Open Access
Add to BibSonomy
Abstract
The complete removal of the impurities like Se-H in Se-based chalcogenide glasses has been challenging in the development of highly transparent chalcogenide glass fiber. In this paper, several purification methods, including dynamic distillation, static distillation, and combined distillation method, were adopted with an aim of purifying arsenic selenide glass with ultra-low content of the impurities. The experimental results demonstrated that the Se-H can be completely eliminated in the arsenic selenide glass host and fiber without the introduction of any chloride. We further explored the applications of such low loss and Se-H-free chalcogenide glass fiber in the mid-infrared. It was found that, using such a Se-H free fiber, a flattened supercontinuum spectrum above the -30 dB level from 1.2 to 13 µm was generated from the Se-H free fiber with a 5.5 µm laser pumping. The sensitivity was found to be improved 5.1 times for CO2 gas in the 3 to 6 µm wavelength range.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The fundamental vibration absorptions of most gas molecules are located at the mid-infrared (MIR) region from 3 to 6 µm. Therefore, air pollutants such as CO, CH4, C6H6, SO2, etc., and greenhouse gases that cause global warming such as carbon dioxide (CO2), methane (CH4), etc. [1]. can be detected in MIR using optical method. For example, CO2 is the most frequently detected greenhouse gas because of its nondegradability and high concentration in the atmosphere [2], and its molecular vibration absorption peaks are located at 4.25 µm (2350 cm-1) and 15 µm (666 cm-1). For gas sensing, a bright and broad light source in MIR is essential [3,4], where a fiber-based supercontinuum (SC) could be one of the ideal choices due to its fine coherence, high brightness and broad bandwidth. Nevertheless, a low-loss MIR optical fiber is essential in both cases of SC source and fiber sensing.
Commercial silica fiber is not suitable for MIR applications due to its strong intrinsic material absorption above 3 µm. Currently, most MIR fibers available are fluoride [5], tellurite [6], and chalcogenide fiber [7]. Compared with the former two fibers, the chalcogenide fiber has the advantages of the broadest transmission window and highest nonlinearity. Depending on the elements contained, S-, Se- and Te-based chalcogenide glasses have a transparent range up to a wavelength of 12, 15 and 20 µm, respectively. As2Se3 fiber is one of the most frequently used MIR fibers in many fields including MIR SC generation [8,9] and fiber sensing [10,11], due to its relatively low optical loss, low material cost and highly chemical stability [12–15]. However, As2Se3 fibers suffer from extrinsic absorption band of Se-H at 4.57 µm associated with residual hydrogen impurities, despite of the fact that 6N elements are used as starting materials for glass preparation. [15,16]. It was clear that the Se-H impurity absorption of arsenic selenide fiber will affect its application on the MIR fiber sensing as well as SC generation. If the concentration of Se-H impurity peak at 4.57 µm is further reduced, the fiber will be benefited much for air pollutants sensing and SC generation. [17,18], static and dynamic distillation have been comprehensively used to prepare high purity glasses. Solid particles such as carbon can be removed by static distillation, while the bubbles and volatile compounds such as C-S, water and hydrocarbons can be reduced by dynamic distillation [10]. Currently, several commercial chalcogenide glass fiber suppliers are CorActive in Canada, IRFlex in America, SelenOptics and Le Verre Fluoré in France, and the minimum Se-H content is 0.2 ppm in the fiber produced by CorActive.
In this work, we reported complete removal of Se-H impurity in As2Se3 bulk glass via a novel glass purification process. The corresponding fiber fabricated by a homemade isolated stacked extrusion facility exhibits a very low value of optical loss which is less than 0.5 dB/m at a wavelength range from 3 to 6 µm without any observable Se-H trace. The fiber shows a broad SC spectrum without any decreasing SC intensity induced by Se-H absorption, and the sensitivity of hydrogen-free fiber for fiber sensing is 5.1 times higher than that of the fiber containing hydrogen impurities. All these demonstrate great potentials of the hydrogen-free fibers in the development of ultra-sensibility gas sensing in the 3.3-6 µm region and high coherent light source.
2. Experiment
In this work, As2Se3 and Ge1As39Se60 glass were selected as the core and cladding materials of the fiber respectively. Ge element was added to the cladding material to reduce the refractive index of the cladding, so as to meet the requirements of the step index fiber. The glass melt was homogenized in a rocking furnace mode at a temperature of 650 °C for 4 h. The melt was then cooled to 500 °C at a rate of 2 deg./min, quenched in air. To reduce or eliminate the permanent stress in the glass, it is necessary to anneal the glass ingots. The ampoule was subsequently returned to the furnace for annealing at ∼170 °C for 15 h, about 10 °C below the glass transition temperature of the glass.
2.1. Glass preparation
(a) Pretreatment of quartz ampoules and raw materials
High purity quartz distillation ampoules were soaked in the aqua regia for 7-8 hours and then cleaned several times with deionized water. The ampoules were dried in a vacuum oven at 700 °C for 12 hours.
Commercially available 99.999% (5N) purity of arsenic and selenium were used in all purification schemes in this work. The raw material of Se was pre-purified under static-vacuum (enclosed vacuum surrounding as shown in Fig. 1(a)) at 650 °C for 8 hours, with appropriate amount of Mg strips were introduced as deoxidizer in the purification process.
(b) Glass purification under Static-vacuum (G0-G2)
Fig. 1. (a) Schematic diagram of raw material and G0-G2 glasses purification based on static vacuum purification; (b) schematic diagram of removing Se-H impurities by dynamic vacuum purification.
Figure 1(a) shows the schematic diagram of As2Se3 glass prepared by static vacuum purification method. The raw materials (As, Se, Mg) were loaded into the reaction tube A in a glove box filled with nitrogen atmosphere. Then the ampules were placed in a two-zone furnace for distilling-purification, and then evacuated down to 5 × 10−5 Torr. The glass melt was distilled from the high temperature zone (tube A: 700 °C) to the low temperature zone (tube B: 400 °C) due to the big difference in their respective saturated vapor pressure. After 20 hours of reaction, the oxides and other impurities were kept at the reaction ampule (tube A) due to their ultra-low saturated vapor pressure.
Sample G0 was thus obtained by the direct static-vacuum glass purification using As and Se elements as starting materials without any pre-treatment. Samples G1 and G2 were prepared by the same static-vacuum glass method, but the Se raw materials used in G1 or G2 were purified by single or twice processes of static distillation, respectively.
(c) Dynamic-vacuum glass purification (G3)
Sample G3 was prepared via a dynamic-vacuum distillation process, and the schematic diagram of the whole process is shown in Fig. 1(b). The Se obtained from the twice static purification and other initial materials (As and Mg) were put into the reacting silica tube B in a glove box filled with nitrogen atmosphere. The batch was preheated at 100 °C for 2 hours to remove moisture from the surface of the raw material. The batch was finally heated to 650 °C and the whole process was sustained by vacuum pumping. The ampule was evacuated at 3.1 × 10−5 Torr and then sealed with hydrogen-oxygen flame.
(d) Dynamic-vacuum distillation + static-vacuum distillation (G4)
Using Se that was treated by twice static distillation, the sample G4 was prepared via static-vacuum distillation after dynamic-vacuum distillation. Mg strips were introduced in both distillation processes.
(e) Static-vacuum distillation + dynamic-vacuum distillation (G5)
Sample G5 was prepared using same conditions as G4 except that the purification process is dynamic-vacuum distillation after static-vacuum distillation.
(f) Dynamic-vacuum distillation + static-vacuum distillation + dynamic-vacuum distillation (G6)
Sample G6 was prepared with additional dynamic-vacuum distillation compared with G4. Purification methods of the several glass samples are listed and described in Table 1.
Table 1. Descriptions of the purification of the several glass samplesa
2.2. Fiber preform fabrication and fiber drawing
The arsenic selenide optical fiber with step index was fabricated by a preform-drawing process. Two glass rods, with diameters of core 9 mm and cladding of 46 mm and lengths of both 15 mm, were extruded via an optimized isolated loop extrusion method [19]. In the first-step extrusion, the core glass was forcibly inserted into the cladding glass under the force of about 5 kN. Then the extrusion stem was pushed down at a steady rate of about 3 mm/min, and the core and the cladding glass were slowly passed together through a tapered outlet of die. Finally, the preform with a good core-cladding interface was obtained, and the core-cladding ratio was about 1:10. Vacuum evacuation and Ar atmosphere protection were used in the extrusion process to ensure high-quality optical fiber preform. In this work, PES (polyethersulfone) polymer film was selected as the protective layer of preform to enhance the flexibility of optical fiber [20]. Then a preform with a length of 5 cm was fixed on the homemade drawing tower for fiber drawing. In order to prevent crystallization, slow drawing (0.20 cm/min) with a relatively low temperature of about 258 °C was adopted. In addition, Ar gas was continuously filled in the whole process of fiber drawing to reduce the risk of preform surface oxidation at high temperature. Finally, a step-index fiber with 350 µm outer-diameter was successfully fabricated. The core diameter of the fiber was about 36 µm.
2.3. Glass and optical fiber properties testing
Crystallization temperature (Tx) and glass transition temperature (Tg) were determined by differential scanning calorimetry (DSC) (TAQ2000). Vis/NIR spectrophotometer (Perkin-Elmer Lambda 950) was used to record the absorption spectra in a wavelength range of 0.4-2.5 µm. Fourier transform infrared spectrometer (Nicolet380) was used to measure the transmission in the wavelength range of 2.5-25 µm. The content of the impurity was extracted from the IR absorption coefficient using the following relationship:
where α is the absorption coefficient (dB/m) and ɛ is the extinction coefficient (dB/m/ppm). In Refs. [15,21] and [22], the extinction coefficient of O-H, Se-H, H2O and As-O was reported to be ${\varepsilon _{OH}}$ = 5.0 dB/m/ppm at 3424 cm-1 (2.92 µm), ${\varepsilon _{SeH}}$ =1.0 dB/m/ppm at 2188 cm-1 (4.57 µm), ${\varepsilon _{H2O}}$ = 34 dB/m/ppm at 1587 cm-1 (6.3 µm), and ${\varepsilon _{A\textrm{s}O}}$= 0.35 dB/m/ppm at 1124 cm-1 (8.9 µm), respectively. The chemical composition of the glass was recorded using an energy dispersive X-ray spectrometer (EDS). The refractive index of the glass was measured using an IR ellipsometer (IR-VASE MARK II, J.A.Woollam) and the dispersion of the fundamental mode (FM) was calculated using a commercial software (RSOFT). Raman spectra were measured by a laser confocal Raman spectrometer (Renishaw in Via, Gloucestershire, UK) using 785 nm excitation.The cut-back method was adopted to measure the fiber loss. Precision fiber cutter (FK11-LDF, Kinetics, Inc.) was used to cut the fiber, and each fiber cross-section was observed under optical microscope (Keyence, VHX-1000) to ensure a smooth surface. The optical fiber loss was calculated by the following formula:
where, P1 is the power injected into the fiber input, P2 is the output power, and L is of the cut fiber length. The near-field output signal was recorded via a near-IR optic fiber field analyzer (Xenics, XEN-000298).For SC measurement, the pump source was femtosecond pulses from an optical parametric amplifier (OPA) system. A ZnSe lens was used to couple the MIR pulses (∼ 150 fs, repetition rate of 1 kHz) from OPA system to the fabricated As2Se3/Ge1As39Se60 fiber. The coupling efficiency was estimated to be around 15%. The output beam was directly injected into a monochromator. Liquid nitrogen cooled HgCdTe detector (spectral response range: 1-16 µm) was used to detect SC signal amplified by a lock-in amplifier.
A home-designed automatic taper platform with electric heating rings was used for fiber tapering. The heating temperature was about 260 °C, and the length of the optical fiber before tapering was 40 cm. The diameter of fiber waist-cone after tapering was 50 µm, and the tapering ratio is 5:1. The broadband infrared light source emitted by the FTIR spectrometer was coupled into the input end of the tapered fiber through the ZnS lens. HgCdTe detector was used to detect optical signal at the output of the fiber. The sensing sensitivity of tapered region is higher than that of non-tapered region, which is due to increasing intensity and number of reflections in tapered region. Therefore, we measured the concentration of carbon dioxide (CO2) gas molecules via the absorption spectra.
3. Results and discussion
Figure 2(a) shows transmission spectra of purified and unpurified As2Se3 glasses prepared at various conditions mentioned above. It can be seen from Fig. 2(a) that, there are some specific absorption bands in the spectral range of 2.5-18 µm, and the intensity of each absorption band is proportional to the content of the corresponding impurity. The estimation of the impurity concentration is listed in Table 2. Upon comparison of the three glasses G0 to G2, it can be found that in the static vacuum distillation system, each impurity peak decreases with an increase in the number of purification processes carried out for the Se raw material. This is due to that, hydrogen, oxygen and carbon impurities contained in commercial raw material elements can be reduced with the addition of Mg strips [23,24]. The Se-H impurity absorption peak at 4.57 µm in G3 glass is completely eliminated in the process of dynamic vacuum distillation. The main form of hydrogen in selenium is hydrogen selenide (H2Se). This result is attributed to the fact that a continuous dynamic vacuum process can remove part of Se-H impurities. On the other hand, due to the high oxidizing ability of SeO2, it can remove hydrogen impurities in the form of hydrogen selenide and organic substances [25]. In addition, the H2O (6.30 µm) and As-O (8.90 µm) impurity concentrations of G3 glass are higher than those of other glasses. The removal of hydrogen impurities and the increase of oxygen-related impurity concentration can be explained by the following chemical reaction [25]:
Fig. 2. (a) Transmission spectra of G0-G6 glasses (∼15 mm) and unpurified As2Se3 glass (∼15 mm); (b) Raman spectra of G0-G6 glasses; (c) differential scanning calorimetry curves of As2Se3 glasses (G0-G6) at a heating rate of 10 K/min.
Table 2. Physical parameters of arsenic selenide glasses and their impurity concentrations
However, a comparison of mid-IR transmission spectra of G4-G6 shows that the combination of dynamic and static vacuum distillation cannot eliminate the Se-H impurities. As long as static vacuum distillation is used, Se-H will be fully integrated into the As-Se glass network and cannot be discharged from the system.
Raman spectroscopy [26] is used to study the molecular vibration and rotation in the glass network. Figure 2(b) shows the normalized Raman spectra of the seven glass samples. It is obvious that the G0-G6 glasses have almost identical Raman profiles. The strongest band centered at 225 cm−1 in these glasses is ascribed to AsSe3/2 pyramidal units [21]. Figure 2(c) shows the DSC (differential scanning calorimetry) characteristic temperature curves of As2Se3 glasses (G0-G6) obtained under different process conditions. The heating rate is 10 K/min. The characteristic temperature parameters of the glasses are summarized in Table 2, including glass transition temperature Tg, initial crystallization temperature Tx and ΔT (the value of Tx - Tg). Tg is the intersection of two tangents as shown in Fig. 2(c), and Tx is the concave point indicated by the arrow. ΔT can be used to feature the thermal stability of the glass, the greater the value of ΔT, the better the thermal stability. From the data in the Table 2, the characteristic temperatures of the glasses prepared by seven different processes are very close. The extrusion temperature of the preform is thus set at 273 °C, which is much lower than the glass crystallization temperature Tx. This relatively low extrusion temperature greatly reduces the crystallization risk of chalcogenide glass and benefits much to realize low-loss optical fibers. The analysis of the chemical compositions of the samples G0-G6 shows negligible difference between the real and nominal compositions, the variation of test between them is less than 1%.
Figure 3(a) shows the transmittance spectra of the core and cladding glass in a wavelength range from 2.5 to 18 µm. Here the cladding glass is G3 glass without Se-H impurity absorption peak. It can be seen from Fig. 3(a), the whole transmission curves are relatively flat, and there are three major absorption peaks centered at 2.92, 6.3 and 7.9 µm, corresponding to the absorption of -OH, H2O and As-O impurities, respectively. The concentrations of impurities associated with absorption peaks in the cladding glass are indicated by the red arrows in the figure.
Fig. 3. (a) Transmission spectra of core glass (∼15 mm) and cladding glass (∼15 mm) (inset is the cross-section image of core and cladding glasses; the minimum scale in the ruler is 1 mm); (b) measured refractive indices of core glass and cladding glass and the calculated NA (inset shows the output light beam-spot).
The refractive index and numerical aperture (NA) of the core and cladding glass are shown in Fig. 3(b). The numerical aperture of the optical fiber is calculated by the following formula:
where ncore is the refractive index of the core glass and ncladding is that of the cladding glasses, respectively. The refractive indices of the two glasses ranges from 2.68 to 2.82 and decrease with the increase of the wavelength. In the whole wavelength range, the calculated numerical aperture (NA) value is between 0.53 and 0.64. Large NA value can ensure the transmitted light to be well confined in the fiber core propagation. The normalized frequency V of the optical fiber is obtained by the following formula:The dispersion of As2Se3/Ge1As39Se60 fiber was analyzed by a commercial software (RSOFT) and the simulated result is shown in Fig. 4(a). It can be seen from the figure that the zero-dispersion wavelength (ZDW) of the fundamental mode in the stepped fiber can be shifted to 4.8 µm from 5.1 µm, which is the zero material dispersion wavelength of the core glass.
Fig. 4. (a) Simulated dispersion of As2Se3/Ge1As39Se60 fiber (core diameter: ∼36 µm). The red solid line is the material dispersion, and the blue dotted line is the fundamental mode (FM) dispersion. (b) The loss comparison between the homemade As2Se3 fiber and the commercial As2Se3 fiber in the wavelength range of 3.3-6 µm.
A 1.8 m long fiber was used to measure the attenuation of the as-prepared As2Se3/Ge1As39Se60 step-index fiber. In order to better characterize the optical properties of the fiber, the fiber loss in the wavelength range of 3.3-6 µm was selected and compared with the loss spectrum of commercial As2Se3 fiber from CorActive as shown in Fig. 4(b). While both fibers exhibit excellent transmission in this wavelength range, it can be found that the commercial As2Se3 fiber has two obvious Se-H impurity absorption peaks at 3.8 µm and 4.57 µm, respectively. The corresponding Se-H impurity concentration at 4.57 µm is 0.2 ppm. In contrast, the loss curve of our As2Se3 fiber has no obvious impurity absorption peak. The loss in this wavelength range is less than 0.5 dB/m, and the minimum loss is 0.2 dB/m@3.6 µm.
We measured the SC generation pumped by OPA laser at different wavelengths and powers, and the results are shown in Fig. 5. Here, a 15 cm As2Se3/Ge1As39Se60 fiber was used for SC measurement. It can be found that the broadening of SC spectrum pumped by different wavelengths is very different at the same power of 20 mW as shown in Fig. 5(a). The spectral broadening is 1.5-12.1 µm, 1.2-13 µm, and 1.6-12.4 µm at -30 dB for the same fiber pumped by different laser wavelength of 5 µm, 5.5 µm and 6 µm, respectively. Since the fiber does not have any impurity absorption peaks in the 3.3-6 µm band, the SC spectra obtained using the three pump wavelengths are very flat in the corresponding wavelength range. All three pump wavelengths are located in the anomalous dispersion regime, which is above the FM ZDW of 4.8 µm. Some pulses are directly located in the anomalous dispersion region and converted into high-order solitons [27]. Under the action of soliton splitting and dispersion wave matching, high-order solitons are split into multiple optical solitons, and phase matched dispersion wave solitons are formed in the positive dispersion region [8]. The red-shifted part of the spectrum still has strong spectral energy before 10 µm, but decreases rapidly after 10 µm, which is limited by multi-phonon vibration absorption [28].
Fig. 5. (a) Experimental SC spectra generation with different pump wavelengths under same input power of 20 mW in the As2Se3 fiber. (b) The SC spectra of the As2Se3 fiber pumped by 5.5 µm laser under different pump powers.
Then we investigated the influence of different pump power on SC output. In order to avoid the difference in the coupling efficiency, an attenuator was adopted to control the pump power to ensure the stable coupling efficiency. The average power from OPA laser after the attenuator is 10 mW, 15 mW and 20 mW, respectively, in Fig. 5(b). The corresponding spectral width is 1.6-11.9 µm, 1.5-12.1 µm, and 1.2-13 µm at -30 dB, respectively. Due to the absence of Se-H impurities in the fabricated As2Se3/Ge1As39Se60 fiber, the obtained SC spectra are relatively flat in the wavelength range of 4-6 µm. The flattest SC spectrum (-30 dB) covering 1.2-13 µm was obtained in the fiber pumped by 5.5 µm laser with a power of 20 mW. Obviously, the output SC spectrum becomes broad with the increase of the pump power.
On the other hand, chalcogenide optical fiber sensing is mainly based on the principle of total internal reflection (TIR) and fiber evanescent wave spectra (FEWS) [29]. The sensing sensitivity of conical region in tapered fiber is higher than that of non-conical region, due to the increase of the total times of internal reflection and density of conical region, and thus the increase of the evanescent wave intensity on the surface of optical fiber [30]. In addition, the penetration depth of evanescent wave is insufficient when entering the optical fiber with core-cladding structure, which will violently reduce the interaction efficiency of evanescent wave. Therefore, in the optical fiber sensing experiment, the mono-index optical fibers are usually used to reduce the fiber size in the form of tapering [31,32]. In this work, Ge1As39Se60 mono-index fiber with a waist cone diameter of 50 µm and a length of 40 cm was selected to ensure that the fiber had better sensing characteristics. The schematic diagram of the optical fiber sensing experiment spectral measuring device is shown in Fig. 6.
Two different types of tapered fibers were used for CO2 sensing. Fiber-I is defined as Se-H free, while fiber-II contains Se-H impurity absorption. The effect of different gas flow rates of CO2 on the infrared absorption spectrum of tapered fiber was studied by using this set-up. In this work, we continuously filled the air chamber with different flow rate of CO2 (0.04 m3/h, 0.06 m3/h, 0.08 m3/h, 0.1 m3/h), and kept the reaction time at 1 hour for all cases. It can be found that the intensity of the absorption peak increases with increasing volume of CO2 injected into the tapered fiber. As can be seen from Fig. 7(a), there is only one obvious characteristic absorption peak near 4.25 µm in a range of the 3.5-5 µm, corresponding to CO2 absorption peak. Considering that the fiber-I has no excess impurity absorption peak in this range, it is very suitable for CO2 gas sensing. In Fig. 7(b), the absorption of Se-H impurity at 4.57 µm affects the signal amplitude of CO2 absorption peak at 4.25 µm. The signal-to-noise ratio (SNR) of two fibers can be obtained according to the following formula [33]:
where ∫ is the integral area of the signal and $\sigma $ is the standard deviation of the noise. The $\sigma $ values of the two optical fibers are 0.00107 (fiber-I) and 0.00296 (fiber-II), respectively, which is sourced from the statistics of the background noise of the fiber loss. The calculated SNR values of the two fibers are 41 and 8, respectively.Fig. 7. Infrared absorption spectra of Ge1As39Se60 tapered fiber for CO2 sensing measurements under different gas flow rates. (a) fiber-I: Se-H free; (b) fiber-II: Se-H impurity contained.
The correlation between carbon dioxide absorption peak integral area and gas flow rate is shown in Fig. 8. The straight lines in the figure are the results of linear fitting of their correlation, with a linear relation y = 0.68953x-0.02759 and y = 0.37671x-0.01482, for fiber-I and fiber-II, respectively. A minimum carbon dioxide detection sensitivity S of the tapered fiber can be defined as [33]:
where Qv is the CO2 gas flow injected into the tapered fiber (m3/h), and k is defined as the ratio of ∫ to Qv. According to the linear fitting results above, the minimum sensitivity detection limit of fiber-I and fiber-II is 1.55 ${\times} $ 10−3 and 7.86 ${\times} $ 10−3, respectively, indicating that the sensitivity of hydrogen-free fiber is about 5.1 times than that of hydrogen-contained fiber. This clearly demonstrates the potential of highly efficient CO2 detection based on a hydrogen-free fiber developed in the present paper.Fig. 8. The dependence of the absorption peak integral area of carbon dioxide on the gas flow rate in two optical fibers.
4. Conclusion
In summary, high purity As-Se glass and Ge-As-Se glass were prepared via various purification processes, the removal of Se-H impurities in Se-based chalcogenide glasses was successfully realized, and finally a low-loss Se-H-free As2Se3 fiber was achieved by the method of glass extrusion and preform-drawing. The fiber exhibits low value of the loss (< 0.5 dB/m) in a wavelength range of 3.3-6 µm without any observable absorption peaks. We have obtained the most flattened SC spectrum covering the wavelength range of 1.2 to 13 µm at -30 dB by 5.5 µm laser pumping, without any decreased SC intensity induced by Se-H absorption. In addition, a mono-index Ge1As39Se60 tapered fiber was demonstrated for CO2 gas detection experiments, the sensitivity of the Se-H-free fiber is 5.1 times higher than commercial fiber. This Se-H-free fiber shows great potentials, especially in the fields of toxic gases detection, greenhouse gas monitoring, supercontinuum generation and IR laser delivery in the wavelength of 3.3-6 µm.
Funding
National Natural Science Foundation of China (61627815, 61705091, 61775109, 61875097, 61935006, 62075107); Natural Science Foundation of Zhejiang Province (LQ21F050005, LR18F050002, LY20F050010); Natural Science Foundation of Ningbo (202003N4101); Ten-Thousands Talents Program of Zhejiang Province; Leading and top-notch personnel training project of Ningbo; K. C. Wong Magna Fund in Ningbo University.
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. A. Schliesser, N. Picqué, and T. W. HÄnsch, “Mid-infrared frequency combs,” Nat. Photonics 6(7), 440–449 (2012). [CrossRef]
2. F. Ahmed, V. Ahsani, K. Nazeri, E. Marzband, C. Bradley, E. Toyserkani, and M. B. G. Jun, “Monitoring of Carbon Dioxide Using Hollow-Core Photonic Crystal Fiber Mach-Zehnder Interferometer,” Sensors (Basel) 19, 3357 (2019). [CrossRef]
3. X. Lu, J. Li, L. Yang, J. Ren, M. Sun, A. Yang, Z. Yang, R. K. Jain, and P. Wang, “Broadband mid-infrared (2.5–5.5 µm) emission from Co2+/Fe2+ codoped chalcogenide glass ceramics,” Opt. Lett. 45(9), 2676–2679 (2020). [CrossRef]
4. J. Ren, X. Lu, C. Lin, and R. Jain, “Luminescent ion-doped transparent glass ceramics for mid-infrared light sources,” Opt. Express 28(15), 21522–21548 (2020). [CrossRef]
5. J. Swiderski and M. Michalska, “High-power supercontinuum generation in a ZBLAN fiber with very efficient power distribution toward the mid-infrared,” Opt. Lett. 39(4), 910–913 (2014). [CrossRef]
6. M. Liao, G. Qin, X. Yan, T. Suzuki, and Y. Ohishi, “A Tellurite Nanowire With Long Suspended Struts for Low-Threshold Single-Mode Supercontinuum Generation,” J. Lightwave Technol. 29(2), 194–199 (2011). [CrossRef]
7. Y. Yu, B. Zhang, X. Gai, C. Zhai, S. Qi, W. Guo, Z. Yang, R. Wang, D. Y. Choi, S. Madden, and B. Luther-Davies, “1.8-10 µm mid-infrared supercontinuum generated in a step-index chalcogenide fiber using low peak pump power,” Opt. Lett. 40(6), 1081–1084 (2015). [CrossRef]
8. C. R. Petersen, U. MØller, I. Kubat, B. Zhou, S. Dupont, J. Ramsay, T. Benson, S. Sujecki, N. Abdel-Moneim, Z. Tang, D. Furniss, A. Seddon, and O. Bang, “Mid-infrared supercontinuum covering the 1.4–13.3 µm molecular fingerprint region using ultra-high NA chalcogenide step-index fibre,” Nat. Photonics 8(11), 830–834 (2014). [CrossRef]
9. L. R. Robichaud, V. Fortin, J. C. Gauthier, S. Chatigny, J. F. Couillard, J. L. Delarosbil, R. Vallee, and M. Bernier, “Compact 3-8 µm supercontinuum generation in a low-loss As2Se3 step-index fiber,” Opt. Lett. 41(20), 4605–4608 (2016). [CrossRef]
10. J. D. C. Jones, L. Brilland, F. Charpentier, J. Troles, B. Bureau, C. Boussard-Plédel, J. L. Adam, D. Méchin, and D. Trégoat, “Microstructured chalcogenide fibers for biological and chemical detection: case study: a CO2 sensor,” presented at the 20th International Conference on Optical Fibre Sensors2009.
11. E. LÉpine, Z. Yang, Y. GuÉguen, J. Troles, X. Zhang, B. Bureau, C. Boussard-Plédel, J. Sangleboeuf, and P. Lucas, “Optical microfabrication of tapers in low-loss chalcogenide fibers,” J. Opt. Soc. Am. B 27(5), 966–971 (2010). [CrossRef]
12. V. Q. Nguyen, G. Drake, G. Villalobos, D. Gibson, S. Bayya, W. Kim, C. Baker, G. Chin, F. H. Kung, M. I. Kotov, L. Busse, and J. S. Sanghera, “Effect of aluminum and tellurium tetrachloride addition on the loss of arsenic selenide optical fiber,” Opt. Mater. 64, 327–333 (2017). [CrossRef]
13. L. Fu, M. Rochette, V. Ta’eed, D. Moss, and B. Eggleton, “Investigation of self-phase modulation based optical regeneration in single mode As2Se3 chalcogenide glass fiber,” Opt. Express 13(19), 7637–7644 (2005). [CrossRef]
14. K. Abedin, “Observation of strong stimulated Brillouin scattering in single-mode As2Se3 chalcogenide fiber,” Opt. Express 13(25), 10266–10271 (2005). [CrossRef]
15. M. Churbanov, I. Scripachev, G. Snopatin, V. Shiryaev, and V. Plotnichenko, “High-purity glasses based on arsenic chalcogenides,” Optoelectron. Adv. Mater. (Russian Academy of Sciences Moscow Institute of Chemistry of High Purity Substances, 2001).
16. W. A. King, A. G. Clare, and W. C. LaCourse, “Laboratory preparation of highly pure As2Se3 glass,” J. Non-Cryst. Solids 181(3), 231–237 (1995). [CrossRef]
17. C. Caillaud, C. Gilles, L. Provino, L. Brilland, T. Jouan, S. Ferre, M. Carras, M. Brun, D. Mechin, and J.-L. Adam, “Highly birefringent chalcogenide optical fiber for polarization-maintaining in the 3-8.5 µm mid-IR window,” Opt. Express 24(8), 7977–7986 (2016). [CrossRef]
18. V. Q. Nguyen, J. S. Sanghera, P. C. Pureza, and I. D. Aggarwal, “Effect of heating on the optical loss in the As-Se glass fiber,” J. Lightwave Technol. 21(1), 122–126 (2003). [CrossRef]
19. Z. Xue, Q. Li, P. Chen, Y. Tian, K. Jiao, X. Wang, Z. Zhao, X. Wang, P. Zhang, S. Dai, R. Wang, and Q. Nie, “Mid-infrared supercontinuum in well-structured As-Se fibers based on peeled-extrusion,” Opt. Mater. 89, 402–407 (2019). [CrossRef]
20. Z. Xue, S. Liu, Z. Zhao, N. Mi, B. Wu, X. Li, P. Zhang, and X. Wang, “Infrared Suspended-Core Fiber Fabrication Based on Stacked Chalcogenide Glass Extrusion,” J. Lightwave Technol. 36(12), 2416–2421 (2018). [CrossRef]
21. A. Yadav, M. Kang, C. Goncalves, C. Blanco, R. Sharma, and K. Richardson, “Impact of raw material surface oxide removal on dual band infrared optical properties of As2Se3 chalcogenide glass,” Opt. Mater. Express 10(9), 2274–2288 (2020). [CrossRef]
22. S. Danto, D. Thompson, P. Wachtel, J. D. Musgraves, K. Richardson, and B. Giroire, “A Comparative Study of Purification Routes for As2Se3 Chalcogenide Glass,” Int. J. Appl. Glass Sci. 4(1), 31–41 (2013). [CrossRef]
23. M. F. Churbanov, “Relevant problems of chemistry of high-purity substances,” Inorg. Mater. 45(9), 955–960 (2009). [CrossRef]
24. L. Mochalov, A. Logunov, A. Kitnis, and V. Vorotyntsev, “Plasma-Chemistry of Arsenic Selenide Films: Relationship Between Film Properties and Plasma Power,” Plasma Chem. Plasma Process. 40(1), 407–421 (2020). [CrossRef]
25. A. P. Velmuzhov, M. V. Sukhanov, N. S. Zernova, V. S. Shiryaev, T. V. Kotereva, L. A. Ketkova, I. I. Evdokimov, and A. E. Kurganova, “Preparation of Ge20Se80 glasses with low hydrogen and oxygen impurities content for middle IR fiber optics,” J. Non-Cryst. Solids 521, 119505 (2019). [CrossRef]
26. Y. Xu, D. Chen, Q. Zhang, H. Zeng, C. Shen, J.-L. Adam, X. Zhang, and G. Chen, “Bright red upconversion luminescence of thulium ion-doped GeS2− In2S3− CsI Glasses,” J. Phys. Chem. C 113(22), 9911–9915 (2009). [CrossRef]
27. A. Husakou and J. Herrmann, “Supercontinuum generation of higher-order solitons by fission in photonic crystal fibers,” Phys. Rev. Lett. 87(20), 203901 (2001). [CrossRef]
28. Z. Zhao, B. Wu, X. Wang, Z. Pan, Z. Liu, P. Zhang, X. Shen, Q. Nie, S. Dai, and R. Wang, “Mid-infrared supercontinuum covering 2.0-16 µm in a low-loss telluride single-mode fiber,” Laser Photonics Rev. 11(2), 1700005 (2017). [CrossRef]
29. K. Michel, B. Bureau, C. Pouvreau, J. C. Sangleboeuf, C. Boussard-Plédel, T. Jouan, T. Rouxel, J. L. Adam, K. Staubmann, H. Steinner, T. Baumann, A. Katzir, J. Bayona, and W. Konz, “Development of a chalcogenide glass fiber device for in situ pollutant detection,” J. Non-Cryst. Solids 326-327, 434–438 (2003). [CrossRef]
30. B. Bureau, C. Boussard, S. Cui, R. Chahal, M. L. Anne, V. Nazabal, O. Sire, O. Loréal, P. Lucas, V. Monbet, J.-L. Doualan, P. Camy, H. Tariel, F. Charpentier, L. Quetel, J.-L. Adam, and J. Lucas, “Chalcogenide optical fibers for mid-infrared sensing,” Opt. Eng. 53, 1 (2014). [CrossRef]
31. X. Jiang and A. Jha, “Engineering of a Ge–Te–Se glass fibre evanescent wave spectroscopic (FEWS) mid-IR chemical sensor for the analysis of food and pharmaceutical products,” Sens. Actuators, B 206, 159–169 (2015). [CrossRef]
32. M. Ahmad and L. L. Hench, “Effect of taper geometries and launch angle on evanescent wave penetration depth in optical fibers,” Biosens. Bioelectron. 20(7), 1312–1319 (2005). [CrossRef]
33. L. Hu, C. Zheng, D. Yao, D. Yu, Z. Liu, J. Zheng, Y. Wang, and F. K. Tittel, “A hollow-core photonic band-gap fiber based methane sensor system capable of reduced mode interference noise,” Infrared Phys. Technol. 97, 101–107 (2019). [CrossRef]
