We report a reproducible fabrication process ensuring repeatable absorption and emission properties of bismuth (Bi) doped silica-glass optical fibers (BDFs) with single-mode operation above ∼800 nm. BDFs were fabricated through a modified chemical vapor deposition process in-line with the solution-doping technique. Lithium-germanosilicate, yttrium-germanosilicate, and yttrium-phosphosilicate glasses were used as hosts for Bi. Bi and other dopant concentrations were measured by EPMA. ∼0.07wt% of Bi content is obtained with a transparent core. The absorptive and emissive properties of BDFs in the VIS-NIR range (targeted onto 600-900 nm) were studied thoroughly. This study is oriented to the development of VIS/NIR lasers and amplifiers, using BDFs of this type.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Since the first report on lasing capability of BDF at ∼1.2 µm , such fibers became an invaluable candidate for optical fiber lasers and amplifiers for a very broadband (1100-1800nm) range, for which no alternative active fibers, say, based on rare-earths, are known [2,3]. Despite notable progress in the area reached during the last decade, there remains a line of problems behind the physics of BDFs, which complicates their practical use. Among these, a key point is the absence of unambiguous knowledge about the nature of optically active Bi-related centers (further – BACs), formed at doping host glass with Bi.
It has not once been revealed that Bi atoms always exist in multiple oxidation states in silica glass and, in turn, the types of BACs strongly depend on the type of silica-glass host and, also, on the fabricating conditions [4–9]. It is a challenging issue inherent to Bi-doped glasses and BDFs that it is impossible to get BACs having a single oxidation state of Bi. However, using a proper selection of glass-host, the activity of BACs can be tuned towards a targeted application, provided proper measures to be applied to get a selected type of BACs.
Dianov et al.  and W. Xu et al.  reported that Bi-doped silica glasses may demonstrate a line of emission bands within a 350-1750 nm range, with many of these capable of lasing. The most part of research on BDFs was performed in the NIR region. Photoluminescence (PL) within this broad range was reported for Bi doped glasses and fibers, with PL bands inherent to BACs of different types and varying in function of fabrication conditions, core-glass compositions, and applied excitations [1,4,6,8,9–14]. Accordingly, a few types of lasers and amplifiers were invented using BDFs for 1.14-1.17 µm, ∼1.2 µm, ∼1.3 µm, ∼1.4 µm, ∼1.5 µm and ∼1.6-1.8 µm [6,9,15–23].
Till now, the reported BDFs were mainly fabricated using Bi-doped pure-silica (of special design), germano-silicate, phospho-silicate, and alumino-silicate glasses, in analogy to well-developed rare-earth doped optical fibers, mostly employing modified chemical vapor deposition (MCVD) process in-line with solution-doping (SD) technique [1,16,18,20,21]. Besides, a few types of multi-component glasses such as alumino-germano-phospho-silicate , germano-phospho-silicate , alumino-germano-silicate , boro-silicate , lithium-alumino-germano-silicate , yttrium-alumino-silicate [11,23], magnesium-alumino-silicate  etc. were explored as hosts of BDFs.
It is known that solubility of Bi is poor in silica (SiO4) network due to its large atomic size. Although Bi solubility can be increased (but a little) in such modified silica-glass as alumino-silicate, it is still not good enough ; a similar situation is with germano- and phospho-silicates. As a result, a yellowish to brown colored core-glass is frequently founded in Bi-doped preforms with a possible reason being Bi clustering. From a general point of view, this problem can be tackled by employing multicomponent glasses, mentioned above [5–7,9–11,22,23]. According to the base glass technology, with increasing content of a structural modifier, such as lithium (Li), yttrium(Y), magnesium (Mg), etc., interstitial gap in glass network is increased via creation of a large amount of non-bridging oxygens (NBOs) [10,11,22,23]. This facilitates dissolving of atoms having large atomic radii, such as Bi atoms, into silica glass. As an example, it was found [10–12,23] that the addition of Y or Li can be helpful for enhancing Bi solubility in some glass structures. Moreover, this can lead to higher probability of creating of such low-valence species as Bi0, Bi+, and Bi2+. Attractivity of this measure is hard to overestimate, given that namely these oxidation states are believed to be the main structural components of BACs with a high emissivity in VIS/NIR.
The main challenging issue of pursuing research work with Bi doped glasses and fibers is search for the ways to access good solubility of Bi in glass structure, along with controllable generation of suitable content of BACs of proper types. It is known that, given inherently low solubility of Bi in silica glass, concentration of Bi and that of BACs in final optical fiber are poor; as a result, long lengths (∼50-100 m) of fiber are usually required to get effective lasing or amplification using BDFs, which is a disadvantage. Also, it is well documented that the low solubility of Bi-salts in water, alcohol etc. at SD is responsible for non-repeatable and vastly incontrollable doping levels of Bi and BACs [11,12].
Here, we report a repeatable fabrication process of Bi-Li / Bi-Y co-doped germano-silicate (BLGS / BYGS) and Bi-Y co-doped phospho-silicate (BYPS) optical fibers with a 500-700 ppm elemental content of Bi and having transparent core. The cutoff wavelengths of these BDFs were maintained to ∼800 nm or less, which is promising for getting efficient low-threshold single-mode lasing or amplification within a 600-900 nm range in future. A thorough study of the waveguide, absorption, and fluorescent properties of the fibers in the VIS-to-NIR spectral range is presented.
Note that our study was focused on future development of fiber lasers and amplifiers based on such BDFs for the 600-900 nm spectral region, an issue never assessed to-date.
Bi doped optical preforms were fabricated through MCVD process in combination with SD technique.
At first, a pure silica tube (suprasil F300), of outer / inner diameter of 30 / 28 mm, was mounted on MCVD lathe. Then, etching was done by flowing of C2F6, O2 and He gases inside the tube and applying uniform heating at 2050°C by graphite tube furnace to clean up the tube’s inner wall. Subsequently, two sintered barrier layers of pure silica were deposited there, using SiCl4 with O2 at 2050°C, followed by deposition of five un-sintered silica-germanium or silica-phosphorus soot layers, applying a reversed deposition process at 1750°C or 1475°C and gaseous flow of SiCl4-GeCl4 or SiCl4-POCl3 with O2 and He. Silica-germanium or silica-phosphorous layers were used for Li or Y co-doped cores. The depositions of multiple core layers were helpful in getting large core diameters in the preforms and drawing very long lengths (several km) of fiber, employing a single preform. After soot layers’ deposition, a reverse pre-sintering pass was done at 1750°C or 1475°C for silica-germanium or silica-phosphorus soot layers, respectively. Then, the furnace power was switched off and leaved for 30 min. for complete cooling. After that, the tube was cut and placed for vertical SD. For SD, a stable and homogeneous solution of Bi(NO3)3 in HNO3 along with LiNO3 (for generating Li-GS core-glass) or Y(NO3)3 (for generating Y-GS and Y-PS core-glasses) was used. The solution was soaked by soot layers for 7-15 min. The latter allowed to effectively incorporate Bi, Li or Y into pores of the soot layers. After completing SD, the solution was draining out and the tube was fixed in a stand vertically and kept in air, ∼1 hour for drying. Subsequently, the tube was mounted on MCVD lathe but without joined the scrubber end and dried by flow of hot mixture of O2 and He gases for ∼30 min. After this, the scrubber end was joined to start oxidation, which was completed at 850°C to 1050°C using O2 flow. Then, stepwise sintering to sealing was done at full He atmosphere with increasing temperature from 1400°C to 2200°C. Note that He atmosphere at this stage was very helpful to mitigate bubbles formation. The fabricated preforms were over-jacketed by silica tubes using the rod-in-tube technique, managed to accomplish cutoff wavelengths to be at ∼800 nm or below. Finally, the jacketed preforms were drawn into fibers with a high refractive-index (RI) polymer, having 125-µm outer diameter.
2.2 Material analysis
The refractive index profiles (RIPs) of the fabricated Bi doped preforms were obtained via standardized procedure using a preform analyzer (‘P104’, Photon Kinetics). Accordingly, the RIPs of the BDFs were measured by a fiber analyzer (‘S14’, Photon Kinetics). The results of RIPs measurements for the fibers are presented in Fig. 1.
To get the radial distributions of dopants in the Bi-doped preforms, electron micro-probe analysis (EPMA) was performed. Cross-sectional slices of the preforms, having thickness of ∼3 mm, were cut by a low-speed shower. Then, the slices were grinded, using silicon carbide powder of different grain-sizes, to reduce their thickness to ∼2 mm and to smoothen their cross-sectional surface. After that, the slices were carefully polished using cerium oxide sol to get the surfaces suitable for EPMA. Finally, one surface of the samples was coated with a thin carbon layer to avoid charging of the samples during the EPMA measurements. The concentration distributions of dopants Ge, P, Y, and Bi, measured by EPMA; using line scans throughout the samples’ core areas, are presented in Fig. 2; note that tolerance of the measurements was insufficient to get such dependence for Li in preform BLGS and, also, to obtain reliable results for the fibers, because of core-area smallness.
The fibers’ cross-sectional images were taken by an optical microscope (Olympus); the results are shown in Fig. 3.
The wave-guiding properties of the BDFs were studied by standard manner. In experiments, we measured spectral attenuations of fiber samples, placed straight and coiled over a small diameter. After processing the obtained data, the oscillating dependences of relative intensity in the VIS-to-NIR region were found, shown in Fig. 4. The oscillations’ stop wavelengths at ∼830 nm (BYGS) and ∼630 nm (BYPS) are the second transversal mode cutoff positions: beyond these wavelengths, the fibers support single-mode regime of propagation. Note that the oscillations observed in the plots to the anti-Stokes side indicate the spectral positions of cutoff wavelengths for the higher-order modes. We do not show here the result for BLGS fiber as it is very similar to that for BYGS (the cutoff wavelength found in this case is ∼820 nm).
These data, together with the data for the fibers’ core diameters (Fig. 3), allowed us to determine their numerical apertures (N.A.) in the step-index approximation, which were further compared with N.A. values, calculated using RI differences obtained from the BDFs RIPs (Fig. 1).
The basic parameters of the fabricated BDFs are summarized in Table 1.
2.3 Absorption and emission spectra
Absorption spectra of BDFs were measured using the cutback method, employing a deuterium-halogen white light source and two USB-based detectors (Ocean Optics), operated within the 250-1650 nm range. The measurements were made by varying fiber length from 1 m to 10 cm, avoiding bending of the samples.
The base losses of the BDFs were measured by the same method using a NIR photo-receiver but in this case ∼50 m and ∼2 m fibers were used as long-length and short-length samples, respectively. In the experiments, BDFs were coiled for ∼30 cm and ∼50 cm in diameter; such large coils were chosen to minimize the effect of bending-induced loss over the base-material loss. Exploring two different coils helped us to understand the effect of bending upon the base loss of BDFS as well as to suggest a mean to reduce the bending induced loss during real applications. The base losses were measured at 1650 nm for BLGS and BYGS fibers and at 1350 nm for BYPS fiber. These wavelengths were selected to avoid any effect of attenuation owing to resonant absorption of BACs. The results of the measurements are demonstrated in Figs. 5 (a) and (b), respectively.
As seen from Fig. 5(a), all basic spectral signatures of BDFs’ attenuation (the presence of wide and complicate absorptions bands) correspond to the ones well-known for BACs of Bi-Ge, Bi-P, and Bi-Si types [10,11,26,27]. Furthermore, as seen from Fig. 5(b), the base-loss is 214 dB/km for 30 cm coil and 60 dB/km for 50 cm coil for BLGS fiber at 1650 nm, 162 dB/km for BYGS fiber at 1650 nm and 524 dB/km at 1350 nm for BYPS fiber. Low absorption at ∼1100 nm by BACs in BLGS, seen from Fig. 5(b), is also affected by the bending effect. The found base-loss values for the BDFs are discussed in more detail in Section 3.
The BDFs’ PL spectra, obtained at different wavelengths of excitation, are summarized in Fig. 6.
To get the PL spectra, we used a line of low-power fiberized laser diodes (LDs, from JDSU Photonics, Thorlabs, and Innolum) for in-core pumping of BDF samples. The pump wavelengths were: 405, 520, 633, 720, and 905 nm; the colors of the spectra in Fig. 6 provide visual congruence with the spectral domains where the excitation lights fall in. We used in experiments the “forward” detection scheme, in which PL plus pump-light remnants outcoming fiber samples are directly registered by OSA (ANDO 6315A). Lengths (Lf) of the samples under study, as specified in insets to each panel in Fig. 6, were chosen by such way that these provided comparable optical densities (OD = α0Lf) of the fibers, in accord with extinctions at the correspondent wavelengths; refer to Fig. 5(a). For the shown PL spectra, OD was around 2 to 3, thus providing effective pump absorption and yet little effect of PL re-absorption and possible excited-state absorption. All spectra shown in Fig. 6 were recorded at the same power (∼4 mW), at all pump wavelengths, which is, on the one hand, a sufficient pump-level to saturate population inversion and hence PL power in the system of BACs and, on the other hand, adheres to the demand of maximal power at OSA safe exploitation (<10 dBm).
The EPMA data in Fig. 2 reveal nearly homogeneous distributions of the dopants (i.e. Ge, P, Y and Bi) in the fabricated preforms. The core-glasses of all samples were transparent, thus pointing on a good solubility of Bi into the glass network as otherwise these would be expected to be slightly purplish colored. The variations in RI difference (Fig. 1) and N.A. (Table 1) of BLGS and BYGS are mainly due to Ge content variation; accordingly, that of BYPS mainly is due to P content variation: compare the RIPs (Fig. 1) with the dopants’ elemental distributions (Fig. 2).
As it stems from the basic chemistry of glass, Si4+, Ge4+, and P4+ act as glass-network formers in Ge doped (BLGS and BYGS) and P doped (BYPS) fibers, respectively, while Li+, Y3+, and Bin+ (n = 0, 1, 2, etc. but of which exact oxidation state(s) are Bi atoms is not known for certain) act as glass-network generators. These classifications are made depending on the field strength of the contributing ions that compose the glass structure. As also known, Li and Y are responsible for creating a large amount of NBOs in silica network. In turn, the presence of NBOs is helpful for dissolving such larger size atoms as Bi into silica glass through creating complexes with NBOs, having 5 to 7 co-ordination numbers [11,13,14,24,25]. This was the main reason of why we obtained transparent core-glasses with 0.05-0.07 wt.% content of Bi (Table 1). Furthermore, Y-doping is known to underlie enhancing of PL in VIS-to-NIR range , which is one of the targets of our research program.
As seen from the data on wave-guiding features of the BDFs (Fig. 4), these fibers are quite promising for laser/amplifying applications in the 600-900 nm range as providing single-mode or two-mode propagation in this spectral range.
The absorption spectra (Fig. 5(a)) demonstrate the absorption bands centered at ∼ 290, 310-330, 355, 370, 405-415, 445, 500, 530, 560, 615, 725, 775, 820, 880-950, 990-1100, and 1200-1500 nm, all inherent to BACs of Bi-Ge, Bi-P, and Bi-Si types. According to [10,11,26,27], we can ascribe these bands to Bi0, Bi+ and Bi2+ oxidation states; the resume of a spectral analysis for these low-valence Bi species is presented in Fig. 7 and discussed below.
It is observed that the absorption bands of BACs are shifted towards UV-side, from BLGS to BYPS. As known, the number of NBOs created in glass network is gained at transition from GS to PS matrix; besides, the number of NBOs is increased in the presence of Y3+. Expectedly, with increasing the content of NBOs the solubility of Bi into glass host increases. Accompanied to the increased content of NBOs, the overall oxidation degree of Bi during fabrication may increase. This scenario may have caused the revealed trend of BACs absorption bands to slightly shift towards UV, from BLGS to BYPS. Note here that multiple low-oxidation states of Bi (Bi0, Bi+ and Bi2+) are always present in the fabricated BDFs (Fig. 5(a)); however, with changing the NBOs content, their relative concentrations may vary.
The BDFs’ base losses (Fig. 5(b)) were higher when the fibers were coiled for a small bending diameter (30 cm) but when the coil diameter was 50 cm the loss was reduced to 60 dB/km. This loss is quite comparable with the desirable one (<50 dB/km) for getting high laser/amplifying efficiency. According to Fig. 4 and Table 1, the cutoff wavelengths of the synthesized BDFs were found to be ∼820 nm (∼830 nm) for BLGS (BYGS) fibers and ∼630 nm for BYPS fiber; that is, the fibers behave as single-mode (supporting the fundamental LP01 mode) above the respective wavelengths. Another notable point is that one needs to measure the fibers’ base losses well far from the cutoff wavelengths, to avoid a contribution of BACs resonant absorption upon base attenuation. However, these spectral zones are very much prone to be suffered from the bending issues due to dispersion of fundamental mode in the cladding, i.e. poor confinement of LP01 within RIP.
To insight more into the bending issue, we made an analysis of the modal behavior of the BDFs using the experimentally measured RIPs (Fig. 1) by means of software Optiwave (Optifiber mode analysis); the simulated results are presented in Fig. 8.
As seen from the results plotted in Fig. 8, the simulated bending attenuation for the BDFs were obtained as 135 dB/km at 1650 nm for BLGS, 88 dB/km at 1650 nm for BYGS, and 446 dB/km at 1350 nm for BYPS, for bending diameter of 30 cm. These data are quite comparable with the experimentally obtained base attenuation values in Fig. 5(b). Therefore, it can be concluded that such high base attenuations appear mostly due to the bend-induced losses along with material absorption and/or scattering-induced loss.
From Fig. 6, it is seen that the most promising for our current research program, i.e. BDFs’ VIS-NIR applications (refer to the dashed boxes for the 600-900 nm range), is excitation of BDFs by “blue-to-green” (400 to 550 nm) light. Regardless deeper NIR PL (beyond 1 µm) is always effectively excited in the BDFs at either pump wavelength (which may draw a limit, or “bottleneck” for such-kind applications), the figured situation seems to be promising; among the studied BDFs, BYGS fiber is likely to be the better choice in this sense.
Returning to the energy-level diagram (Fig. 7), note that VIS-to-NIR emissions (within the 600-900 nm range) can be tentatively ascribed to the transitions originating from 2P1/2, 2D5/2 manifolds of Bi0, 1D2 or 3P2 manifolds of Bi+, and 2P3/2(1) manifolds of Bi2+ species to their respective ground states. Certainly, the attributions of the transitions imply their entire electronic nature, which is, generally to say, a hypothesis only. In the reality, the species responsible for PL can be more complicate complexes, or “defect” centers, involving nearby locating defects of core-glass, for example, such ones that are inherent to co-doping with Ge, P, Li, and Y in the current (BLGS, BYGS, and BYPS) fibers. Till now, the types of PL-active Bi-related centers, at least, BACs fluorescing beyond 1 µm in BDFs are not identified with certainty (the issue is yet disputing). However, we believe that PL spectrally covering the 600-900 nm domain can be treated with a high confidence as stemming from purely electronic transitions of low-valence oxidation states of Bi (Fig. 7).
A reproducible optimized fabrication process of bismuth doped Li-germanosilicate, Y-germanosilicate, and Y-phosphosilicate (BLGS, BYGS, BYPS) optical fibers is reported, confirmed by the repeatable concentration distributions of dopants along with repeatable absorption and fluorescent properties of Bi active centers (BACs). The good transparency of core area of the fabricated fibers suggests good solubility of Bin+ into the silica network at a relatively high Bi concentration of 0.05-0.07 wt%. The BDFs’ absorption bands in the VIS/NIR domain are argued to arise due to low oxidation states of Bi, viz. Bi0, Bi+, and Bi2+. The shift of resonant absorption as the whole towards UV (from BLGS to BYPS) is inferred to arise owing to increased content of non-bridging oxygen (NBO) in silica network. This tendency may draw a guidance for modifying BACs according to a desirable application by means of tuning the NBOs content in core-glass. The relatively high base attenuations in the BDFs are found to arise mainly due to the bending effect, which is confirmed by experiment and modelling. The developed BDFs have cutoffs at ∼630 nm (BYPS), ∼820 nm and ∼830 nm (BLGS and BYGS, respectively) and hence support single-mode operation beyond these wavelengths. This study may be helpful for further developing single- or few-mode lasers and amplifiers for ∼650 nm and / or ∼830 nm wavelengths using BDFs of these or similar types.
Ministry of Education and Science of the Russian Federation (Minobrnauka) (K3-2018-23).
The authors A. Halder, E. Sekiya and K. Saito acknowledge the help and support of all group members, stuffs of Toyota Technological Institute. The authors are also thankful to the Director of Toyota Technological Institute for supporting this work. A. Halder acknowledges the support of all stuffs of Fiber Optics and Photonics Division, CSIR-Central Glass and Ceramic Research Institute, India, for measuring RIPs of the fibers. A.V. Kir’yanov acknowledges financial support via the Increase Competitiveness Program of NUST “MISIS” of the Ministry of Education and Science (Russian Federation) under Grant K3-2018-23. All the authors acknowledge the efforts of stuff members of Kohoku Kogyo Co., Ltd., Japan, for drawing the fibers, using Toyota Technological Institute’s in-house fabricated preforms.
1. E. M. Dianov, V. V. Dvoyrin, V. M. Mashinsky, A. A. Umnikov, M. V. Yashkov, and A. N. Gur’yanov, “CW bismuth fibre laser,” Quantum Electron. 35(12), 1083–1084 (2005). [CrossRef]
2. I. A. Bufetov and E. M. Dianov, “Bi-doped fiber lasers,” Laser Phys. Lett. 6(7), 487–504 (2009). [CrossRef]
3. 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]
4. V. V. Dvoyrin, A. V. Kir’yanov, V. M. Mashinsky, O. I. Medvedkov, A. A. Umnikov, A. N. Guryanov, and E. M. Dianov, “Absorption, gain, and laser action in Bismuth-doped aluminosilicate optical fibers,” IEEE J. Quantum Electron. 46(2), 182–190 (2010). [CrossRef]
5. 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]
6. I. A. Bufetov, S. V. Firstov, V. F. Khopin, O. I. Medvedkov, A. N. Guryanov, and E. M. Dianov et al., “Bi-doped fiber lasers and amplifiers for a spectral region of 1300-1470 nm,” Opt. Lett. 33(19), 2227–2229 (2008). [CrossRef]
7. M. P. Kalita, S. Yoo, and J. Sahu, “Bismuth doped fiber laser and study of unsaturable loss and pump induced absorption in laser performance,” Opt. Express 16(25), 21032–21038 (2008). [CrossRef]
8. E. M. Dianov, “Bismuth-doped optical fibers: a challenging active medium for near-IR lasers and optical amplifiers,” Light: Sci. Appl. 1(12), e12 (2012). [CrossRef]
9. W. Xu, M. Peng, Z. Ma, G. Dong, and J. Qiu, “A new study on bismuth doped oxide glasses,” Opt. Express 20(14), 15692–15702 (2012). [CrossRef]
10. N. Zhang, J. Qiu, G. Dong, Z. Yang, Q. Zhang, and M. Peng, “Broadband tunable near-infrared emission of Bi-doped composite germanosilicate glasses,” J. Mater. Chem. 22(7), 3154–3159 (2012). [CrossRef]
11. A. Halder, S. K. Bhadra, S. Bysakh, M. C. Paul, and S. Das, “Strong and broad visible emission of Bismuth doped nano-phase separated yttria-alumina-silica optical fibers,” Curr. Nanosci. 12(3), 309–315 (2016). [CrossRef]
12. A. Halder, E. H. Sekiya, and K. Saito, “Fabrication and optical analysis of bismuth doped germanosilicate preforms to develop ∼830 nm fiber laser,” in Advanced Photonics 2018 (BGPP, IPR, NP, NOMA, Sensors, Networks, SPPCom, SOF), OSA Technical Digest, JTu5A.62 (2018).
13. V. Parihar, B. Bhatia, and S. Singh, “Optical absorption study of different concentration of Nd3+ ion in Zinc Lithium Bismuth Borate glass,” Int. Arch. App. Sci. Technol. 5, 65–69 (2014).
14. H. M. Oo, H. Mohamed-Kamari, and W. M. D. Wan-Yusoff, “Optical properties of Bismuth Tellurite based glass,” Int. J. Mol. Sci. 13(4), 4623–4631 (2012). [CrossRef]
15. Y. Fujimoto and M. Nakatsuka, “Optical amplification in bismuthdoped silica glass,” Appl. Phys. Lett. 82(19), 3325–3326 (2003). [CrossRef]
16. E. M. Dianov, A. A. Krylov, V. V. Dvoyrin, V. M. Mashinsky, P. G. Kryukov, O. G. Okhotnikov, and M. Guina, “Mode-locked Bi-doped fiber laser,” J. Opt. Soc. Am. B 24(8), 1807–1808 (2007). [CrossRef]
17. E. M. Dianov, A. V. Shubin, M. A. Melkumov, O. I. Medvedkov, and I. A. Bufetov, “High-power cw bismuth-fiber lasers,” J. Opt. Soc. Am. B 24(8), 1749–1755 (2007). [CrossRef]
18. E. M. Dianov, S. V. Firstov, V. F. Khopin, O. I. Medvedkov, A. N. Guryanov, and I. A. Bufetov, “Bi-doped fiber lasers operating in the range 1470-1550 nm,” Quantum Electron. 39(4), 299–301 (2009). [CrossRef]
19. M. A. Melkumov, I. A. Bufetov, A. V. Shubin, S. V. Firstov, V. F. Khopin, A. N. Guryanov, and E. M. Dianov, “Laser diode pumped bismuth-doped optical fiber amplifier for 1430 nm band,” Opt. Lett. 36(13), 2408–2410 (2011). [CrossRef]
20. A. V. Shubin, I. A. Bufetov, M. A. Melkumov, S. V. Firstov, O. I. Medvedkov, V. F. Khopin, A. N. Guryanov, and E. M. Dianov, “Bismuth-doped silica-based fiber lasers operating between 1389 and 1538 nm with output power of up to 22 W,” Opt. Lett. 37(13), 2589–2591 (2012). [CrossRef]
21. Y. Fujimoto and M. Nakatsuka, “Infrared luminescence from bismuthdoped silica glass,” Jpn. J. Appl. Phys. 40(Part 2), L279–L281 (2001). [CrossRef]
22. B. Denker, B. Galagan, V. Osiko, I. Shulman, S. Sverchkov, and E. Dianov, “Absorption and emission properties of Bi-doped Mg–Al–Si oxide glass system,” Appl. Phys. B 95(4), 801–805 (2009). [CrossRef]
23. M. A. Hughes, T. Suzuki, and Y. Ohishi, “Compositional optimization of bismuth-doped yttria–alumina–silica glass,” Opt. Mater. 32(2), 368–373 (2009). [CrossRef]
24. T. Schaller and J. F. Stebbins, “The structural role of lanthanum and yttrium in aluminosilicate glasses: A 27Al and 17O MAS NMR study,” J. Phys. Chem. B 102(52), 10690–10697 (1998). [CrossRef]
25. A. Halder, M. C. Paul, S. W. Harun, S. K. Bhadra, S. Bysakh, S. Das, and M. Pal, “Visible and near infrared up-conversion luminescence in Yb3+/Tm3+ co-doped yttria-alumino-silicate glass based optical fibers,” J. Lumin. 143, 393–401 (2013). [CrossRef]
26. M. Qian, C. Yu, J. Cheng, K. Li, and L. Hua, “The broadband NIR emission properties of Bi doped La2O3–Al2O3–SiO2 glass,” J. Lumin. 132(10), 2634–2638 (2012). [CrossRef]
27. V. O. Sokolov, V. G. Plotnichenko, and E. M. Dianov, “The origin of near-IR luminescence in bismuth-doped silica and germania glasses free of other dopants: First-principle study,” Opt. Mater. Express 3(8), 1059–1074 (2013). [CrossRef]