- Email: [email protected]
Broadband infrared luminescence in Bi-doped silicate glass
Journal of Non-Crystalline Solids 464 (2017) 34–38
Contents lists available at ScienceDirect
Journal of Non-Crystalline Solids journal homepage: www.elsevier.com/locate/jnoncrysol
Broadband infrared luminescence in Bi-doped silicate glass Pingsheng Yu a,⁎, Liangbi Su b, Wei Guo a, Jun Xu c a b c
School of Materials Engineering, Yancheng Institute of Technology, Yancheng, Jiangsu 224051, China Key Laboratory of Transparent and Opto-Functional Inorganic Materials, Chinese Academy of Sciences, Shanghai Institute of Ceramics, Shanghai 200050, China School of Physics & Engineering, Tongji University, Shanghai 200092, China
a r t i c l e
i n f o
Article history: Received 16 November 2016 Received in revised form 9 March 2017 Accepted 15 March 2017 Available online xxxx Keywords: Bi SiO2-CaO-MgO Glass Photoluminescence
a b s t r a c t Glasses having composition (0.5–5)Bi2O3-40SiO2-40CaO-20MgO were synthesized by the conventional melt quenching technique. The amorphous nature of the glasses was confirmed by XRD studies. Absorption, transmittance, fluorescence, and XPS spectra were measured. Heavily Bi-doped (Bi ions up to 8 mol%) SiO2-CaO-MgO glass can exhibit broadband NIR (Near-infrared) emission at about 1245 nm (under 808 nm excitation), and the FWHM (Full Width at Half Maximum) reach up to 300 nm. The measured fluorescent lifetime is several hundred microseconds. The XPS measurements of the Bi 4f region and decay time detection for the glass samples reveal that the NIR emission should be related to lower valence Bi ions (Bi+). After the Bi-doped glass was annealed in air atmosphere, the amount of luminescent active ions decreased. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Bismuth doped fiber/glass has been paid much attention because it was found to be a new medium for ultra-broadband optical fiber amplifier covering the entire telecommunications window [1–3]. The ultrabroadband infrared luminescence in Bi-doped glasses and fibers have been assigned to electronic transition of Bi5+, Bi2+, Bi+, Bi clusters [1, 4–7], or color centers [8], however, the origin of NIR emission in Bidoped glasses is still controversial. The NIR broadband luminescence has been observed in bismuth-doped germanate [3,9,10], silicate [11– 13], borate [14,15], and phosphate [16,17] glasses. Silica glass is a good host material for Bi luminescent centers due to the advantages of its chemical and thermal stability, appropriate mechanical properties and wide range of optical transmittance. However, pure silica or high silica content glass production requires very high temperatures (1600 °C or more), at so high temperature, the bismuth oxide is highly volatile and the glass composition should be non-stoichiometric. In previous publications, NIR broadband luminescence was observed in (10-40)Bi2O3–(90-60)GeO2 binary glasses, and the emission intensity increases monotonically until the ratio of Bi2O3 is up to 30 mol% (under 808 nm excitation), and then quenches rapidly [18], which indicates that concentration quenching of Bi ions might happen in Bi containing materials. In this work, we plan to study the NIR spectroscopic properties of Bidoped SiO2-CaO-MgO (Bi up to 10 mol%, relative to sum of SiO2 + CaO ⁎ Corresponding author. E-mail address: [email protected] (P. Yu).
http://dx.doi.org/10.1016/j.jnoncrysol.2017.03.017 0022-3093/© 2017 Elsevier B.V. All rights reserved.
+ MgO, unit: mol) glass. Adding CaO and MgO into silica glass, on one hand, the added oxides can decrease the production temperature; on the other hand, CaO and MgO have been used to avoid the quenching of luminescence [19]. (However, some scholars consider that the alkali and alkali earth metal oxides lead to quenching of the NIR luminescence [20]). The NIR emission and luminescence mechanism of Bi-doped SiO2CaO-MgO glass will be studied in this work. 2. Experiments 2.1. Sample preparation The Bi-doped SiO2-CaO-MgO glass samples were prepared by the conventional melt quenching procedure. The starting materials were 99.99% pure Bi2O3, SiO2, CaO and MgO powders. We prepared 5 series of Bi-doped SiO2-CaO-MgO glass raw materials with different components, 0.5Bi2O3-40SiO2-40CaO-20MgO glass marked as G-Bi1, and 1.5Bi2O3-40SiO2-40CaO-20MgO glass marked as G-Bi3, and so on, as presented in Table 1. The raw materials were thoroughly mixed in a mortar, and then melted at 1320 °C (the heating rate is 10 °C/min) in platinum crucible (with the lid on top) for 45 min in air atmosphere. Consequently, the molten materials were pouring onto brass mold (which preheated at about 400 °C), and then slow cooling in air to the room temperature [21]. We observed that the Bi-doped SiO2-CaO-MgO glasses were colored, and the color become darker with the increasing Bi content, but still transparent, see Fig. 1. The glass samples were of same thickness (3 mm) and high surface quality.
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
35
Table 1 Component of Bi-doped SiO2-CaO-MgO samples (raw materials). Sample
Bi2O3 (mol%)
SiO2 (mol%)
CaO (mol%)
MgO (mol%)
1, G-Bi1 2, G-Bi3 3, G-Bi6 4, G-Bi8 5, G-Bi10
0.5 1.5 3 4 5
40 40 40 40 40
40 40 40 40 40
20 20 20 20 20
2.2. Measurements We have studied the Bi-doped SiO2-CaO-MgO glass samples by means of the XRD (Rigaku D/max 2550 V X-ray diffractometer; radiation at 60 kV, 450 mA; resolution: 0.002°). A Perkin Elmer Lambda 900 UV/VIS spectrometer was used to measure the visible absorption spectra for the glass samples. The transmittance spectra were measured with a Spectrum 400 FT-IR and FT-NIR spectrometer (Perkin Elmer). The NIR emission spectra were taken with Princeton Instruments Trivista 557 (600 grooves/mm grating, electric cooling InGaAs pointtype detector). XPS (X-ray photoelectron spectroscopy) spectra were recorded on multifunctional imaging electron spectrometer (Thermo Fisher Scientific, ESCALAB 250XI). The XPS had a monochromatic Al Kα (hν = 1486.6 eV) source, and the peak positions were calibrated by carbon C1s line with a binding energy 284.8 eV.
3. Results and discussion The Bi concentrations in the raw materials of samples no. 1, 2, 3, 4 and 5 were 1.0 mol%, 3.0 mol%, 6.0 mol%, 8.0 mol% and 10 mol% relative to sum of SiO2 + CaO + MgO (unit: mol), while the concentrations of Bi in the obtained glasses were checked out 0.842 mol%, 2.51 mol%, 5.07 mol%, 6.81 mol% and 8.43 mol%, respectively, by ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry, Agilent 725 ICPOES). Fig. 2 displays the XRD patterns of Bi-doped SiO2-CaO-MgO glasses. The broad humps (centered at 2θ = 28.5°) confirm the amorphous nature of the prepared glasses. The glasses are remaining amorphous during incorporation of Bi ions (up to 10 mol%) without crystallization of other phases, which indicates that Bi doping has no distinct effect on SiO2-CaO-MgO glass structure. The visible absorption spectra of Bi-doped SiO2-CaO-MgO glasses are shown in Fig. 3. These glass samples have the same absorption band at about 360 nm. The sample G-Bi8 exhibits absorption band peaking at about 470 nm, while the G-Bi10 has the similar absorption band shifting to 510 nm. These absorption bands (at 360 nm, 470 nm and 510 nm) lead to the color change of glass samples, and the similar absorption is usually observed in Bi-doped IR-luminescent materials [22,23]. As Fig. 3 shows, the visible absorption coefficient become bigger with the higher of Bi doping, which is the one reason why the color of the glass samples varied from light color to dark brown with increasing Bi content. The NIR transmission spectra of Bi-doped SiO2-CaO-MgO glasses are shown in Fig. 4. For all samples, there are broad band absorption peaking at 2120 cm−1 and 3550 cm−1, and the absorption intensity enhanced with the increasing Bi content. These NIR absorption bands were
Fig. 1. Photos of the Bi-doped SiO2-CaO-MgO glasses (from left: samples 2, 3, 4, 5).
Fig. 2. X-ray diffraction patterns of Bi-doped SiO2-CaO-MgO glasses.
likely to be the characteristic absorption of the Bi-doped SiO2-CaO-MgO glasses (the absorption at 3550 cm−1 may also be superposition of the OH group band in this region). NIR emission are detected under 808 nm, 940 nm and 980 nm LDs excitation in (before and after annealing) Bi-doped SiO2-CaO-MgO glass samples, as shown in Figs. 5,6 and 8. The emission measurements were performed at room temperature and all the emission spectra were corrected for the setup characteristic. NIR emission spectra under excitation at 808 nm are shown in Fig. 5. The broadband emission peaking at 1245 nm can be observed, and the FWHM reach up to 300 nm (sample G-Bi8). To some extent, the emission intensity increases with increasing Bi content (up to 8 mol%). The emission at 1245 nm should be attributed to Bi ions [24]. Some researchers had studied various Bi-doped (with low concentrations) glasses, and found that the NIR luminescence is undoubtedly caused by the presence of bismuth in the glass composition [25,26], however, its lineshape of emission spectra differs from that of our results. In previous publications, the NIR luminescence of Bi-doped glass is often ascribed to reduced Bi states (e.g. Bi+/Bi2+ and various clusters [27–29]). As Fig. 6 shows, sample G-Bi8 exhibits a noticeable emission band centered at about 1235 nm and 1460 nm under 940 nm LD excitation, and only an emission peaking at 1235 nm was observed under 980 nm LD excitation. Furthermore, the NIR emission intensity was weaker than the intensity under 808 nm LD pumping (in Figs. 5,6, the
Fig. 3. Absorption spectra of Bi-doped SiO2-CaO-MgO glasses.
36
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
Fig. 4. Transmission spectra of Bi-doped SiO2-CaO-MgO glass samples.
Fig. 6. NIR emission spectra of sample G-Bi8 under excitation at 940 nm and 980 nm.
emission measurements under same condition). The results suggest that 940 nm and 980 nm LDs are not the suitable pumping source in our experiments. Fig. 7 shows the typical XPS spectra of Bi ions in the glass samples of G-Bi8 (with intensive NIR luminescence, see Fig. 5) and G-Bi3 (with moderate NIR luminescence, see Fig. 5). The Bi 4f signal consists of the 4f5/2 and 4f7/2 components. The peaks at 164.9 eV (4f5/2) and 159.5 eV (4f7/2) in sample G-Bi8 might be partially attributed to the lower Bi valence state. In previous publications, the researchers pointed out that the peaks at 165.6 eV (4f5/2) and 160.3 eV (4f7/2) may be ascribed to the Bi3+ 4f in xBi2O3(100-x)SiO2 (x = 0.1–2) glasses, and the lower binding energy of 164.2 eV (4f5/2) and 158.8 eV (4f7/2) corresponds to a reduced Bi valence state [24,30]. In this work, the 164.9 eV fall in between 165.6 eV and 164.2 eV, and the 159.5 eV fall in between 160.3 eV and 158.8 eV, which suggests that the sample G-Bi8 contains Bi3+ and lower valence Bi. Some researchers believed that the lower valence Bi (Bi+/Bi2+) contributes to the NIR emission, and they thought Bi3+ does not exhibit NIR luminescence in Bi2O3-GeO2 glass [18]. Therefore, the NIR emission might be attributed to the reduced Bi valence state in our work. (The peak positions were calibrated by carbon C1s line with a binding energy 285 eV in reference [24], and the calibrated carbon C1s line binding energy is 284.8 eV in our work).
As for sample G-Bi3, the binding energies locate at 165.2 eV and 159.9 eV, which suggests that the amount of the Bi3+ becomes larger, and the amount of Bi2 +/Bi+ becomes smaller, however, only the lower valence Bi (Bi+/Bi2+) can emit NIR fluorescence, so the NIR emission intensity of sample G-Bi3 weaker than that of sample G-Bi8. After annealing at 400 °C for 2 h in air atmosphere, sample G-Bi8 exhibits emission at 1250 nm, and the emission intensity are getting weaker than that of unannealed sample. However, after annealing at 400 °C for 24 h, we cannot observe any emission. Annealing in air is almost equal to annealing in oxidation atmosphere, so some lower valence Bi ions might be oxidized to Bi3+ ions, and the Bi3+ ions do not be considered to contribute to the NIR emission [18]. If the oxidation time is enough, the lower valence Bi ions must be oxidized to Bi3 + ions. The results suggest that only lower valence Bi ions lead to the NIR emission. The oxidation process could be described as:
Fig. 5. NIR emission spectra of Bi-doped SiO2-CaO-MgO glasses under excitation at 808 nm.
Biþ
Oxidation
→
Bi2þ
Oxidation
→
Bi3þ
Fig. 9 presents the XPS spectra of the Bi 4f region for the glass sample of G-Bi8. After annealing in air at 400 °C for 2 h, the peaks at 165.2 eV and 160.0 eV might be mainly attributed to the Bi3+ ions, only a small portion should be ascribed to lower Bi valence state [24,30], so the NIR emission intensity of annealed glass weaker than that of unannealed sample. However, after annealing in air at 400 °C for 24 h, the binding energies locate at 165.4 eV and 160.1 eV, which indicates that the
Fig. 7. XPS spectra of the Bi 4f region for the glass samples of G-Bi8 and G-Bi3.
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
37
Fig. 8. NIR emission spectra of sample G-Bi8 under excitation at 808 nm (after annealing at 400 °C in air).
Fig. 10. Luminescence decay time spectra of sample G-Bi8 (monitored emission at 1245 nm, 808 nm LD pump).
amount of the Bi3+ becomes biggest in sample G-Bi8, while the amount of lower valence Bi (Bi2+/Bi+) change to zero, so we cannot observe any emission after annealing for 24 h (see Fig. 8). The decay curves of sample G-Bi8 (monitored emission at 1245 nm, under excitation at 808 nm) are shown in Fig. 10 [fitting to a first order exponential decay equation, I = A ∗ exp(−t/τ) + I0]. Before and after annealing, the glass sample G-Bi8 exhibits lifetimes of 0.489 ms and 0.461 ms, respectively, which corresponding to the emission of Bi ions (see Figs. 5 and 8). The lifetime detected on emission at 1245 nm in G-Bi8 (4Bi2O3-40SiO2-40CaO-20MgO) glass is close to that of Pynenkov's work (about 100 mol%GeO2–0.002 mol%Bi2O3 glass) [25], and also close to that of Guo's study (xBi2O3-(100-x)GeO2 glass, in mol%, x = 10, 20, 30, 40) [18]. The detected lifetimes are one or two orders of magnitude longer than that of Bi3+ and Bi2+ [1], which means NIR centers are not connected with the Bi3+ and Bi2+. Therefore, the Bi+ ions probably act as the NIR luminescence centers in Bi-doped SiO2-CaO-MgO glasses (this work), however, we should discover more evidence of these Bi+ ions in the future. The NIR emission of Bi-doped SiO2-CaO-MgO glasses may be useful in some fields [used as effective amplifier for signals in the low loss O band (1260–1360 nm), E band (1360–1460 nm), and S band (1460– 1530 nm) of optical fiber, and so on].
4. Conclusions In summary, Bi-doped SiO2-CaO-MgO glasses were produced with different Bi concentration, and their photoluminescence properties in NIR range were studied. Adding CaO and MgO into silica glass can dramatically decrease the production temperature. The present results indicate that heavily Bi-doped (Bi ions up to 8 mol%) glass can exhibit strong broadband NIR emission, and the FWHM reach up to 300 nm, which also reveals that concentration quenching of Bi ions did not happen in this experiment. The XPS measurements of the Bi 4f region and the decay time detection for the glass samples were carried out, the results suggest that lower valence Bi (Bi+) contribute to the NIR emission. After annealing in air atmosphere for a period of time, the NIR emission become weaker, and then disappeared in the end, the quenching of the luminescence was attributed to the decreased/disappeared of the lower valence state of Bi (which responsible for the NIR emission). The bandwidth of the emission is so broad (FWHM: 300 nm), which offered the possible applications in low loss O band, E band, and S band of optical fiber.
Acknowledgements This work was supported by National Natural Science Foundation of China under the No. 61422511, and by Natural Science Foundation of Jiangsu Province, China (No. BK20141263). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
[13] [14] [15] Fig. 9. XPS spectra of the Bi 4f region for the sample G-Bi8 (after annealing at 400 °C in air).
Y. Fujimoto, M. Nakatsuka, Jpn. J. Appl. Phys. 40 (2001) 279. I.A. Bufetov, E.M. Dianov, Laser Phys. Lett. 6 (2009) 487. R.F. Wang, J. Liu, Z. Zhang, J. Alloys Compd. 688 (2016) 332. J. Ren, J. Qiu, D. Chen, C. Wang, X. Jiang, C. Zhu, J. Mater. Res. 22 (2007) 1954. V.G. Truong, L. Bigot, A. Lerouge, M. Douay, I. Razdobreev, Appl. Phys. Lett. 92 (2008) 041908. M. Peng, J. Qiu, D. Chen, X. Meng, C. Zhu, Opt. Lett. 30 (2005) 2433. Z.W. Jiang, N.L. Dai, L.Y. Yang, J.G. Peng, H.Q. Li, J.Y. Li, W. Liu, J. Non-Cryst. Solids 383 (2014) 196. M.Yu. Sharonov, A.B. Bykov, V. Petricevic, R.R. Alfano, Opt. Lett. 33 (2008) 2131. X.G. Meng, M.Y. Peng, D.P. Chen, L.Y. Yang, X.W. Jiang, C.S. Zhu, J.R. Qiu, Chin. Phys. Lett. 22 (2005) 615. G. Lakshminarayana, J. Ruan, J.R. Qiu, J. Alloys Compd. 476 (2009) 878. Y. Fujimoto, M. Nakatsuka, J. Non-Cryst. Solids 352 (2006) 2254. V.V. Dvoyrin, V.M. Mashinsky, L.I. Bulatov, I.A. Bufetov, A.V. Shubin, M.A. Melkumov, E.F. Kustov, E.M. Dianov, A.A. Umnikov, V.F. Khopin, M.V. Yashkov, A.N. Guryanov, Opt. Lett. 31 (2006) 2966. J.J. Ren, J.R. Qiu, D.P. Chen, X. Hu, X.W. Jiang, C.S. Zhu, J. Alloys Compd. 463 (2008) L5. M.Y. Peng, X.G. Meng, J.R. Qiu, Q.Z. Zhao, C.S. Zhu, Chem. Phys. Lett. 403 (2005) 410. B. Denker, B. Galagan, V. Osiko, S. Sverchkov, E. Dianov, Appl. Phys. B Lasers Opt. 87 (2007) 135.
38
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
[16] A.M. Mogus-Milankovic, A. Santic, V. Licina, D.E. Day, J. Non-Cryst. Solids 351 (2005) 3235. [17] Q.C. Sheng, S. Liu, W.T. Li, L.F. Wang, C.M. Tang, D.P. Chen, J. Lumin. 144 (2013) 26. [18] X. Guo, H.J. Li, L.B. Su, P.S. Yu, H.Y. Zhao, J.F. Liu, J. Xu, Laser Phys. 21 (2011) 901. [19] Masoud Eslami, Zohreh Hamnabard, Ali Nemati, J. Rare Earths 31 (2013) 595. [20] X. Meng, J. Qiu, M. Peng, D. Chen, Q. Zhao, X. Jiang, et al., Opt. Express 13 (2005) 1628. [21] S. Polosan, J. Non-Cryst. Solids 355 (2009) 1900. [22] Y.Q. Qui, Q.J. Kang, C.X. Li, X.Y. Dong, C.L. Zhao, Laser Phys. 20 (2010) 487. [23] N. Zhang, J. Qiu, G. Dong, Z. Yang, Q. Zhang, M.J. Peng, Mater. Chem. 22 (2012) 3154. [24] O.V. Usovich, L.A. Trusov, V.V. Lennikov, G. de la Fuente, A.A. Veber, P.E. Kazin, Mater. Lett. 85 (2012) 44.
[25] A.A. Pynenkov, S.V. Firstov, A.A. Panov, E.G. Firstova, K.N. Nishchev, I.A. Bufetov, E.M. Dianov, Quantum Electron. 43 (2013) 174. [26] I.A. Bufetov, M.A. Melkumov, S.V. Firstov, K.E. Riumkin, A.V. Shubin, V.F. Khopin, A.N. Guryanov, E.M. Dianov, IEEE J. Sel. Top. Quantum 20 (2014) 0903815. [27] H.T. Sun, Y. Sakka, M. Fujii, N. Shirahata, H. Gao, Opt. Lett. 36 (2011) 100. [28] A.N. Romanov, Z.T. Fattakhova, A.A. Veber, O.V. Usovich, E.V. Haula, V.N. Korchak, et al., Opt. Express 20 (2012) 7212. [29] Q.C. Sheng, X.L. Wang, D.P. Chen, J. Quant. Spectrosc. Radiat. Transf. 141 (2014) 9. [30] Unifit universal spectrum processing and analysis program for ESCA—spectra version, 2006.
Contents lists available at ScienceDirect
Journal of Non-Crystalline Solids journal homepage: www.elsevier.com/locate/jnoncrysol
Broadband infrared luminescence in Bi-doped silicate glass Pingsheng Yu a,⁎, Liangbi Su b, Wei Guo a, Jun Xu c a b c
School of Materials Engineering, Yancheng Institute of Technology, Yancheng, Jiangsu 224051, China Key Laboratory of Transparent and Opto-Functional Inorganic Materials, Chinese Academy of Sciences, Shanghai Institute of Ceramics, Shanghai 200050, China School of Physics & Engineering, Tongji University, Shanghai 200092, China
a r t i c l e
i n f o
Article history: Received 16 November 2016 Received in revised form 9 March 2017 Accepted 15 March 2017 Available online xxxx Keywords: Bi SiO2-CaO-MgO Glass Photoluminescence
a b s t r a c t Glasses having composition (0.5–5)Bi2O3-40SiO2-40CaO-20MgO were synthesized by the conventional melt quenching technique. The amorphous nature of the glasses was confirmed by XRD studies. Absorption, transmittance, fluorescence, and XPS spectra were measured. Heavily Bi-doped (Bi ions up to 8 mol%) SiO2-CaO-MgO glass can exhibit broadband NIR (Near-infrared) emission at about 1245 nm (under 808 nm excitation), and the FWHM (Full Width at Half Maximum) reach up to 300 nm. The measured fluorescent lifetime is several hundred microseconds. The XPS measurements of the Bi 4f region and decay time detection for the glass samples reveal that the NIR emission should be related to lower valence Bi ions (Bi+). After the Bi-doped glass was annealed in air atmosphere, the amount of luminescent active ions decreased. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Bismuth doped fiber/glass has been paid much attention because it was found to be a new medium for ultra-broadband optical fiber amplifier covering the entire telecommunications window [1–3]. The ultrabroadband infrared luminescence in Bi-doped glasses and fibers have been assigned to electronic transition of Bi5+, Bi2+, Bi+, Bi clusters [1, 4–7], or color centers [8], however, the origin of NIR emission in Bidoped glasses is still controversial. The NIR broadband luminescence has been observed in bismuth-doped germanate [3,9,10], silicate [11– 13], borate [14,15], and phosphate [16,17] glasses. Silica glass is a good host material for Bi luminescent centers due to the advantages of its chemical and thermal stability, appropriate mechanical properties and wide range of optical transmittance. However, pure silica or high silica content glass production requires very high temperatures (1600 °C or more), at so high temperature, the bismuth oxide is highly volatile and the glass composition should be non-stoichiometric. In previous publications, NIR broadband luminescence was observed in (10-40)Bi2O3–(90-60)GeO2 binary glasses, and the emission intensity increases monotonically until the ratio of Bi2O3 is up to 30 mol% (under 808 nm excitation), and then quenches rapidly [18], which indicates that concentration quenching of Bi ions might happen in Bi containing materials. In this work, we plan to study the NIR spectroscopic properties of Bidoped SiO2-CaO-MgO (Bi up to 10 mol%, relative to sum of SiO2 + CaO ⁎ Corresponding author. E-mail address: [email protected] (P. Yu).
http://dx.doi.org/10.1016/j.jnoncrysol.2017.03.017 0022-3093/© 2017 Elsevier B.V. All rights reserved.
+ MgO, unit: mol) glass. Adding CaO and MgO into silica glass, on one hand, the added oxides can decrease the production temperature; on the other hand, CaO and MgO have been used to avoid the quenching of luminescence [19]. (However, some scholars consider that the alkali and alkali earth metal oxides lead to quenching of the NIR luminescence [20]). The NIR emission and luminescence mechanism of Bi-doped SiO2CaO-MgO glass will be studied in this work. 2. Experiments 2.1. Sample preparation The Bi-doped SiO2-CaO-MgO glass samples were prepared by the conventional melt quenching procedure. The starting materials were 99.99% pure Bi2O3, SiO2, CaO and MgO powders. We prepared 5 series of Bi-doped SiO2-CaO-MgO glass raw materials with different components, 0.5Bi2O3-40SiO2-40CaO-20MgO glass marked as G-Bi1, and 1.5Bi2O3-40SiO2-40CaO-20MgO glass marked as G-Bi3, and so on, as presented in Table 1. The raw materials were thoroughly mixed in a mortar, and then melted at 1320 °C (the heating rate is 10 °C/min) in platinum crucible (with the lid on top) for 45 min in air atmosphere. Consequently, the molten materials were pouring onto brass mold (which preheated at about 400 °C), and then slow cooling in air to the room temperature [21]. We observed that the Bi-doped SiO2-CaO-MgO glasses were colored, and the color become darker with the increasing Bi content, but still transparent, see Fig. 1. The glass samples were of same thickness (3 mm) and high surface quality.
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
35
Table 1 Component of Bi-doped SiO2-CaO-MgO samples (raw materials). Sample
Bi2O3 (mol%)
SiO2 (mol%)
CaO (mol%)
MgO (mol%)
1, G-Bi1 2, G-Bi3 3, G-Bi6 4, G-Bi8 5, G-Bi10
0.5 1.5 3 4 5
40 40 40 40 40
40 40 40 40 40
20 20 20 20 20
2.2. Measurements We have studied the Bi-doped SiO2-CaO-MgO glass samples by means of the XRD (Rigaku D/max 2550 V X-ray diffractometer; radiation at 60 kV, 450 mA; resolution: 0.002°). A Perkin Elmer Lambda 900 UV/VIS spectrometer was used to measure the visible absorption spectra for the glass samples. The transmittance spectra were measured with a Spectrum 400 FT-IR and FT-NIR spectrometer (Perkin Elmer). The NIR emission spectra were taken with Princeton Instruments Trivista 557 (600 grooves/mm grating, electric cooling InGaAs pointtype detector). XPS (X-ray photoelectron spectroscopy) spectra were recorded on multifunctional imaging electron spectrometer (Thermo Fisher Scientific, ESCALAB 250XI). The XPS had a monochromatic Al Kα (hν = 1486.6 eV) source, and the peak positions were calibrated by carbon C1s line with a binding energy 284.8 eV.
3. Results and discussion The Bi concentrations in the raw materials of samples no. 1, 2, 3, 4 and 5 were 1.0 mol%, 3.0 mol%, 6.0 mol%, 8.0 mol% and 10 mol% relative to sum of SiO2 + CaO + MgO (unit: mol), while the concentrations of Bi in the obtained glasses were checked out 0.842 mol%, 2.51 mol%, 5.07 mol%, 6.81 mol% and 8.43 mol%, respectively, by ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry, Agilent 725 ICPOES). Fig. 2 displays the XRD patterns of Bi-doped SiO2-CaO-MgO glasses. The broad humps (centered at 2θ = 28.5°) confirm the amorphous nature of the prepared glasses. The glasses are remaining amorphous during incorporation of Bi ions (up to 10 mol%) without crystallization of other phases, which indicates that Bi doping has no distinct effect on SiO2-CaO-MgO glass structure. The visible absorption spectra of Bi-doped SiO2-CaO-MgO glasses are shown in Fig. 3. These glass samples have the same absorption band at about 360 nm. The sample G-Bi8 exhibits absorption band peaking at about 470 nm, while the G-Bi10 has the similar absorption band shifting to 510 nm. These absorption bands (at 360 nm, 470 nm and 510 nm) lead to the color change of glass samples, and the similar absorption is usually observed in Bi-doped IR-luminescent materials [22,23]. As Fig. 3 shows, the visible absorption coefficient become bigger with the higher of Bi doping, which is the one reason why the color of the glass samples varied from light color to dark brown with increasing Bi content. The NIR transmission spectra of Bi-doped SiO2-CaO-MgO glasses are shown in Fig. 4. For all samples, there are broad band absorption peaking at 2120 cm−1 and 3550 cm−1, and the absorption intensity enhanced with the increasing Bi content. These NIR absorption bands were
Fig. 1. Photos of the Bi-doped SiO2-CaO-MgO glasses (from left: samples 2, 3, 4, 5).
Fig. 2. X-ray diffraction patterns of Bi-doped SiO2-CaO-MgO glasses.
likely to be the characteristic absorption of the Bi-doped SiO2-CaO-MgO glasses (the absorption at 3550 cm−1 may also be superposition of the OH group band in this region). NIR emission are detected under 808 nm, 940 nm and 980 nm LDs excitation in (before and after annealing) Bi-doped SiO2-CaO-MgO glass samples, as shown in Figs. 5,6 and 8. The emission measurements were performed at room temperature and all the emission spectra were corrected for the setup characteristic. NIR emission spectra under excitation at 808 nm are shown in Fig. 5. The broadband emission peaking at 1245 nm can be observed, and the FWHM reach up to 300 nm (sample G-Bi8). To some extent, the emission intensity increases with increasing Bi content (up to 8 mol%). The emission at 1245 nm should be attributed to Bi ions [24]. Some researchers had studied various Bi-doped (with low concentrations) glasses, and found that the NIR luminescence is undoubtedly caused by the presence of bismuth in the glass composition [25,26], however, its lineshape of emission spectra differs from that of our results. In previous publications, the NIR luminescence of Bi-doped glass is often ascribed to reduced Bi states (e.g. Bi+/Bi2+ and various clusters [27–29]). As Fig. 6 shows, sample G-Bi8 exhibits a noticeable emission band centered at about 1235 nm and 1460 nm under 940 nm LD excitation, and only an emission peaking at 1235 nm was observed under 980 nm LD excitation. Furthermore, the NIR emission intensity was weaker than the intensity under 808 nm LD pumping (in Figs. 5,6, the
Fig. 3. Absorption spectra of Bi-doped SiO2-CaO-MgO glasses.
36
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
Fig. 4. Transmission spectra of Bi-doped SiO2-CaO-MgO glass samples.
Fig. 6. NIR emission spectra of sample G-Bi8 under excitation at 940 nm and 980 nm.
emission measurements under same condition). The results suggest that 940 nm and 980 nm LDs are not the suitable pumping source in our experiments. Fig. 7 shows the typical XPS spectra of Bi ions in the glass samples of G-Bi8 (with intensive NIR luminescence, see Fig. 5) and G-Bi3 (with moderate NIR luminescence, see Fig. 5). The Bi 4f signal consists of the 4f5/2 and 4f7/2 components. The peaks at 164.9 eV (4f5/2) and 159.5 eV (4f7/2) in sample G-Bi8 might be partially attributed to the lower Bi valence state. In previous publications, the researchers pointed out that the peaks at 165.6 eV (4f5/2) and 160.3 eV (4f7/2) may be ascribed to the Bi3+ 4f in xBi2O3(100-x)SiO2 (x = 0.1–2) glasses, and the lower binding energy of 164.2 eV (4f5/2) and 158.8 eV (4f7/2) corresponds to a reduced Bi valence state [24,30]. In this work, the 164.9 eV fall in between 165.6 eV and 164.2 eV, and the 159.5 eV fall in between 160.3 eV and 158.8 eV, which suggests that the sample G-Bi8 contains Bi3+ and lower valence Bi. Some researchers believed that the lower valence Bi (Bi+/Bi2+) contributes to the NIR emission, and they thought Bi3+ does not exhibit NIR luminescence in Bi2O3-GeO2 glass [18]. Therefore, the NIR emission might be attributed to the reduced Bi valence state in our work. (The peak positions were calibrated by carbon C1s line with a binding energy 285 eV in reference [24], and the calibrated carbon C1s line binding energy is 284.8 eV in our work).
As for sample G-Bi3, the binding energies locate at 165.2 eV and 159.9 eV, which suggests that the amount of the Bi3+ becomes larger, and the amount of Bi2 +/Bi+ becomes smaller, however, only the lower valence Bi (Bi+/Bi2+) can emit NIR fluorescence, so the NIR emission intensity of sample G-Bi3 weaker than that of sample G-Bi8. After annealing at 400 °C for 2 h in air atmosphere, sample G-Bi8 exhibits emission at 1250 nm, and the emission intensity are getting weaker than that of unannealed sample. However, after annealing at 400 °C for 24 h, we cannot observe any emission. Annealing in air is almost equal to annealing in oxidation atmosphere, so some lower valence Bi ions might be oxidized to Bi3+ ions, and the Bi3+ ions do not be considered to contribute to the NIR emission [18]. If the oxidation time is enough, the lower valence Bi ions must be oxidized to Bi3 + ions. The results suggest that only lower valence Bi ions lead to the NIR emission. The oxidation process could be described as:
Fig. 5. NIR emission spectra of Bi-doped SiO2-CaO-MgO glasses under excitation at 808 nm.
Biþ
Oxidation
→
Bi2þ
Oxidation
→
Bi3þ
Fig. 9 presents the XPS spectra of the Bi 4f region for the glass sample of G-Bi8. After annealing in air at 400 °C for 2 h, the peaks at 165.2 eV and 160.0 eV might be mainly attributed to the Bi3+ ions, only a small portion should be ascribed to lower Bi valence state [24,30], so the NIR emission intensity of annealed glass weaker than that of unannealed sample. However, after annealing in air at 400 °C for 24 h, the binding energies locate at 165.4 eV and 160.1 eV, which indicates that the
Fig. 7. XPS spectra of the Bi 4f region for the glass samples of G-Bi8 and G-Bi3.
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
37
Fig. 8. NIR emission spectra of sample G-Bi8 under excitation at 808 nm (after annealing at 400 °C in air).
Fig. 10. Luminescence decay time spectra of sample G-Bi8 (monitored emission at 1245 nm, 808 nm LD pump).
amount of the Bi3+ becomes biggest in sample G-Bi8, while the amount of lower valence Bi (Bi2+/Bi+) change to zero, so we cannot observe any emission after annealing for 24 h (see Fig. 8). The decay curves of sample G-Bi8 (monitored emission at 1245 nm, under excitation at 808 nm) are shown in Fig. 10 [fitting to a first order exponential decay equation, I = A ∗ exp(−t/τ) + I0]. Before and after annealing, the glass sample G-Bi8 exhibits lifetimes of 0.489 ms and 0.461 ms, respectively, which corresponding to the emission of Bi ions (see Figs. 5 and 8). The lifetime detected on emission at 1245 nm in G-Bi8 (4Bi2O3-40SiO2-40CaO-20MgO) glass is close to that of Pynenkov's work (about 100 mol%GeO2–0.002 mol%Bi2O3 glass) [25], and also close to that of Guo's study (xBi2O3-(100-x)GeO2 glass, in mol%, x = 10, 20, 30, 40) [18]. The detected lifetimes are one or two orders of magnitude longer than that of Bi3+ and Bi2+ [1], which means NIR centers are not connected with the Bi3+ and Bi2+. Therefore, the Bi+ ions probably act as the NIR luminescence centers in Bi-doped SiO2-CaO-MgO glasses (this work), however, we should discover more evidence of these Bi+ ions in the future. The NIR emission of Bi-doped SiO2-CaO-MgO glasses may be useful in some fields [used as effective amplifier for signals in the low loss O band (1260–1360 nm), E band (1360–1460 nm), and S band (1460– 1530 nm) of optical fiber, and so on].
4. Conclusions In summary, Bi-doped SiO2-CaO-MgO glasses were produced with different Bi concentration, and their photoluminescence properties in NIR range were studied. Adding CaO and MgO into silica glass can dramatically decrease the production temperature. The present results indicate that heavily Bi-doped (Bi ions up to 8 mol%) glass can exhibit strong broadband NIR emission, and the FWHM reach up to 300 nm, which also reveals that concentration quenching of Bi ions did not happen in this experiment. The XPS measurements of the Bi 4f region and the decay time detection for the glass samples were carried out, the results suggest that lower valence Bi (Bi+) contribute to the NIR emission. After annealing in air atmosphere for a period of time, the NIR emission become weaker, and then disappeared in the end, the quenching of the luminescence was attributed to the decreased/disappeared of the lower valence state of Bi (which responsible for the NIR emission). The bandwidth of the emission is so broad (FWHM: 300 nm), which offered the possible applications in low loss O band, E band, and S band of optical fiber.
Acknowledgements This work was supported by National Natural Science Foundation of China under the No. 61422511, and by Natural Science Foundation of Jiangsu Province, China (No. BK20141263). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
[13] [14] [15] Fig. 9. XPS spectra of the Bi 4f region for the sample G-Bi8 (after annealing at 400 °C in air).
Y. Fujimoto, M. Nakatsuka, Jpn. J. Appl. Phys. 40 (2001) 279. I.A. Bufetov, E.M. Dianov, Laser Phys. Lett. 6 (2009) 487. R.F. Wang, J. Liu, Z. Zhang, J. Alloys Compd. 688 (2016) 332. J. Ren, J. Qiu, D. Chen, C. Wang, X. Jiang, C. Zhu, J. Mater. Res. 22 (2007) 1954. V.G. Truong, L. Bigot, A. Lerouge, M. Douay, I. Razdobreev, Appl. Phys. Lett. 92 (2008) 041908. M. Peng, J. Qiu, D. Chen, X. Meng, C. Zhu, Opt. Lett. 30 (2005) 2433. Z.W. Jiang, N.L. Dai, L.Y. Yang, J.G. Peng, H.Q. Li, J.Y. Li, W. Liu, J. Non-Cryst. Solids 383 (2014) 196. M.Yu. Sharonov, A.B. Bykov, V. Petricevic, R.R. Alfano, Opt. Lett. 33 (2008) 2131. X.G. Meng, M.Y. Peng, D.P. Chen, L.Y. Yang, X.W. Jiang, C.S. Zhu, J.R. Qiu, Chin. Phys. Lett. 22 (2005) 615. G. Lakshminarayana, J. Ruan, J.R. Qiu, J. Alloys Compd. 476 (2009) 878. Y. Fujimoto, M. Nakatsuka, J. Non-Cryst. Solids 352 (2006) 2254. V.V. Dvoyrin, V.M. Mashinsky, L.I. Bulatov, I.A. Bufetov, A.V. Shubin, M.A. Melkumov, E.F. Kustov, E.M. Dianov, A.A. Umnikov, V.F. Khopin, M.V. Yashkov, A.N. Guryanov, Opt. Lett. 31 (2006) 2966. J.J. Ren, J.R. Qiu, D.P. Chen, X. Hu, X.W. Jiang, C.S. Zhu, J. Alloys Compd. 463 (2008) L5. M.Y. Peng, X.G. Meng, J.R. Qiu, Q.Z. Zhao, C.S. Zhu, Chem. Phys. Lett. 403 (2005) 410. B. Denker, B. Galagan, V. Osiko, S. Sverchkov, E. Dianov, Appl. Phys. B Lasers Opt. 87 (2007) 135.
38
P. Yu et al. / Journal of Non-Crystalline Solids 464 (2017) 34–38
[16] A.M. Mogus-Milankovic, A. Santic, V. Licina, D.E. Day, J. Non-Cryst. Solids 351 (2005) 3235. [17] Q.C. Sheng, S. Liu, W.T. Li, L.F. Wang, C.M. Tang, D.P. Chen, J. Lumin. 144 (2013) 26. [18] X. Guo, H.J. Li, L.B. Su, P.S. Yu, H.Y. Zhao, J.F. Liu, J. Xu, Laser Phys. 21 (2011) 901. [19] Masoud Eslami, Zohreh Hamnabard, Ali Nemati, J. Rare Earths 31 (2013) 595. [20] X. Meng, J. Qiu, M. Peng, D. Chen, Q. Zhao, X. Jiang, et al., Opt. Express 13 (2005) 1628. [21] S. Polosan, J. Non-Cryst. Solids 355 (2009) 1900. [22] Y.Q. Qui, Q.J. Kang, C.X. Li, X.Y. Dong, C.L. Zhao, Laser Phys. 20 (2010) 487. [23] N. Zhang, J. Qiu, G. Dong, Z. Yang, Q. Zhang, M.J. Peng, Mater. Chem. 22 (2012) 3154. [24] O.V. Usovich, L.A. Trusov, V.V. Lennikov, G. de la Fuente, A.A. Veber, P.E. Kazin, Mater. Lett. 85 (2012) 44.
[25] A.A. Pynenkov, S.V. Firstov, A.A. Panov, E.G. Firstova, K.N. Nishchev, I.A. Bufetov, E.M. Dianov, Quantum Electron. 43 (2013) 174. [26] I.A. Bufetov, M.A. Melkumov, S.V. Firstov, K.E. Riumkin, A.V. Shubin, V.F. Khopin, A.N. Guryanov, E.M. Dianov, IEEE J. Sel. Top. Quantum 20 (2014) 0903815. [27] H.T. Sun, Y. Sakka, M. Fujii, N. Shirahata, H. Gao, Opt. Lett. 36 (2011) 100. [28] A.N. Romanov, Z.T. Fattakhova, A.A. Veber, O.V. Usovich, E.V. Haula, V.N. Korchak, et al., Opt. Express 20 (2012) 7212. [29] Q.C. Sheng, X.L. Wang, D.P. Chen, J. Quant. Spectrosc. Radiat. Transf. 141 (2014) 9. [30] Unifit universal spectrum processing and analysis program for ESCA—spectra version, 2006.