Surface Treatment of Fluorozirconate Glass Fiber Preform Using Nonaqueous Solution
Chen Huiyu,Wang Xin,Xue Tianfeng,Meng Tao,He Hong,Liao Meisong,Hu Lili,Chen Wei
DOI: https://doi.org/10.3788/cjl202148.0703001
2021-01-01
Abstract:Objective Fluorozirconate optical glasses are widely used as fibers because of their excellent optical properties, such as their transparency in the mid-infrared wavelength range around 8 ttm. These glass fibers show a wide range of military and industrial applications, such as in the ultra-low loss, long, and repeaterless optical communication links. They are typically fabricated from the fluorozirconate glass preform by the rod-in-tube method that allows for precise control of fiber dimension and geometry. To a large degree, the fiber quality depends on the quality of the preform. Before fabrication of the fluorozirconate glass fiber, the preform must undergo optical cold processing, including grinding and polishing. After shown in the structure model of the glass surface in Fig. 1, defects, such as impurities, scratches and microcracks, are introduced on the surface and subsurface of the fiber preform during this processing. During the fiber fabrication process, the defects end up at the core/cladding interface, decrease the strength of the fiber, and increase the optical losses. An acid treatment is an effective method to enhance the mechanical strength of the optical glass surface and eliminate the surface and subsurface defects. In general, the preform surface of the fluorozirconate glass fiber is treated with an aqueous acid solution, such as hydrochloric acid or boracic acid. However, the fluorozirconate glass surface degrades rapidly upon exposure to aqueous media or humid atmospheric environments because of its poor chemical stability. Therefore, the conventional surface treatment of fluorozirconate glass fiber preform by etching with an aqueous acid solution increases the surface roughness and creates a hydrated surface layer. It also causes precipitation of crystalline dissolution products on the surface, which increases the optical losses, decreases the tensile strength of the fiber, and increases the risk of devitrification during fabrication of the fiber. Therefore, it is more optimal to treat the fluorozirconate glass surface using an nonaqueous solution. Methods An nonaqueous organic solution composed of organic solvent, mixed acid, and additives ( Table 1) was formulated to treat the surface of an fluorozirconate glass fiber preform. The organic solvent was composed of ethyl alcohol and amyl alcohol, which cannot corrode the fluorozirconate glass surface. Therefore, a corrosion layer cannot be created. The mixed acid component was a mixture of hydrochloric acid and oxalic acid with a small amount of water, which was further diluted with an organic solvent. The amount of water used did not cause chemical etching of the surface. The additives used were zirconyl chloride and ethylene diamine tetraacetic acid, which are metal complexing agents that increase the solubility of the etching products and prevent their precipitation on the surface. The nonaqueous organic solution could more effectively remove the surface and subsurface defects caused by optical cold processing, without creating a corrosion layer. Results and Discussions To compare the two different surface treatments (using an aqueous acid solution and an nonaqueous organic solution, respectively), we have studied the surface compositions of the fluorozirconate glass using X-ray photoelectron spectroscopy ( XPS). The composition variation of the fluorozirconate glass surfaces treated by these two surface treatments is shown in Fig. 2. It is evident that the compositions of the fluorozirconate glass surface etched by the nonaqueous organic solution were closer to those of bulk glass for the elements, Zr, Na, and Ba. Furthermore, it is clear that dissolution of the surface compositions etched by the aqueous acid solution was quicker than that by the nonaqueous organic solution, and the chemical etching by the nonaqueous organic solution was more uniform and did not create a corrosion layer or precipitation leading to opaque crystalline surface deposits. By using atomic force microscopy ( AFM), we have studied the surface morphology of the fluorozirconate glass surface treated with these two methods. The surface root-mean-square (RMS) roughness of the fluorozirconate glass surfaces was 0.925 nm after polishing with nano-Ce02 ( Fig. 3), 8.971 nm after etching with the aqueous acid solution ( Fig. 4), and 2.152 nm after etching with the nonaqueous organic solution. The surface roughness after etching with the nonaqueous organic solution was lower than that after etching with the aqueous acid solution. Furthermore, the morphology of the surface etched by the nonaqueous organic solution was better than that etched by the aqueous acid solution. This means that the chemical reaction between the nonaqueous organic solution and the fluorozirconate glass surface was more stable, and with the lack of precipitation of the reaction products on the surface to affect the etching, the surface was smoother. Etching of the fluorozirconate glass fiber preform surface with the nonaqueous organic solution results in a higher surface quality, which is useful for subsequent fiber fabrication. Figure 6 shows the optical losses of the fluorozirconate optical glass fibers fabricated from preforms with the different surface treatments. The optical losses of the fluorozirconate optical glass fiber fabricated from the preform etched by the nonaqueous organic solution were lower at different wavelengths than those fabricated from the preform etched by the aqueous acid solution. It is evident that the fluorozirconate glass surface treatment with the nonaqueous organic solution was more effective at eliminating surface and subsurface defects and removing impurities, which increases the scattering and absorption losses compared to the traditional surface treatment using aqueous acid solution. The weibull failure probabilities of the fluorozirconate glass fibers fabricated from preforms with the two surface treatments are displayed in Fig. 7. The median tensile stress at failure is about 300 MPa for the fiber fabricated from the preform etched by aqueous acid solution and about 450 MPa for the fiber fabricated from the preform etched by nonaqueous organic solution. Therefore, it is shown that fluorozirconate glass preform surface treatment with nonaqueous organic solution was more effective in removing failure -producing defects than that with aqueous acid solution. Conclusions In order to decrease optical losses at the fluorozirconate glass fiber core/cladding interface, the method to treat the preform surface using nonaqueous organic solvents is investigated. When comparing the surface quality of the preform and the performance of the resulted fiber after treatment with nonaqueous organic solvents or traditional aqueous acid solution, it is evident that the treatment with the former allows for the fabrication of an fluorozirconate glass fiber with lower optical losses and higher strengths than that with the latter.