba JobinYvon FluoroMax–4 P spectrofluorimeter. The light source was an xenon lamp (150 Watts) coupled to a singlegrating monochromator operating in a wavelength range of 240–800 nm. Emission spectra were measured using a monochromator and side-on type Hamamatsu photomultiplier model R928. The experimental setup for luminescence kinetics and time-resolved emission spectra consists of a PL 2143 A/ SS laser and a PG 401/SH parametric optical generator as the excitation source. The system can generate 30 ps laser pulses, with the frequency of 10 Hz with wavelengths ranging from 220 nm to 2200 nm. The emission signal was analyzed by a Bruker Optics spectrometer model 2501 S equipped with a 50 grooves mm−1 grating and a Hamamatsu Streak Camera model C4334-01. The time-resolved spectra were collected by integration of the streak camera images over the time intervals, whereas luminescence decays were obtained by the integration of streak camera images over the wavelength intervals. Details of the experimental setup are described in [28]. The sample was cooled by a closed-cycle optical cryostat, which allows the temperature to be varied between 15 K and 600 K. The system consists of a water-cooled helium compressor model ARS-4HW and expander model DE-204 SI produced by Advanced Research System, Inc and a LakeShore temperature controller Model 336.
Figure 2. Room temperature emission and luminescence excitation spectra of the LiMgPO4:Eu2+, Eu3+. (a) Emission spectra excited at 260 nm (38462 cm–1) – curve 1, excited at 325 nm (30769 cm–1) – curve 2 and excited at 394 nm (25381 cm–1) – curve 3. (b) Excitation spectra of the emission monitored at 378 nm (26455 cm–1) related to the 4f65d1 → 4f7 (8S7/2) transition of Eu2+ – curve 1, monitored at 490 nm (20408 cm–1) related to ETE emission – curve 2 and monitored at 588 nm (17007 cm–1) related to the 5D0 → 7F1 transition of Eu3+ – curve 3. In the insert of figure 2(b) the terminal levels are indicated.
3. Spectroscopic results 3.1. Ambient temperature luminescence and luminescence excitation spectra
Luminescence and luminescence excitation spectra of the LiMgPO4 doped with Eu3+ and Eu2+ are presented in figure 2(a) and (b), respectively. It is seen in figure 2(a) that, depending on the excitation wavelength, LiMgPO4:Eu yields emission related to Eu2+ and/or Eu3+. Under excitation with 260 nm, the luminescence spectrum (see curve 1 in figure 2(a)) consists of two distinct bands peaking at 380 nm (20400 cm–1) and 490 nm (26300 cm−1), attributed to the 4f65d1 → 4f7 (8S7/2) transitions in Eu2+ ion and recombination of europium-trapped exciton (ETE) [29], respectively and several sharp lines, observed in the spectral region between 580 and 710 nm, ascribed to the 5 D0 → 7FJ (J = 0, 1, 2, 3, 4) transitions of Eu3+. The terminal
levels of the transitions label the spectral lines in figure 2(a). Under excitation with 325 nm, only luminescence related to the 4f65d1 → 4f7 transition in the Eu2+ and ETE emission is observed (curve 2 in figure 2(a)), whereas for excitation with 394 nm (curve 3 in figure 2(a)) only emission from Eu3+ is obtained. The excitation spectra are presented in figure 2(b). Curves 1 and 2 represent the excitation spectra of the emission monitored at 378 nm (26450 cm−1) and 490 nm (20400 cm−1), corresponding to the 4f65d1 → 4f7(8S7/2) transition in the Eu2+ and ETE emission, respectively. In both cases, spectra consist of 3
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 3. Luminescence of the LiMgPO4: Eu2+, Eu3+ under excitation with 260 nm obtained for different temperatures, collected in time
interval 0 – 10 μs (a) and (b) and collected in time interval 0–5 ms (c).
the structured bands with maximum at 275 nm (36360 cm−1) and the bump at 320 nm (31200 cm−1). The relative intensity of these features depends on the monitored wavelength. The bump at 320 nm is weaker and whole spectrum is shifted to the blue when violet emission (378 nm) is monitored. Usually, the ionization transition and creation of ETE are rather weakly represented in the excitation spectrum of Ln2+/3+ ions [30, 31], therefore both bands seen in the spectra should be attributed to the transition from the ground state to the excited state belonging to the 4f65d1 electronic configuration of Eu2+ ions. The excitation spectrum of the emission monitored at 588 nm (17000 cm–1) related to the 5D0 → 7F1 transition in the Eu3+ is represented by curve 3 in figure 2(b). The spectrum consists of a broad band which peaked at 253 nm (29500 cm−1) related to the charge transfer (CT) transition and the sharp lines which peaked at 270–320 nm, related to the 4f → 4f transitions in Eu3+. The respective 4f → 4f transitions are labelled in the inset of figure 2(b). One notices that weak bands attributed to excitation of Eu2+ are also seen in the excitation spectrum of
the Eu3+ emission that can be related to energy transfer from Eu2+ to Eu3+ ions. 3.2. Temperature dependence of the emission spectra and emission kinetics
The streak camera system was used to analyze temperature dependence of the luminescence spectra and luminescence kinetics. In figures 3(a–c),the luminescence spectra obtained under excitation 260 nm at different temperatures are presented. In figures 3(a) and (b) spectra collected in the time interval 0–10 μs are presented. In this time scale, emission related to Eu3+ is too weak to be observed, therefore only the features related to the Eu2+ are seen. In figure 3(a) the normalized spectra are presented. At temperatures lower than 200 K, the emission consists of three bands which peaked at 383 nm (26100 cm−1), 421 nm (23750 cm−1) and 512 nm (19530 cm−1). The bands which peaked at 383 nm and 512 nm are related to the 4f65d1 → 4f7 (8S7/2) transition 4
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Emission decay profiles of the luminescence excited with 260 nm, observed in the wavelength 370–390 nm, corresponded to the 4f65d1 → 4f7 (8S7/2) and ETE emission monitored at 480–510 nm, obtained at different temperatures, and are presented in figure 5(a) and (b), respectively. The decay profiles of the Eu3+ emission monitored at the 5D0 → 7F1 transition obtained at different temperatures are presented in figure 5(c). It is seen that decays of the 4f65d1 → 4f7 (8S7/2) emission presented in figure 5(a) are not single exponential. The profile changes in a non-monotonic way when the temperature increases. We decomposed these decays into two exponential components; the shorter is characterized by τ1 and the longer is characterized by τ2 using the formula:
Figure 4. Temperature dependence of IEu2+/IEu3+ emission for
LiMgPo4: Eu2+, Eu3+ obtained under excitation 260 nm.
AQ4
⎡ −t ⎤ ⎡ −t ⎤ I (t ) = A1exp ⎢ ⎥ + A2 exp ⎢ ⎥ ⎣ τ1 ⎦ ⎣ τ2 ⎦
(1)
The results of fitting the decay curves to relation (1) are presented in figure 6. It is seen that the shorter component τ1 decreases with increasing temperature, whereas the longer component τ2 increases in the temperature range between 100 K and 300 K from 1.8 µs to 3 µs, respectively and then decreases with increasing temperature for T > 300 K. The unusual temporal behavior of the 4f65d1 → 4f7 (8S7/2) luminescence can be related to the existence of shallow electron traps or the influence of higher-lying 6P7/2 state of the Eu2+ [15]. The time evolution of the ETE emission for different temperatures is presented in figure 5(b). For all temperatures, the decays are single exponential. The respective lifetime is equal 3.5 µs between 15 K and 300 K and then decreases with temperature to 0.25 µs at 600 K. It is seen that for temperatures lower than 300 K the ETE emission is delayed by time approximately equal to 1 µs. The luminescence rise is observed. For temperatures higher than 300 K the ETE emission appears almost at once after excitation. We described the ETE emission kinetics using the formula:
and recombination of the ETE [15, 29]. The origin of the emission band which peaked at 421 nm cannot be clearly assigned since such an emission was not reported before for the LiMgPO4:Eu2+ system. It can be related to the structure of the 4f65d1 electronic configuration as well as to the ETE. The two emission bands 383 nm and 421 nm could be also related to the existence of two different Eu2+ sites that replace Li+ and Mg2+, but we did not find strong arguments for such an assignation. The relative intensity of the band which peaked at 412 nm increases with temperature for the range from 15 K to 100 K and then diminishes and disappears at ambient temperature. When the temperature increased the ETE band peaked at 512 nm shifts to the shorter wavelengths, from 512 nm at 15 K to 490 nm at 500 K. Below 300 K, relative intensity of the ETE emission (with respect to the 4f65d1 → 4f7 (8S7/2) emission) does not depend on temperature, whereas for temperatures higher than 300 K the intensity of the ETE luminescence decreases. One can see this effect in figure 3(b) where non-normalized spectra measured at the temperature range 300–600 K are presented. In the inset of figure 3(b) the integrated intensity of the emission related to Eu2+ ions (4f65d1 → 4f7 and ETE) versus temperature is presented. The effect of the ETE emission quenching is related to the non-radiative recombination of ETE activated at higher temperatures. Luminescence spectra obtained at different temperatures collected in time interval 0–5 ms are presented in figure 3(c). The spectra are a superposition of Eu2+ luminescence represented by broad bands and Eu3+ luminescence represented by sharp lines. The spectra were normalized to the intensity of the 5D0 → 7F1 line. In figure 4 the ratio of the integrated luminescence intensities of Eu2+ to Eu3+ is presented. An evident effect of increasing the intensity of Eu2+ emission with respect to Eu3+ emission is observed for temperature range 50–250 K. The nature of this effect will be discussed later. The nonradiative intersystem crossing responsible for decreasing the relative intensity of the Eu2+ luminescence for temperature range 300–500 K is also responsible for decreasing the Eu2+ to Eu3+ luminescence intensity ratio for temperatures higher than 250 K.
⎡ −t ⎤ ⎡ −t ⎤ I (t ) = Adec exp ⎢ ⎥ − Arise exp ⎢ ⎥ ⎣ τdec ⎦ ⎣ τrise ⎦
(2)
The fitted quantities of the decay time τdec and the rise time τrise for different temperatures are presented in figure 6. Since the ETE emission rise time is approximately equal to the shorter component of the 4f65d1 → 4f7 (8S7/2) luminescence decay, we can consider the effect of nonradiative energy transfer from the 4f65d1 state to ETE state. The energy transfer process is seen more evidently when one compares the decay profiles of the emission monitored at 480–510 nm (ETE) and 370–390 nm (4f65d1 → 4f7) obtained for different temperatures (see figure 7). For 15 K the decay of the band 410–430 nm is additionally presented. One notices that decay of the emission band 410–430 nm is more similar to the decay of the ETE than to the decay of the 4f65d1 → 4f7 (8S7/2) luminescence. From figure 7 it is seen that at low temperature the ETE decay time is longer than longer component of the 4f65d1 → 4f7 (8S7/2) luminescence decay whereas at 300 K we have the opposite situation. Similar dependence of the Eu2+ luminescence on temperature was obtained for luminescence excited under 5
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 5. Decay profiles of emission for selected temperatures obtained under excitation λexc = 260 nm; (a) 4f65d luminescence (b) the ETE luminescence; (c) the Eu3+ luminescence related to the 5D0 → 7F1 transition. Results of fitting are presented by solid curves.
Figure 6. Dependence of the emission parameters of the luminescence kinetics, the decay and rise times, on temperature.
Figure 7. Comparison of the decays of the ETE emission and 4f65d1 → 4f7 (8S7/2) emission obtained at 15 K, 300 K and 600 K. For 15 K the decay of the band 410–430 nm is additionally presented.
320 nm and 355 nm. The only difference was that contribution from the band which peaked at 421 nm observed at temperature range between 100 K and 200 K diminishes with increasing excitation wavelength. The luminescence decay profiles of the 5D0 → 7F1 transition in Eu3+ obtained under 260 nm excitation are presented in figure 5(c). The decays are almost single exponential below 50 K and above 350 K. At medium temperatures the decays are nonexponential and faster. Under excitation 394 nm, which corresponds to the 7F0 → 5L6 transition in Eu3+, only the Eu3+ emission is seen. The emissions obtained at different temperatures are presented in figure 8(a). The luminescence decay profiles of the emission related to 5D0 → 7F1 transition obtained at different temperatures under excitation with 394 nm are presented in figure 8(b). One notices that luminescence decay does not depend on temperature. This is a different situation to when Eu3+ luminescence was excited through the CT transition with 260 nm. The specific temporal behavior of the luminescence related to the 5D0 → 7F1 transition under excitation by the CT process may be related to thermally-induced conversion excitation of the Eu3+ which involves the higher excited states 5GJ, 5HJ. However, the existence of such effects should be additionally confirmed.
4. Discussion The ETE luminescence under excitation with 260 nm is delayed in time. The same is observed under excitation with 325 nm. The emission related to the 4f65d1 → 4f7 (8S7/2) transition appears just after excitation also independently on excitation wavelength. Such phenomena can take place when ETE state is not excited directly, but obtains the excitation energy from the 4f65d1 state. To analyze such a process we constructed the one-dimensional configurational coordinate diagram describing the ground state 4f7 (8S7/2), the 4f65d1 state of Eu2+ and the ETE state. The diagram is presented in figure 9(a). To present the electronic energies of the Eu2+ system in the ETE, in the 4f65d1 and in the ground state we used following formulas:
Ei (Q ) = Ei0 + k
[Q − Qi ]2 2
(3)
where subscript i = g, ETE and d, labels the ground state, the ETE state and the 4f65d1 state, respectively, k is effective elastic constant of the lattice. The quantity of the configurational 6
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 9. (a) Configurational coordinate diagram of Eu2+ in one
dimensional approximation obtained from the emission spectra. (b) Possible relative positions of the minima of the excited electronic manifolds representing 4f65d1 state and ETE state in the many-dimensional configurational space.
the strength of the electron-lattice coupling. We used emission/absorption spectra to quantity the diagram and obtained −1 0 = 26900 cm−1 and S following data EETE ETE ℏω = 6500 cm , −1 −1 0 Ed = 28200 cm and SETE ℏω = 1500 cm and 0 = 1300 cm−1. The feeding of the ETE state ΔE = Ed0 − EETE from the excited 4f65d1 state is presented by dashed arrows. One notices that when one-dimensional lattice distortion is assumed the shift in the configurational coordinate space between the minima of the ETE and the 4f65d1 electronic ⎞ ⎛ manifolds ΔQ = Qd − QETE ⎜where Q i = 2Siℏω k ⎟ is too ⎝ ⎠ small to create the energy barrier which is responsible for delaying the ETE emission in time. On the other hand, the single configurational coordinate model probably does not work since the electron in the d state and in the ETE state have probably different symmetry. As a result, different vibration modes are involved in lattice distortion after excitation of the system into the ETE and the 4f65d1 states. When these modes are an orthogonal quantity ΔQ can be expressed as follows:
Figure 8. (a) The luminescence spectra of the LiMgPo4: Eu2+,
Eu3+, under excitation λexc = 394 nm at different temperatures; (b) Decay profiles of emission related to the 5D0 → 7F1 transition of Eu3+ for selected temperatures, under excitation λexc = 394 nm
coordinate Q represents the changes of the distance between Eu ion and ligands, Qi is the value of configurational coordinates that corresponds to the minimum energy of the respective state. Further, we assumed that Qg = 0 and E0g = 0. Under this assumption, one obtained that energy of lattice relaxation of the system in the state i is equal to:
Siℏω = k
Qi2 2
(4)
and the zero – phonon energy of i → g of radiative transition is equal to Ei0, Si is the Huang-Rhys Factor that determines 7
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
ΔQ =
2 Qd2 − QETE
transition and (plus) energy of ETE emission (2.5 eV). We estimated the bandgap of the LiMgPO4, Eg > 7 eV. Europium is located in the Li+ site octahedrally coordinated by oxygen ions and can exist as Eu2+ or Eu3+. Such a feature needs the creation of charge compensation defects. Usually, it is assumed that they are vacancies of Li+ or Mg2+ or Li1+ in Mg2+ site. It is a very reasonable assumption that independent of where Eu2+ and Eu3+ are located in LiMgPO4:Eu2+, Eu3+ all charges are compensated totally. One can consider two possible ways of compensation: the case when all Eu2+ and Eu3+ ions in Li+ sites are correlated spatially with respective compensation defects and the case when impurity ions and compensation defects are randomly distributed in crystal. As has been shown in [16] for the β-Ca2SiO4:Eu2+, Eu3+ when Eu3+ in Ca2+ site is compensated at the short distance, Eu3+ ions and compensation defects are correlated spatially and temperature does not influence the Eu2+/ Eu3+ emission intensity ratio. In the case of the LiMgPO4:Eu2+, Eu3+ system intensity of the Eu2+ increases with increasing temperature. One should consider this dependence by means of the Fermi energy. It has been discussed in paper [16] that when Eu2+ occupies the Me2+ site in perfect crystal the Fermi energy is located in the bandgap above energy of the Eu2+ ground state. In the case when Eu2+ ions occupy Li+ the Fermi energy can be even higher than ground state of Eu2+ when energy of compensation defect (Li+ in Mg2+ site or Li+ vacancy) is higher than energy of the Eu2+. In the case when Eu3+ replaces Me2+ in the lattice creation of compensation defects, the Fermi energy diminishes by the quantity:
(5)
The above relation was used to draw the diagram presented in figure 9(b). As can be seen from this figure, the nonradiative transition from the 4f65d1 state to the ETE state needs the activation energy and therefore the emission from the ETE state is delayed in time. Here it should be noted that analysis of the luminescence kinetics and dependence of the luminescence spectra on temperature allow us to conclude that we have two different Eu2+ sites. In the first type of site, the ETE state is located below the 4f65d1 state and the 4f65d1 state is additionally non-radiatively depopulated to the ETE state, as it is presented in figure 9 (a) and (b). These sites contribute to the fast component of the 4f65d1 luminescence and the ETE luminescence. In the second type of site, the energy of the ETE is higher than the energy of the 4f65d1 or ETE does not exist. These sites yield the slower component of the 4f65d1 emission. Also, only these sites are observed at higher temperatures for which the ETE luminescence is quenched. Although the ion-ligand distance is different in the Li+ and Mg2+ sites, we do not observe two different bands related to the 4f65d1 → 4f7 transitions in different sites. It can mean that only the Li+ site is occupied. Such occupation needs the creation of a compensation defect, which can be Li+ ion in Mg2+ site (Li/Mg) or Li vacancy VLi. The location of the compensation defect with respect to Eu2+ ion in lattice can influence the energy and also the existence of the ETE states. When the compensation defect is located near the Eu2+ site the ETE is created as the Eu3+ and electron bounded by long distance Coulomb potential created by difference V(Eu3+)–V(Li+ +Li/ Mg). In this case, the potential at a distance r from the system is equal −e/εr, where ε is static dielectric constant and e is electron charge. Such relatively weak potential can bind electrons which are located outside the first coordination sphere (see the model presented in [33]). If the compensation defect is far from Eu2+, substituting the respective Coulomb potential for Li+, V(Eu3+)–V(Li+) is equal to −2e/εr (it is two times stronger) and therefore the system probably will not create the stable state with electrons outside the first coordination sphere of Eu. In such a situation, the existence of stable ETE state is doubtful. It is observed that intensity of the Eu2+ luminescence increases with respect to the Eu3+ luminescence when the temperature increases from 100 K to 250 K. To understand this effect one should consider the energetic structure of the LiMgPO4:Eu system. The energy of CT transition seen in the excitation spectrum of Eu3+ luminescence allows us to estimate that the ground state of Eu2+ is located 4.5 eV above the top of the valence band. The ground state of the Eu3+ should be located between 5 eV and 8 eV below the ground state of Eu2+ [32]. The ground state and the excited state 4f65d1 of Eu2+ are located below the conduction band. The separation between the ground state of the Eu2+ and the conduction band is equal to the ionization energy and is greater than the energy of the ETE. One can estimate the lower limit of the energy bandgap of the LiMgPO4 as equal to energy of CT (4.5 eV)
ΔEi = E (Eu2 +) − E (Eu3 +) − Ui 2+
3+
(6)
where ΔE = E (Eu ) − E (Eu ) is the difference between energies of the ground states of Eu2+ and Eu3+ and Ui is the energy necessary for formation of i-th defect compensation. For short distance compensation, when the defect is located near the Eu3+ ion the compensation defect charge causes the bands to bend and leave the Eu2+ unoccupied [16]. For long distance compensation energy of the ground state of Eu2+ with respect to the Fermi level depends on the distance from the compensation defect(s). When Eu3+ replaces Li+ the situation is more complicated. The additional charge can be compensated by two Li+ ions replacing Mg2+ two Li+ vacancy or single Mg2+ vacancy. The energy of the Eu2+ and Eu3+ ground states with respect to the Fermi energy can be different for different sites and depends on actual spatial distribution of compensation defects. In figure 10 the energetic structure of the LiMgPO4:Eu2+, Eu3+ system for the assumption of random distribution of the compensation defects is presented. Here, three compensation defects (one VLi and two VMg) with potential to cause the band bends and four Eu ions placed in Li+ sites are presented. In sites labelled (a) energy of the ground state of Eu2+ is lower than the Fermi energy; these sites contribute always to the Eu2+ luminescence. Eu in the site labelled (b) near the VMg compensation defect and therefore the energy of the Fermi level is much lower than energy of the ground state of Eu2+, ΔE1 >> kT, (for ambient temperature); therefore, Eu2+ state is unoccupied and this site contributes always to the Eu3+ luminescence. Eu in the site labelled (c) is further from the 8
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
In the previous paper, it has been discussed that Eu3+ ions in LiMgPO4 substitute for Li+ and form two different sites [26]. Presented results have shown also that Eu2+ substitutes for Li+ in two different sites. It has been discussed that Eu2+ in Li+ site with short distance charge compensation contributes to the luminescence due to the 4f65d1 → 4f7 (8S7/2) and anomalous emission related to ETE recombination, whereas Eu2+ in Li+ sites with long distance compensation yields only the 4f65d1 → 4f7 (8S7/2) emission. The configurational coordinate diagram has been developed which describes the luminescence kinetics of the Eu2+ in Li+ site by means of electronlattice coupling which takes place with different lattice modes for the system in the 4f65d1 and ETE states. We also discussed that charge compensation defects necessary for the existence of Eu2+ in Li+ sites as well as Eu3+ in Li+ sites are not associated with individual europium ions but are rather randomly distributed in the lattice. The energies of Eu ions in the Eu2+ and Eu3+ charge state and energies of the band edges respect energy of the Fermi level in the specific Eu site depend on the potential of randomly distributed compensation defects ( VLi, VMg and Li/Mg). As a result, for part of Eu ions the Fermi energy is slightly below the energy of the Eu2+ ground state. For these sites, the actual charge of Eu ions is the result of a thermodynamic process (thermal excitation of Eu3+ to Eu2+). As a result of increasing the population of the Eu2+, the intensity of the Eu2+ luminescence increases with an increase in temperature.
Figure 10. Energetic structure of the LiMgPO4:Eu2+, Eu3+
system obtained under assumption of random distribution of the compensation defects (a), (b) and (c) represent the Eu ion in Li+ sites at different distance from compensation defect (for details see the main text).
compensation defect VMn than site (b), therefore the energy of the Fermi level is not far from the ground state of Eu2+. For temperature T ≈ ΔE2/k the mission from this site will depend on temperature. The total intensity of the Eu2+ luminescence is proportional to the following sum over all the Eu ions
IEu2+ = ∑ m
1 ⎡ (E m 2+ − EF ) ⎤ ⎥ 1 + exp ⎢ Eu kT ⎢⎣ ⎥⎦
Acknowledgements (7)
This study was supported by the POIG.01.01.02-02-006/09 project co-funded by the European Regional Development Fund within the Innovative Economy Program (Priority I, Activity 1.1. Sub-activity 1.1.2), which is gratefully acknowledged. A B was additionally supported by the European Social Fund as part of the project ‘Educators for the elite - integrated training program for PhD students, post-docs and professors as academic teachers at University of Gdansk’ within the framework of Human Capital Operational Program, Action IV and the grant PMN No. 538-5200-B160-13. This work was also supported under the framework of the international cooperation program managed by the National Research Foundation of Korea (2013K2A1A2055224).
m − E is the difference between energy of Eu2+ where EEu 2+ F ground state and energy of the Fermi level near m-th site. m The sites where this energy is negativeEEu 2 + − EF < 0 yield 2+ always the Eu emission (like sites labelled by a in figure 10), 3+ m the sites where EEu lumines2 + − EF ≫ kT yield always Eu cence (like site b in figure 10), whereas emission of the sites m − E is comparable to kT (like site c in \ 10) for which EEu 2+ F depend on temperature and for these sites intensity of the Eu2+ luminescence increases when temperature increases. One notices that at temperatures higher than ambient the ETE emission is quenched, whereas the 4f65d1 → 4f7 (8S7/2) emission is more stable. The temperature quenching of the ETE luminescence, accompanied by shortening of the luminescence lifetime, is related to nonratiative intersystem crossing the ETE and the ground electronic manifold caused by large electron-lattice coupling. It is not the case of the 4f65d1 → 4f7 (8S7/2) luminescence which weakly depends on temperature. It should be noted here that temperature stability of the 4f65d1 → 4f7 (8S7/2) emission concerns only the sites for which ETE state has higher energy than the 4f65d1 state or the ETE state does not exist.
References [1] Amara M B, Vlasse M, Flem G L and Hagenmuller P 1983 Acta Crystallogr. C39 1483 [2] Heine F, Handlovic M, Burdova K and Mailing J 1982 J. Crystallogr. Spectrosc. Res. 12 99 [3] Goňi A, Bonagamba T J, Silva M A, Panepucci H, Rojo T and Barberisa G E 1998 J. Appl. Phys. 84 416 [4] Dhabekar B, Menon S N, Alagu Raja E, Bakshi A K, Singh A K, Chougaonkar M P and Mayya Y S 2011 Nucl. Instrum. Methods Phys. Res. B 269 1844 [5] Kumar M, Dhabekar B, Menon S N, Chougaonkar M P and Mayya Y S 2011 Nucl. Instrum. Methods Phys. Res. B 269 1849
5. Conclusions Luminescence spectra and luminescence kinetics of the LiMgPO4:Eu2+, Eu3+ have been measured and discussed. 9
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
[21] Baby T and Nampoori V P N 1992 Solid State Commun. 81 367 [22] Lian Z, Wang J, Lv Y, Wang S and Su Q 2007 J. Alloys Compounds 430 257 [23] Dobrowolska A and Zych E 2012 MRS Symp. Proc. 1342 55 [24] Dobrowolska A and Zych E 2012 J. Phys. Chem. C 116 25493 [25] Baran A, Mahlik S, Grinberg M and Zych E 2013 J. Phys.: Condens. Matter 25 02560 [26] Zhang S, Huang Y, Shi L and Seo H J 2010 J. Electrochem. Soc. 157 J186 [27] Hanic F, Handlovic M, Burdova K and Majling J 1982 J. Crystallogr. Spectrosc. Res. 12 99 [28] Kubicki A, Bojarski P, Grinberg M, Sadownik M, Kukliński B 2006 Opt. Commun. 269 275 [29] Seo H J 2013 J. Ceram. Process. Res. 14 22 [30] Mahlik S, Lazarowska A, Grobelna B and Grinberg M 2012 J. Phys.: Condens. Matter 24 485501 [31] Mahlik S, Lazarowska A, Speghini A, Bettinelli M and Grinberg M 2014 J Lumin. 152 62 [32] Dorenbos P 2003 J. Phys.: Condens. Matter 15 841 [33] Grinberg M 2013 Opt. Mater. 35 2006
[6] Gai M, Chen Z, Fan Y and Wang J 2013 J. Rare Earth 31 551 [7] Dorenbos P 2003 J. Lumin. 104 239 [8] Vestegen J M P J and Sommerdijk J L 1974 J. Lumin. 9 297 [9] Hoffman M 1972 J. Electrochem. Soc. 119 905 [10] Mahlik S, Grinberg M, Shi L and Seo H J 2009 J. Phys.: Condens. Matter 21 235603 [11] Su H, Jia Z, Shi C, Xin J and Reid S A 2001 Chem. Phys. Lett. 335 17–22 [12] Latourrette B, Guillen F and Fouassier C 1979 Matter Res. Bull. 14 865 [13] Yamashita N, Yamamoto I, Ninugawa K, Wada T, Yamashita Y and Nakao Y 1985 Japan. J. Appl. Phys. 24 1174 [14] Shannon R D 1976 Acta Crystallogr. A32 751 [15] Zhang S, Huang Y, Shi L, Shi Land Seo H J 2010 J. Phys.: Condens. Matter 22 235402 [16] Baran A, Barzowska J, Grinberg M, Mahlik S, Szczodrowski K and Zorenko Y 2013 Opt. Mater. 35 2107 [17] Biswas K, Sontakke A D and Annapurna K 2012 Int. J. Appl. Glass Sci. 3 154 [18] Reddy A, Das S, Ahmad S, Babu SS, Ferreira J M F and Prakash G V 2012 RSC Adv. 2 8768 [19] Pei Z, Su Q and Zhang Z 1993 J. Alloys Compounds 198 51 [20] Hong B C and Kawano K 2006 J. Alloys Compounds 408 838
10
Search
Collections
Journals
About
Contact us
My IOPscience
Luminescence properties of different Eu sites in LiMgPO4:Eu2+, Eu3+
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 J. Phys.: Condens. Matter 26 385401 (http://iopscience.iop.org/0953-8984/26/38/385401) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 130.92.9.56 This content was downloaded on 12/09/2014 at 16:07
Please note that terms and conditions apply.
Journal of Physics: Condensed Matter J. Phys.: Condens. Matter 26 (2014) 385401 (10pp)
doi:10.1088/0953-8984/26/38/385401
Luminescence properties of different Eu sites in LiMgPO4:Eu2+, Eu3+ A Baran1, S Mahlik1, M Grinberg1, P Cai2, S I Kim2 and H J Seo2 1
Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80–952 Gdansk, Poland Department of Physics and Center for Marine-Integrated Biomedical Technology, Pukyong National University, Busan 608–737, Republic of Korea 2
E–mail: [email protected] Received 3 June 2014, revised 23 July 2014 Accepted for publication 5 August 2014 Published 2 September 2014 Abstract
The effect of temperature on the luminescence properties of LiMgPO4 doped with Eu3+ and Eu2+ are presented. Depending on the excitation wavelength, luminescence spectra consist of two distinct broad emission bands peaking at 380 nm and 490 nm related to 4f65d1 → 4f7 (8S7/2) luminescence of Eu2+ and to europium-trapped exciton, respectively, and/or several sharp lines between the 580 nm and 710 nm region, ascribed to the 5D0 → 7FJ (J = 0, 1, 2, 3 and 4) transitions in Eu3+. To explain all the features of the Eu2+ and Eu3+ luminescence we discussed the existence of two different Eu sites substituting for Li+, with short and long distance compensation. The evident effect of increasing the intensity of the Eu2+ luminescence with increasing temperature was observed. It was considered that the charge compensation mechanism for Eu3+ and Li+ as well as Eu2+ replacing Li+ in the LiMgPO4 is a long distance compensation that allows for the existence of some of the europium ions either as Eu3+ at low temperature or as Eu2+ at high temperature. We concluded that Eu2+ in the Li+ site with long distance compensation yields only 4f65d1 → 4f7 luminescence, whereas Eu2+ in the Li+ site with short distance compensation yields 4f65d1 → 4f7 luminescence and europium-trapped exciton emission. Keywords: LiMgPO4:Eu2+, Eu3+ phosphor, luminescence properties, time-resolved spectroscopy, long distance compensation model, the fermi level (Some figures may appear in colour only in the online journal)
typically originates from the intra-configurational 4f → 4f transitions which are almost independent of ligand field strength and even the effect of the local symmetry. Mostly, the room temperature emission of Eu3+ originates from the 5D0 state and locates about 595 nm (5D0 → 7F1) and 610‒620 nm (5D0 → 7F2). As a result, Eu3+ ion provides an orange-red luminescence. The luminescence properties of the Eu2+ depend strongly on the matrix. The energy of the lowest state of the 4f65d1 electronic configuration of Eu2+ in lattice is diminished with respect to the free ion by quantity of the centroid shift and crystal field splitting, therefore the energy of the 4f65d1 state is usually lower than the energy of the first excited state of the ground 4f7 electronic configuration 6P7/2 [7]. In such a case, the luminescence spectrum of the Eu2+ ions consists of the broad band related to the parity allowed 4f65d1 → 4f7 (8S7/2) transition. In lattices where
1. Introduction LiMgPO4 belongs to a large family of phosphates with formula ABPO4 where A and B are mono and divalent cations, respectively, with structures dependent on the relative size of the ions [1]. If the size of A and B are small, such as Li1+ and Mg2+, the resulting compound adopts the olivine structure [2]. The mono-phosphates, e.g. LiMPO4 (M = Mg, Mn, Co, Ni) have attracted attention due to their magnetoelectric properties [3]. LiMgPO4 doped with Tb and B have been investigated due to their potential application as dosimetry phosphors [4–6]. Among the rare earth elements, europium is an exceptional dopant because it can exist in divalent and trivalent states. Both these ions show many different characteristics as emitting centers. In the case of Eu3+, photoluminescence 0953-8984/14/385401+10$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401 6
6
1
energy of the P7/2 state is lower than the 4f 5d state, the emission spectrum consists of the sharp lines related to the 6 P7/2 → 8S7/2 transition. Such emission has been observed in some silicates [8], fluorites [9–12] and sulphates [13]. Eu2+ can potentially replace Li+ and Mg2+ in the LiMgPO4 lattices. The large difference between ionic radiuses of sixfold coordinated Eu2+ and Mg2+/Li+, which are equal to 1.17 Å and 0.76 Å/ 0.72 Å, respectively [14], causes the large lattice distortion [15], or can move the Eu2+ ions to the interstitial position. On the other hand, Eu3+ ion has an ionic radius equal to 0.947 Å and therefore it fits better to the Mg2+ /Li+ sites than Eu2+. As one expects, LiMgPO4 can contain simultaneously the Eu2+ and Eu3+ ions and the relative amount depends on synthesis conditions. The co-existence of Eu2+ and Eu3+ can be obtained by annealed material containing Eu2+ in ambient atmosphere, when energy necessary for the creation of defects compensating the Eu3+ in divalent or monovalent metal sites is smaller than the difference between energies of the ground states of Eu2+ (4f7 electronic configuration) and Eu3+ (4f6 electronic configuration [16]). Luminescence of Eu2+ and Eu3+ in the same materials has been observed in β-Ca2SiO4 [16], oxifluoride glass ceramics [17], borates [18, 19] and fluorides [20, 21]. In both glassy and crystalline materials, Eu2+ is achieved from the reduction of Eu3+ to Eu2+ [19, 22]. It has been shown that Eu2+ and Eu3+ coexist in Ca3Y2(SiO4)2 [23–25] where Eu2+ and Eu3+ occupy Ca2+ and Y3+ sites, respectively. Generally, luminescence of Eu3+ provides the red component to the blue or green luminescence of Eu2+. Therefore, materials doped simultaneously with Eu2+ and Eu3+ can be good candidates for white phosphors for light-emitting diodes (LEDs). This prompts us to study LiMgPO4 doped with Eu2+ and Eu3+. In this contribution, spectroscopic properties and the effect of temperature on the luminescence of the LiMgPO4 doped with Eu3+ and Eu2+ (mol 3%) are presented. We discuss the fundamental aspects of the existence of Eu3+ and Eu2+ in the LiMgPO4 related to the compensation mechanism, which in our opinion, different from the case of β-Ca2SiO4 [16], is also the long distance compensation.
Figure 1. (a) XRD patterns of the LiMgPO4; (b) crystal structure of the LiMgPO4.
in the reducing atmosphere were mixed and placed in the covered corundum crucible and the crucible was then buried in carbon sticks. Subsequently, the crucible was heated at 800 °C for 20 h with a reduction atmosphere. The final product was obtained after cooling to ambient temperature naturally. More details on the synthesis procedure can be found in [15, 26]. Sintered samples were characterized by x-ray powder diffraction (XRD) patterns using a Philips Xpert/MPD diffraction system with Cu Kα (λ = 1.5405 Å). By comparison of the XRD patterns of the samples with Joint Committee on Powder Diffraction Standards (JCPDS no. 32–0574), no impurities were detected and all the reflections could be well indexed to a LiMgPO4 single-phase, olivine-type orthorhombic structure with the space group Pnma and lattice parameters of a = 10.147 Å, b = 5.909 Å, c = 4.692 Å and Z = 4 [27]. The XRD pattern is presented in figure 1(a). LiMgPO4 crystal tetrahedral PO4 and octahedral LiO6, and MgO6 groups constitute the structure forming a 3D network with perpendicular tunnels along the [0 1 0] and [0 0 1] directions occupied by Li ions. The average bond distances Li-O and Mg-O are 2.143 and 2.105 Å within polyhedra LiO6 and MgO6, respectively. The Eu2+/Eu3+ ions are prone to occupy
2. Experimental techniques 2.1. Synthesis and structure
The samples of the LiMgPO4:Eu2+, Eu3+ were synthesized by the conventional solid-state reaction. The raw materials were LiCO3 (99.9%), 4(MgCO3), Mg(OH)2, 5H2O (magnesium carbonate basic pentahydrate, 99.9%), NH4H2PO4 (99.9%) and Eu2O3 (99.9%). The starting materials with stoichiometric amounts were mixed in a small covered corundum crucible. The Eu2O3 was added in an amount to obtain 0.5, 1 and 3 mol% of Eu in the sample. For further spectroscopic analysis, the sample containing 3 mol % of Eu was selected. The mixture was first heated to 300 °C and kept at this temperature for 5 h in air. After a second homogenization in the mortar with acetone, the sample was heated to 600 °C and kept at this temperature for 10 h in air. The samples prepared 2
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401 +
the Li sites on the account of the longer Li-O distance in LiMgPO4 and the fact that Li+ occupy the structural tunnels in the 3D network [15]. The crystal structure of the LiMgPO4 is presented in figure 1(b). 2.2. Spectroscopy
To perform spectroscopic experiments, a powdered sample was placed in a quartz cuvette type 210.003-QS produced by Hellma Analytics GmbH & Co. KG. The dimensions of the cell are as follows: height – 24 mm, width – 24 mm, light path – 0,1 ± 0,005 mm, volume – 18 μl. Steady-state luminescence and luminescence excitation spectra at room temperature were obtained using a Horiba JobinYvon FluoroMax–4 P spectrofluorimeter. The light source was an xenon lamp (150 Watts) coupled to a singlegrating monochromator operating in a wavelength range of 240–800 nm. Emission spectra were measured using a monochromator and side-on type Hamamatsu photomultiplier model R928. The experimental setup for luminescence kinetics and time-resolved emission spectra consists of a PL 2143 A/ SS laser and a PG 401/SH parametric optical generator as the excitation source. The system can generate 30 ps laser pulses, with the frequency of 10 Hz with wavelengths ranging from 220 nm to 2200 nm. The emission signal was analyzed by a Bruker Optics spectrometer model 2501 S equipped with a 50 grooves mm−1 grating and a Hamamatsu Streak Camera model C4334-01. The time-resolved spectra were collected by integration of the streak camera images over the time intervals, whereas luminescence decays were obtained by the integration of streak camera images over the wavelength intervals. Details of the experimental setup are described in [28]. The sample was cooled by a closed-cycle optical cryostat, which allows the temperature to be varied between 15 K and 600 K. The system consists of a water-cooled helium compressor model ARS-4HW and expander model DE-204 SI produced by Advanced Research System, Inc and a LakeShore temperature controller Model 336.
Figure 2. Room temperature emission and luminescence excitation spectra of the LiMgPO4:Eu2+, Eu3+. (a) Emission spectra excited at 260 nm (38462 cm–1) – curve 1, excited at 325 nm (30769 cm–1) – curve 2 and excited at 394 nm (25381 cm–1) – curve 3. (b) Excitation spectra of the emission monitored at 378 nm (26455 cm–1) related to the 4f65d1 → 4f7 (8S7/2) transition of Eu2+ – curve 1, monitored at 490 nm (20408 cm–1) related to ETE emission – curve 2 and monitored at 588 nm (17007 cm–1) related to the 5D0 → 7F1 transition of Eu3+ – curve 3. In the insert of figure 2(b) the terminal levels are indicated.
3. Spectroscopic results 3.1. Ambient temperature luminescence and luminescence excitation spectra
Luminescence and luminescence excitation spectra of the LiMgPO4 doped with Eu3+ and Eu2+ are presented in figure 2(a) and (b), respectively. It is seen in figure 2(a) that, depending on the excitation wavelength, LiMgPO4:Eu yields emission related to Eu2+ and/or Eu3+. Under excitation with 260 nm, the luminescence spectrum (see curve 1 in figure 2(a)) consists of two distinct bands peaking at 380 nm (20400 cm–1) and 490 nm (26300 cm−1), attributed to the 4f65d1 → 4f7 (8S7/2) transitions in Eu2+ ion and recombination of europium-trapped exciton (ETE) [29], respectively and several sharp lines, observed in the spectral region between 580 and 710 nm, ascribed to the 5 D0 → 7FJ (J = 0, 1, 2, 3, 4) transitions of Eu3+. The terminal
levels of the transitions label the spectral lines in figure 2(a). Under excitation with 325 nm, only luminescence related to the 4f65d1 → 4f7 transition in the Eu2+ and ETE emission is observed (curve 2 in figure 2(a)), whereas for excitation with 394 nm (curve 3 in figure 2(a)) only emission from Eu3+ is obtained. The excitation spectra are presented in figure 2(b). Curves 1 and 2 represent the excitation spectra of the emission monitored at 378 nm (26450 cm−1) and 490 nm (20400 cm−1), corresponding to the 4f65d1 → 4f7(8S7/2) transition in the Eu2+ and ETE emission, respectively. In both cases, spectra consist of 3
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 3. Luminescence of the LiMgPO4: Eu2+, Eu3+ under excitation with 260 nm obtained for different temperatures, collected in time
interval 0 – 10 μs (a) and (b) and collected in time interval 0–5 ms (c).
the structured bands with maximum at 275 nm (36360 cm−1) and the bump at 320 nm (31200 cm−1). The relative intensity of these features depends on the monitored wavelength. The bump at 320 nm is weaker and whole spectrum is shifted to the blue when violet emission (378 nm) is monitored. Usually, the ionization transition and creation of ETE are rather weakly represented in the excitation spectrum of Ln2+/3+ ions [30, 31], therefore both bands seen in the spectra should be attributed to the transition from the ground state to the excited state belonging to the 4f65d1 electronic configuration of Eu2+ ions. The excitation spectrum of the emission monitored at 588 nm (17000 cm–1) related to the 5D0 → 7F1 transition in the Eu3+ is represented by curve 3 in figure 2(b). The spectrum consists of a broad band which peaked at 253 nm (29500 cm−1) related to the charge transfer (CT) transition and the sharp lines which peaked at 270–320 nm, related to the 4f → 4f transitions in Eu3+. The respective 4f → 4f transitions are labelled in the inset of figure 2(b). One notices that weak bands attributed to excitation of Eu2+ are also seen in the excitation spectrum of
the Eu3+ emission that can be related to energy transfer from Eu2+ to Eu3+ ions. 3.2. Temperature dependence of the emission spectra and emission kinetics
The streak camera system was used to analyze temperature dependence of the luminescence spectra and luminescence kinetics. In figures 3(a–c),the luminescence spectra obtained under excitation 260 nm at different temperatures are presented. In figures 3(a) and (b) spectra collected in the time interval 0–10 μs are presented. In this time scale, emission related to Eu3+ is too weak to be observed, therefore only the features related to the Eu2+ are seen. In figure 3(a) the normalized spectra are presented. At temperatures lower than 200 K, the emission consists of three bands which peaked at 383 nm (26100 cm−1), 421 nm (23750 cm−1) and 512 nm (19530 cm−1). The bands which peaked at 383 nm and 512 nm are related to the 4f65d1 → 4f7 (8S7/2) transition 4
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Emission decay profiles of the luminescence excited with 260 nm, observed in the wavelength 370–390 nm, corresponded to the 4f65d1 → 4f7 (8S7/2) and ETE emission monitored at 480–510 nm, obtained at different temperatures, and are presented in figure 5(a) and (b), respectively. The decay profiles of the Eu3+ emission monitored at the 5D0 → 7F1 transition obtained at different temperatures are presented in figure 5(c). It is seen that decays of the 4f65d1 → 4f7 (8S7/2) emission presented in figure 5(a) are not single exponential. The profile changes in a non-monotonic way when the temperature increases. We decomposed these decays into two exponential components; the shorter is characterized by τ1 and the longer is characterized by τ2 using the formula:
Figure 4. Temperature dependence of IEu2+/IEu3+ emission for
LiMgPo4: Eu2+, Eu3+ obtained under excitation 260 nm.
AQ4
⎡ −t ⎤ ⎡ −t ⎤ I (t ) = A1exp ⎢ ⎥ + A2 exp ⎢ ⎥ ⎣ τ1 ⎦ ⎣ τ2 ⎦
(1)
The results of fitting the decay curves to relation (1) are presented in figure 6. It is seen that the shorter component τ1 decreases with increasing temperature, whereas the longer component τ2 increases in the temperature range between 100 K and 300 K from 1.8 µs to 3 µs, respectively and then decreases with increasing temperature for T > 300 K. The unusual temporal behavior of the 4f65d1 → 4f7 (8S7/2) luminescence can be related to the existence of shallow electron traps or the influence of higher-lying 6P7/2 state of the Eu2+ [15]. The time evolution of the ETE emission for different temperatures is presented in figure 5(b). For all temperatures, the decays are single exponential. The respective lifetime is equal 3.5 µs between 15 K and 300 K and then decreases with temperature to 0.25 µs at 600 K. It is seen that for temperatures lower than 300 K the ETE emission is delayed by time approximately equal to 1 µs. The luminescence rise is observed. For temperatures higher than 300 K the ETE emission appears almost at once after excitation. We described the ETE emission kinetics using the formula:
and recombination of the ETE [15, 29]. The origin of the emission band which peaked at 421 nm cannot be clearly assigned since such an emission was not reported before for the LiMgPO4:Eu2+ system. It can be related to the structure of the 4f65d1 electronic configuration as well as to the ETE. The two emission bands 383 nm and 421 nm could be also related to the existence of two different Eu2+ sites that replace Li+ and Mg2+, but we did not find strong arguments for such an assignation. The relative intensity of the band which peaked at 412 nm increases with temperature for the range from 15 K to 100 K and then diminishes and disappears at ambient temperature. When the temperature increased the ETE band peaked at 512 nm shifts to the shorter wavelengths, from 512 nm at 15 K to 490 nm at 500 K. Below 300 K, relative intensity of the ETE emission (with respect to the 4f65d1 → 4f7 (8S7/2) emission) does not depend on temperature, whereas for temperatures higher than 300 K the intensity of the ETE luminescence decreases. One can see this effect in figure 3(b) where non-normalized spectra measured at the temperature range 300–600 K are presented. In the inset of figure 3(b) the integrated intensity of the emission related to Eu2+ ions (4f65d1 → 4f7 and ETE) versus temperature is presented. The effect of the ETE emission quenching is related to the non-radiative recombination of ETE activated at higher temperatures. Luminescence spectra obtained at different temperatures collected in time interval 0–5 ms are presented in figure 3(c). The spectra are a superposition of Eu2+ luminescence represented by broad bands and Eu3+ luminescence represented by sharp lines. The spectra were normalized to the intensity of the 5D0 → 7F1 line. In figure 4 the ratio of the integrated luminescence intensities of Eu2+ to Eu3+ is presented. An evident effect of increasing the intensity of Eu2+ emission with respect to Eu3+ emission is observed for temperature range 50–250 K. The nature of this effect will be discussed later. The nonradiative intersystem crossing responsible for decreasing the relative intensity of the Eu2+ luminescence for temperature range 300–500 K is also responsible for decreasing the Eu2+ to Eu3+ luminescence intensity ratio for temperatures higher than 250 K.
⎡ −t ⎤ ⎡ −t ⎤ I (t ) = Adec exp ⎢ ⎥ − Arise exp ⎢ ⎥ ⎣ τdec ⎦ ⎣ τrise ⎦
(2)
The fitted quantities of the decay time τdec and the rise time τrise for different temperatures are presented in figure 6. Since the ETE emission rise time is approximately equal to the shorter component of the 4f65d1 → 4f7 (8S7/2) luminescence decay, we can consider the effect of nonradiative energy transfer from the 4f65d1 state to ETE state. The energy transfer process is seen more evidently when one compares the decay profiles of the emission monitored at 480–510 nm (ETE) and 370–390 nm (4f65d1 → 4f7) obtained for different temperatures (see figure 7). For 15 K the decay of the band 410–430 nm is additionally presented. One notices that decay of the emission band 410–430 nm is more similar to the decay of the ETE than to the decay of the 4f65d1 → 4f7 (8S7/2) luminescence. From figure 7 it is seen that at low temperature the ETE decay time is longer than longer component of the 4f65d1 → 4f7 (8S7/2) luminescence decay whereas at 300 K we have the opposite situation. Similar dependence of the Eu2+ luminescence on temperature was obtained for luminescence excited under 5
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 5. Decay profiles of emission for selected temperatures obtained under excitation λexc = 260 nm; (a) 4f65d luminescence (b) the ETE luminescence; (c) the Eu3+ luminescence related to the 5D0 → 7F1 transition. Results of fitting are presented by solid curves.
Figure 6. Dependence of the emission parameters of the luminescence kinetics, the decay and rise times, on temperature.
Figure 7. Comparison of the decays of the ETE emission and 4f65d1 → 4f7 (8S7/2) emission obtained at 15 K, 300 K and 600 K. For 15 K the decay of the band 410–430 nm is additionally presented.
320 nm and 355 nm. The only difference was that contribution from the band which peaked at 421 nm observed at temperature range between 100 K and 200 K diminishes with increasing excitation wavelength. The luminescence decay profiles of the 5D0 → 7F1 transition in Eu3+ obtained under 260 nm excitation are presented in figure 5(c). The decays are almost single exponential below 50 K and above 350 K. At medium temperatures the decays are nonexponential and faster. Under excitation 394 nm, which corresponds to the 7F0 → 5L6 transition in Eu3+, only the Eu3+ emission is seen. The emissions obtained at different temperatures are presented in figure 8(a). The luminescence decay profiles of the emission related to 5D0 → 7F1 transition obtained at different temperatures under excitation with 394 nm are presented in figure 8(b). One notices that luminescence decay does not depend on temperature. This is a different situation to when Eu3+ luminescence was excited through the CT transition with 260 nm. The specific temporal behavior of the luminescence related to the 5D0 → 7F1 transition under excitation by the CT process may be related to thermally-induced conversion excitation of the Eu3+ which involves the higher excited states 5GJ, 5HJ. However, the existence of such effects should be additionally confirmed.
4. Discussion The ETE luminescence under excitation with 260 nm is delayed in time. The same is observed under excitation with 325 nm. The emission related to the 4f65d1 → 4f7 (8S7/2) transition appears just after excitation also independently on excitation wavelength. Such phenomena can take place when ETE state is not excited directly, but obtains the excitation energy from the 4f65d1 state. To analyze such a process we constructed the one-dimensional configurational coordinate diagram describing the ground state 4f7 (8S7/2), the 4f65d1 state of Eu2+ and the ETE state. The diagram is presented in figure 9(a). To present the electronic energies of the Eu2+ system in the ETE, in the 4f65d1 and in the ground state we used following formulas:
Ei (Q ) = Ei0 + k
[Q − Qi ]2 2
(3)
where subscript i = g, ETE and d, labels the ground state, the ETE state and the 4f65d1 state, respectively, k is effective elastic constant of the lattice. The quantity of the configurational 6
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
Figure 9. (a) Configurational coordinate diagram of Eu2+ in one
dimensional approximation obtained from the emission spectra. (b) Possible relative positions of the minima of the excited electronic manifolds representing 4f65d1 state and ETE state in the many-dimensional configurational space.
the strength of the electron-lattice coupling. We used emission/absorption spectra to quantity the diagram and obtained −1 0 = 26900 cm−1 and S following data EETE ETE ℏω = 6500 cm , −1 −1 0 Ed = 28200 cm and SETE ℏω = 1500 cm and 0 = 1300 cm−1. The feeding of the ETE state ΔE = Ed0 − EETE from the excited 4f65d1 state is presented by dashed arrows. One notices that when one-dimensional lattice distortion is assumed the shift in the configurational coordinate space between the minima of the ETE and the 4f65d1 electronic ⎞ ⎛ manifolds ΔQ = Qd − QETE ⎜where Q i = 2Siℏω k ⎟ is too ⎝ ⎠ small to create the energy barrier which is responsible for delaying the ETE emission in time. On the other hand, the single configurational coordinate model probably does not work since the electron in the d state and in the ETE state have probably different symmetry. As a result, different vibration modes are involved in lattice distortion after excitation of the system into the ETE and the 4f65d1 states. When these modes are an orthogonal quantity ΔQ can be expressed as follows:
Figure 8. (a) The luminescence spectra of the LiMgPo4: Eu2+,
Eu3+, under excitation λexc = 394 nm at different temperatures; (b) Decay profiles of emission related to the 5D0 → 7F1 transition of Eu3+ for selected temperatures, under excitation λexc = 394 nm
coordinate Q represents the changes of the distance between Eu ion and ligands, Qi is the value of configurational coordinates that corresponds to the minimum energy of the respective state. Further, we assumed that Qg = 0 and E0g = 0. Under this assumption, one obtained that energy of lattice relaxation of the system in the state i is equal to:
Siℏω = k
Qi2 2
(4)
and the zero – phonon energy of i → g of radiative transition is equal to Ei0, Si is the Huang-Rhys Factor that determines 7
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
ΔQ =
2 Qd2 − QETE
transition and (plus) energy of ETE emission (2.5 eV). We estimated the bandgap of the LiMgPO4, Eg > 7 eV. Europium is located in the Li+ site octahedrally coordinated by oxygen ions and can exist as Eu2+ or Eu3+. Such a feature needs the creation of charge compensation defects. Usually, it is assumed that they are vacancies of Li+ or Mg2+ or Li1+ in Mg2+ site. It is a very reasonable assumption that independent of where Eu2+ and Eu3+ are located in LiMgPO4:Eu2+, Eu3+ all charges are compensated totally. One can consider two possible ways of compensation: the case when all Eu2+ and Eu3+ ions in Li+ sites are correlated spatially with respective compensation defects and the case when impurity ions and compensation defects are randomly distributed in crystal. As has been shown in [16] for the β-Ca2SiO4:Eu2+, Eu3+ when Eu3+ in Ca2+ site is compensated at the short distance, Eu3+ ions and compensation defects are correlated spatially and temperature does not influence the Eu2+/ Eu3+ emission intensity ratio. In the case of the LiMgPO4:Eu2+, Eu3+ system intensity of the Eu2+ increases with increasing temperature. One should consider this dependence by means of the Fermi energy. It has been discussed in paper [16] that when Eu2+ occupies the Me2+ site in perfect crystal the Fermi energy is located in the bandgap above energy of the Eu2+ ground state. In the case when Eu2+ ions occupy Li+ the Fermi energy can be even higher than ground state of Eu2+ when energy of compensation defect (Li+ in Mg2+ site or Li+ vacancy) is higher than energy of the Eu2+. In the case when Eu3+ replaces Me2+ in the lattice creation of compensation defects, the Fermi energy diminishes by the quantity:
(5)
The above relation was used to draw the diagram presented in figure 9(b). As can be seen from this figure, the nonradiative transition from the 4f65d1 state to the ETE state needs the activation energy and therefore the emission from the ETE state is delayed in time. Here it should be noted that analysis of the luminescence kinetics and dependence of the luminescence spectra on temperature allow us to conclude that we have two different Eu2+ sites. In the first type of site, the ETE state is located below the 4f65d1 state and the 4f65d1 state is additionally non-radiatively depopulated to the ETE state, as it is presented in figure 9 (a) and (b). These sites contribute to the fast component of the 4f65d1 luminescence and the ETE luminescence. In the second type of site, the energy of the ETE is higher than the energy of the 4f65d1 or ETE does not exist. These sites yield the slower component of the 4f65d1 emission. Also, only these sites are observed at higher temperatures for which the ETE luminescence is quenched. Although the ion-ligand distance is different in the Li+ and Mg2+ sites, we do not observe two different bands related to the 4f65d1 → 4f7 transitions in different sites. It can mean that only the Li+ site is occupied. Such occupation needs the creation of a compensation defect, which can be Li+ ion in Mg2+ site (Li/Mg) or Li vacancy VLi. The location of the compensation defect with respect to Eu2+ ion in lattice can influence the energy and also the existence of the ETE states. When the compensation defect is located near the Eu2+ site the ETE is created as the Eu3+ and electron bounded by long distance Coulomb potential created by difference V(Eu3+)–V(Li+ +Li/ Mg). In this case, the potential at a distance r from the system is equal −e/εr, where ε is static dielectric constant and e is electron charge. Such relatively weak potential can bind electrons which are located outside the first coordination sphere (see the model presented in [33]). If the compensation defect is far from Eu2+, substituting the respective Coulomb potential for Li+, V(Eu3+)–V(Li+) is equal to −2e/εr (it is two times stronger) and therefore the system probably will not create the stable state with electrons outside the first coordination sphere of Eu. In such a situation, the existence of stable ETE state is doubtful. It is observed that intensity of the Eu2+ luminescence increases with respect to the Eu3+ luminescence when the temperature increases from 100 K to 250 K. To understand this effect one should consider the energetic structure of the LiMgPO4:Eu system. The energy of CT transition seen in the excitation spectrum of Eu3+ luminescence allows us to estimate that the ground state of Eu2+ is located 4.5 eV above the top of the valence band. The ground state of the Eu3+ should be located between 5 eV and 8 eV below the ground state of Eu2+ [32]. The ground state and the excited state 4f65d1 of Eu2+ are located below the conduction band. The separation between the ground state of the Eu2+ and the conduction band is equal to the ionization energy and is greater than the energy of the ETE. One can estimate the lower limit of the energy bandgap of the LiMgPO4 as equal to energy of CT (4.5 eV)
ΔEi = E (Eu2 +) − E (Eu3 +) − Ui 2+
3+
(6)
where ΔE = E (Eu ) − E (Eu ) is the difference between energies of the ground states of Eu2+ and Eu3+ and Ui is the energy necessary for formation of i-th defect compensation. For short distance compensation, when the defect is located near the Eu3+ ion the compensation defect charge causes the bands to bend and leave the Eu2+ unoccupied [16]. For long distance compensation energy of the ground state of Eu2+ with respect to the Fermi level depends on the distance from the compensation defect(s). When Eu3+ replaces Li+ the situation is more complicated. The additional charge can be compensated by two Li+ ions replacing Mg2+ two Li+ vacancy or single Mg2+ vacancy. The energy of the Eu2+ and Eu3+ ground states with respect to the Fermi energy can be different for different sites and depends on actual spatial distribution of compensation defects. In figure 10 the energetic structure of the LiMgPO4:Eu2+, Eu3+ system for the assumption of random distribution of the compensation defects is presented. Here, three compensation defects (one VLi and two VMg) with potential to cause the band bends and four Eu ions placed in Li+ sites are presented. In sites labelled (a) energy of the ground state of Eu2+ is lower than the Fermi energy; these sites contribute always to the Eu2+ luminescence. Eu in the site labelled (b) near the VMg compensation defect and therefore the energy of the Fermi level is much lower than energy of the ground state of Eu2+, ΔE1 >> kT, (for ambient temperature); therefore, Eu2+ state is unoccupied and this site contributes always to the Eu3+ luminescence. Eu in the site labelled (c) is further from the 8
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
In the previous paper, it has been discussed that Eu3+ ions in LiMgPO4 substitute for Li+ and form two different sites [26]. Presented results have shown also that Eu2+ substitutes for Li+ in two different sites. It has been discussed that Eu2+ in Li+ site with short distance charge compensation contributes to the luminescence due to the 4f65d1 → 4f7 (8S7/2) and anomalous emission related to ETE recombination, whereas Eu2+ in Li+ sites with long distance compensation yields only the 4f65d1 → 4f7 (8S7/2) emission. The configurational coordinate diagram has been developed which describes the luminescence kinetics of the Eu2+ in Li+ site by means of electronlattice coupling which takes place with different lattice modes for the system in the 4f65d1 and ETE states. We also discussed that charge compensation defects necessary for the existence of Eu2+ in Li+ sites as well as Eu3+ in Li+ sites are not associated with individual europium ions but are rather randomly distributed in the lattice. The energies of Eu ions in the Eu2+ and Eu3+ charge state and energies of the band edges respect energy of the Fermi level in the specific Eu site depend on the potential of randomly distributed compensation defects ( VLi, VMg and Li/Mg). As a result, for part of Eu ions the Fermi energy is slightly below the energy of the Eu2+ ground state. For these sites, the actual charge of Eu ions is the result of a thermodynamic process (thermal excitation of Eu3+ to Eu2+). As a result of increasing the population of the Eu2+, the intensity of the Eu2+ luminescence increases with an increase in temperature.
Figure 10. Energetic structure of the LiMgPO4:Eu2+, Eu3+
system obtained under assumption of random distribution of the compensation defects (a), (b) and (c) represent the Eu ion in Li+ sites at different distance from compensation defect (for details see the main text).
compensation defect VMn than site (b), therefore the energy of the Fermi level is not far from the ground state of Eu2+. For temperature T ≈ ΔE2/k the mission from this site will depend on temperature. The total intensity of the Eu2+ luminescence is proportional to the following sum over all the Eu ions
IEu2+ = ∑ m
1 ⎡ (E m 2+ − EF ) ⎤ ⎥ 1 + exp ⎢ Eu kT ⎢⎣ ⎥⎦
Acknowledgements (7)
This study was supported by the POIG.01.01.02-02-006/09 project co-funded by the European Regional Development Fund within the Innovative Economy Program (Priority I, Activity 1.1. Sub-activity 1.1.2), which is gratefully acknowledged. A B was additionally supported by the European Social Fund as part of the project ‘Educators for the elite - integrated training program for PhD students, post-docs and professors as academic teachers at University of Gdansk’ within the framework of Human Capital Operational Program, Action IV and the grant PMN No. 538-5200-B160-13. This work was also supported under the framework of the international cooperation program managed by the National Research Foundation of Korea (2013K2A1A2055224).
m − E is the difference between energy of Eu2+ where EEu 2+ F ground state and energy of the Fermi level near m-th site. m The sites where this energy is negativeEEu 2 + − EF < 0 yield 2+ always the Eu emission (like sites labelled by a in figure 10), 3+ m the sites where EEu lumines2 + − EF ≫ kT yield always Eu cence (like site b in figure 10), whereas emission of the sites m − E is comparable to kT (like site c in \ 10) for which EEu 2+ F depend on temperature and for these sites intensity of the Eu2+ luminescence increases when temperature increases. One notices that at temperatures higher than ambient the ETE emission is quenched, whereas the 4f65d1 → 4f7 (8S7/2) emission is more stable. The temperature quenching of the ETE luminescence, accompanied by shortening of the luminescence lifetime, is related to nonratiative intersystem crossing the ETE and the ground electronic manifold caused by large electron-lattice coupling. It is not the case of the 4f65d1 → 4f7 (8S7/2) luminescence which weakly depends on temperature. It should be noted here that temperature stability of the 4f65d1 → 4f7 (8S7/2) emission concerns only the sites for which ETE state has higher energy than the 4f65d1 state or the ETE state does not exist.
References [1] Amara M B, Vlasse M, Flem G L and Hagenmuller P 1983 Acta Crystallogr. C39 1483 [2] Heine F, Handlovic M, Burdova K and Mailing J 1982 J. Crystallogr. Spectrosc. Res. 12 99 [3] Goňi A, Bonagamba T J, Silva M A, Panepucci H, Rojo T and Barberisa G E 1998 J. Appl. Phys. 84 416 [4] Dhabekar B, Menon S N, Alagu Raja E, Bakshi A K, Singh A K, Chougaonkar M P and Mayya Y S 2011 Nucl. Instrum. Methods Phys. Res. B 269 1844 [5] Kumar M, Dhabekar B, Menon S N, Chougaonkar M P and Mayya Y S 2011 Nucl. Instrum. Methods Phys. Res. B 269 1849
5. Conclusions Luminescence spectra and luminescence kinetics of the LiMgPO4:Eu2+, Eu3+ have been measured and discussed. 9
A Baran et al
J. Phys.: Condens. Matter 26 (2014) 385401
[21] Baby T and Nampoori V P N 1992 Solid State Commun. 81 367 [22] Lian Z, Wang J, Lv Y, Wang S and Su Q 2007 J. Alloys Compounds 430 257 [23] Dobrowolska A and Zych E 2012 MRS Symp. Proc. 1342 55 [24] Dobrowolska A and Zych E 2012 J. Phys. Chem. C 116 25493 [25] Baran A, Mahlik S, Grinberg M and Zych E 2013 J. Phys.: Condens. Matter 25 02560 [26] Zhang S, Huang Y, Shi L and Seo H J 2010 J. Electrochem. Soc. 157 J186 [27] Hanic F, Handlovic M, Burdova K and Majling J 1982 J. Crystallogr. Spectrosc. Res. 12 99 [28] Kubicki A, Bojarski P, Grinberg M, Sadownik M, Kukliński B 2006 Opt. Commun. 269 275 [29] Seo H J 2013 J. Ceram. Process. Res. 14 22 [30] Mahlik S, Lazarowska A, Grobelna B and Grinberg M 2012 J. Phys.: Condens. Matter 24 485501 [31] Mahlik S, Lazarowska A, Speghini A, Bettinelli M and Grinberg M 2014 J Lumin. 152 62 [32] Dorenbos P 2003 J. Phys.: Condens. Matter 15 841 [33] Grinberg M 2013 Opt. Mater. 35 2006
[6] Gai M, Chen Z, Fan Y and Wang J 2013 J. Rare Earth 31 551 [7] Dorenbos P 2003 J. Lumin. 104 239 [8] Vestegen J M P J and Sommerdijk J L 1974 J. Lumin. 9 297 [9] Hoffman M 1972 J. Electrochem. Soc. 119 905 [10] Mahlik S, Grinberg M, Shi L and Seo H J 2009 J. Phys.: Condens. Matter 21 235603 [11] Su H, Jia Z, Shi C, Xin J and Reid S A 2001 Chem. Phys. Lett. 335 17–22 [12] Latourrette B, Guillen F and Fouassier C 1979 Matter Res. Bull. 14 865 [13] Yamashita N, Yamamoto I, Ninugawa K, Wada T, Yamashita Y and Nakao Y 1985 Japan. J. Appl. Phys. 24 1174 [14] Shannon R D 1976 Acta Crystallogr. A32 751 [15] Zhang S, Huang Y, Shi L, Shi Land Seo H J 2010 J. Phys.: Condens. Matter 22 235402 [16] Baran A, Barzowska J, Grinberg M, Mahlik S, Szczodrowski K and Zorenko Y 2013 Opt. Mater. 35 2107 [17] Biswas K, Sontakke A D and Annapurna K 2012 Int. J. Appl. Glass Sci. 3 154 [18] Reddy A, Das S, Ahmad S, Babu SS, Ferreira J M F and Prakash G V 2012 RSC Adv. 2 8768 [19] Pei Z, Su Q and Zhang Z 1993 J. Alloys Compounds 198 51 [20] Hong B C and Kawano K 2006 J. Alloys Compounds 408 838
10
