This paper investigates the chemical durability of a fluoroindate (IZSBGC) glass (developed by our previous research for low-loss fluoroindate fiber production) compared to the widely studied fluorozirconate (ZBLAN) system via leaching of glass samples in deionized water. The chemical stability of both glass systems is probed using a series of analytical techniques such as FTIR, XPS and SEM to study the sample surfaces (before and after leaching) and hydrated layer products, both of which reflected the nature of the leaching process. Our experimental results suggest that IZSBGC glass presented better chemical stability in water than ZBLAN. The absorption due to both OH- stretching and HOH bending vibrations for both glass types increased with increasing amounts of hydrated layers formed during the leaching. The investigation of hydrated layers using SEM suggests that the NaF content in fluoride glass accelerated the leaching significantly. XPS analyses suggest that (hydr)oxyfluorides and hydroxides formed on both fluorozirconate and fluoroindate glass surfaces after leaching, respectively. The degradation of fiber breaking strain in NaF-free IZSBGC glass is less than that of NaF-containing ZBLAN glass.
© 2014 Optical Society of America
Fluoride glasses are attractive for optical applications offering excellent transmission characteristics from visible to infrared wavelengths [1,2] and are of significant interest for a range of applications such as high power delivery fibers for mid-IR lasers . However, fluoride glasses have relatively poor chemical durability when compared to oxide glasses (e.g., silicate  and tellurite glasses ), as they are reactive in aqueous environments. The presence of water or an acidic solution leads to dissolution of the fluoride glasses and subsequent formation of surface layers. In general, the corrosion of heavy metal fluoride (HMF) glasses in un-buffered water is extremely rapid compared to silicate glasses, with dissolution reactions more than 10-100 times faster than those for silicate glasses . This is not surprising since the structure of fluoride glasses is quite different from that of the silicates. Chemical corrosion can result in high transmission losses in fluoride fibers as well as have a significant negative impact on fiber strength [7,8], posing potential limitations to the application of fluoride fibers. Therefore, it is essential to understand the corrosion process and chemical durability of these glasses to guide the development of improved materials or methods for the protection of these materials.
To date many types of HMF glasses have been synthesized, of which the fluorozirconate glasses are most widely studied . Leaching behaviors of HMF glasses (e.g., fluorozirconate) have been studied in earlier published work [10–12]. The chemical durability of these NaF-containing fluorozirconate glasses (e.g., ZBLAN) in aqueous environments was, however, found to be extremely poor and rather sensitive to water, while other fluoride glasses without NaF (e.g., ZBL and ZBLA [4,13]) are relatively stable in aqueous solutions. Similarly, Moynihan et al.  compared two fluorozirconate glasses (ZBLAN and ZBLA) and found that the ZBLA glass without NaF has better chemical durability in deionized water than the ZBLAN glass containing 20 mol% NaF.
Compared with the fluorozirconate glasses, fluoroindate glasses are characterized by transmittance extending further into the IR region due to their lower maximum phonon energy of ~510 cm−1 . Fluoroindate glasses were said be relatively stable against atmospheric moisture, as compared to fluorozirconate glasses [15,16], possibly due to lower solubility of InF3 (0.04 g/100ml in water at room temperature)  compared to ZrF4 (1.32-1.39 g/100ml in water at room temperature) . However, no actual experimental data are given in refs [15,16] regarding the stability of the fluoroindate glasses under atmospheric moisture conditions. Indeed, there are no research studies reported on the chemical durability of fluoroindate glasses.
Our previous study  and others [3,20] on low loss ZBLAN have shown that fluorozirconate glasses with a composition of 53ZrF4-20BaF2-4LaF3-3AlF3-20NaF (in mol %) (fluorozirconate; ZBLAN) can be fabricated into low-loss fibers. Our recent work  has also shown another candidate of fluoride glass with a composition of 32InF3-20ZnF2-20SrF2-18BaF2-8GaF3-2CaF2 (in mol %) (fluoroindate; IZSBGC) suitable for making low-loss fiber. Here we report recent results of the leaching of these two glasses in deionized water to gain insights into their corrosion behaviors. The reactions and possible formation of hydroxyl species on their glass surfaces during corrosion were studied by Fourier transform infrared spectroscopy, scanning electron microscopy and X-ray photoelectron spectroscopy. In addition, the fiber breaking strains of fibers before and after leaching for both glass types were also compared. The aim of this study was to test the chemical durability of both bulk glasses and glass fibers, the latter of which is particularly important as poor chemical durability of the fibers can limit its practical application.
2. Experimental procedure
IZSBGC and ZBLAN were chosen as starting materials for the leaching study in this work. These glasses were synthesized from high purity raw materials (5N for SrF2, CaF2, BaF2 and AlF3; 4N for InF3·3H2O, LaF3 and NaF; 3N for ZnF2 and ZrF4; 2N for GaF3) that were thoroughly mixed into 30~60 g batches and then melted in a platinum alloy crucible containing 5% gold (internal volume: 100 ml). The glasses were melted using the conventional melting-quenching method in a controlled dry N2 atmosphere (99.99%) melting facility. As H2O in InF3·3H2O reacts with InF3, producing In(OH)3 and gaseous HF, a higher amount of excess NH4HF2 (3N, 6.7 wt% of the batch weight) was added into glass batches containing InF3·3H2O to fluorinate the oxide impurities produced from the raw materials and compensate for the loss of fluorine due to the formation and loss of HF. The fabrication of fluoroindate glass included batching, fluorination with NH4HF2 at 235 °C (52 mins), melting (3 h) and casting at 900 °C, while the preparation of fluorozirconate glass included batching, melting at 850 ° C (2.5 h) and casting at 650 °C. All the melts were cast into preheated brass moulds.
In this work, the two preforms used for ZBLAN and IZSBGC fiber drawing were 110 mm and 140 mm long, respectively. The fiber diameter was 156 ± 8 µm for ZBLAN and 176 ± 13 µm for IZSBGC. Both fibers were drawn with a preform feed rate at 1.6 mm/min. The draw tower furnace temperatures, which are higher than the glass temperatures , were maintained at 665 °C and 725 °C for ZBLAN and IZSBGC fiber drawing, respectively.
Leaching experiments were conducted in HDPE (high density polyethylene) containers using only deionized water (pH ~5.60) at the beginning of the leaching experiments at 25 ± 2 °C. Fluoroindate and fluorozirconate glass samples were cut into ~2 mm thick slides which were then polished using ~3 μm diamond paste and ~40 nm colloidal silica to obtain scratch free surfaces .
IR transmission measurements were conducted using a commercial Fourier transform infra-red (FTIR) spectrometer (PerkinElmer Spectrum 400) before and after leaching in deionized water for various reaction times. The absorption coefficients were calculated from the transmission data, where background absorptions due to the Fresnel reflection at the interfaces were subtracted.
The cross-sections of the glass slides after leaching were imaged by a scanning electron microscope (SEM; Philips XL30 field emission SEM), equipped with energy dispersive spectrometers (EDS). One glass slide for both ZBLAN and IZSBGC glasses was fractured in a glove box filled with N2 (99.99%). One part of the fractured glass slide was stored in an air lock container (filled with N2) prior to analysis of the fractured surface using X-ray photoelectron spectroscopy (XPS). The second part of the fractured slide was immersed in deionized water for 30 mins and then stored in an air lock container (filled with N2), prior to XPS analysis on the fractured/leached surface. XPS measurements were performed in an ultra high vacuum apparatus built by SPECS (Berlin, Germany) using a non-monochromatic Mg Kα X-ray source (1253.6 eV) and a hemispherical Phoibo100 energy analyser from SPECS. Charge compensation was performed by an electron flood gun SPECS FG20 at 1 eV and 1 μA. Calibrations of the XPS binding energies were made against adventitious carbon (284.8 eV). The Zr 3d spectrum was fitted using the 3d3/2 and 3d5/2 doublet which has a 2:3 ratio with an energy separation of 2.43 eV. The FWHM (full width at half maximum) for Zr 3d3/2 was set to the same value as that for Zr 3d5/2.
During the course of the leaching experiments, the pH of the deionized water solutions changed. pH measurements were performed using a temperature-corrected pH Meter (Metrohm 827), with errors of ~0.05. Calibrations were carried out using standard buffers from Metrohm.
To obtain the mechanical breaking strain of the fibers, we used an in-house fiber bending test setup shown in Fig. 1 (refer to  for detailed description). The mechanical breaking strain of each fluorozirconate and fluoroindate fiber piece under test (before and after leaching for 45 mins) (e.g., four sets of measurements) was determined by measuring the fiber bending radius at fiber breakage (minimum bending radius) and using Eq. (1),
3.Results and Discussion
3.1 Infrared spectroscopic studies
The water-glass interaction for HMF glasses can be studied by IR absorption and assessed by the change in OH- stretching (2.9 μm) and HOH bending (6.1 μm) vibrations . The bending mode of HOH at 6.1 μm can be attributed to water molecules present within a thin hydrated layer on the surface . The stretching mode of OH- at 2.9 μm is thought to be due to OH- in both bulk glass and glass surface [4,24].
Figure 2 shows the absorption coefficients of ZBLAN and IZSBGC glasses exposed to water at 25 ± 2 °C for various reaction times. After reaction with water, both the ZBLAN and IZSBGC glass surfaces were found to be covered by opaque films to varying degrees, under which somewhat cloudy but still translucent surface layers could be discerned. Both OH- stretching and HOH bending vibrations were observed with an increase in absorption for ZBLAN and IZSBGC glasses from ~30 mins leaching onwards (up to 90 mins) in Fig. 2a and 2b. This indicates that hydrated layers formed in the leaching processes and altered the glass surface properties . The OH- stretching vibration peak grew slightly faster than the HOH bending peak for both glass types after leaching. However, the absorption coefficient of IZSBGC glass both at 2.9 and 6.1 μm increased more slowly than that of ZBLAN glass (Fig. 2c and 2d), suggesting that hydrolysis reaction of the former was slower. In previous studies [12,25–27], incorporation of alkali glass modifiers into fluoride glasses such as Li and Na fluorides with high aqueous solubilities increased their leaching rates. For example, the high content of NaF in fluorozirconate glasses increased their glass corrosion rate for all glass components  compared to other fluorozirconate glasses without NaF (e.g., ZBLA). As discussed above, the slower increase in absorption coefficient for IZSBGC glass at 2.9 and 6.2 μm in this work, compared to ZBLAN glass, is likely due to the absence of alkali glass modifiers (as opposed to 20 mol% NaF in the ZBLAN glass). Another possible reason is that zirconium fluoride is an undesirable component from the viewpoint of glass durability, because of its higher solubility in water as discussed in the Introduction.
3.2 pH study of corrosion solution
During the course of the leaching experiments, the pH of the aqueous solutions (originally at ~5.60 for deionized water prior to leaching) all decreased (Fig. 3). The pH drifts were attributed primarily to hydrolysis reactions . In the very early stage of corrosion in stagnant solution (within 0.5 day), fluorides near the fresh glass surface undergo an appreciable hydrolysis associated with ion exchange between F- and OH- . As a consequence, OH- substitutes for F- near the glass surface structure , leading rapidly to a pH decrease. In this work, after leaching for 0.5 day (Fig. 3), the pH drifted from 5.60 to 2.93 and 3.20 for leaching of ZBLAN and IZSBGC glasses, respectively, indicating that the pH decreased significantly into a more acidic range for both glass types after the initial leaching period (within 0.5 day). Similar results were found in an earlier study of hydrolysis reaction of fluorozirconate by Le Toullec et al. . The reduced pH increased markedly the solubility of the fluoride glass components (i.e., metal fluorides), resulting in the leaching process occurring primarily via matrix dissolution as described in a study of dissolution of fluorozirconate glasses in Eq. (2) :10].
3.3 Scanning electron microscopy study
The SEM images below (Figs. 4 and 5) show the formation of corrosion layers on ZBLAN and IZSBGC glasses after leaching in deionized water for different reaction times. SEM results showed that the leached ZBLAN glasses exhibited multi-layered products with different compositions after leaching for > 0.5 day (Fig. 4). Spherical particles observed from 4 days onwards (Fig. 4) were made up of a multitude of thin platelets containing Zr and Ba, as confirmed by EDS analysis (Table 1). The spherical particles formed within Layer 1 were also found on the outermost surfaces after leaching for 4 and 6 days (Fig. 4). Polyhedral and spherical particles formed on the outermost surface were previously identified as ZrF4XH2O and ZrBaF6XH2O crystals, respectively , consistent with the compositions found in this work (Table 1). Compared to ZBLA glass after leaching in deionized water reported in , ZBLAN glasses studied in this work produced more porous hydrated layers (Fig. 4). In a previous study by Simmons et al. , it was proposed that a glass with high alkali content formed a thick and porous de-alkalized layer that can be easily dissolved. Similarly, a reduction in alkali content (e.g., ZBLA glass with no NaF vs. ZBLAN with high content of NaF) resulted in a decrease in the porosity of the hydrated layer . In this work, the leach rate for IZSBGC without NaF slowed down compared to ZBLAN glass with high content of NaF. Clearly the nature of the hydrated layers, depending on glass compositions, strongly affects the subsequent leach rate.
The hydrated layers formed in the IZSBGC glass (Fig. 5 and 6) had similar cation compositions to that of the original un-leached glass. Leaching for 2 days or longer resulted in an obvious gap between the formed layer and the reacted glass (e.g., 4 days in Fig. 5). This size of this gap increased as the leaching progressed. The formed thin layers on IZSBGC glasses had numerous pores on the micron scale. The increased Ba content at Position 10 in Fig. 6 was probably due to the formation of a very thin Ba-rich layer (~1 µm) on the etched glass surface .
The thickness of the hydrated layer for both glass types generally increased with reaction time (Fig. 7). The results determined in a previous dissolution study of ZBLA glass  were employed in this study for comparison purpose. For all three glasses, the thickness increased rapidly at the beginning but much more slowly in the subsequent leaching. Within the experimental time scale investigated, no equilibrium state was reached for ZBLAN glass (i.e., thickness of the hydrated layer increased throughout the leaching in Fig. 7), suggesting that leaching was progressing, while equilibrium was almost established for IZSBGC and ZBLA glasses (i.e., thickness of the hydrated layer ‘levelled off’). The leaching of IZSBGC glass slowed down from 4 days (Fig. 7), which may be partly due to the slight increases in pH (Fig. 3). In comparison, the relative large decreases (Fig. 3) in pH for the leaching of the ZBLAN glass at 7 days may result in the marked increase in its leach rate (a sharp increase in the hydrated layer thickness; Fig. 7). The porosity in the hydrated layer facilitated the leaching solution to pass through the layer to further react with the glass surface. However, the IZSBGC glass layers after leaching appeared to be less porous compared to the ZBLAN glass layers. This is possibly because the hydrated layers of IZSBGC glass may play a role in preventing the molecular water from diffusing through the layer into the unreached glass surface. This could also be another possible reason as to why the leaching of the IZSBGC glasses in deionized water occurred more slowly. Measurement of the thickness of the hydrated layers showed that the hydrated layer was less than 20 µm after leaching for 7 days for the IZSBGC glasses (Fig. 7), while the hydrated layer for ZBLAN was greater than 50 µm. Compared to ZBLA glass with no NaF, ZBLAN with a high NaF content had a much thicker hydrated layer (due to the higher solubility of NaF in water  than other fluoride components), while IZSBGC glass had a slightly thinner porous hydrated layer structure. Therefore, the main difference in the hydrated layer between IZSBGC and ZBLAN glasses is likely attributed to the presence of NaF in ZBLAN glass.
3.4 X-ray photoelectron spectroscopic study
Fractured surfaces of fluoroindate and fluorozirconate glasses before and after leaching in deionized water were studied using X-ray photoelectron spectroscopy. An attempt was made to fit the Zr oxyfluoride (for ZBLAN glass) and In(OH)3 (for IZSBGC glass) peaks into the Zr 3d and In 3d spectra for fresh surfaces, but it was unsuccessful. This suggests the oxyfluoride/hydroxide content was negligible or not present on the fresh surfaces. The high resolution Zr and In 3d XPS spectra before and after leaching in deionized water (Figs. 8 and and 9) revealed the most important aspects of the glass surface chemistry. It can be seen that the Zr 3d peaks (3d5/2 and 3d3/2) in Fig. 8 (top) observed on the reacted surface were broadened and somewhat shifted towards the lower binding energy, as compared to the fresh surface before leaching in deionized water. Fittings of the high resolution Zr 3d spectrum indicated formation of a new Zr(hydr)oxyfluoride species on the surface. The Zr 3d binding energy of this (hydr)oxyfluoride species was interpreted on the basis of its Pauling charge . It is proposed that Zr(hydr)oxyfluoride is produced in aqueous solutions via attack of Zr-F bonds by molecular water. Similar to the Zr 3d peaks in ZBLAN glass, the In3d5/2 peak was broadened and a shoulder peak appeared at the lower binding energy after leaching in the deionized water (Fig. 9). This shoulder peak was thought to be In(OH)3 at a lower bonding energy (444.9 eV ), a hydrolysis product from reaction with deionized water. The results in Figs. 8 and 9 further confirmed that hydrolysis reaction occurred in the leaching process, in good agreement with the FTIR and SEM findings. It is interesting to note that XPS analyses found no Na on the reacted surface of ZBLAN glass. Similarly, no Ca was found on the reacted IZSBGC glass surface. This is presumed to be due to the Na and Ca being preferentially dissolved during the leaching processes. The formation of hydroxide and oxyfluoride on the glass surfaces after leaching in deionized water for 30 mins also supports the concept of the ion exchange between F- and OH- in the initial stages of leaching (Section 3.2).
3.5 ZBLAN and IZSGBG fiber strain and SEM study before and after corrosion
3.5.1 Fiber strain before and after corrosion
Chemical corrosion processes exert a strong influence on the strength and fracture behavior of fluoride glasses, not only in bulk samples but also in fiber forms. Previously published works have shown that surface and bulk crystal formation during fabrications of both preforms and fibers cause degradation in fiber strength [7,8,31].
Fiber breaking strain measurements were conducted before and after leaching in deionized water. We found that the fiber breaking strain after leaching in deionized water only changed slightly after leaching for 30 mins and that the fibers were too fragile for breaking strain measurements due to the heavy surface crystals after leaching in deionized water for 90 mins. Therefore, the leaching time (in deionized water) was set to 45 mins for the fiber strain measurements. The ZBLAN fibers exhibited heavier opaque films on the surfaces than those on the IZSBGC fiber surfaces after leaching in deionized water. The fiber breaking strain results shown in Fig. 10 and Table 2 demonstrated that the average breaking strain of the IZSBGC fiber was reduced by 32%, which was lower than that of the ZBLAN fiber (39%). It suggests that the degradation in fiber strain for ZBLAN was greater than that for IZSBGC fiber, in agreement with the results in the preceding sections showing that ZBLAN has a relatively poor chemical stability.
3.5.2 SEM study of fiber surfaces before and after corrosion
At a low fiber strength level, fracture surfaces of broken fibers with mirror-mist-hackle patterns have been observed [2,31–33]. It is likely that this fracture pattern arises from surface defects and propagated rapidly when the fiber is fractured . Compared to the fiber surface before leaching in deionized water, the obvious mirror-mist-hackle patterns occurred on both ZBLAN and IZSBGC fiber surface after leaching in deionized water for 45 mins, which suggested the formation of surface product layers or porous defects as reflected by our findings in Figs. 4 and 5. In Fig. 11, the fiber surfaces after leaching showed obvious defects, in contrast to the original glasses (Fig. 11(a) vs. Figure 11(b); Fig. 11(c) vs. Figure 11(d)). These microscopic flaws led to the low breaking strains of the fibers, confirmed by our fiber breaking strain measurements.
As a summary, the negative effect of NaF on the chemical durability was clearly demonstrated by the much more rapid leaching of ZBLAN compared to ZBLA (Fig. 7), supported by XPS observations where no Na was found on the ZBLAN surfaces after leaching. FTIR results (Fig. 2) showed that the increase in absorption due to OH- was faster for ZBLAN than that for the IZSBGC glass, suggesting a more rapid leaching process for ZBLAN. This is consistent with the thickness measurements by SEM (Fig. 7), showing the thicknesses of the hydrated product layers for ZBLAN increased more rapidly. Both FTIR and SEM observations indicate that the IZSBGC glass and its fibers have a better chemical stability against deionized water compared to ZBLAN. For example, the stronger OH absorptions after leaching of the ZBLAN glass, compared to the IZSBGC glass, indicate a higher degree of hydrolysis reaction, consistent with the greater thicknesses of hydrated layers determined by SEM.
The chemical durability and the evolution of the hydrolysis reaction of IZSBGC and ZBLAN glasses in deionized water have been investigated by Fourier transform infrared spectroscopic, scanning electron microscopic and X-ray photoelectron spectroscopic analyses. A comparison between absorption coefficients at 2.9 and 6.2 μm for both glass types after leaching in deionized water showed that the intensities of OH stretching and HOH bending vibration peaks of IZSBGC glass increased more slowly than those of ZBLAN glass. Our results suggest that dissolution of the glasses occurred initially via an ion exchange between F- and OH-, decreasing the pH of the leaching solution. Subsequently, the solubility of metal fluorides increased dramatically due to the decreasing pH, leading to dissolution of the glass matrices. SEM studies suggested that the reactions occurred on the glass surfaces and that hydrated layers developed on both IZSBGC and ZBLAN glass surfaces. The NaF-free fluoride glass, IZSBGC, had thinner and less porous hydrated layers than the NaF-containing fluoride ZBLAN glass. This indicates that dissolution of NaF-free fluoride glass was slower than NaF-containing fluoride glass, in agreement with previous findings on ZBLA glasses . Therefore, the fluoride glasses IZSBGC studied here were shown to be more stable against deionized water compared to NaF-containing glasses. XPS studies suggested that (hydr)oxyfluoride or hydroxide species were formed on the ZBLAN and IZSBGC glass surfaces after leaching in deionized water, respectively, confirming the presence of OH- on the glass surfaces; the resulting increases in absorption at both 2.7 and 6.2 μm observed in the FTIR spectra also supported XPS results. FTIR, SEM and fiber breaking strain measurements showed that NaF-free fluoride glass IZSBGC presented better mechanical and chemical stability against water than NaF-containing ZBLAN glass. Although ZBLAN glass presented poorer stability against water than ZBLA glass due to the content of NaF, the former has been widely studied and used as fiber material because of its better crystallization stability than ZBLA glass . Our NaF-free IZBSGC glass investigated here, however, exhibited not only extended IR transparency but also improved resistance against deionized water compared to the NaF-containing ZBLAN glass. The work presented in this paper advances our understanding of the corrosion behavior of our IZSBGC glass relative to the established ZBLAN glass in deionized water, and will assist with optimization of the glass compositions to enable low-loss fiber with good chemical durability. In the future, we will explore new fluoride glass compositions with not only optimum optical properties but also good chemical durability. For example, fluorozirconate glasses with low content of alkali components (e.g., NaF) or fluoroindate glasses with low content of alkaline earth components (e.g., CaF2) would be of interest.
We acknowledge support from DSTO for funding this work. We also acknowledge the facilities, and the scientific and technical assistance of the Australian Microscopy & Microanalysis Research Facility at Adelaide Microscopy Centre of The University of Adelaide. This work was performed in part at the Optofab node of the Australian National Fabrication Facility utilizing Commonwealth and SA State Government funding. We thank Associate Professor Gunther Andersson (The Flinders University of South Australia) for assistance with XPS measurements. Jiafang Bei acknowledges the International Postgraduate Research Scholarship (IPRS) and discipline of physics supplementary scholarship supported by The University of Adelaide. H. Ebendorff-Heidepriem acknowledges the support of a DSTO Fellowship and T.M. Monro acknowledges the support of an ARC Laureate Fellowship.
References and links
1. J. Bei, T. M. Monro, A. Hemming, and H. Ebendorff-Heidepriem, “Fabrication of extruded fluoroindate optical fibers,” Opt. Mater. Express 3(3), 318–328 (2013). [CrossRef]
2. J. Bei, T. M. Monro, A. Hemming, and H. Ebendorff-Heidepriem, “Reduction of scattering loss in fluoroindate glass fibers,” Opt. Mater. Express 3(9), 1285–1301 (2013). [CrossRef]
3. X. Zhu and N. Peyghambarian, “High-power ZBLAN glass fiber lasers: review and prospect,” Adv. Optoelectron. 2010, article ID 501956, 23 pages (2010).
4. M. Le Toullec, C. J. Simmons, and J. H. Simmons, “Infrared spectroscopic studies of the hydrolysis reaction during leaching of heavy-metal fluoride glasses,” J. Am. Ceram. Soc. 71(4), 219–224 (1988). [CrossRef]
5. J. S. Wang, E. M. Vogel, and E. Snitzer, “Tellurite glass: a new candidate for fiber devices,” Opt. Mater. 3(3), 187–203 (1994). [CrossRef]
6. C. G. Pantano and R. K. Brow, “Hydrolysis reactions at the surface of fluorozirconate glass,” J. Am. Ceram. Soc. 71(7), 577–581 (1988). [CrossRef]
7. S. Sakaguchi, Y. Terunuma, Y. Ohishi, and T. Kanamori, “Fluoride fibre drawing with improved tensile strength,” J. Mater. Sci. Lett. 6(9), 1063–1065 (1987). [CrossRef]
8. H. W. Schneider, A. Schoberth, A. Staudt, and C. Gerndt, “Fluoride glass etching method for preparation of infra-red fibres with improved tensile strength,” Electron. Lett. 22(18), 949–950 (1986). [CrossRef]
9. M. Saad, “Fluoride glass fiber: state of the art,” Proc. SPIE 7316, 73160N (2009). [CrossRef]
10. D. G. Chen, C. J. Simmons, and J. H. Simmons, “Corrosion layer formation of ZrF4-based fluoride glasses,” Mater. Sci. Forum 19–20, 315–320 (1987). [CrossRef]
11. C. T. Moynihan and S. R. Loehr, “Chemical durability of fluoride glasses,” Mater. Sci. Forum 32–33, 243–253 (1988). [CrossRef]
12. C. J. Simmons and J. H. Simmons, “Chemical durability of fluoride glasses: I, reaction of fluorozirconate glasses with water,” J. Am. Ceram. Soc. 69(9), 661–669 (1986). [CrossRef]
13. A. J. Bruce, S. R. Loehr, R. Mossadegh, R. H. Doremus, and C. T. Moynihan, “IR spectroscopy studies of attack of liquid water on ZrF4-based glasses,” Mater. Sci. Forum 5–6, 311–322 (1985).
14. A. S. Oliveira, E. A. Gouveia, M. T. de Araujo, A. S. Gouveia-Neto, C. B. de Arau’jo, and Y. Messaddeq, “Twentyfold blue upconversion emission enhancement through thermal effects in Pr3 + /Yb3 +-codoped fluoroindate glasses excited at 1.064 µm,” J. Appl. Phys. 87(9), 4274–4278 (2000). [CrossRef]
15. L. E. E. De Araújo, A. S. L. Gomes, C. B. De Araújo, Y. Messaddeq, A. Florez, and M. A. Aegerter, “Frequency upconversion of orange light into blue light in Pr3+-doped fluoroindate glasses,” Phys. Rev. B Condens. Matter 50(22), 16219–16223 (1994).
16. Y. Messaddeq, A. Delben, M. Boscolo, M. A. Aegerter, A. Soufiane, and M. Poulain, “New fluoroindate glass compositions,” J. Non-Cryst. Solids 161, 210–212 (1993). [CrossRef]
19. H. Ebendorff-Heidepriem, T.-C. Foo, R. C. Moore, W. Zhang, Y. Li, T. M. Monro, A. Hemming, and D. G. Lancaster, “Fluoride glass microstructured optical fiber with large mode area and mid-infrared transmission,” Opt. Lett. 33(23), 2861–2863 (2008). [CrossRef] [PubMed]
20. T. Kanamori and S. Sakaguchi, “Preparation of Elevated NA Fluoride Optical Fibers,” Jpn. J. Appl. Phys. 25(2-6), L468–L470 (1986). [CrossRef]
21. R. Kostecki, H. Ebendorff-Heidepriem, S. C. Warren-Smith, and T. M. Monro, “Predicting the drawing conditions for Microstructured Optical Fiberfabrication,” Opt. Mater. Express 4(1), 29–40 (2014). [CrossRef]
22. G. Qian, F. Xia, J. Brugger, W. M. Skinner, J. Bei, G. Chen, and A. Pring, “Replacement of pyrrhotite by pyrite and marcasite under hydrothermal conditions up to 220 °C: An experimental study of reaction textures and mechanisms,” Am. Mineral. 96(11-12), 1878–1893 (2011). [CrossRef]
23. I. D. Aggarwal and G. Lu, Fluoride Glass Fiber Optics (Academic, 1991).
24. E. O. Gbogi, K. H. Chung, C. T. Moynihan, and M. G. Drexhage, “Surface and bulk -OH infrared absorption in ZrF4- and HfF4-based glasses,” J. Am. Ceram. Soc. 64(3), 51–53 (1981). [CrossRef]
25. N. N. Sagamihara, “Fluoride glass,” U. S. Patent 4 358 543, Nov 9, 1982.
27. T. Iqbal, M. R. Shahriari, G. Merberg, and G. H. Sigel, “Synthesis, characterization, and potential application of highly chemically durable glasses based on AlF3,” J. Mater. Res. 6(02), 401–406 (1991). [CrossRef]
28. J. Lucas, “Fluoride glasses,” J. Mater. Sci. 24(1), 1–13 (1989). [CrossRef]
30. C. Donley, D. Dunphy, D. Paine, C. Carter, K. Nebesny, P. Lee, D. Alloway, and N. R. Armstrong, “Characterization of Indium−Tin oxide interfaces using X-ray photoelectron spectroscopy and redox processes of a chemisorbed probe molecule: effect of surface pretreatment conditions,” Langmuir 18(2), 450–457 (2002). [CrossRef]
31. S. Sakaguchi and S. Mitachi, “Strength and fatigue of fluoride glass optical fibers,” J. Am. Ceram. Soc. 66(9), c151–c152 (1983). [CrossRef]
32. T. Shibata, H. Takahashi, M. Kimura, T. Ijichi, K. Takahashi, Y. Sasaki, and S. Yoshida, “Fabrication of high-strength, low-loss fluorozirconate glass optical fibers,” Mater. Sci. Forum 5–6, 379–385 (1985). [CrossRef]
33. P. C. Pureza, P. H. Klein, W. I. Roberts, and I. D. Aggarwal, “Influence of preform surface treatments on the strength of fluorozirconate fibres,” J. Mater. Sci. 26(19), 5149–5154 (1991). [CrossRef]
34. R. Lebullenger, S. Benjaballah, C. Le Deit, and M. Poulain, “Systematic substitutions in ZBLA and ZBLAN glasses,” J. Non-Cryst. Solids 161, 217–221 (1993). [CrossRef]