Open Access
Add to BibSonomyAbstract
We have studied trap centers and nonradiative (NRR) recombination centers in a Ba3Si6O12N2:Eu2+ (BSON), one of promising materials for efficient and stable phosphors in white LED lamp applications. The energy distribution of four trap centers was obtained by thermo-luminescence (TL) with the excitation energy of 5.59eV. By superposing a below-gap excitation light of 1.77eV and observing the intensity change of the 5d-4f emission of Eu2+ centered at 2.36eV in our two-wavelength excited photoluminescence (TWEPL) measurement, both transient and steady state enhancement were observed. Such peculiar behavior of photo-stimulation is attributed to the coexistence of trap centers and NRR centers: the photoexcitation of electrons from trap centers generates the transient component, while that from NRR centers maintains the steady state component. An optical detection of relatively faint contribution of defects became possible in order to improve further the reliability and efficiency of phosphor materials.
© 2015 Optical Society of America
1. Introduction
White light-emitting diode (LED) lamps are attracting a lot of attention due to their excellent properties of low energy consumption, high luminous efficacy, high chromatic stability, and long lifetime without mercury [1–3]. However, the white LED lamp produced by a blue LED and only a yellowish phosphor shows an insufficient color rendering index due to a limited spectral width [4]. The role of efficient and stable phosphors with broader spectral width is becoming more and more important. Oxynitride/nitride phosphors perform good thermal and chemical stability and have attracted attention as such phosphors for white LEDs. One promising oxynitride phosphor material, BSON, has been studied by many researchers [5–13] for its splendid efficiency, stability and spectral features. The 4f-5d transition energy in the actual lattice is understood by subtracting both a centroid shift and a crystal field splitting from the transition energy of an isolated ion. An empirical formula, a relation to optical basicity [14], dielectric constant of the host [15] were reported. Though there still exists a possibility to improve internal quantum efficiency (IQE) and reliability by reducing NRR centers and trap centers, detailed characterization of such defect levels have not yet been accomplished in BSON phosphors.
Previous investigations based on photoluminescence (PL) spectroscopy have been focusing on radiative recombination process and relative comparison among different samples phenomenologically. They bring little information about the properties of trap centers and NRR centers at each material. The former bring a long persistent phosphorescence and/or a reduction of reliability, while the latter deteriorate the IQE directly at actual applications. In order to understand carrier recombination mechanisms of BSON phosphors and optimize each process condition, it is indispensable for us to discriminate carrier trapping/de-trapping processes by the trap centers and NRR kinetics via the NRR centers by optical means without electrode.
In the present work, a non-contacting and non-destructive method of TWEPL [16–18] was used to examine the BSON phosphor by changing the energy of the above-gap excitation (AGE) light. A distribution of NRR centers was obtained in addition to the trap centers previously revealed by TL measurement. Due to the coexistence of the trap centers and the NRR centers, we obtained a unique behavior of the PL intensity increase as a function of the AGE density [19]. Due to the coexistence of the trap centers and the NRR centers, we obtained a unique behavior of the PL intensity increase with both transient and steady state components. These defect levels were activated by the AGE of 5.59eV, not by that of 3.31eV.
2. Experiments and characterization
The BSON phosphor was prepared from BaCO3, Si3N4, SiO2 and Eu2O3 by a solid state reaction method like [13]. The raw materials were weighted out stoichiometrically, mixed and grinded using an agate mortar, placed into an alumina crucible and sintered under N2/H2 reducing atmosphere at 1050 °C during 12h. The flake was ground finally for characterization.
For the TL measurements, the BSON sample was placed into a cryostat and cooled down by liquid nitrogen, illuminated by an excimer lamp of 222nm wavelength (5.59eV) or an LED of 375nm (3.31eV) for 30 min. After turning off the light and keeping in the dark for 15 minutes, it was heated from 80K to 573K with constant temperature gradient of either 5K/min, 10K/min, 15K/min or 20K/min. The PL and PL excitation (PLE) spectra were measured by a spectrophotometer (Horiba Jovin-Yvon: FluoroMax-3). The IQE was measured by the combination of an integrating sphere and a multichannel analyzer (Systems Engineering: QEMS-2000).
In our TWEPL, the combination of a halogen lamp and a monochromator (Shimadzu SPG-120IR) was used as the below-gap excitation (BGE) light source. Its photon energy is lower than the band gap (hνB < Eg), but it modulates the electronic population of a localized level whose energy matches to the hνB [16]. Both the AGE and the BGE light beams were focused to the same spot on the sample surface at room temperature. The PL from the sample was fed into a photomultiplier through a monochromator and recorded in a photon counting analyzer. During the measurement of PL under the AGE, the BGE light was switched on and off. We define the normalized PL intensity by the ratio of the PL intensity with and without the BGE after each intensity reached a steady state, IA + B and IA, respectively, as
Here IB and ID are stray light components at the PL peak wavelength measured under only the BGE and without any excitation, respectively. These stray light components were subtracted from IA + B and IA, respectively in Eq. (1).1), for improving accuracy.3. Results and discussion
3.1 XRD, PL and PLE
The crystallinity of the sample was investigated by powder X-ray diffraction (XRD) as shown in Fig. 1(a). The diffraction peaks of the sample agree well with that of the standard data of Ba3Si6O12N2 phosphor (ICSD #42-1332). The PL and PLE spectra of BSON phosphor at room temperature are shown in Fig. 1(b). The 4f65d-4f7 transition of Eu2+ ion shows an efficient emission from 450 to 620nm with an emission peak at 525nm (2.36eV) and the FWHM of 70nm. Meanwhile, three bands centered at about 320, 365 and 430nm due to the 4f-5d transition of Eu2+ were found in this broad excitation spectrum 250-470nm which matches well with near UV-emitting or blue-emitting LEDs. The IQE at room temperature was 78.5% under the excitation of 375nm wavelength. Considering the high efficiency and the excitation spectrum, the BSON phosphor is considered to be a promising candidate for a green phosphor in white LED lamps.
3.2 TL and activation energies of trap centers
A main peak and three sub-peaks were resolved in the TL glow curves at different temperature gradients as shown in Fig. 2(a). All the glow curves showed at least four peaks which correspond to four trap centers excited by the AGE of 5.59eV. After changing the excitation energy to 3.31eV (375nm), on the other hand, at least five closely located peaks appeared in the glow curves as shown in Fig. 2(b) at different temperatures from the former. We noticed on a distinct difference in the TL intensities between the two excitation conditions roughly by a factor of 40. The number of captured carriers by 5.59eV excitation was much higher than those by 3.31eV. This difference gives a clue for the energy distribution of trap centers.
Fig. 2 TL spectra of BSON excited by 5.59eV (a), those excited by 3.31eV (b), and the Hoogenstraaten plot of both cases (c).
Their thermal activation energies ɛ can be calculated by Hoogenstraaten method based on the Eq. (1).2),
where βi represents the temperature gradient, Tmi is the peak temperature of the m-th level, and κ is the Boltzmann constant. By plotting the curves 1/Tmi vs. ln (Tmi /βi), as shown in Fig. 2(c) and calculating each slope, we obtained the activation energies of four levels as sub 1 (0.23eV), main (0.45eV), sub 2 (0.73eV) and sub 3 (0.77eV) for 5.59eV excitation [Fig. 2(a)], and peak 1 (0.31eV), peak 2 (0.33eV), peak 3 (0.45eV) and peak 4 (0.84eV) for 3.31eV excitation [Fig. 2(b)], respectively.3.3 TWEPL
During PL measurement under the AGE of 3.31eV, no PL intensity change was observed when the BGE of 1.77eV was added as shown by the black line in Fig. 3. In case the AGE energy was turned to 5.59eV, however, the PL intensity increased when the BGE of 1.77eV was superposed as shown by the red curve in the figure. The BGE effect under the AGE of 5.59eV comprises two parts. Firstly, the PL intensity showed sudden increase just after turning on the BGE and then gradually decreased to a steady value (transient BGE effect). The steady value of the PL intensity after 100s of initializing the BGE irradiation was still higher than that without the BGE (steady BGE effect). With the above results of TL, both the transient BGE effect and the steady BGE effect were considered to be partly from radiative recombination of electrons from the trap centers. We treat the latter steady BGE effect in this paper to calculate the normalized PL intensity.
The dependence of the BGE photon number density is summarized in Fig. 4(a) for the BGE energies of 1.24eV (1000nm), 1.38eV (900nm), 1.55eV (800nm), 1.68eV (740nm), 1.77eV (700nm) and 1.91eV (650nm), respectively. While the AGE of 5.59eV was used at the excitation density of 3.35 × 1016 cm−2s−1, all the PL intensities of the sample increased by the BGE. Except the influence of the trap centers, based on Shockley-Read-Hall (SRH) statistics [20,21], the increase in the PL intensity due to the BGE partly can be explained by a simple electronic excitation from the 4f level to a NRR center and/or from the NRR center to the upper levels. These BGE processes decrease the electron population in the 4f level or increase that in the upper levels, giving rise to the enhanced 5d-4f radiative recombination [19]. Figure 4(b) clearly shows the normalized PL intensity as a function of the BGE energy at the same BGE photon number density of 2.6 × 1018 cm−2s−1. The energy of the highest IN is around 1.77eV. There have been no reports on such optical detection in relatively efficient phosphor materials.
Fig. 4 The BGE photon number density dependence of IN for BSON phosphor (a) and the BGE photon energy dependence of IN at the same photon number density (b).
As can be seen in Fig. 4(a), the normalized PL intensity increased with increasing the BGE density. This is interpreted as our one-level model based on the SRH statistics. Except the curve of 1.24eV (1000nm) BGE, the averaged slopes of the rest IN curves above the BGE photon density of 1 × 1016 (cm−2s−1) were smaller than those below. When a saturating tendency in the BGE density dependence is confirmed by further increasing the BGE density, it can be attributed to the trap filling effect of the NRR center selected by the BGE energy. It leads to the quantitative determination of the NRR parameters by virtue of the trap filling effect [16–18].
Above experimental results can be summarized schematically as shown in Fig. 5 [10,15]. The bandgap of the BSON phosphor was reported as 6.79eV [15]. The absorption spectrum relevant to the experiment is attributed to the Eu-4f originated absorption process. Its final states are Eu-5d level, distribution of mixed Eu-N ligand levels and Eu-O clusters (Eu-originated band), and the conduction band. The AGE of 3.31eV excites electrons to the Eu-originated band, while that of 5.59eV excites electrons to the conduction band of BSON matrix. Our TL and TWEPL results imply that there are at least four dominant trap centers and one NRR center between the two AGE energies, i.e. 3.31eV and 5.59eV. The discussion on their origins need further study, but they accept carriers via the conduction band and are considered to be defects in BSON matrix. As electrons are captured by the trap levels during the AGE irradiation, the transient PL intensity increase takes place at the moment of the BGE irradiation. The steady state component of the BGE effect is due mainly to NRR centers, though there is a possibility of a combined effect including the trap centers. In order to further improve the reliability and efficiency of phosphors for white LED lamps, it is crucial for us to detect defect levels and clarify the way of reducing their incorporation.
4. Conclusion
We have studied trap centers and NRR centers in a Ba3Si6O12N2:Eu2+ prepared by the solid state reaction having the IQE of 78.5%. Four dominant trap levels whose activation energies range from 0.23 to 0.84eV were resolved from TL glow curves under the 5.59eV excitation. Within the BGE energy range from 1.24eV to 1.91eV, an increase of the PL intensity due to the addition of the BGE was observed at the AGE of 5.59eV but not at the AGE of 3.31eV. The PL intensity increase consists of both transient and steady state components. The BGE energy and photon number density dependence of the latter was measured systematically. It was shown that the distribution of the detected defect centers are activated at around 1.77eV and possibly originating from defects in BSON matrix. The present result opens a possibility to utilize the optical method for detecting relatively faint contribution of defects in phosphor materials in order to improve reliability and efficiency of phosphor materials.
References and links
1. S. Nakamuru and G. Fasol, The Blue Laser Diode: GaN based Light Emitters and Lasers (Springer-Verlag, 1997).
2. M. Shang, C. Li, and J. Lin, “How to produce white light in a single-phase host?” Chem. Soc. Rev. 43(5), 1372–1386 (2014). [CrossRef] [PubMed]
3. D. Geng, H. Lian, M. Shang, Y. Zhang, and J. Lin, “Oxonitridosilicate Y10(Si6O22N2)O2:Ce3+,Mn2+ phosphors: a facile synthesis via the soft-chemical ammonolysis process, luminescence, and energy-transfer properties,” Inorg. Chem. 53(4), 2230–2239 (2014). [CrossRef] [PubMed]
4. X. Q. Piao, K. Machida, T. Horikawa, H. Hanzawa, Y. Shimomura, and N. Kijima, “Preparation of CaAlSiN3: Eu3+phosphors by the self-propagating high-temperature synthesis and their luminescent properties,” Chem. Mater. 19(18), 4592–4599 (2007). [CrossRef]
5. Y. H. Song, M. O. Kim, M. K. Jung, K. Senthil, T. Masaki, K. Toda, and D. H. Yoon, “Photoluminescence properties of Ba3Si6O12N2:Eu2+ oxynitride phosphor synthesized by liquid phase precursor,” Mater. Lett. 77, 121–124 (2012). [CrossRef]
6. G. Chen, W. Zhuang, R. Liu, Y. Liu, Y. Hu, and H. He, “Photoluminescent properties of tunable green-emitting oxynitride (Ba3-xSrx) Si6O12N2:Eu2+ phosphor and its application in white LEDs,” J. Rare Earths 31(10), 944–949 (2013). [CrossRef]
7. R. Zhang, T. Maeda, R. Maruta, S. Kusaka, B. Ding, K. Murai, and T. Moriga, “Luminescence enhancement of Eu2+, Ce3+ co-doped Ba3Si5O13-δNδ phosphors,” J. Solid State Chem. 183(3), 620–623 (2010). [CrossRef]
8. C. Li, H. Chen, Y. Hua, L. Yu, Q. Jiang, D. Deng, S. Zhao, H. Ma, and S. Xu, “Photoluminescence and energy transfer studies on Ce3+/Eu2+ co-doped Ba3Si6O12N2 phosphor for white light-emitting-diodes,” Opt. Commun. 295, 129–133 (2013). [CrossRef]
9. G. Denis, P. Deniard, X. Rocquefelte, M. Benabdesselam, and S. Jobic, “The thermally connected traps model applied to the thermoluminescence of Eu2+ doped Ba13-xAl22-2xSi10+2xO66 materials (x~0.6),” Opt. Mater. 32(9), 941–945 (2010). [CrossRef]
10. J. Y. Tang, J. Chen, L. Hao, X. Xu, W. J. Xie, and Q. X. Li, “Green Eu2+-doped Ba3Si6O12N2 phosphor for white light-emitting diodes: Synthesis, characterization and theoretical simulation,” J. Lumin. 131(6), 1101–1106 (2011). [CrossRef]
11. E. Kang, S. Choi, and S. Hong, “Synthesis of Ba3Si6O12N2:Eu2+ green phosphors using Ba3SiO5:Eu2+ precursor and their luminescent properties,” ECS J. Solid State Science and Technology 1(1), R11–R14 (2012). [CrossRef]
12. M. Mikami, “Response function calculations of Ba3Si6O12N2 and Ba3Si6O9N4for the understanding of the optical properties of the Eu-doped phosphors,” Opt. Mater. 35(11), 1958–1961 (2013). [CrossRef]
13. M. Mikami, S. Shimooka, K. Uheda, H. Imura, and N. Kijima, “New green phosphor Ba3Si6O12N2:Eu for white LED: Crystal structure and optical properties,” Key Eng. Mater. 403, 11–14 (2009). [CrossRef]
14. P. Dorenbos, “A review on how lanthanide impurity levels change with chemistry and structure of inorganic compounds,” ECS J. Solid State Science and Techn. 2(2), R3001–R3011 (2013). [CrossRef]
15. M. Mikami, “Computational chemistry approach for white LED (oxy)nitride phosphors,” ECS J. Solid State Science and Techn. 2(2), R3048–R3058 (2013). [CrossRef]
16. E. Kanoh, K. Hoshino, N. Kamata, K. Yamada, M. Nishioka, and Y. Arakawa, “Saturation of photolumine-scence quenching under below-gap excitation in a GaAs/AlGaAs quantum well,” J. Lumin. 63(5-6), 235–240 (1995). [CrossRef]
17. N. Kamata, J. M. Z. Ocampo, K. Hoshino, K. Yamada, M. Nishioka, T. Someya, and Y. Arakawa, “Below-gap spectroscopy of semiconductor quantum wells by two-wavelength excited photoluminescence (TWEPL),” Recent Research Developments in Quantum Electronics 1, 123–135 (1999).
18. A. Z. M. Touhidul Islam, N. Murakoshi, T. Fukuda, H. Hirayama, and N. Kamata, “Optical detection of nonradiative recombination centers in AlGaN quantum wells for deep UV region,” Phys. Status Solidi C 11(3-4), 832–835 (2014). [CrossRef]
19. T. Li, Y. Kotsuka, T. Fukuda, T. Kurushima, and N. Kamata, “Nonradiative centers in Ba3Si6O12N2:Eu2+ phosphors observed by below-gap excitation method,” Mater. Lett. 145, 158–161 (2015). [CrossRef]
20. R. N. Hall, “Electron-hole recombination in germanium,” Phys. Rev. 87(2), 387 (1952). [CrossRef]
21. W. Shockley and W. T. Read Jr., “Statistics of the recombinations of holes and electrons,” Phys. Rev. 87(5), 835–842 (1952). [CrossRef]
