For application of bismuth laser glasses in either fiber amplifier or laser, their performance stability in long run should be understood especially in extreme conditions. However, so far, there are few reports on it. Here, we found, after the cycle experiments on heating and cooling, that the proper increase of lithium content in lithium tantalum silicate laser glass can lead to unusual anti-thermal degradation of bismuth NIR luminescence, which completely differs from the scenario in germanate glass. FTIR, 29Si MAS NMR spectra, absorption and dynamic photoluminescence spectra are employed to unravel how this happens. The results illustrate that it should be due to the decrease of polymerization of silicate glass network, which in turn allows the regeneration at 250°C, and therefore, the content increase of bismuth NIR emission centers. In the meanwhile, we noticed though Bi luminescence can be thermally quenched its peak does not shift along with temperature, which seldom appears in laser materials. The unique property might guarantee the unshift of Bi fiber laser wavelength once such glass was made into fiber devices even as the environmental temperature changes. The role of lithium is discussed in the evolution of glass structures, the suppression of glass heterogeneity, and the thermal stability of Bi luminescence, and it should be helpful to design homogeneous silicate laser glass with outstanding thermal stability.
© 2016 Optical Society of America
Bismuth doped laser glasses were recently recognized as a new member of laser materials with extraordinary luminescence in 1000 to 1800nm, and even to 2000nm where traditional rare earth cannot make [1–12]. Because of this, the research on it was very hot and intensive in the decade, and it leads to the fast emergence of bismuth fiber laser and amplifier [13–16]. Despite of the rapid progress in this, few knowledges are available on the stability of performance of the glasses and fiber devices on long run especially in extreme conditions for instance high temperature or high humidity in Guangzhou, South China . This, however, is very important for practical application. It determines whether we should isolate bismuth fiber or devices from ambient and whether we should cool the gain medium efficiently. The latter becomes more significant for high power lasers because the pump wavelength is usually shorter than the laser wavelength, and the quantum defect between them will eventually evolve into heats, and they can heat the devices up if the devices are not cooled well [1, 13–16, 18]. Since bismuth glasses are one of the key components for fiber devices, it is essential to study the performance of these glasses as exposed to either higher temperatures or humidity, and the knowledge on it should be instructive to design the glass for subsequent fiber drawings and devices.
Recently, we investigated the dependence of thermal degradation of bismuth luminescence on glass compositions in lithium tantalum germanate glasses . As the glasses are heated up to higher temperature and cooled afterwards, no permanent degradation is led as the contents of bismuth or tantalum change. However, addition of lithium will produce permanent irreversible degradation and the degradation becomes worse as lithium content increases in the glass. The loss of luminescence was found due to partial oxidation of bismuth near infrared emission center. It is also noticed that bismuth luminescence exhibits blueshift at higher temperature .
Here, in this work, we found a very different situation in lithium tantalum silicate glass. The proper addition of lithium content can lead to unusual anti-thermal degradation of bismuth NIR luminescence. The luminescence is enhanced gradually rather than permanently weakened after the yoyo experiment of heating and cooling processes. To understand why this could happen, FTIR, 29Si MAS NMR spectra, absorption and dynamic photoluminescence spectra have been measured. We also noticed that as the temperature rises up to 250°C, the emission of bismuth does not move though it is quenched somehow. The role of lithium will be discussed accordingly in the evolution of glass structures, the suppression of glass heterogeneity, and the thermal stability of Bi luminescence.
2. Experimental procedure
All glass samples with molar compositions of (90-x) SiO2-9Ta2O5-1Bi2O3-xLi2O (x = 10, 15, 20, 25, 30, 35, 40) were prepared by a conventional melting and quenching technique. For that, analytical grade reagents SiO2, Li2CO3, and 99.99% Ta2O5 and Bi2O3 were employed as raw materials. 20 g batches which correspond to each glass composition were mixed homogeneously in an agate mortar, and afterwards melted in high pure corundum crucibles at 1580 °C for 20 min in air. The melt was subsequently cast onto a stainless steel plate and pressed with another plate immediately to improve the cooling rate. The obtained glass samples were transparent, and cut into a size of 10 × 10 × 1 mm and polished for consequent optical measurements. Since it has been found that increase of bismuth content can increase the dissolution of aluminum into glass , we kept the content of bismuth constant during this study to suppress the aluminum dissolution and melted the glasses in the same conditions. In this way, the influence of aluminum could be minimized during the investigation on the effect of lithium content.
Optical absorption spectra were recorded on a Perkin Elmer Lambda-900 UV/Vis/NIR spectrophotometer. NIR luminescence spectra were taken on a Zolix Omni λ3007 spectrometer equipped with an InGaAs photodetector and a SR830 Stanford Research lock-in amplifier in a temperature range of 50 to 250°C. The dynamic emission spectra were measured on an Edinburgh FLS 920 spectrofluorometer, using a photomultiplier tube for light detection (Hamamatsu R5509-72). Here, a microsecond-pulsed xenon flashlamp (μF900) with an average power of 60 W was used for the dynamic analyses, while a 450 W Xe lamp was used as the excitation source for steady-state spectra. All the curves were corrected for the spectral response of the detector. 29Si MAS NMR spectra were acquired at 9.4 T at a frequency of 79.49 MHz using a 2.5 μs pulse with a 4 mm MAS probe and 15-20 s recycle delays to prevent saturation. FTIR spectra were recorded on a Bruker Vector 33 spectrometer on samples dispersed homogenously in KBr pellets. Unless otherwise specified, the spectroscopic measurements were performed at room temperature.
3. Results and discussion
Figure 1 illustrates the luminescence spectra of (90-x) SiO2-9Ta2O5-1Bi2O3-xLi2O (x = 10, 15, 20) at high temperatures (50-250°C) during different rounds of heating and cooling upon the excitation of 808 nm. As temperature rises from 50°C to 250°C, the luminescence quenches, and it, however, restores to the initial intensity as the sample is cooled to 50°C. Another round of heating and cooling shows that the intensity restores to 99.4% of the initial state and 0.6% loss of intensity is produced. This happens to the 10% Li2O sample. As lithium content increases to 15%, different scenario occurs. That is, intensity is increased rather than decreased. First round of yoyo experiment leads to 4.4% net gain of intensity as compared to the original states, while it reaches 8.5% after the second round of heating and cooling experiments. Similar situation happens as lithium keeps increasing to 20%. For 20% Li2O sample, the first cycle produces 9.7% net gain of luminescence intensity and the second cycle produces 15.0% gain. This is very different from what we found recently in bismuth doped germanate glass .
In order to understand the mechanism for the anti-thermal degradation, we prepared the glass samples with a broader distribution of lithium contents, and measured FTIR, 29Si MAS NMR, and absorption spectra, and listed as Fig. 2(a). We wished to see the evolution of glass structure along lithium content and tried to find its relationship to what we found above.
For all samples, longitudinal and transverse optic modes of the asymmetric stretching of SiO4 group always dominate the FTIR spectra. Another dominant peak was found at about 640cm−1 in the 10% or 15% Li2O samples and it can be assigned to symmetric stretching of Si-O-Si or Ta-O-Ta [20, 21]. Because of the spectral overlaps, it’s very difficult to distinguish precisely which bonds contribute more to the peak. At least, it should be partially contributed by Ta-O-Ta vibrations. These vibrations also suggest the heterogeneity of these glasses. This consists with the white thin layer we noticed on surface of the samples, and it can be removed after polish, and it should be due to partial phase separation. As lithium content increases, the peak intensity at ~640cm−1 is weakened greatly and it means the phase separation can be suppressed efficiently perhaps by formation of new bonds such as Si-O-Ta, which shows stretching vibration at 990cm−1 . Perhaps, that’s why as Li2O %> 15% we can get the very clear pink glasses where streaks gradually disappear, bismuth can be homogeneously distributed at least visually.
29Si MAS NMR spectra show more clear evolution of silicon tetrahedra (see Fig. 2(a)). Qn stands for a SiO4 group with n bridging oxygen atoms, where n can change between 0 and 4 . According to refs [23, 24], we marked out the locations of Q2 to Q4 with dashed lines to help read the changes along with lithium content. In 10% Li2O sample, Q3 and Q4 are dominant species. As lithium content increases, Q4 disappears gradually while Q3 and Q2 take the dominance in turn (see Fig. 2(a)). The increase of non-bridge oxygen along with lithium content leads to depolymerization of glass network. This is not beneficial to stabilize bismuth NIR emission centers. It can be evidenced by the absorption spectra of glass samples of (90-x) SiO2-9Ta2O5-1Bi2O3-xLi2O (x = 10, 15, 20, 25, 30, 35, 40).
As lithium content increases the strength of typical absorptions [1, 2, 8, 12, 19] of bismuth NIR emission center decreases obviously, meaning the decrease of concentration of bismuth NIR emission center. Correspondingly, we found the photoluminescence is weakened gradually (see Figs. 2(b) and 2(c)). As Li% is higher than 35 or 40, no obvious absorptions can be traced due to these typical emission centers. It can also be noticed that the emission peak gradually shifts to longer wavelength as lithium increases. This should be due to the formation of different type of bismuth NIR emission centers as the absorption spectra depict, where the strongest absorption peaks shift towards shorter wavelength (see Fig. 2(b)). The lifetimes were measured upon different excitations such as 380, 470, 540 and 720nm. They lie between 200 and 400µs and they, however, do not show clear dependence on lithium. In addition, we found the full width at half maximum (FWHM) increases from 220 to 340nm as lithium increases to 30%.
We found the optimal concentration of activator is 1.5% Bi2O3. Here we selected 1% rather than 1.5% for glass melting because the luminescence intensity is proportional to the content of bismuth NIR emission center as the content of bismuth is lower than the critical content. So, the changes of Bi luminescence may reflect the change of bismuth content. Looking back on Fig. 1, we see after the heating and cooling processes, the loss of luminescence intensity is 0.6% for the 10% Li2O glass while the gains are 8.5 and 15.0% for the 15% and 20% Li2O glasses, respectively. This means that the content of bismuth NIR emission should decrease in the 10% Li2O glass, but increase in the latter glasses. We, therefore, measured the absorption spectra of these glasses before and after annealed at 250°C for 30 min, respectively, which consists with the expectation (see Fig. 2(d)). The different impacts lithium has on the thermal degradation of Bi doped germanate and silicate glasses should be partially due to the lower polymerization in germanate because of the preference for Q2 formation and presence of GeO6 units .
Previous works have shown that Bi2O3 which we used as bismuth source can be decomposed thermally at melting temperature, and it can be converted into lower valent ions, atoms, clusters or even nanoparticles of Bi [17, 25]. These particles have been proven not the bismuth NIR emission centers. As for why the content of bismuth NIR emission center increases in the 15% and 20% Li2O samples, it might be partially due to slowly oxidizing them into Bi NIR emission centers. The oxidation reaction can be facilitated by lowering the connectivity of glass network, and, therefore, the increase of oxygen diffuse rate. We unexpectedly noticed as Fig. 2(a) shows, after annealed, that the content of Q4 decreases obviously in 10% Li2O sample but increases in 15% and 20% Li2O samples. Higher polymerization of glass network may stop the oxidization reaction by improved isolation of Bi ions from ambient before they could be oxidized into higher valence states, which are not NIR emission active. In contrast, connectivity lowing might be one of reasons why the intensity is decreased slightly after the yoyo experiments in 10% Li2O sample (see Fig. 1).
Normally for optical materials, emission peak redshifts as the temperature rises up, which should be due to the enhanced interaction between activator and host. What Fig. 1 shows is very different from this. As the temperature increases, the three glass samples behavior in the same way. The emission peak does not shift along with temperature though the intensity is lost in all of them. This has seldom been reported previously. The property may guarantee the stability of Bi output laser wavelength even as the environmental temperature changes if the glass can be made into fiber devices. The thermal quenching of bismuth luminescence can be understood in configurational coordinate diagram similar to ref . The resistance to thermal quenching can also be tailored by management of network polymerization as Fig. 1 reveals.
In all, we observed an unusual anti-thermal degradation of bismuth NIR luminescence in bismuth doped lithium tantalum silicate laser glass upon introduction of proper amount of lithium into it. This is different from the situation in bismuth doped germanate glass. Structural and optical analyses on basis of FTIR, 29Si MAS NMR spectra, absorption and dynamic photoluminescence spectra unravel that it should be due to the regeneration of bismuth NIR emission centers by oxidizing bismuth species which are not NIR emissive. The reaction is promoted by the decrease of glass network connectivity by lithium and inhibited by polymerization increase after annealed. As temperature increases, bismuth emission peak does not shift. This has seldom been reported before. We also noticed that beside the above, proper addition of lithium into the glass can suppress glass heterogeneity and improve the bismuth distribution. We believe this work should be helpful for design of Bi laser glass in future.
National Natural Science Foundation of China (51322208); Guangdong Natural Science Foundation for Distinguished Young Scholars (S20120011380); Department of Education of Guangdong Province (2013gjhz0001), Fundamental Research Funds for the Central Universities, Key Program of Guangzhou Scientific Research Special Project (201607020009), Hundred, Thousand and Ten Thousand Leading Talent Project in Guangdong Program for Special Support of Eminent Professionals.
References and links
1. M. Peng, J. Qiu, D. Chen, X. Meng, I. Yang, X. Jiang, and C. Zhu, “Bismuth- and aluminum-codoped germanium oxide glasses for super-broadband optical amplification,” Opt. Lett. 29(17), 1998–2000 (2004). [CrossRef] [PubMed]
3. L. Su, H. Zhao, H. Li, L. Zheng, G. Ren, J. Xu, W. Ryba-Romanowski, R. Lisiecki, and P. Solarz, “Near-infrared ultrabroadband luminescence spectra properties of subvalent bismuth in CsI halide crystals,” Opt. Lett. 36(23), 4551–4553 (2011). [CrossRef] [PubMed]
4. J. Xu, H. Zhao, L. Su, J. Yu, P. Zhou, H. Tang, L. Zheng, and H. Li, “Study on the effect of heat-annealing and irradiation on spectroscopic properties of Bi:α-BaB2O4 single crystal,” Opt. Express 18(4), 3385–3391 (2010). [CrossRef] [PubMed]
7. Z. Yang, Z. Liu, Z. Song, D. Zhou, Z. Yin, K. Zhu, and J. Qiu, “Influence of optical basicity on broadband near infrared emission in bismuth doped aluminosilicate glasses,” J. Alloys Compd. 509(24), 6816–6818 (2011). [CrossRef]
9. Y. Tian, R. Xu, L. Hu, and J. Zhang, “Effect of chloride ion introduction on structural and 1.5 um emission properties in Er3+-doped fluorophosphate glass,” J. Opt. Soc. Am. B 28(7), 2701 (2011). [CrossRef]
10. D. Chen, Y. Wang, Y. Yu, E. Ma, F. Bao, Z. Hu, and Y. Cheng, “Luminescence at 1.53 um for a new Er3+-doped transparent oxyfluoride glass ceramic,” Mater. Res. Bull. 41(6), 1112–1117 (2006). [CrossRef]
11. I. Bufetov and E. Dianov, “Bi-doped fiber lasers,” Laser Phys. Lett. 6(7), 487–504 (2009). [CrossRef]
12. Q. Sheng, Q. Zhou, and D. Chen, “Efficient methods of obtaining good optical properties in Yb-Bi co-doped phosphate glasses,” J. Mater. Chem. C Mater. Opt. Electron. Devices 1(18), 3067–3071 (2013). [CrossRef]
13. 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, and A. N. Guryanov, “Bismuth-doped-glass optical fibers--a new active medium for lasers and amplifiers,” Opt. Lett. 31(20), 2966–2968 (2006). [CrossRef] [PubMed]
14. E. M. Dianov, S. L. Semjonov, and I. A. Bufetov, “New generation of optical fibres,” Quantum Electron. 46(1), 1–10 (2016). [CrossRef]
15. I. Razdobreev, L. Bigot, V. Pureur, A. Favre, G. Bouwmans, and M. Douay, “Efficient all-fiber bismuth-doped laser,” Appl. Phys. Lett. 90(3), 031103 (2007). [CrossRef]
16. A. A. Pynenkov, S. V. Firstov, A. A. Panov, E. G. Firstova, K. N. Nishchev, I. A. Bufetov, and E. M. Dianov, “IR luminescence in bismuth-doped germanate glasses and fibres,” Quantum Electron. 43(2), 174–176 (2013). [CrossRef]
17. M. Peng, G. Dong, L. Wondraczek, L. Zhang, N. Zhang, and J. Qiu, “Discussion on the origin of NIR emission from Bi-doped materials,” J. Non-Cryst. Solids 357(11), 2241–2245 (2011). [CrossRef]
18. X. Guo, H. Li, L. Su, P. Yu, H. Zhao, Q. Wang, J. Liu, and J. Xu, “Study on multiple near-infrared luminescent centers and effects of aluminum ions in Bi2O3–GeO2 glass system,” Opt. Mater. 34(4), 675–678 (2012). [CrossRef]
19. L. Wang, Y. Zhao, S. Xu, and M. Peng, “Thermal degradation of ultrabroad bismuth NIR luminescence in bismuth-doped tantalum germanate laser glasses,” Opt. Lett. 41(7), 1340–1343 (2016). [CrossRef] [PubMed]
20. D. Pickup, G. Mountjoy, M. Holland, G. Wallidge, R. Newport, and M. Smith, “Structure of (Ta2O5)x(SiO2)1-x xerogels (x = 0.05, 0.11, 0.18, 0.25 and 1.0) from FTIR, 29Si and 17O MAS NMR and EXAFS,” J. Mater. Chem. 10(8), 1887–1894 (2000). [CrossRef]
21. M. Augsburger, E. Strasser, E. Perino, R. Mercader, and J. Pedregosa, “FTIR and Mössbauer investigation of a substituted palygorskite: silicate with a channel structure,” J. Phys. Chem. Solids 59(2), 175–180 (1998). [CrossRef]
22. M. Moesgaard, R. Keding, J. Skibsted, and Y. Yue, “Evidence of intermediate-range order heterogeneity in calcium aluminosilicate glasses,” Chem. Mater. 22(15), 4471–4483 (2010). [CrossRef]
23. P. Dirken, M. Smith, and H. Whitfield, “17O and 29Si solid state NMR study of atomic scale structure in sol-gel-prepared TiO2-SiO2 materials,” J. Phys. Chem. 99(1), 395–401 (1995). [CrossRef]
24. L. Soltay and G. Henderson, “The structure of lithium containing silicate and germanate glasses,” Can. Mineral. 43(5), 1643–1651 (2007). [CrossRef]
25. M. Peng, C. Zollfrank, and L. Wondraczek, “Origin of broad NIR photoluminescence in bismuthate glass and Bi-doped glasses at room temperature,” J. Phys. Condens. Matter 21(28), 285106 (2009). [CrossRef] [PubMed]