Charge Carrier Trapping Processes in RE2O2S (RE = La, Gd, Y, and Lu).

Two different charge carrier trapping processes have been investigated in RE2O2S:Ln3+ (RE = La, Gd, Y, and Lu; Ln = Ce, Pr, and Tb) and RE2O2S:M (M = Ti4+ and Eu3+). Cerium, praseodymium and terbium act as recombination centers and hole trapping centers while host intrinsic defects provide the electron trap. The captured electrons released from the intrinsic defects recombine at Ce4+, Pr4+, or Tb4+ via the conduction band. On the other hand, Ti4+ and Eu3+ act as recombination centers and electron trapping centers while host intrinsic defects act as hole trapping centers. For these codopants we find evidence that recombination is by means of hole release instead of electron release. The released holes recombine with the trapped electrons on Ti3+ or Eu2+ and yield broad Ti4+ yellow-red charge transfer (CT) emission or characteristic Eu3+ 4f-4f emission. We will conclude that the afterglow in Y2O2S:Ti4+, Eu3+ is due to hole release instead of more common electron release.

Introduction

Charge carrier trapping and detrapping processes are of great interest in the luminescence research field both for an application and for a theoretical point of view.[1] Afterglow phosphors require that the captured electrons or holes are spontaneously released at room temperature to recombine at the luminescence center. Neither a too shallow nor a too deep trap will produce room temperature afterglow.[2,3] For storage materials used in X-ray imaging, deeper traps are needed to prevent thermal fading at room temperature.[4]

The lanthanide dopant can either act as an electron or as a hole trapping center. Such electron trapping was reported as early as in the 1960s by McClure et al., who found that trivalent lanthanides in CaF2 can be reduced to divalent under γ-irradiation.[5] In 2005, Dorenbos proposed that when the divalent lanthanide 4f ground state levels are below the conduction band (CB) the corresponding trivalent ions may act as electron trapping centers and as a function of type of lanthanide codopant there is a predictable variation in trap depth.[6] Later, this hypothesis was experimentally confirmed by thermoluminescence (TL) studies of YPO4:Ce3+, Ln3+ (Ln = Pr, Nd, Sm, Dy, Ho, Er, Tm, and Yb) by Bos et al.[7] Here, Ce3+ acts as the hole trapping center as well as the recombination (luminescence) center while the selected lanthanide codopants are the electron trapping centers. During the TL readout, the trapped electrons are released and move freely in the CB to eventually recombine at Ce4+. Different lanthanide codopants have different TL glow peak maxima indicating different trap depth. The same phenomenon has been reported later in Sr3AlSi1–O5:Ce3+, Ln3+ (Ln = Er, Nd, Sm, Dy, and Tm),[8] Y3Al5O12: Ln3+, RE3+ (Ln = Ce3+, Pr3+ and Tb3+; RE = Eu3+ and Yb3+)[9] and GdAlO3:Ce3+, Ln3+ (Ln= Pr, Er, Nd, Ho, Dy, and Tm).[10]

When the trivalent lanthanide 4f ground state levels are close above the VB, these ions may act as hole trapping centers. The captured holes can be released to recombine with a luminescence center via the VB or as a migrating Vk center. Compared to the many reports on electron trapping and detrapping processes, there are much less reports that discuss hole trapping and detrapping processes. One of the few is by Chakrabarti et al. in the 1980s who found that during UV irradiation of MgS:Ce3+, Sm3+ the holes are captured by cerium and electrons by samarium. After hole release, they recombine with samarium producing Sm3+ characteristic emission during the TL readout.[11] The other example is from our own studies on Gd1–LaAlO3:Eu3+,Tb3+ where Tb3+ acts as the hole trapping center and Eu3+ as the electron trapping center. The captured holes release from Tb4+ earlier than electrons from Eu2+ and recombine with Eu2+ producing Eu3+ characteristic 4f-4f emission. Another example of hole detrapping is given by Bos et al. in YPO4:Tb3+, RE3+ (RE3+ = Nd, Ho, and Dy) where again Tb3+ is acting as a hole trapping center and RE3+ as electron tapping center.[7]

Eu2+ and Ce3+ are the most widely used recombination (luminescence) centers in afterglow materials. For instance, SrAl2O4:Eu2+,Dy3+,[12] CaAl2O4:Eu2+,Nd3+,[13] CaS: Eu2+,Dy3+,[14] Ca2Si5N8:Eu2+,Tm3+[15] and Y3Al5-xGaO12:Ce3+,Cr3+.[16] The 5d excited levels of divalent europium or trivalent cerium are located very close to the CB in those compounds, and therefore excited electrons are easily released into the CB and subsequently caught by a trivalent lanthanide or Cr3+ cation.[1] The trapped electrons are released slowly and recombine with the europium or cerium recombination center to generate Eu2+ or Ce3+ emission. In these cases, the afterglow mechanism is due to the electron trapping and electron release.

In 2003, Kang et al. reported on the afterglow material of Y2O2S:Mg2+, Ti4+ that shows a unique orange broad band persistent luminescence centered at ∼595 nm after 380 nm UV excitation.[17] After that, dozens of reports were published to modify or improve this material. For instance, Y2O2S:Eu3+,Mg2+,Ti4+ [18] and Gd2O2S:Eu3+,Mg2+,Ti4+ [19] were synthesized and show afterglow emission both from Eu3+ and Ti4+.

Different to Eu2+ and Ce3+, Eu3+ can only act as an electron acceptor. The same applies to Ti4+ with the 3d0 electron configuration. If neither Eu3+ nor Ti4+ can be an electron donor then what is the electron donor in phosphors like Y2O2S:Eu3+,Mg2+,Ti4+ and Gd2O2S:Eu3+,Mg2+,Ti4+? Where and how are the electrons captured? How are these electrons released and why is the afterglow from Ti4+ and Eu3+?

Several studies were carried out to analysis the afterglow mechanism of Y2O2S:Eu3+,Mg2+,Ti4+. Hölsä et al. found that the afterglow emission is from Eu3+ and Ti3+ but did not explain the afterglow mechanism.[20] Zhou et al. observed that the afterglow is from Eu3+ and Ti4+, and the traps that contribute to the afterglow are complex Ti related traps.[18] Lei et al. studied the thermoluminescence of Gd2O2S:RE3+, Ti, Mg (RE = Ce, Pr, Nd, Sm, Eu, Tb, Dy, Ho, Er, Tm, and Yb).[21] Lei et al. proposed that UV-light exposure causes an electronic transition from the ground state of the RE3+ to the excited state, and simultaneously electrons and holes are created in the host. Then the captured electrons return back to the exited states of RE3+ at room temperature resulting in characteristic f–f persistent afterglow emission. Since the 4f states of the lanthanide ions are localized impurity states, after excitation of RE3+ to the excited state one may not interpret that the 4f state leaves a hole in the host lattice that can be filled by another electron.[22] Therefore, the transition suggested by Lei et al. is highly unlikely.

The objective of this study is to reveal the trapping and detrapping processes of electrons and holes in RE2O2S:M (RE= La, Gd, Y and Lu; M= Ce3+, Pr3+, Tb3+, Eu3+, and Ti4+) materials. Photoluminescence emission (PL) and excitation (PLE) spectra of Eu3+ or Ti4+ single-doped samples have been measured to construct the vacuum referred binding energy (VRBE) diagrams showing the lanthanide and titanium levels within the band gap. Thermoluminescence emission (TLEM) and thermoluminescence (TL) measurements were performed to identify the recombination centers and to derive the trap depths. Thermoluminescence excitation (TLE) spectra were measured to analyze the charging process of Eu3+ or Ti4+ single-doped samples. Finally, to show how the obtained knowledge can be applied, the results are used to propose the persistent luminescence mechanism of Y2O2S: Ti4+, Eu3+.

Experimental Section

All starting materials were purchased from Sigma-Aldrich and used without further treatment. The materials were synthesized by mixing of 5 N (99.999%) purity rare earth oxides, S (99.5%), TiO2 (99.99%) and Na2CO3 (99.99%) and fired from 1150 to 1250 °C during 4–8 h one or two times in a corundum crucible in CO atmosphere. The obtained compounds were washed by deionized water a couple of times to remove the Na2CO3 flux. The content of Na2CO3 is 7% by weight.

All powders were checked with a PANalytical XPert PRO X-ray diffraction system with a Co Kα (λ = 0.178901 nm) X-ray tube (45 kV, 40 mA). The PL spectra of Ti-doped samples were measured by a UV to VIS spectrometer (Ocean Optics, QE65000) with a UV LED (365 nm, 780 mW) excitation. The PLE spectra of Ti-doped samples and the PLE and PL spectra for all the Eu3+-doped samples were measured with a setup that consists of an UV/vis branch with a 500W Hamamatsu CW Xe lamp and Gemini 180 monochromator. The PerkinElmer MP-1913 photomultiplier was exploited as a detector connected at the exit slit of a Princeton Acton SP2300 monochromator.

Low-temperature TL measurements (90–450 K) were recorded with a sample chamber operating under vacuum (P = 10–7 mbar), a 90Sr/90Y β irradiation source having a dose rate of ∼0.4 mGy s–1 and a PerkinElmer channel PM tube (MP-1393). Liquid nitrogen was used as a cooling medium. A 600 nm bandpass filter (600FS40–50, Andover Corporation) was placed between the sample and the PMT during the measurements of Ce3+, Pr3+, Eu3+, and Ti4+ singly doped samples to transmit the red emission from the above dopants. For the Tb3+-doped samples, a 550 nm bandpass filter (550FS40–50, Andover Corporation) was placed between the sample and the PMT. TLEM spectra were measured using an UV to VIS spectrometer (Ocean Optics, QE65000) with a HR composite grating (300 lines/mm) and an entrance aperture of 100 μm resulting in a 3.3 nm (fwhm) wavelength resolution. Samples were irradiated with a 60Co gamma source to an absorbed dose of ∼1.6 kGy.[23]

The TL excitation spectra (TLE) were measured by first illuminating the samples during 600 s with monochromatic photons from a 150 W xenon arc lamp (Hamamatsu L2273) filtered by a 1/8 monochromator (Oriel Cornerstone 130) with wavelength resolution of 0.8 nm/0.1 mm slit width. The slit width was selected as 1 mm and the wavelength step was fixed as 10 nm. Next, the system is programmed by LabVIEW to record all the TL glow curves from room temperature to 350 °C for excitation wavelengths between 200 and 450 nm. The plot of the integrated TL glow peaks versus the excitation wavelength is called a TL excitation spectrum.[24] The TL spectra were all recorded by a RISØ TL/OSL reader model DA-15 and a controller model DA-20. The same 600 nm bandpass filter (600FS40-50) was placed between the sample and PMT.

Results

X-ray Diffraction Spectra and Photoluminescence Spectroscopy

The X-ray diffraction (XRD) patterns of as-prepared RE2O2S materials are shown in Figure . All samples are of single phase and match with the Y2O2S reference card (No. 382242) due to the same crystal structure (space group: P3m̅1). A slight shift of the XRD patterns can be observed in Figure b due to different lattice parameters.

Figure 1

(a) XRD patterns of the as-prepared samples of La2O2S, Gd2O2S, Y2O2S, and Lu2O2S. (b) Detailed XRD patterns in the range from 28 to 38°.

Figure displays the PLE (a) and PL (b) spectra of Eu3+ single-doped RE2O2S. All the samples have the characteristic Eu3+ 4f–4f emission. The host exciton creation energy increases from 4.57 eV (271 nm) for La2O2S to 4.66 eV (266 nm) for Lu2O2S. Those exciton energies are similar to that in previous reports, i.e., 4.60 eV for La2O2S[25] and 4.71 eV for Lu2O2S.[26] The broad excitation band near 320–400 nm originates from electron transfer from the valence band (VB) to Eu3+, also called the charge transfer (CT) band. It increases from 3.61 eV (343 nm) for La2O2S to 3.77 eV (329 nm) for Lu2O2S and shows the same tendency as the host exciton creation energy. The CT bands of Eu3+ in RE2O2S have fwhm (full width at half-maximum) around 0.8 eV which is quite typical for Eu3+ CT bands.[27,28]

Figure 2

Room temperature PLE (a) and PL spectra (b) of La2O2S:0.01Eu3+, Gd2O2S:0.01Eu3+, Y2O2S:0.01Eu3+, and Lu2O2S:0.01Eu3+. The excitation spectra were recorded at 627 nm emission. The emission spectra were excited at the charge transfer peak maxima, which are labeled in the legend of part b. The numbers in part a show the host exciton excitation maxima (left) and the charge transfer maxima (right).

Figure illustrates the PLE (a) and PL (b) spectra of Ti4+ single-doped RE2O2S. All samples show a broad band emission with fwhm around 0.5 eV. The broad emission bands originate from the Ti4+ charge transfer emission. The Ti4+ emission red shifts from 555 nm for La2O2S:Ti4+ to 635 nm for Lu2O2S:Ti4+. Here a calibrated CCD spectrometer was used to measure the emission spectra since the PerkinElmer MP-1913 photomultiplier we used is not sensitivity to the red light. A comparison of Y2O2S:Ti4+ emission spectra measured by the PerkinElmer MP-1913 PMT and the CCD are shown in Figure S1 in the Supporting Information.

Figure 3

Room temperature PLE (a) and PL spectra (b) of La2O2S:0.01Ti4+, Gd2O2S:0.01Ti4+, Y2O2S:0.01Ti4+ and Lu2O2S:0.01Ti4+. The excitation spectra were measured at the emission maxima. The CT maxima are shown in the legend of part a. The emission spectra were recorder by a calibrated CCD spectrometer (Ocean Optics, QE65000) under the 365 nm UV-LED excitation.

The excitation spectra of Ti4+ are shown in the Figure a. Similar to that for Eu3+, the broad band near 265 nm is the host exciton creation band and the one near 320–380 nm is the VB → Ti4+ charge transfer. The relative intensity of the Ti4+ CT excitation band (the ratio of Ti4+ CT intensity to the host exciton intensity) increases from La2O2S to Lu2O2S. The Ti4+ CT excitation band of La2O2S:Ti4+ is weak at room temperature. Therefore, the low-temperature (10 K) photoluminescence excitation spectrum of La2O2S:Ti4+ was measured and shown in Figure S2.

Figure S2 shows that at 10 K Ti4+ CT excitation band locates at ∼327 nm (3.79 eV). Figure S3 shows that the temperature T0.5 where Ti4+ emission intensity is quenched by 50% is at ∼165 K. The activation energy for thermal quenching can be derived from[29]where I(T) and I(0) is the luminescence intensity at temperature T and 0 and E indicates the activation energy. A fit through the data in Figure S3, as indicated by the solid curve provides the activation energy E = 0.05 eV.

Figure a shows that the Ti4+ CT excitation bands shift to longer wavelength from La2O2S (3.79 eV) to Lu2O2S (3.44 eV). Here we take the CT excitation band maxima as the Ti4+ CT energy and the numbers are displayed on the figure legend (Figure a). For Lu2O2S:Ti4+, we use the centroid of the band near 360 nm (3.44 eV) as the Ti4+ CT energy.

Thermoluminescence Emission Spectra

Thermoluminescence emission (TLEM) spectra were measured for RE2O2S:Ti4+ and RE2O2S:Eu3+ in order to identify the luminescence and recombination center during the TL readout. Figure shows two typical TLEM spectra of Y2O2S:Ti4+ and Y2O2S:Eu3+. Similar figures for RE = La, Gd and Lu can be found in Figure S4.

Figure 4

Thermoluminescence emission (TLEM) spectra of (a) Y2O2S:0.01Ti4+ and (b) Y2O2S:0.01Eu3+. The heating rate is 1 K/s after and each sample has been exposed to an irradiation dose of 1.6 kGy from a 60Co source.

Two broad TL glow curves centered at ∼350 and ∼455 K can be observed for Y2O2S:Ti4+ (Figure a). The TL emission spectra centered at ∼615 nm matches with the photoluminescence emission spectra shown in Figure b, indicating that Ti acts as the recombination center leading to Ti4+ charge transfer emission. The Ti4+ TL emission can also be observed in the Gd2O2S:Ti4+ (Figure S4a) and Lu2O2S:Ti4+ (Figure S4b) although the later one shows much weaker Ti4+ TL intensity than the others. No Ti4+ TL emission was observed in La2O2S:Ti4+, which is attributed to the almost complete thermal quenching of Ti4+ emission above room temperature (Figure S3). Characteristic red Eu3+ TL emission can be observed for Y2O2S:Eu3+ (Figure b), La2O2S:Eu3+ (Figure S4c), Gd2O2S:Eu3+ (Figure S4d) and Lu2O2S:Eu3+ (Figure S4e). The observation of Eu3+ TL emission evidence that, like Ti4+, Eu3+ acts as the recombination center.

Low-Temperature Thermoluminescence

Figure displays the low-temperature TL glow curves of Y2O2S single-doped with Tb3+, Pr3+, Ce3+, Eu3+, or Ti4+.

Figure 5

Low temperature thermoluminescence glow curves of (a) Y2O2S:0.01Tb3+, Y2O2S:0.0Pr3+, and Y2O2S:0.0Ce3+ and (b) Y2O2S:0.01Ti4+ and Y2O2S:0.01Eu3+. The heating rate was 1 K/s for all TL-recordings. The peak intensities are normalized by the mass of the sample.

Tb3+, Pr3+ and Ce3+ single-doped Y2O2S (Figure a) have the same glow peaks (herein referred to peaks 1 and 2) at ∼115 K and ∼182 K with different relative intensity. This implies that charge carriers are not released from Ce, Pr or Tb but from other trapping centers. The TL intensity of Ce3+ is around 3 orders of magnitude lower than that of Tb3+ which is due to the significant thermal quenching of Ce3+ emission at this temperature. Figure S5 shows that the temperature T0.5 where Ce3+ emission intensity is quenched by 50% is at ∼63 K. An Arrhenius fit of the quenching curve provides a 30 meV quenching energy barrier. The rising glow above 350 K in the Y2O2S:Pr3+ and Y2O2S:Ce3+ TL glow curves are due to blackbody radiation. No TL glow peaks are observed above 225 K in Figure a.

Figure b displays the TL glow curves for Ti4+ and Eu3+ single-doped Y2O2S. Between 90 and 260 K both samples share the same TL glow peaks (numbered 3, 4, 5, 6, 7 and 8) indicating that charge carriers are released from the same type of trapping centers not related to Eu or Ti. One observes a very broad TL glow starting from ∼260 K and maximum at ∼350 K in the Ti4+-doped sample which matches with that in the TLEM spectra (Figure a), and it contributes to the Ti4+ CT–luminescence afterglow. The Eu3+-doped sample shows like in Figure b the same glow at ∼350 K that contributes to the afterglow although it is 1 order of magnitude less intense than that of the Ti4+-doped sample.

The trap depth E corresponding with the TL glow peaks numbered in Figure was roughly estimated using the temperature T at the maximum of the glow and employing the first order kinetics equationwhere β = 1 K s–1 is the heating rate, k is the Boltzmann constant (8.62 × 10–5 eV/K), and s is the frequency factor (s–1).[30] The frequency factor s, which is related to the host lattice vibrational mode, is estimated using the 444 cm–1 (1.3 × 1013 s–1)[31] R3 line from Y2O2S Raman spectroscopy and assumed to be the same for all the Y2O2S samples with different dopants. The TL parameters (frequency factor s, peak position T and trap depth E) are listed in Table .

Table 1

Peak Number, Frequency Factor s (s–1), Peak Maxima T (K), and Trap Depth E (eV) of the TL Glow Peaks Recorded at β = 1 K/s from RE2O2S (RE = La, Gd, Y, and Lu)

		1	2	3	4	5	6	7	8	9	10	11
La2O2S	s	1.1 × 1013
	Tm	105	185	240	110	145	165	190	208	245	312	360
	E	0.28	0.51	0.66	0.30	0.39	0.45	0.52	0.57	0.68	0.87	1.0
Gd2O2S	s	1.3 × 1013
	Tm	115	115	130	180	203	220	250	288	335	370
	E	0.31	0.31	0.35	0.50	0.56	0.61	0.70	0.80	0.94	1.0
Y2O2S	s	1.3 × 1013
	Tm	115	182	113	134	181	210	240	278	350
	E	0.31	0.50	0.31	0.37	0.50	0.58	0.67	0.78	0.98
Lu2O2S	s	1.4 × 1013
	Tm	108	170	267	105	160	200	223	320
	E	0.29	0.47	0.75	0.29	0.44	0.61	0.62	0.90

Figure a, 7a, and 8a display the low-temperature TL for La2O2S, Gd2O2S, and Lu2O2S with different dopants. As in Figure a, with Tb3+ and Pr3+ doping, TL glow peaks at the same temperature are observed. With Ce3+ doping, the same TL glow peak temperature as with Tb3+ and Pr3+ doping in Lu2O2S is observed in Figure a. For Ce3+ doping in La2O2S:Ce3+ and Gd2O2S:Ce3+ the Ce3+ emission totally quenched[26] and no TL glow peaks were measured.

Figure 6

Low-temperature thermoluminescence glow curves of (a) La2O2S:0.01Tb3+ and La2O2S:0.0Pr3+ and (b) La2O2S:0.01Ti4+ and La2O2S:0.01Eu3+. The heating rate was 1 K/s for all TL-recordings. The peak intensities are normalized by the mass of the sample.

Figure 7

Low-temperature thermoluminescence glow curves of (a) Gd2O2S:0.01Tb3+ and Gd2O2S:0.0Pr3+, (b) Gd2O2S:0.01Ti4+ and Gd2O2S:0.01Eu3+. The heating rate was 1 K/s for all TL-recordings. The peak intensities are normalized by the mass of the sample.

Figure 8

Low-temperature thermoluminescence glow curves of (a) Lu2O2S:0.01Tb3+, Lu2O2S:0.0Pr3+ and Lu2O2S:0.0Ce3+ (b) Lu2O2S:0.01Ti4+ and Lu2O2S:0.01Eu3+. The heating rate was 1 K/s for all TL-recordings. The peak intensities are normalized by the mass of the sample.

Figure b illustrates the low temperature TL for Ti4+ and Eu3+ single-doped La2O2S. We observe that between 90 to 400 K both samples share almost the same TL glow peaks except for peaks 7 and 11. The same TL peak position implies that charge carriers are released from the same type of trapping centers not related to Eu or Ti. The absence of TL glow above 325 K in La2O2S:Ti4+ is probably due to the almost complete thermal quenching of Ti4+ emission above room temperature (Figure S3).

Almost the same TL glow peak positions are observed for Gd2O2S:Ti4+ and Gd2O2S:Eu3+ shown in Figure b with peak numbers 2, 3, 4, 5, 6, 7, 8, and 9. An extra peak (peak 10) is observed at ∼375 K for Gd2O2S:Ti4+.

Lu2O2S:Ti4+ in Figure b shows an extremely broad Ti4+ TL glow curve that begins at ∼135 K and reaches maximum glow at ∼320 K. It appears that Ti4+-doped RE2O2S shows broader TL glow peaks than when Eu3+ is the dopant. This may indicate a trap depth distribution[32] caused by the need for charge compensating defects.

Table lists all thermoluminescence parameters. All the frequency factors s are from the R3 line of RE2O2S Raman spectroscopy,[31] which is 1.1 × 1013 (365 cm–1) for La2O2S, 1.3 × 1013 (428 cm–1) for Gd2O2S, 1.3 × 1013 (444 cm–1) for Y2O2S and 1.4 × 1013 (472 cm–1) for Lu2O2S and all the trap depths (E) were calculated by eq .

So far, we found that in Tb3+, Pr3+, and Ce3+ single-doped RE2O2S, the charge carriers are not released from Ce, Pr, or Tb but from other trapping centers. The same applies for Ti4+ and Eu3+ single-doped RE2O2S. Here we conclude that for Tb3+-, Pr3+- and Ce3+-doped RE2O2S the TL glow curves are from host related electron traps while for Ti4+- and Eu3+-doped samples the TL bands are from host related hole trapping centers. The reasons will be discussed in detail in the Discussion.

Thermoluminescence Excitation Spectra

Figure a shows the thermoluminescence excitation (TLE) spectra of Eu3+ single-doped RE2O2S. A comparison with the Eu3+ excitation spectra (PLE) from Figure a can be seen in Figures S7–S10 in the Supporting Information. For each sample two broad bands centered near 260 and 330 nm can be observed that matches with the host exciton creation bands and Eu3+ CT-bands, respectively.

Figure 9

Thermoluminescence excitation (TLE) spectra of (a) RE2O2S:0.01Eu3+, (b) Gd2O2S:0.01Ti4+ and Y2O2S:0.01Ti4+. The samples have been excited by an Xe lamp with wavelengths ranging from 200 to 450 nm during 600 s before TL glow curve recording. The slit width was set at 1 mm leading to a spectrum resolution of 8 nm. The thermoluminescence excitation spectra were obtained by plotting the integrated TL from 300 to 600 K as a function of the excitation wavelength. The heating rate for TL readout is 1 K/s, and the wavelength step is 10 nm. The sample was excited at room temperature.

Figure b shows the TLE spectra of Gd2O2S:Ti4+ and Y2O2S:Ti4+. Comparison with the PLE spectra from Figure a can be seen in Figures S8 and S9. The TLE band between 260 to 280 nm is the host exciton creation band and the one between 300 to 400 nm is the Ti4+ CT-band, similar to the Ti4+ PLE spectra shown in Figure a. Gd2O2S:Ti4+ shows a very weak TLE band near 350 nm with intensity much lower than in the PLE spectrum. However, the band still exists indicating that this sample can be charged by 350 nm UV light. No TLE spectra could be recorded for La2O2S:Ti4+ and Lu2O2S:Ti4+ which is probably related to the very weak excitation efficiency of Ti4+ CT–luminescence in Figure .

Discussion

Vacuum Referred Binding Energy (VRBE) Diagram of RE2O2S and Ti4+ Charge Transfer Bands

To discuss the trapping and detrapping processes of electrons and holes in RE2O2S, we will first construct and exploit the VRBE diagram. The VRBE stands for vacuum referred binding energy that is defined as the energy needed to bring an electron from a level in the diagram to the vacuum outside the sample. The energy at rest in vacuum or vacuum level is then defined as energy zero. The reason to choose the VRBE diagram is because the binding energy of an electron in a lanthanide defect (both divalent and trivalent) states within the bandgap can be compared in different materials with respect to the same energy reference. Further details about how to construct the VRBE diagrams from spectroscopic data can be found in refs (33 and 34).

Figure shows the stacked VRBE diagrams of RE2O2S with location of Pr3+, Tb3+, Eu2+ and Ti3+ levels. The detailed VRBE diagrams with all lanthanide impurities level locations can be found in Figure S11. All the data needed and used to construct the VRBE diagrams are listed in Table . We adopted for all four samples a value of 6.37 eV for the so-called U-parameter of the chemical shift model. The reason for adopting the same U-parameter is due to the similar chemical environment surrounding Eu3+ in RE2O2S. This value defines within the chemical shift model a VRBE of −3.77 eV in the ground state of Eu2+ in the four RE2O2S samples. The Pr3+ and Tb3+ grounds states are then fixed at the same energy for all of the samples, with the values of −6.76 eV and −6.57 eV, respectively.

Figure 10

Stacked VRBE diagrams of RE2O2S (RE = La, Gd, Y, and Lu) with the VRBE in the ground states of Pr3+, Tb3+, Eu2+, and Ti3+.

Table 2

Parameters Used to Construct the VRBE Diagram for RE2O2S with U = 6.37 eVa

sample	Eex	ECT(Eu3+)	ECT(Ti4+)	EV	EC	ETb3+	EPr3+
La2O2S	4.57	3.61	3.79	–7.38	–2.44	–6.57	–6.75
Gd2O2S	4.66	3.72	3.65	–7.49	–2.46	–6.57	–6.75
Y2O2S	4.67	3.75	3.53	–7.52	–2.48	–6.57	–6.75
Lu2O2S	4.66	3.77	3.44	–7.54	–2.51	–6.57	–6.75

All numbers are in eV.

The top of the valence band is obtained from the VB → Eu3+ CT energy in Figure a and Table . The increase of that CT energy with smaller RE implies that the valence band maximum moves downward. The conduction band bottom is obtained from the exciton creation energy (Figure a) plus 8% of that to account for the electron–hole binding energy.

The Ti4+-doped oxysulfides show very broad excitation (fwhm ∼0.8 eV) and emission (fwhm ∼0.5 eV) bands in Figure . The broad excitation band between 320 and 380 nm is due to the VB → Ti4+ charge transfer, which means that electrons in the anions are excited to Ti4+ forming Ti3+ in its lowest 3d1 state. Here we assume that the energy at the maximum of the CT-band corresponds with the location of the Ti3+/4+ level above the VB-top. Therefore, the VRBE in the ground states of Ti3+ can be obtained as shown in Figure .

Rogers et al. compiled the VRBE in the Ti3+ ground state levels derived from different Ti4+-doped materials. They found that the VRBE in the Ti3+ lowest 3d1 state (E3d1) appears always near −4 ± 1 eV and the compound to compound variation of VRBE is attributed to the crystal field splitting(CFS).[35]Figure shows that the Ti3+ 3d1 states are near −4 eV and decrease from La2O2S to Lu2O2S. It was empirically found that the size of the CFS for the 5d-levels of the lanthanides is inversely propositional to the square of the bond length.[36] Ti3+ has one electron in the d-orbital like the 5d excited states of the lanthanides that shows the same CFS tendency as the 5d-levels of lanthanides.[35,37]Table S1 shows that the RE-(O,S) bond lengths decrease from La2O2S to Lu2O2S.[38−41] Therefore, the CFS of the Ti3+ 3d-levels is expected to increase from La2O2S to Lu2O2S. Such increased CFS will then reduce the VRBE in the lowest 3d1 state of Ti3+. This forms then our explanation for the red-shift from 327 nm (3.79 eV) in La2O2S:Ti4+ to 360 nm (3.44 eV) in Lu2O2S:Ti4+ of the VB → Ti4+ charge transfer excitation and the red-shift of the Ti4+ CT–luminescence in Figure b.

Trapping and Detrapping

The stacked VRBE diagrams of Figure show that the divalent Eu ground state is about 1.3 eV below the CB in RE2O2S which implies that the corresponding trivalent Eu can act as an electron trapping center. The same applies to Ti which has Ti3+ ground state location about 1.2 to 1.5 eV below the CB as illustrated in Figure which means that Ti4+ also acts as the electron trapping center.

During γ-ray irradiation in the TLEM spectra and β-ray irradiation in the low temperature TL spectra, the free charge carriers are generated that can move freely through the CB and the VB. For the Eu3+ or Ti4+ single-doped RE2O2S, the electrons will be trapped in either Eu3+ or Ti4+ forming Eu2+ or Ti3+, and the holes must be trapped somewhere else. In Figure b–8b, some common TL glow peaks at the same temperature can be observed. The temperatures at the maxima of glow peaks 3, 4, 5, 6, 7, and 9 listed in Table are about the same but with different relative intensities in Y2O2S:Eu3+/Ti4+. La2O2S:Eu3+/Ti4+, Gd2O2S:Eu3+/Ti4+, and Lu2O2S:Eu3+/Ti4+ also show the common TL glow peaks at the same temperature. These suggest that the Eu3+ or Ti4+ single-doped samples have the same type of hole trapping centers.

The thermoluminescence excitation (TLE) spectra for each Eu3+-doped sample in Figure a shows a broad band that coincides with the VB → Eu3+ CT excitation in Figure a. During CT-band excitation electrons are excited from the valence band to the Eu2+8S7/2 ground state leaving a hole in the valence band. It was demonstrated by Struck et al. that during the Eu3+ CT excitation a hole can dissociate from the CT-state.[42] Also p-type photoconductivity was observed by Dobrov et al. in La2O2S:Eu3+ during VB → Eu3+ CT excitation.[43] So, during CT-band excitation, part of the holes are released and trapped in the hole trapping centers. Then during the TL readout, the captured holes release again to recombine with Eu2+ producing Eu3+ characteristic emission as shown in Figure and Figure S4. The same conclusion was also suggested by Forest et al.[25] and Fonger et al.[44,45] by studying the thermoluminescence after CT excitation in La2O2S:Eu3+ and Y2O2S:Eu3+.

The TLE spectra for Ti4+-doped Y2O2S and Gd2O2S in Figure b again shows a broad band that coincides with the VB → Ti4+ CT-bands in Figure a. During the CT-band excitation, Ti3+ is formed and holes are released to the VB to be captured by the hole trapping center. Similar as for Eu doping, during the TL readout, the captured holes are released again and recombine at Ti3+ producing Ti4+ CT-luminescence as shown in Figure and Figure S4.

Figure a and Table show that the Ce3+, Pr3+, and Tb3+ single-doped Y2O2S all have the same glow peaks. From the stacked VRBE diagrams in Figure and Figure S11, we observe that the trivalent Ce, Pr, and Tb ground states are 2.6, 0.77, and 0.95 eV above the VB and these trivalent ions can act as hole trapping center during β irradiation. Then the electrons must be captured by the host lattice itself. Now the question turns to whether the captured electrons release earlier or the trapped holes release earlier.

If the holes from Ce4+, Pr4+, or Tb4+ release earlier than electrons, one can estimate according to eq with a heating rate of 1K/s that the TL peak positions (T) due to hole release from Ce4+, Pr4+ and Tb4+ to the VB in Y2O2S are expected at ∼900, 276, and 339 K, respectively. This means that the TL peak temperature for Y2O2S:Ce3+, Y2O2S:Pr3+ and Y2O2S:Tb3+ should be different and much higher than the observed TL temperature shown in Figure a. Similarly, one observes that with Pr3+ and Tb3+ doping in La2O2S, Gd2O2S and Lu2O2S, the TL glow peaks for each sample are at the same temperature shown in Figure a, 7a and 8a, respectively. Therefore, in Ce3+, Pr3+, and Tb3+ single-doped RE2O2S, the TL glow curves originate from electrons released from host lattice related trapping centers and recombine at the Ce4+, Pr4+ or Tb4+ hole trapping center providing Ce3+, Pr3+, and Tb3+ emission.

The Afterglow Mechanism of Y2O2S:Ti4+,Eu3+

On the basis of the above discussion, the afterglow mechanism of Y2O2S:Ti4+,Eu3+ can be proposed as illustrated in Figure . Upon UV excitation by day light, electrons are excited from the VB to Ti4+ and Eu3+ forming Ti3+ and Eu2+ in the ground states (arrows 1). The holes released to the VB are captured by the hole trapping centers (arrow 2) although it is still not clear what are those hole trapping centers. Then, the hole trapping center with a shallow trap depth enables spontaneously release of holes at the room temperature (arrow 3). It travels as a free hole via the VB or as a self-trapped hole or Vk center to recombine with Ti3+ and Eu2+ producing Ti4+ CT emission and Eu3+ 4f–4f emission (arrows 4). We conclude that the afterglow of Y2O2S:Ti4+,Eu3+ is due to the hole release instead of the more common electron release. However, further research needs to be performed to identify the nature of the hole trapping centers.

Figure 11

Proposed afterglow mechanism for Y2O2S:Ti4+,Eu3+. The filled circle stands for electrons and the open circle stands for hole.

Conclusion

Photoluminescence spectroscopy, thermoluminescence and the chemical shift model have been combined to study the trapping and detrapping processes of the charge carriers in RE2O2S. Photoluminescence spectroscopy shows that Ti4+ CT-luminescence provides the orange-red emission in RE2O2S:Ti4+. The red-shift of the Ti4+ CT-excitation and emission from La2O2S:Ti4+ to Lu2O2S:Ti4+ is attributed to the increased crystal field splitting of the Ti3+ 3d levels with smaller size of the site occupied. The TLEM spectra confirm that Ti4+ and Eu3+ act as the recombination center. The low temperature TL measurements reveal that for Tb3+-, Pr3+-, and Ce3+-doped RE2O2S the TL glow curves are from host related electron traps while for Ti4+- and Eu3+-doped RE2O2S the TL bands are from the host related hole trapping centers. The TL excitation spectra show that the electrons captured by Ti4+ and Eu3+ originate from the VB. Finally, the afterglow mechanism of Y2O2S:Ti4+,Eu3+ were derived that is due to the hole release instead of the more common electron release based on the above information.