The potential of clear Ga2S3-GeS2-CsCl based sulfide glasses transparent up to 11.5 μm to be used as new optical material for multispectral applications has been investigated. The addition of large amount of chlorine ions – above 40 mol.% of CsCl – into the chalcogenide vitreous network in order to produce colorless glasses results in a drastic increase of their water contamination. We report for the first time, to the best of our knowledge, the purification of cesium chloride CsCl by dynamic distillations under vacuum in order to reduce water and hydroxyl group contamination before complete melting of the glass. Besides, sulfur purification by dynamic and static distillations was also performed in the implemented method. The obtained glasses were then characterized by UV-visible and infrared (FTIR) spectroscopies, by electron probe microanalysis (EPMA), thermal analysis (DSC), and their refractive indices in the visible and near infrared ranges were also measured. A large improvement of the glass transmission spectrum has been achieved with an estimated reduction of about 45 times of the OH and H2O content and 60 times of the SH content. The glass thermal molding ability and chemical durability with and without protective coating have been tested to probe their potential for fabrication of complex optics.
© 2014 Optical Society of America
The multispectral imaging technology is undergoing a rapid development since the last decade due to the advances in detection/acquisition systems, spectral measurements and data processing as well, offering new opportunities for a wide variety of applications in research and analysis based on remote sensing [1–3]. This technology takes advantage of the different spectral bands where the earth atmosphere is transparent, typically in the visible (0.4 to 0.7 μm), the near infrared range (NIR, 0.7 to 1 μm), the short-wave infrared range (SWIR, 1 to 2.7 μm), the mid-wave infrared (MWIR, 3 to 5 μm) and the long-wave infrared (LWIR, 8 to 14 μm), as represented in the Fig. 1.The demand for new single systems operating in two or more of these bands is continuously increasing for intelligence, surveillance and reconnaissance applications [1–3]. Such systems usually require optics and sensors adapted for each spectral band, limiting therefore the packaging and/or increasing the weight of the complete system and then its practical utilization. One of the avenues explored today to simplify and improve the capabilities of such systems is to replace their complex combination of reflective/refractive optics by more efficient ones operating in multiple spectral bands. The development of new optical materials transparent in multiple spectral bands is thus of great interest in view to share common optics and then simplify systems dedicated to multispectral applications.
The commercially available optical materials possessing a wide transmission window covering the multiple spectral bands from the visible up to the LWIR band are limited to: (i) polycrystalline or single crystal alkali-earth fluorides such as MgF2, CaF2 or even BaF2, transparent from the deep-ultraviolet (~100-200 nm) up to 6.5-11 μm depending on the alkali-earth; (ii) alkali chlorides such as sodium chloride NaCl and potassium chloride KCl that transmit light from 0.25 to 16 μm and 0.3 to 20 μm, respectively; (iii) polycrystalline Zinc sulfide ZnS or selenide ZnSe transparent from 0.4 to 12 μm and 0.6 to 20 μm, respectively. Germanium single crystal is also widely used for infrared optics but is opaque in both visible and NIR bands. The selection of an optical material for a given application is usually based on a compromise between its specific properties (e.g. thermal expansion coefficient CTE within temperature range of using, chemical and/or mechanical resistance, optical anisotropy, etc) versus the conditions of its practical utilization and its cost as well. Polycrystalline zinc sulfide or selenide (yellow/orange color) is usually a material of choice for multispectral applications, while CaF2 is usually preferred when UV/visible transparency is required. However, the fabrication of these materials with optical grade purity is very costly (e.g. CVD process), as well as their shaping in complex optics (e.g. diamond turning). To respond to the increasing demand for such specific optics and reduce the related manufacturing costs, the development of new vitreous materials appears thus as an alternative and competitive avenue, as it was successfully implemented in the last decade for mainstream infrared imaging .
Vitreous materials are widely used for a large variety of optical applications (as optics, optical fibers, thin films, etc) but only few of them exhibit a wide transmission window including the above cited Visible, NIR, SWIR, MWIR and LWIR bands. Among those reported in the literature, we can cite: (i) the heavy metal fluoride glasses based on fluoroindate which transmit light up to 8-9 μm ; (ii) some glasses belonging to Ga2S3-Na2S-CsCl  and Ga2S3-SrS systems transparent up to 11.5 μm ; (iii) the cadmium fluoro-chloride glasses transparent up to 13-14 μm . Barium gallo-germanate BGG glasses transparent up to 5-6 μm  were also intensively investigated and have already found applications where LWIR is not required. Recently, we have reported on the fabrication and characterization of colorless Ga2S3-GeS2-CsCl glasses exhibiting very wide transparency ranging from the near ultraviolet UV (~390 nm) up to the mid-infrared (11.5μm), covering thus almost all the bands of interest, as presented in Fig. 1 [10, 11]. More generally, the chalco-halide glass systems have raised great interest over the past years due to their better glass-forming ability if compared to that of the chalcogenide glasses well-known for their infrared applications [12–18]. Their extended composition ranges (glass-forming region) allow more flexibility to tailor the glass physical and chemical properties, making this class of vitreous materials a serious candidate not only for multispectral optical applications but also for various active/passive optic and photonic applications.
In the present work, we report on a new method for the purification and preparation of colorless Ga2S3-GeS2-CsCl glasses in order to reduce their content of extrinsic impurities and improve their transmission window. The high glass-forming ability of this system, firstly studied by Tveryanovich et al. , allows the incorporation of large amount of chlorine ions into the chalcogenide network, resulting in a colorless glass while typical color of sulfide glasses is yellowish or reddish. The addition of chlorine ions is accompanied by a contamination of the glass by water molecules, marked by the presence of specific extrinsic absorption bands on its optical transmission curve, as observed in Fig. 1. Large volume (150 grams) glass samples have been prepared and characterized; the repeatability of the fabrication process has been demonstrated. Finally, the prepared glasses were tested (thermal molding, chemical durability, protective coating) in view to probe their potential for fabrication of complex optics.
2. Samples preparation and experimental setup
2.1 Glass synthesis
The Ga2S3-GeS2-CsCl glass rods have been prepared following the classical melt-quenching route usually used for chalcogenide glass synthesis, i.e. into fused silica ampoule sealed under vacuum. The used precursors are metallic gallium (99.999%, Cerac Inc., USA), metallic germanium (99.999%, Hefa Rare Earth Canada Co. Ltd), sulfur (99.999%, Strem Chemicals Inc. USA) and cesium chloride (99.999%, Strem Chemicals Inc., USA). The synthesis fused silica ampoules containing the batches of about 150 g weight were heated up to 800-850°C at a rate of 1°C/min into a rocking tubular furnace; maintained at this temperature for at least 12h to provide a good fining; then quenched in water at room temperature and finally annealed during 5h at 250 °C and slowly cooled down to room temperature. The same procedure was utilized for all syntheses.
2.2 Starting materials purification
The main purpose of the present work has consisted in the purification of the raw materials in order to remove their contaminants causing strong absorption in the infrared range of the optical transmission window of the obtained glasses. It is important to mention that the above cited levels of purity (all of 99.999%), which are provided by the chemicals suppliers, are all based on metal content, and thus do not account for their water and/or hydrocarbons content. These contaminating agents are the main cause of extrinsic absorptions in specialty optical glasses, such as chalcogenide, fluoride or heavy metal oxide glasses, and their removal still remains challenging even after decades of research.
Here, the polycrystalline sulfur and cesium chloride are expected to be the main source of such contamination among the used precursors. Indeed, the cesium chloride CsCl, as an inorganic salt with a high water solubility of 191g/100g H2O at 298K  may carry large amount of water molecules to the glass, mainly in the form of hydroxyl groups (CsOH). Besides, the sulfur, which is mainly produced from the extraction of petroleum and other hydrocarbon fossil resources through the Claus process  may also contain large amount of hydrocarbons CxHy, H2S, H2O and OH hydroxyl groups. On the other hand, these contaminants are expected to be inexistent for both metallic gallium and germanium raw materials for which only surface oxidation has to be considered.
The experimental setup implemented to purify both sulfur and cesium chloride is schematized in Fig. 2.The entire complex system tubing is made of one single piece of fused silica (i.e. without any ground glass connection and vacuum grease) and designed in such a way to avoid any contamination from ambient air from the beginning of the process, i.e. the loading of the raw materials, up to the removal of the annealed bulk glass from its synthesis ampoule. Prior to the loading of the system with the weighed raw materials, the silica tubing was assembled, cleaned with hydrofluoric acid, rinsed with deionized water and finally heated at 800°C for several hours under dry nitrogen gas flow to remove adsorbed –OH groups from the inner walls of silica tubing.
The stoichiometric composition chosen for this study was (Ga2S3)25 - (GeS2)30 - (CsCl)45, in molar %. First, a blank glass sample has been prepared by direct loading of the commercial starting materials into the synthesis ampoule without performing any treatment for comparison purposes. Then, for the subsequent purified glass samples, gallium and germanium have been loaded into the synthesis ampoule, while the sulfur and cesium chloride have been loaded into their respective first tank, as depicted in Fig. 2. After loading, the system has been sealed and connected to vacuum. In a first step, the cesium chloride has been distilled 3 times under dynamic vacuum at 800-900°C. The silica tube between two consecutive tanks has been sealed off after each complete distillation in order to avoid backward of cesium salt vapor during process. Then the purification process of sulfur was started once the loading of CsCl into the synthesis ampoule was completed. The purification of sulfur was done by two successive dynamic distillations at 250°C followed by a static distillation to load the synthesis ampoule. As for the CsCl distillations, silica tubing was sealed off after each completed distillation while for the last one, the synthesis ampoule was sealed off from the pumping system, as indicated by position 8 in Fig. 2. The last distillation of sulfur was thus performed in vacuum static mode, providing a better control of loaded quantity of sulfur.
2.3 Glass characterizations
The UV-visible transmission spectra have been recorded on a double beam Cary 500 spectrophotometer while a Perkin Elmer Frontier FTIR spectrometer was used for the infrared transmission/absorption spectra. Polished sliced samples of about 2 mm thickness were used to record the transmission spectra while thicker samples (5-10 mm thickness) were preferred to get more accurate data from absorption spectra with the view to estimate the OH/H2O and SH impurities concentration.
Refractive index has been measured by employing the M-lines prism coupling technique (Metricon 2010) at 532, 632.8, 972, 1308 and 1538 nm in transverse electric (TE) and transverse magnetic (TM) modes. The Sellmeier function was then used to fit the experimental data recorded at room temperature as a function of wavelength and reconstruct the dispersion curve of the glass according to the procedure described in .
The differential scanning calorimetric (DSC) measurements were performed by using a Netzsch DSC Pegasus 404F3 apparatus on glass pieces into aluminum pans at a heating rate of 10°C/min up to 550°C. The thermo-mechanical analyses (TMA) were performed by using a Netzsch TMA Hyperion 402F1 equipment on glass rods of 20 mm length at a heating of 5°C/min and a load of 0.02N. The linear thermal expansion coefficient was then determined in the temperature ranging from 100 to 250°C.
The density ρ has been determined by the Archimedes’ method with an Alfa-Mirage MD-300S densimeter using deionized water as buoyant liquid. Large bulk pieces of about 40g weight were used to obtain accurate density measurements.
The elemental analysis measurements have been performed thanks to a CAMECA-SX100 electron probe micro-analyzer (EPMA). At least five quantitative measurements have been recorded for each sample to get an accurate mean value.
3. Results and discussion
3.1 Reduction of optical losses related to extrinsic impurities
The transmission spectra of the (Ga2S3)25-(GeS2)30-(CsCl)45 glass samples are presented in Fig. 3.Three main absorption bands are observed in the spectra, centered at 2.79 μm, 3.98 μm, and 6.24 μm and respectively ascribed to the OH, SH and H2O extrinsic absorptions. Other minor bands can also be noted on the spectra in Fig. 3, especially on the spectrum of the blank sample (without purification). Their position and respective attribution are resumed in Table 1.The infrared multiphonon cut-off is observed for these glasses near 11.5 μm.
One can observe from Fig. 3 a strong decrease of intensity of all the extrinsic absorption bands from the blank to the purified glass samples, indicating a strong improvement of the glass optical quality thanks to the implemented purification setup.
Furthermore, for comparison purposes, one has also included in Fig. 3 the transmission spectrum of the (Ga2S3)25-(GeS2)25-(CsCl)50 glass reported in our anterior work . In the latter work, the sulfur was purified by separate distillation prior to the glass synthesis. Once purified, it was stored in a dry glove box before to be weighed and loaded into the synthesis ampoule. Although the glass studied in  has a slightly different CsCl content, i.e. 50 mol.% vs 45 mol.% in the present work, both compositions are similar enough to be compared, and interestingly, one can observe in Fig. 3 similar large absorption bands due to the presence of OH/H2O and SH groups for both glasses. This indicates thus the inefficiency of sulfur purification by distillation prior to the glass synthesis if no precaution is taken to purify CsCl as well. Indeed, the addition of undried CsCl into the glass results in its contamination not only by forming OH/H2O groups but also SH ones, as shown in Fig. 3.
To assess the achieved purification, we have calculated the impurities concentrations from the absorption coefficient spectra by using the corresponding extinction absorption coefficients ε(SH) = 2.5 dB/m/ppm, ε(OH) = 4.6 dB/m/ppm, ε(H2O) = 9.2 dB/m/ppm. While the SH extinction absorption coefficient ε(SH) was determined in rather similar arsenic sulfide glass in  and references therein, there was no available literature on values for ε(OH) and ε(H2O) neither in chalcogenide glasses nor chalco-halide glasses, probably due to the high difficulty in determining accurately such values . For example, values of molar extinction coefficients for OH absorption ranging from about 40 to 90 L.mol−1.cm−1 and from 8 to 38 L.mol−1.cm−1 in silicate and fluoride glasses, respectively, were reported in  and references therein. Here we have arbitrarily chosen a molar extinction coefficient of 25 and 50 L.mol−1.cm−1 for OH and H2O absorption bands, respectively, considering that these calculations aim to quantify the reduction of impurity content between the two glass samples. With a measured glass density of ρ = 3.125 g/cm3, this gives values of ε(OH) = 4.6 dB/m/ppm wt. and ε(H2O) = 9.2 dB/m/ppm wt. The as-obtained impurity contents in ppm weight (1 mg of OH or H2O per 1 kg of glass) are reported in Table 2.One can observe from Table 2 that the purification setup has dramatically reduced by a factor of about 45 the concentrations of OH and H2O and by a factor of about 60 that of the SH content.
Nevertheless, one has to note that the purification setup has probably also resulted in some contamination of the glass, as suggested by the broad absorption band near 8.99 μm in the spectrum of the purified glass (inset of Fig. 3). Such absorption is also detectable on the blank glass spectrum, as well as on the transmission spectrum presented in Fig. 1. This absorption band is attributed to the presence Si-O groups and the contamination source is clearly the silica tubing walls. This was already well documented in literature and can be explained by the reactivity of some compounds with silica walls as well as using of high melting temperature (close to that of silica glass transition temperature Tg ≈1100°C). As the blank and purified glasses were melted and homogenized under the same conditions, we assume that this silica incorporation originated from the purification process and particularly the successive silica tubing seals off. Indeed, the sublimation of silica usually occurs at working temperature, which is a well-known issue in glass-blowing.
Regarding the metallic gallium and germanium used raw materials, no particular purification process was performed. First, even if germanium-containing chalcogenide glasses can be distillated under vacuum , such purification way is not achievable with gallium, due to its very low vapor pressure. To minimize contamination by surface oxides, freshly broken germanium metal pieces and melted metal gallium were used for the synthesis. The infrared absorption bands for GaO and GeO bonds are usually located near 12.8 and 7.9 μm, respectively . No evidence of presence of GeO bonds is observed on the spectra of Fig. 3, while it is not possible to conclude about the existence GaO bonds.
3.2 Repeatability of fabrication process - Control of stoichiometry
Large volume of (Ga2S3)25 - (GeS2)30 - (CsCl)45 glass was prepared for each purification process, typically for 100 to 150 g weight of material. This allows to minimize experimental error by better controlling the material losses – particularly sulfur and cesium chloride – induced by the multiple dynamic distillations. Photographs of a (Ga2S3)25 - (GeS2)30 - (CsCl)45 glass rod with about 20 cm length and 150g weight are presented in Fig. 4.
In order to control the conservation of stoichiometry between nominal and experimental compositions along with the purification process, the electron probe microanalysis (EPMA) tool was routinely used, providing valuable information about the sulfur S and cesium chloride CsCl losses during their multiple dynamic distillations. Thanks to this monitoring, it was thus possible to adjust the loading of raw materials in order to prepare the desired stoichiometry. Averaged values (from at least five measurements) from the blank glass and from 3 glass samples (labeled 2a, 2b and 2c) prepared according to the developed procedure are reported in Table 3.No significant difference of elemental composition is observed between the blank glass and purified glasses, demonstrating the excellent procedure repeatability.
3.3 Glass thermal properties and molding ability
The glass transition temperature determined after DSC analysis is Tg = 282°C while the measured onset crystallization temperature is Tx = 479°C. Their difference ΔT = 197°C indicates a relatively high thermal stability against crystallization and thus a high potential for these glasses to be thermally shaped without suffering devitrification.
The thermo-mechanical analysis (TMA) was performed on a 20 mm length (Ga2S3)25 - (GeS2)30 - (CsCl)45 glass rod and has permitted to determine a coefficient of thermal expansion CTE = 28 ppm.K−1 between 100 and 250°C. The dilatometric glass transition temperature measured is Tg = 278°C, which is in very good agreement with the Tg value determined by calorimetric analysis. The determined softening temperature for this glass is Ts = 340°C.
To assess the potential of this glass composition to be used as new material for optical lenses for instance and thus exploiting its multiband transmission window, simple tests by hot-pressing technique were carried out. This method is largely used to manufacture vitreous optical lenses and consists in heating a thick glass disc, a glass ball or bead around its softening temperature and pressing it within a mold to get the desired optics shape [4, 25]. Two similar experiments based on the scheme presented in Fig. 5(a) were conducted here. For the first one, whose aim is to show the ability of the glassy material to be thermally shaped without crystallizing, a large bulk glass piece, as shown in Fig. 5(b), was simply squeezed between two silica plates under a 600g weight load upon heat-treatment at 350°C for 2h. The sample thickness was thus reduced, as shown in Fig. 5(c), without altering its transparency. For the second experiment, the objective is to show the feasibility of imprinting a complex pattern at the surface of the glass.
The same experimental set-up is used but here a stamp with a relief on its surface is used and the pressing time is shortened to 30 min. Typically, in this type of test, simple metal coins are used as stamp to imprint their pattern on the glass surface . However, due to the reactivity of the metal alloys from which are produced the coins with our samples at such temperature, probably due to their high chlorine content, an alternative stamp had to be found. Assuming that the glass under study will not react with fused silica at high temperature, the metal coin was simply replaced by a silica stamp machined with a proprietary CO2 laser process by Laserax Company . They imprinted the negative of their company logo on the stamp, as presented in Fig. 5(e). Images of the glass sample before and after the hot-pressing test are presented in Figs. 5(d) and 5(f). One can observe in Fig. 5(f) the Laserax logo imprinted on the glass surface, without altering the glass transparency. All these experiments were conducted in ambient atmosphere. To better assess the achieved resolution of imprinting, 3D topography measurement was performed by using a stylus profilometer Dektak Veeco 150 equipment. The 3D image presented in Fig. 5(g) corresponds to the scanning of the fused silica stamp employed for this test. A maximum depth of about 250-260 μm is measured for the contours of the imprinted pattern. Note that the surface roughness observed in Fig. 5(e) (blue – violet color in Fig. 5(g)) corresponds to the region ablated by the CO2 laser during the stamp fabrication process. Figure 5(h) presents the 3D scan of the hot-pressed sample (Fig. 5(f)). One has to note that scanning entirely the logo from the stamp and imprinted sample shown respectively in Figs. 5(e) and 5(f), was not possible due to its too large size vs the stamp and sample dimension. Nevertheless, one can clearly observe in Fig. 5(h) an excellent contours precision of the imprinted pattern, with similar maximum depth around 250 μm.
In summary, one can clearly observe from the different images of the pressed samples that: (i) no glass devitrification nor deterioration is observed after thermal molding, ensuring the conservation of its optical transparency and; (ii) complex shapes and/or complex patterns can be obtained by thermal molding from this glass. Although all these performed tests are preliminary, the obtained results are promising for practical applications requiring molded optics of complex shapes such as lenses for multispectral imaging .
3.4 Glass optical properties
The glass UV cut-off wavelength λ0 was determined from the absorption coefficient spectra for a linear absorption coefficient value of α = 10 cm−1. Such arbitrarily selected value of α = 10 cm−1 is commonly used in literature to report the glass transmission window edge at short wavelength. Cut-off wavelengths λ0 ranging from 386 to 391 nm were determined for the prepared (Ga2S3)25-(GeS2)30-(CsCl)45 glass samples, ensuring the colorlessness of the samples with thicknesses up to 15-20 mm. Thicker samples may result in very slightly yellowish color due to the weak Urbach tail absorption at the beginning of the visible range (around 400 nm), as sometimes observed for heavy-metal oxide glasses based on tellurite.
The measured refractive indices of (Ga2S3)25-(GeS2)30-(CsCl)45 glass are presented in Fig. 6 as a function of wavelength and fitted with the Sellmeier function . Note that the values reported here were obtained in transverse electric (TE) mode with the Metricon M2010 Prism coupler. The transverse magnetic mode measurements which were also carried out (not presented here) have given very close index values – maximum measured difference of 1.10−3 – indicating a negligible birefringence in these glasses. The measured values are lower to those measured in germanium gallium sulfide glasses (n ~2.2) due to the addition of CsCl into the glass network, as previously discussed elsewhere [28, 29].
3.5 Glass chemical durability
Due to the high chlorine content of the Ga2S3-GeS2-CsCl glass required to guarantee its clearness, a relatively poor moisture resistance of the material is expected. Such issue was previously reported in  and other works on fluoro-chloride glasses [30, 31]. Whereas pure chalcogenide glasses are known to be insensitive to moisture, the progressive addition of alkali halide into chalcogenide glass results in a proportional decrease of its moisture resistance, as reported in . The chlorine-containing glasses usually suffer surface hydrolysis and water absorption when they are stored under ambient atmospheric conditions, resulting in a rising of the water-related absorption bands along with time. Figure 7(a) shows the transmission spectra of (Ga2S3)25-(GeS2)30-(CsCl)45 glass slice freshly polished and recorded after being stored in open ambient air for different durations. Although a dramatic decrease of the glass transmission is observed with time, a simple quick polishing of both surfaces allows the glass to recover almost entirely its initial transmission, as observed in Fig. 7(a). Furthermore, after being stored in dry atmosphere (i.e. in glove box with water level below 1 ppm) for more than 9 months, no significant alteration of the (Ga2S3)25-(GeS2)30-(CsCl)45 glass sample transmission was observed, as shown in Fig. 7(b). This clearly confirms the poor moisture resistance at glass surface but also demonstrates the good glass stability with time if it is kept away from any contact with moisture.
The deposition on the glass surface of a protective coating appears thus as a practical avenue in view to encapsulate and protect the material from the environment to extend its life time. The technology of coating deposition for optics is well developed, for instance to minimize surface reflections or improve abrasion resistance. Here, an appropriate coating material fulfilling the following requirements is required: (i) high resistance and impermeability to moisture, i.e. a non-porous/dense and hydrophobic coating; (ii) high adhesion to glass surface and; (iii) optical transmission window compatible with that of the glassy material.
Preliminary trials with deposition of YbF3 coating were performed. The ytterbium fluoride shows a very high adhesion to the glass (passing the tape-test traditionally used in thin-film deposition technology) and is colorless. The chemical durability tests carried out by water immersion of a glass slice with 400 nm-thick YbF3 film deposited on both surfaces have shown a strong improvement if compared to the uncoated glass  but more extensive efforts are required to develop an efficient protective coating in view to find practical applications.
3.6 Summary of glass characteristics and comparison with existing glassy materials
The main physical characteristics measured on the (Ga2S3)25-(GeS2)30-(CsCl)45 are summarized in Table 4 and compared with those of well-known glasses as fluorozirconate, fluoroindate, gallo-germanate and gallo-germanium sulfide which exhibit optical transparency from the UV-visible range up the mid-infrared.
One can observe from Table 4 that the Ga2S3-GeS2-CsCl glass offers one of the widest optical transmission range, slightly smaller than that of cadmium fluoro-chloride glasses and slightly larger than that of As2S3 and Ga2S3-GeS2 glasses. Despite a quite high CTE value, the Ga2S3-GeS2-CsCl glass possesses an excellent thermal stability against crystallization if compared to others, for instance the cadmium fluoro-chloride glasses. The BaO-Ga2O3-GeO2 glasses exhibit very good thermo-mechanical properties and have already found practical applications, but they are opaque in the 8-12 μm band. In summary, the Ga2S3-GeS2-CsCl glass appears to be an excellent compromise as vitreous multiband optical material, transparent in the 3 transmission bands of atmosphere and free of toxic elements like cadmium, arsenic or lead.
In this work, we report on the optical transmission enhancement achieved on the (Ga2S3)25-(GeS2)30-(CsCl)45 glass achieved thanks to a new complex purification and synthesis experimental setup based on multiple distillations of sulfur and cesium chloride raw materials. Comparison of estimated concentrations of OH/ H2O and SH impurities contained in unpurified and purified glasses has shown a reduction by a factor of about 45 and 60, respectively. The repeatability of the procedure described here has been carefully verified through stoichiometry conservation and glass properties as well. Nevertheless, the proposed method has also resulted in silica incorporation within the glass network, resulting in a new absorption band near 9.0 μm.
Besides the extra-wide optical transparency from 380 nm up to 11.5 μm of the (Ga2S3)25-(GeS2)30-(CsCl)45 glass, covering thus the 3 transmission bands of atmosphere, high thermal stability against crystallization (about 200°C) and thermal molding ability were demonstrated. The main drawback of this glass composition lies in its surface moisture sensitivity related to its high content of chlorine ions. However, different approaches exist to minimize this effect and deserve to be explored: (i) firstly, by depositing appropriate protective coating, like YbF3; (ii) by modifying the glass nominal composition and reducing the CsCl concentration, even this will intrinsically shift its optical band gap and then lead to pale yellow glass and; (iii) by combining both approaches.
Finally, it is worth mentioning that the method proposed here can be applied to any composition within the Ga2S3-GeS2-CsCl, with potentially even better results in terms of extrinsic impurities removal by reducing the CsCl concentration. The unique properties reported here make this material a promising candidate for the development of new optics for multiband optical applications.
This research was supported by the Canadian Excellence Research Chair program (CERC) on Enabling Photonic Innovations for Information and Communication. The authors are also grateful to the Natural Sciences and Engineering Research Council of Canada (NSERC), the Fonds de Recherche Québecois sur la Nature et les Technologies (FRQNT) and the Canadian Foundation for Innovation (CFI) for the financial support. Special thanks to Laserax Company for providing the laser-machined silica stamp.
References and links
1. M. T. Eismann, Hyperspectral Remote sensing (SPIE Press Book, 2012).
2. G. A. Shaw and H.-K. Burke, “Spectral imaging for remote sensing,” Lincoln Lab. J. 14, 3–28 (2003).
3. V. C. Coffey, “Multispectral imaging moves into the mainstream,” Opt. & Phot. News 23(4), 18–24 (2012). [CrossRef]
4. X. H. Zhang, Y. Guimond, and Y. Bellec, “Production of complex chalcogenide glass optics by molding for thermal imaging,” J. Non-Cryst. Solids 326-327, 519–523 (2003). [CrossRef]
5. Y. Messaddeq, A. Delben, M. A. Aegerter, A. Soufiane, and M. Poulain, “New fluoroindate glass compositions,” J. Mater. Res. 8(04), 885–889 (1993). [CrossRef]
6. L. C. Barbosa, C. L. Cesar, I. O. Mazali, and O. L. Alves, “Spectroscopic and thermal properties of Ga2S3-Na2S- CsCl glasses,” J. Am. Ceram. Soc. 89(3), 1037–1041 (2006). [CrossRef]
7. T. Ishibashi, H. Takebe, and K. Morinaga, “Glass forming region and structure of vitreous RS-Ga2S3 (R = Ca, Sr, Ba),” J. Ceram. Soc. Jpn. 111(1293), 308–311 (2003). [CrossRef]
8. M. Poulain, “Halide glasses,” J. Non-Cryst. Solids 56(1-3), 1–14 (1983). [CrossRef]
10. Y. Ledemi, L. Calvez, M. Roze, X. H. Zhang, B. Bureau, M. Poulain, and Y. Messaddeq, “Totally visible transparent chloro-sulphide glasses based on Ga2S3-GeS2-CsCl,” J. Optoelectron. Adv. Mater. 9, 3751–3755 (2007).
11. Y. Ledemi, B. Bureau, L. Calvez, M. Le Floch, M. Rozé, C. G. Lin, X. H. Zhang, M. Allix, G. Matzen, and Y. Messaddeq, “Structural investigations of glass ceramics in the Ga2S3-GeS2-CsCl system,” J. Phys. Chem. B 113(44), 14574–14580 (2009). [CrossRef] [PubMed]
12. B. Bureau, X. H. Zhang, F. Smektala, J. L. Adam, J. Troles, H. L. Ma, C. Boussard-Pledel, J. Lucas, P. Lucas, D. Le Coq, M. R. Riley, and J. H. Simmons, “Recent advances in chalcogenide glasses,” J. Non-Cryst. Solids 345-356, 276–283 (2004). [CrossRef]
13. J. S. Sanghera, L. B. Shaw, and I. D. Aggarwal, “Applications of chalcogenide glass optical fibers,” C. R. Chim. 5(12), 873–883 (2002). [CrossRef]
14. G. E. Snopatin, V. S. Shiryaev, V. G. Plotnichenko, E. M. Dianov, and M. F. Churbanov, “High-purity chalcogenide glasses for fiber optics,” Inorg. Mater. 45(13), 1439–1460 (2009). [CrossRef]
15. J. S. Sanghera, J. Heo, and J. D. Mackenzie, “Chalcohalide glasses,” J. Non-Cryst. Solids 103(2-3), 155–178 (1988). [CrossRef]
16. R. Balda, M. Sanz, A. Mendioroz, J. Fernandez, L. S. Griscom, and J. L. Adam, “Infrared-to-visible upconversion in Nd3+-doped chalcohalide glasses,” Phys. Rev. B 64(14), 144101 (2001). [CrossRef]
17. X. H. Zhang, L. Calvez, V. Seznec, H. L. Ma, S. Danto, P. Houizot, C. Boussard-Pledel, and J. Lucas, “Infrared transmitting glasses and glass-ceramics,” J. Non-Cryst. Solids 352(23-25), 2411–2415 (2006). [CrossRef]
18. Y. S. Tveryanovich, E. G. Nedoshovenko, V. V. Aleksandrov, E. Y. Turkina, A. S. Tveryanovich, and I. A. Sokolov, “Chalcogenide glasses containing metal chlorides,” Glass Phys. Chem. 22, 9–14 (1996).
19. http://www.hbcpnetbase.com/, CRC Handbook of Chemistry and Physics, 93rd Edition.
20. J. S. Eow, “Recovery of sulfur from sour acid gas: A review of the technology,” Environ. Prog. 21(3), 143–162 (2002). [CrossRef]
21. G. Ghosh, M. Endo, and T. Iwasaki, “Temperature-dependent Sellmeier coefficients and chromatic dispersions for some optical-fiber glasses,” J. Lightwave Technol. 12(8), 1338–1342 (1994). [CrossRef]
22. H. Ebendorffheidepriem and D. Ehrt, “Determination of the OH content of glasses,” Glastech. Ber. Glass Sci. and Technol. 68, 139–146 (1995).
23. S. Mitachi, G. Fonteneau, P. S. Christensen, and J. Lucas, “Molar extinction coefficients of OH in various types of fluoride glasses,” J. Non-Cryst. Solids 92(2-3), 326–336 (1987). [CrossRef]
24. V. S. Shiryaev, J. Troles, P. Houizot, L. A. Ketkova, M. F. Churbanov, J. L. Adam, and A. A. Sibirkin, “Preparation of optical fibers based on Ge-Sb-S glass system,” Opt. Mater. 32(2), 362–367 (2009). [CrossRef]
25. M. Schaub, J. Schwiegerling, E. C. Fest, A. Symmons, and R. H. Shepard, Molded optics, design and manufacture (CRC press, 2011).
26. M. Rozé, L. Calvez, J. Rollin, P. Gallais, J. Lonnoy, S. Ollivier, M. Guilloux-Viry, and X. H. Zhang, “Optical properties of free arsenic and broadband infrared chalcogenide glass,” Appl. Phys., A Mater. Sci. Process. 98(1), 97–101 (2010). [CrossRef]
28. P. Masselin, D. Le Coq, L. Calvez, E. Petracovschi, E. Lepine, E. Bychkov, and X. Zhang, “CsCl effect on the optical properties of the 80GeS2-20Ga2S3 base glass,” Appl. Phys., A Mater. Sci. Process. 106(3), 697–702 (2012). [CrossRef]
29. Y. Ledemi, M. El-Amraoui, L. Calvez, X. H. Zhang, B. Bureau, and Y. Messaddeq, “Colorless chalco-halide Ga2S3-GeS2-CsCl glasses as new optical material,” Proc. SPIE 8847, 884704 (2013). [CrossRef]
30. J. L. Adam, C. Ricordel, and J. Lucas, “New compositions of low phonon energy fluoride and chloro-fluoride glasses,” J. Non-Cryst. Solids 213-214, 30–35 (1997). [CrossRef]
31. M. Matecki and M. Poulain, “Composition adjustments in cadmium fluorochloride glasses,” J. Non-Cryst. Solids 140, 82–86 (1992). [CrossRef]
32. L. Calvez, P. Lucas, M. Roze, H. L. Ma, J. Lucas, and X. H. Zhang, “Influence of gallium and alkali halide addition on the optical and thermo-mechanical properties of GeSe2-Ga2Se3 glass,” Appl. Phys., A Mater. Sci. Process. 89(1), 183–188 (2007). [CrossRef]
33. Y. Ledemi, S. H. Messaddeq, I. Skhripachev, S. J. L. Ribeiro, and Y. Messaddeq, “Influence of Ga incorporation on photoinduced phenomena in Ge-S based glasses,” J. Non-Cryst. Solids 355(37-42), 1884–1889 (2009). [CrossRef]
34. M. Saad, “High Purity Fluoride Glass Synthesis a review,” Proc. SPIE 7228, 72280G (2009). [CrossRef]
35. P. L. Higby and I. D. Aggarwal, “Properties of barium gallium germanate glasses,” J. Non-Cryst. Solids 163(3), 303–308 (1993). [CrossRef]