Open Access
Add to BibSonomy
Abstract
In this study, a Si defect structure was added into the silica network in order to activate the bismuth and silica structure active center. TD-DFT theoretical simulations show that the Bi and Si ODC(I) models can excite the active center of the E-band at 1408 nm. Additionally, the Bi-doped silica fiber (BDSF) with improved fluorescence was fabricated using atomic layer deposition (ALD) combined with the modified chemical vapor deposition (MCVD) technique. Some tests were used to investigate the structural and optical properties of BDSF. The UV-VIS spectral peak of the BDSF preform is 424 cm−1, and the binding energy of XPS is 439.3 eV, indicating the presence of Bi° atom in BDSF. The Raman peak near 811 cm−1 corresponds to the Bi-O bond. The Si POL defect lacks a Bi-O structure, and the reason for the absence of simulated active center from the E-band is explained. A fluorescence spectrometer was used to analyze the emission peak of a BDSF at 1420 nm. The gain of the BDSF based optical amplifier was measured 28.8 dB at 1420 nm and confirmed the effective stimulation of the bismuth active center in the E-band.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The number of worldwide internet users is increasing at a stable compound rate, as is the use and development of online multimedia applications, there is a large capacity demand for transmission bandwidth. Extending the amplification bandwidth from the current C–L band (1530–1620 nm) to the O–U band (1260–1675 nm) is a potentially simple solution. Numerous attempts have been made to achieve optical amplification of glass hosts incorporating with various dopants such as praseodymium (Pr) [1], neodymium (Nd) [2], thulium (Tm) [3,4] and bismuth (Bi) [5–7]. Especially, luminescence characteristics of bismuth doped silica-based fibers span the broadband near-infrared (NIR) band across O, E, S, and U-bands, which makes bismuth a viable dopant for developing ultra-wideband optical amplifiers [6,8].
In 1999, k. Murara et al first discovered near-infrared fluorescence from Bi and Al co-doped silica glass with a half-height full width of 150 nm and a peak wavelenth of 1150 nm [9]. In 2005, The first bismuth-doped fibers were fabricated by using the traditional modified chemical vapor deposition (MCVD) and solution doping technology, and they were regarded as a starting point for more thorough research in this area [10]. Many reports have shown that bismuth doped on different substrates has different bands of luminescence [11–16]. Among them, the bismuth active centers (BAC) of BAC-Al, BAC-P, BAC-Si, and BAC-Ge are located at 1150,1300,1400, and 1700nm, respectively. Recently, many researchers believe that bismuth with certain valence states induces near-infrared luminescence, such as Bi5+ [17], Bi3+, Bi2+, Bi1+ [18,19] ions and Bi° interstitial molecules [20], Bi2 interstitial dimers and other bismuth clusters [21,22]. Meanwhile, there has been much evidence that Bi2+ fluoresces in the red spectrum range [23] and Bi3+ fluoresces in the ultraviolet to visible spectrum range [24]. Furthermore, Sokolov et al. confirmed that subvalent Bi centers in hosts doped with bismuth valence states lower than the positive tri-valence, which is quite likely the source of the broadband NIR luminescence [25]. On the one hand, the complex luminescence mechanism of Bi materials may exist, and on the other hand, the bismuth doping concentration is low, which cannot be measured by the accuracy of existing instruments [10].
Bi-doped silica fibers have become the most promising gain medium for E-band amplifiers. Yet scientific debate on understanding the nature of bismuth active centers, and mechanisms of bismuth local structure formation in the silica network structure still exist. Benefiting from the rapid development of density functional theory (DFT), it provides a theoretical approach for comprehending the atomic-level interaction between bismuth materials and silica structure. Atomic layer deposition (ALD) has significant benefits in the preparation of bismuth doped silica fibers owing to its high uniformity and conformity in the deposited nano-films [26,27]. This is of great significance to avoid the formation of bismuth doping clusters and achieve a high concentration of bismuth doping.
In this paper, the local structural models of Bi-doped Si and defect Si fiber materials were established for the first time. Based on DFT theory, the interaction between Bi and silica network, and the local field effect of bismuth active center on fluorescence properties of different Si doped and defective Si doped local structures were studied. Subsequently, the Bi doped silica fiber (BDSF) was obtained by the ALD combined with the MCVD technology, the E-band active center of the bismuth microstructure model is analyzed from the structure and optical properties of the bismuth-doped fiber. Based on BDSF, the amplification performance of BDSF is investigated in terms of gain and noise figure.
2. Study on simulation modeling of bismuth silicon active center
2.1 Local structural models of Bi and Si co-doped fiber materials
Some optical and structural properties of BDSF can be determined by sample testing during fiber preparation. However, the conversion fiber fabircated process cannot be observed experimentally in terms of the formation of bismuth active center and the mechanism of light radiation. Therefore, density functional theory (DFT) based on first principles is selected for simulation calculation and analysis [28]. Silica glass possesses an amorphous network structure characterized by both long-range disorder and short-range order. The microstructure of silica glass is predominantly composed of three to six-membered rings (MRs), and other hybrid ring units [29]. Optimizing computational complexity is crucial for handling large atomic–molecular structures. 3MRs as the simplest and most prevalent network structure, play a substantial role and are often utilized in defining the structure and optical properties of silica-doped materials, aligning with economic calculation. Hence, the 3MR structure serves as the foundational unit of the silica glass network for investigating the interaction between BAC and the Si network. Additionally, its influence on the local structure and fluorescence properties is analyzed using DFT. Despite the limited number of atoms in the atomic–molecular structure, it can provide an approximate representation of the band gap distribution range in Bi-doped materials [30,31]. Finally, the energy level structure of Bi-doped material was established.
Initially, the optimization of ground-state local structural models was conducted employing the Becke-3- Lee-Yang-Parr (B3LYP) hybrid function in the Gaussian-09 program [32]. The 6-311G* basis sets were applied for H, O, and Si elements, while the Lanl2dz basis sets were used for Bi [33,34]. Subsequently, the excited state characteristics of the local structures were analyzed utilizing time-dependent density functional theory (TD-DFT) [28].
The interaction of Bi with the silica network can disrupt or rearrange the silica lattice structure, leading to the formation of numerous non-bridging oxygen (NBO) sites in the host glass. Three distinct local structural models for the integration of Bi into silica glass were established based on the presence of 3MRs, as illustrated in Fig. 1. In model (a), Bi functions as a modifier and becomes embedded within the 3MR. In model (b), the Si tetrahedron and Bi collaborate to form a new ring distinct from the 3MR. Model (c) entails direct connection between Bi and the Si tetrahedron through bridging oxygen (BO), resulting in the generation of NBO. The ground-state energy levels of these three structural models were determined through Density Functional Theory (DFT) optimization. This, coupled with the equation for calculating bonding energy (eV), was utilized for further analysis [35]:
Here, n, m, and k denote the number of Si, O, and H atoms in the models, respectively. As shown in Table 1, the calculated bond energies for models (a), (b), and (c) are 6.7776, 6.3231 and 6.7815 eV, respectively. The relatively small bonding energies of the obtained structures can be attributed to the incorporation of Bi into the silica network, rendering the local structures unstable. Notably, model (c) displays the highest bonding energy, signifying its relative stability. Subsequent simulation models are also based on model (c).
Fig. 1. Local structural models of Bi-doped 3MR. (a) Bi embedded into the 3MR; (b) Bi combined with Si tetrahedron; (c) Bi connected with Si tetrahedron by BO.
Table 1. Energy parameters of different doping structures in the 3MR microstructural models
To substantiate the validity of the Bi-doped 3MR, absorption characteristics are scrutinized through TD-DFT analysis. In the bismuth-doped structure model (c), based on the calculations of the ground and excited states, the calculated absorption peak of the Bi-doped 3MR structure appears at ultraviolet wavelength far away from the near infrared active center, as shown in Table 2. There is a discrepancy in the ultraviolet absorption wavelength compared to the earlier experimental findings [36]. We have obtained stable Bi and Si doping local structural model by comparing different structures, but the excited state data are in ultraviolet, and we expect to introduce defects to increase the possibility of excitation of active centers in near infrared.
Table 2. Excited states parameters of Bi and Si doping local structural model
2.2 Local structural models of Bi and defective Si co-doped fiber materials
Based on the stable local structure model (c), POL and ODC(I), two classical silica defect structures, are introduced to further activate the BAC in the NIR. Some researchers also simulated Bi and Si defect structure models to further calculate the excited state structure [37,38]. Compared with the equilibrium state, the defect state has a better activation effect on the NIR active center. Therefore, the addition of bismuth ions will break the original network, causing defects in the Si structure and forming connections. To balance the local electronic valence, Bi ions also change valence states in the coupling process with silica structure. Nevertheless, the incorporation of Bi in the optimal defect structure results in an increased length of the BO bond between the Si tetrahedron and Bi ion. The conventional Si defect manifests as ODC(I) as opposed to POL, indicating the connectivity of Bi to O ion. These two defect structures may destroy the local structures around Bi and Si ions through the production or not of NBO. Therefore, two defect optimization structure model (a) and model (b) of BDSF materials were established, as shown in Fig. 2. Due to defects, the valence state of Bi in the structure tends to zero, which is well consistent with the doping model of Bi° in amorphous Si structures reported in the literature [39].
Fig. 2. Local structural models of Bi and Si defect doped 3MR. (a) Bi connected with Si POL; (b) Bi connected with Si ODC(I).
Excited state parameters for models (a) and (b) of BDSF are computed using TD-DFT. Tables 2 and 3 present electronic transitions, oscillator strengths, and configurations for these models. For model (a), there are mainly excitation energy levels of 0.9198, 1.0062, 1.3787, 1.6548, 2.3519 and 2.8865 eV, and the corresponding oscillator strengths (ƒ) are 0.1539, 0.0492, 0.0816, 0.0046, 0.0473 and 0.0220, respectively (Table 4). The incorporation of Bi and Si into the POL defect structure, based on the 3MR structure, is found not to alter the excited state characteristics of Bi ions near 500, 700, and 800 nm. However, the 1347 nm wavelength in the excited state energy level is not encompassed within the E-band, deviating from the conventional BAC-Si active center.
Table 3. Excited states parameters of Bi and Si POL local structural model
Table 4. Excited states parameters of Bi and Si ODC(I) local structural model
For model (b), there are mainly excitation energy levels of 0.8803, 1.0054, 1.3984, 1.5630, 2.3761 and 2.8313 eV, and the corresponding oscillator strengths (f) are 0.0135, 0.0071, 0.0132, 0.0145, 0.0080 and 0.0044, respectively. Bi ion and Si ODC(I) defect, the excited state energy level becomes 1408 nm in the E-band range. Some researchers have reported that gap Bi° interacts more stable with Si defect ODC(I) than POL, which further supports our study [40].The obvious activation of the active center in the E-band is mainly due to the combination of Bi and Si ODC(I) defect structure, which is the balance between the Si defect structure and the change of Bi ion valence state, thus affecting the surrounding local field. Moreover, the excited state transitions to a wavelength of 1408 nm, with the oscillator strength reaching 0.0135. The results show that the local coordination environment of Si ions can be changed by introducing bismuth into silica glass materials. This results in the coverage of spectral bandwidth in the E-band and the improvement of excitation efficiency. Local structure model (b) is a reliable and valid choice for investigating the impact of Bi co-doping on the local structure and energy levels of Si ions.
2.3 Frontier molecular orbital
In order to further explain the regulating effect of Bi material on silica material, the band structure and electron cloud density distribution of model (a) and model (b) were compared and analyzed, as shown in Fig. 3. The frontier molecular orbital is defined as the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO), and the energy difference between the two is the band gap.
Fig. 3. Energy eigenvalues and electron density distributions of the HOMO and LUMO; (a) Bi and Si POL doped 3MR structural model; (b) Bi and Si ODC(I) doped 3MR structural model.
The corresponding bandgaps of models (a) and (b) are 3.783 and 0.444 eV, respectively, due to the doping of Bi materials [26]. In addition, no matter HOMO or LUMO, the electron cloud distribution density of Bi ion and Si ion binding part in model (b) is significantly higher than that in model (a). The results further show that the introduction of Bi nanomaterials combined with Si ODC(I) defects can not only provide a better valence equilibrium environment, but also enable the charge of local structures to interact with 3MR around Bi ions. As a result, Bi ion activity increases and acts as an active center in conjunction with Si defect structures throughout the local structure. Importantly, in the LUMO state of model (b), after the introduction of ODC(I) defects of Bi and Si, the Si ions are completely covered by electron clouds, which increases the strength of the local crystal field and the active center of Bi and Si activation.
2.4 Energy level
Based on the calculated excited state levels and their corresponding oscillator strengths, the energy level structure of the BDSF material is established, as shown in Fig. 4. The excitation energy is represented by the left axis, which corresponds to the excitation wavelength. The excitation intensity is represented by the oscillator strength f. The black arrow line in the figure represents the absorption process of Bi ions. The solid red lines represent the energy levels emitted by the local structure in the E-band due to the co-doping of the Bi material. The dashed line with the purple arrow shows the excited state transition process. The solid blue arrow line shows the transition luminescence process of Bi and Si local structure. It provides a theoretical foundation for the development of ultra-broad and high-performance Bi-doped silica fiber amplifier.
Further optimization of the BDSF excited state models above will produce fluorescence spectrum characteristics. The conditions of the basis group should be maintained consistent throughout this process. The fluorescence spectra of Bi-doped Si ODC(I) structures were further drawn based on simulation results. When a molecule is excited to a highly excited singlet Sn, the energy level structure data of emission are simplified from Sn to S1, according to the Kasha rule [41]. The energy level data is shown as a spectrum graph when combined with the vibrator intensity, as shown in Fig. 5. Emission characteristics are investigated using TD-DFT. The excited absorption of Bi and the Si-O-Bi-O-Si structure allow the absorption range of Bi-doped 3MR to occur at 432, 500, 700, 800, and 1150-1450 nm, according to computations of the ground and excited states, as shown in Fig. 5(a). Furthermore, fluorescence spectra emerge under excitations at 437, 521, 793, 1123, 1233, and 1408 nm. For short-wavelength excitation, transitioning to the E-band may necessitate multiple energy level transitions instead of a single energy transition, as shown in Fig. 5(b).
Fig. 5. (a) Excited spectrum of Bi and Si ODC(I) doped 3MR structural model; (b) Emission spectrum of Bi and Si ODC(I) doped 3MR structural model
3. Fabrication and characterization of optical fiber samples
3.1 Structural properties
In this work, The BDSF samples were prepared using ALD combined with MCVD technique [27]. Firstly, the loose soot layer containing silica was deposited with MCVD to increase the surface area of bismuth deposition. Then using ALD, the sample of bismuth coated slica sheet was prepared, and the coated base tube in the actual fiber preparation process was dark green, as shown in Fig. 6(a). The high uniformity and conformity bismuth dopants are deposited at low temperature, resulting in the increment of bismuth concentration. Secondly, the deposition of an inner core layer containing silica and phosphorus promotes further binding of bismuth and silica, as shown in Fig. 6(b), which contributes to the formation of an E-band active center.
Fig. 6. Schematic diagram of fiber fabrication process, (a) silica tube after ALD process; (b) Graphite furnace collapses to form optical fiber preforms after deposition of MCVD process; (c) The fiber sample is formed after drawing the fiber, and the cross section of BDSF.
Finally, the fiber preform was put into the drawing tower to make the BDSF samples with core and cladding diameters of approximately 9.7 and 125.5 µm, respectively, as shown in Fig. 6(c). The core composition of the manufactured fiber were Si, O, P, Ge and Bi ions. The relative refractive index difference (RID) of the fiber samples was ascertained using the fiber refractive index analyzer (S14, Photon Kinetics, Beaverton, OR, USA). The RIDs of BDSF was 0.73%. The absorption coefficient at 1310 nm pump wavelength was 2.1 dB/m, and the background loss at 1550 nm was 0.5 dB/m.
Because BDSF preform have a larger core diameter than optical fibers. Some properties can be measured in BDSF preform that cannot be measured in the optical fiber core. Finally, the substrate tube was collapsed at high temperature to form a fiber preform, and a fiber drawing tower was used to draw it into optical fibers. The gain and noise characteristics of the amplifier based on BDSF can directly display the excitation of the active center in E-band.
Using X-ray photoelectron spectroscopy (XPS, Thermo ESCALAB 250 Xi, USA), the chemical composition of BDSF preform samples was studied. A monochromatized Al (Kα) radiation source (hν = 1486.6 eV) was used to excite photoelectrons in an ultrahigh vacuum. The power of the X-ray beam was 200 W, and the spot size was 650 µm. There are two samples, a bismuth doped silica plate sample of 10*10*1 mm and a BDSF preform cut into discs with a diameter of 14 mm and a thickness of 3 mm. The most essential aspect of XPS test spectrum analysis is determining that the carbon peak C(1s) is at 284.8 eV, as shown in Fig. 7. The XPS spectrum shows a typical Bi3+ binding energy peak of 442.0 eV in the Bi(4d) region of the bismuth doped silica sample. The binding energy peak in BDSF preform goes blue to 439.3 eV corresponding to Bi° after being produced into optical fiber preform, as shown in Fig. 7(c) [20]. This difficulty is resolved due to the high doping concentration, making the active center luminescence mechanism of bismuth-doped silica fiber more evident. According to the above study, a reduction process forms the bismuth active center, and Bi° is required in the bismuth active center.
Fig. 7. XPS profile of the Bi materials deposited. (a) Bi deposited at silicon substrate after ALD process; (b) BDSF preform slice after MCVD process; (c) Changes of binding energy in two samples Bi(4d); (d) Changes of binding energy in two samples Bi(4f).
A Lambda 1050 + spectrophotometer (Perkin-Elmer, USA) was used to measure the UV-VIS absorbance spectra. Bismuth oxide samples generated over 1400, 2800, and 4000 ALD cycles, as well as bismuth doped fiber preform samples created using ALD and MCVD techniques, were studied, as shown in Fig. 8(a). The equivalent absorption peak of the bismuth oxide layer is observed at 232 nm [42], matching to Bi3+ in Bi2O3, which agrees with the XPS data. The absorption peak of the bismuth-doped fiber preform sample is 424 nm, which corresponds to the bismuth nanoparticles with a diameter less than 20 nm after the Bi3+ reduction reported in the literature [43], as shown in Fig. 8(b). The valence state of bismuth ions in a Bi doped fiber sample shifts from Bi3+ to Bi°. The results agree with those obtained by XPS.
Fig. 8. The UV-VIS spectrum of (a) Bi deposited sample after ALD and (b) BDSF preform after MCVD process.
Raman spectrometer (LabRam HR800, HORIBA Jobin Yvon, France) was utilized to characterize the vibrations of the molecular structure in BDSF samples. The samples were excited with a 532 nm laser light source with a power of 4.5 mW. The Raman spectra of bismuth-doped silica optical fiber preform was compared to those of BDSF samples. Measure the BDSF preform sample and the 1 m length of BDSF sample stated previously, as shown in Fig. 9. Because the BDSF sample is impacted by fluorescence during the Raman test, the test data was subjected to baseline processing. The Raman spectra of typical silica glass is characterized by a reasonably strong diffuse band at 440 cm−1, with lesser characteristics between 800, 1060 and 1190 cm−1. Furthermore, there are two relatively sharp bands at 492 and 605 cm−1, represented by the defect bands D1 and D2, respectively [44]. The vibrations of the classical Raman peaks of the Bi-O structure are represented by peaks at 811 cm−1 [45], however, because the Raman peak associated with Si-O is likewise at 800 cm−1, the peak value is moved after doping Bi, and the peak band is broader. The peaks at 1000 and 1323 cm−1 [46,47] represent the hydroxyl (-OH) and P = O structures, respectively. The Raman peaks on both Bi doped sample materials are consistent, indicating that the optical properties of the bismuth-doped silica fiber preform were effectively preserved even after the fiber drawing process.
3.2 Fluorescence properties
The photoluminescence (PL), PL excitation (PLE), and PL decay data were obtained using a spectrophotometer (FLS-980, Edinburgh Instruments, UK) with a Xe lamp with a power of 450 W. Measure the BDSF preform sample as well as fifty 5 cm removed coated BDSF samples. FLS-980 is equipped with visible-photomultiplier (PMT) and NIR-PMT detection range at at 200∼900 nm and 500∼1700nm photomultiplier tube detectors. The excitation peaks of both optical fiber preform and optical fiber samples occur at 1420 nm. Additionally, peaks are observed at 424, 432, and 820 nm in the 300-900 nm range, and at 1075, 1183, and 1264 nm in the 900-1300 nm region, as shown in Fig. 10(a) and (b). The largest intensity excitation wavelength for 1420 nm emission wavelength is 424 nm, while the excitation wavelength closest to the emission wavelength is 1264 nm. The emission spectrum of the preform is identical to that of the fiber sample at 424 nm excitation, covering the range of 1260-1460 nm. E-band emission intensity is greater, as shown in Fig. 10(c). Furthermore, for an excitation peak of 1264 nm, both samples had emission centers at 1420 nm. At 1420 nm, the fluorescence lifetime recorded with a 424 nm excitation was 674 µs, which was slightly longer than the reference parameter of 630 µs [36], it indicating more efficient excitation and emission. According to the results of the testing, the E-band bismuth active centers were activated. Meanwhile, the BDSF preform and BDSF sample have consistent structural and optical properties, as shown in Fig. 10(d).
Fig. 10. Fluorescence and excitation fluorescence spectrum of BDSF preform after MCVD process and fiber sample. (a) The emission spectra of 372, 424 and 825 nm excited; (b) The emission spectra of 1075, 1183 and 1264 nm excited; (c) The BAC-P and BAC-Si of emission spectra; (d) Comparison of the emission of BDSF preform after MCVD process and fiber sample.
Since it is difficult to obtain optical fiber XPS test data, only XPS results for BDSF preform samples with higher cross-sectional areas may be examined. However, when the Raman, excitation, and emission spectra of the samples were analyzed, it was determined that the structural properties of the BDSF preform and the BDSF sample were consistent. Through XPS and UV-VIS testing, we speculate that the valence state of bismuth is reduced from Bi3+ to Bi°. To summarize, the bismuth in the optical fiber also has a 0 valence state.
3.4 Gain properties
To further validate the optical performance of BDSF, its optical amplification was thoroughly measured. The test system resembled the Ref. [48]. As pump sources, 1310 nm fiber laser devices (LDs) with an output power of 1059 mW were utilized. A tunable laser source (TLS-550, Santec, Japan) with a linewidth of 200 kHz and a wavelength range of 1360-1485 nm was used to generate the signal. ISOs were employed to protect the pump and the tunable laser source. The pump and signal wavelengths were combined and separated using wavelength division multiplexers (WDMs). The single pump double pass arrangement was built using two circulators working in the E-band.
BDSF fluorescence was initially measured. The BDSF with a length of 30 m has good fluorescence properties in the E-band. At 1420 nm, the fluorescence intensity reached a maximum of -6 dBm at 2 nm resolution. Furthermore, the influence of the pump and signal power on the gain characteristics of BDSF was explored, as shown in Fig. 11(a). Notably, The bandwidth of BDSF with a gain exceeding 25 dB occurs at signal optical powers of -23 dBm and -10 dBm, corresponding to spectral widths of 40 nm and 25 nm, respectively. The gain at 1420 nm with varied input signal powers was also measured using a pump power of 1059 mW. A maximum gain of 28.8 dB was obtained for an input signal strength of -38.1 dBm, and the corresponding noise figure was measured 5.6 dB, as shown in Fig. 11(b). Considering both gain and gain bandwidth, this observation further underscores the excitation of the BAC-Si.
Fig. 11. (a) Fluorescence and gain characteristics of BDSF for signal powers of -10, and -23 dBm; (b) Gain and noise figure characteristics of BDSF at the signal wavelength of 1420 nm for different signal powers.
4. Conclusion
In summary, we used a Si defect structure to activate the near infrared BAC-Si. We calculate the local TD-DFT and discover that defects cause the bismuth valence state to be changed into Bi° in Bi and Si ODC(I) models, which can excite the E-band active center at 1408 nm. In addition, we have prepared BDSF with excellent fluorescence and gain characteristics by combining ALD with MCVD technology. We measured the peaks of UV-VIS absorption spectrum and XPS combined energy spectrum are 424 cm−1 and 439.3 eV, respectively. Moreover, the absence of the active center of the POL defect of Si in the E-band is corroborated by the Bi-O bond in the Raman test, indirectly affirming the reliability of the simulated local structure of Bi-doped silica fiber. Finally, in the measured BDSF sample, the excitation of multiple wavelengths was found to be in the active center of 1420 nm at E-band. To further verify the effective excitation of bismuth active center in E-band. The optical amplifier of BDSF obtained a small signal gain of 28.8 dB at 1420 nm. Analyzing and validating the BAC-Si contributes to an enhanced comprehension of its characteristics, thereby facilitating the manufacturing and utilization of BDSF.
Funding
National Key Research and Development Program of China (2020YFB1805800).
Acknowledgment
We appreciate the High Performance Computing Center of Shanghai University, and Shanghai Engineering Research Center of Intelligent Computing System. (No. 19DZ2252600).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Y. Ohishi, T. Kanamori, T. Kitagawa, et al., “Pr 3+-doped fluoride fiber amplifier operating at 1.31 µm,” Opt. Lett. 16(22), 1747–1749 (1991). [CrossRef]
2. R.-H. Cheng, Z.-H. Qi, T.-Q. Wu, et al., Amplification characteristics of Nd-doped fluoride glass fiber at 1.3-um wavelength, Photonics West ‘97(SPIE, 1997), Vol. 2996.
3. T. Sakamoto, M. Shimizu, M. Yamada, et al., “35-dB gain Tm-doped ZBLYAN fiber amplifier operating at 1.65 µm,” IEEE Photon. Technol. Lett. 8(3), 349–351 (1996). [CrossRef]
4. S. Aozasa, H. Masuda, H. Ono, et al., “1480-1510 nm-band Tm doped fiber amplifier (TDFA) with a high power conversion efficiency of 42%,” in OFC 2001. Optical Fiber Communication Conference and Exhibit. Technical Digest Postconference Edition (IEEE Cat. 01CH37171), (IEEE, 2001), PD1-PD1.
5. E. M. Dianov, “Amplification in Extended Transmission Bands Using Bismuth-Doped Optical Fibers,” J. Lightwave Technol. 31(4), 681–688 (2013). [CrossRef]
6. N. K. Thipparapu, Y. Wang, S. Wang, et al., “Bi-doped fiber amplifiers and lasers [Invited],” Opt. Mater. Express 9(6), 2446 (2019). [CrossRef]
7. Y. Wang, N. K. Thipparapu, D. J. Richardson, et al., “Ultra-Broadband Bismuth-Doped Fiber Amplifier Covering a 115-nm Bandwidth in the O and E Bands,” J. Lightwave Technol. 39(3), 795–800 (2021). [CrossRef]
8. Y. Luo, J. Wen, J. Zhang, et al., “Bismuth and erbium codoped optical fiber with ultrabroadband luminescence across O-, E-, S-, C-, and L-bands,” Opt. Lett. 37(16), 3447–3449 (2012). [CrossRef]
9. K. Murata, Y. Fujimoto, T. Kanabe, et al., “Bi-doped SiO2 as a new laser material for an intense laser,” Fusion Eng. Des. 44(1-4), 437–439 (1999). [CrossRef]
10. V. Dvoyrin, V. Mashinsky, E. Dianov, et al., “Absorption, fluorescence and optical amplification in MCVD bismuth-doped silica glass optical fibres,” in 2005 31st European Conference on Optical Communication, ECOC 2005, (IET, 2005), 949–950.
11. I. A. Bufetov, M. A. Melkumov, S. V. Firstov, et al., “Optical gain and laser generation in bismuth-doped silica fibers free of other dopants,” Opt. Lett. 36(2), 166–168 (2011). [CrossRef]
12. B. Denker, B. Galagan, V. Osiko, et al., “Luminescent properties of Bi-doped boro-alumino-phosphate glasses,” Appl. Phys. B 87(1), 135–137 (2007). [CrossRef]
13. T. Haruna, A. K. Dutta, J. Iihara, et al., “Bismuth-doped silicate glass fiber for ultra-broadband amplification media,” in Active and Passive Optical Components for Communications VI, (2006).
14. S. Zhou, H. Dong, H. Zeng, et al., “Broadband optical amplification in Bi-doped germanium silicate glass,” Appl. Phys. Lett. 91(6), 1 (2007). [CrossRef]
15. J. Bao, S. Zhou, G. Feng, et al., “Luminescence properties of nickel and bismuth co-doped barium aluminosilicate glasses,” J. Alloys Compd. 456(1-2), 239–242 (2008). [CrossRef]
16. S. V. Firstov, S. V. Alyshev, K. E. Riumkin, et al., “A 23-dB bismuth-doped optical fiber amplifier for a 1700-nm band,” Sci. Rep. 6(1), 28939 (2016). [CrossRef]
17. Y. Fujimoto and M. Nakatsuka, “27Al NMR structural study on aluminum coordination state in bismuth doped silica glass,” J. Non-Cryst. Solids 352(21-22), 2254–2258 (2006). [CrossRef]
18. X. G. Meng, J. R. Qiu, M. Y. Peng, et al., “Near infrared broadband emission of bismuth-doped aluminophosphate glass,” Opt. Express 13(5), 1628–1634 (2005). [CrossRef]
19. V. G. Truong, L. Bigot, A. Lerouge, et al., “Study of thermal stability and luminescence quenching properties of bismuth-doped silicate glasses for fiber laser applications,” Appl. Phys. Lett. 92(4), 1 (2008). [CrossRef]
20. J. Ren, L. Yang, J. Qiu, et al., “Effect of various alkaline-earth metal oxides on the broadband infrared luminescence from bismuth-doped silicate glasses,” Solid State Commun. 140(1), 38–41 (2006). [CrossRef]
21. S. Khonthon, S. Morimoto, Y. Arai, et al., “Luminescence characteristics of Te-and Bi-doped glasses and glass-ceramics,” J. Ceram. Soc. Japan 115(1340), 259–263 (2007). [CrossRef]
22. M. Peng, J. Qiu, D. Chen, et al., “Superbroadband 1310 nm emission from bismuth and tantalum codoped germanium oxide glasses,” Opt. Lett. 30(18), 2433–2435 (2005). [CrossRef]
23. M. de Jong, A. Meijerink, R. A. Gordon, et al., “Is Bi2+ Responsible for the Red-Orange Emission of Bismuth-Doped SrB4O7?” J. Phys. Chem. C 118(18), 9696–9705 (2014). [CrossRef]
24. J. Li, J. Liu, and X. Yu, “Synthesis and luminescence properties of Bi3+-doped YVO4 phosphors,” J. Alloys Compd. 509(41), 9897–9900 (2011). [CrossRef]
25. 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, 1059 (2013). [CrossRef]
26. X. Pan, Y. Dong, J. Zheng, et al., “Enhanced FBG Temperature Sensitivity in PbS-Doped Silica Optical Fiber,” J. Lightwave Technol. 37(18), 4902–4907 (2019). [CrossRef]
27. Y. Chen, W. Wang, Y. Yang, et al., “Near 0.5 dB gain per unit length in O-band based on a Bi/P co-doped silica fiber via atomic layer deposition,” Opt. Express 31(1), 1 (2023). [CrossRef]
28. M. Atanasov, C. Daul, H.-U. Güdel, et al., “Ground States, Excited States, and Metal− Ligand Bonding in Rare Earth Hexachloro Complexes: A DFT-Based Ligand Field Study,” Inorg. Chem. 44(8), 2954–2963 (2005). [CrossRef]
29. T. Y. Wang, J. X. Wen, W. Y. Luo, et al., “Influences of irradiation on network microstructure of low water peak optical fiber material,” J. Non-Cryst. Solids 356(25-27), 1332–1336 (2010). [CrossRef]
30. M. C. Weidman, “Controlling nanomaterial self-assembly for next generation optoelectronic applications,” (Massachusetts Institute of Technology, 2016).
31. A. M. Smith and S. J. A. O. C. R. Nie, “Semiconductor nanocrystals: structure, properties, and band gap engineering,” Acc. Chem. Res. 43(2), 190–200 (2010). [CrossRef]
32. J. L. F. Da Silva, M. V. Ganduglia-Pirovano, J. Sauer, et al., “Hybrid functionals applied to rare-earth oxides: The example of ceria,” Phys. Rev. B 75(4), 045121 (2007). [CrossRef]
33. Y. Dong, J. Wen, F. Pang, et al., “Formation and photoluminescence property of PbS quantum dots in silica optical fiber based on atomic layer deposition,” Opt. Mater. Express 5(4), 712 (2015). [CrossRef]
34. L. Maron and P. Sabatier, “Do f electrons play a role in the lanthanide− ligand bonds? A DFT study of Ln (NR2) 3; R = H, SiH3,” J. Phys. Chem. A 104(30), 7140–7143 (2000). [CrossRef]
35. X. Sun, J. Wen, Q. Guo, et al., “Fluorescence properties and energy level structure of Ce-doped silica fiber materials,” Opt. Mater. Express 7, 751–759 (2017). [CrossRef]
36. Y. F. Y. Fujimoto and M. N. M. J. J. J. O. A. P. Nakatsuka, “Infrared luminescence from bismuth-doped silica glass,” J. Appl. Phys. 40, L279 (2001). [CrossRef]
37. X. Li, B. Wang, T. Song, et al., “Aggregation of ODC(I) and POL Defects in Bismuth Doped Silica Fiber,” Micromachines 14(2), 358 (2023). [CrossRef]
38. S. Sun, B. Jia, B. Yan, et al., “Optical absorption of Bi2+-ODC(II) active center in Bi-doped silica optical fiber,” J. Lumin. 213, 304–309 (2019). [CrossRef]
39. J. Zhang, L. Han, Z. Guan, et al., “Electronic and luminescence characteristics of interstitial Bi0 atom in bismuth-doped silica optical fiber,” J. Lumin. 207, 346–350 (2019). [CrossRef]
40. L. Han, W. Xiao, J. Zhang, et al., “Investigations of interstitial Bi0 interacting with intrinsic defects in bismuth-doped silica optical fiber,” Appl. Phys. A 125(9), 612 (2019). [CrossRef]
41. M. J. D. O. T. F. S. Kasha, “Characterization of electronic transitions in complex molecules,” Discuss. Faraday Soc. 9, 14–19 (1950). [CrossRef]
42. C. Li, K. P. O’Halloran, H. Ma, et al., “Multifunctional multilayer films containing polyoxometalates and bismuth oxide nanoparticles,” J. Phys. Chem. B 113(23), 8043–8048 (2009). [CrossRef]
43. Z. Wang, C. Jiang, R. Huang, et al., “Investigation of optical and photocatalytic properties of bismuth nanospheres prepared by a facile thermolysis method,” J. Phys. Chem. C 118(2), 1155–1160 (2014). [CrossRef]
44. R. J. Hemley, H. K. Mao, P. M. Bell, et al., “Raman spectroscopy of SiO2 glass at high pressure,” Phys. Rev. Lett. 57(6), 747–750 (1986). [CrossRef]
45. Q. Chen, B. Miao, and Q. Ma, “Sb2O3 functionalized Plasmon, photoluminescence and Faraday rotation in glass,” J. Non-Cryst. Solids 545, 120251 (2020). [CrossRef]
46. Z. Q. Du, Y. Xu, and W. D. Yu, “Analysis of Characters and Structures of Polylactic Fiber,” Adv. Mater. Res. 236-238, 1089–1092 (2011). [CrossRef]
47. P. Mapara, A. Godbole, R. Swarup, et al., “Extraction of uranium and plutonium from oxalate bearing solutions using phosphonic acid: Solvent extraction, extraction chromatography and infrared studies,” J. Radioanal. Nucl. Chem. 240(2), 631–635 (1999). [CrossRef]
48. N. K. Thipparapu, Y. Wang, A. A. Umnikov, et al., “40 dB gain all fiber bismuth-doped amplifier operating in the O-band,” Opt. Lett. 44(9), 2248–2251 (2019). [CrossRef]
