Open Access
Add to BibSonomyAbstract
We report laser-induced modification of SbSI glass as a proof-of-concept of fabricating ferroelectric architectures in chalcogenide glasses. We observe structural (crystallization), chemical (evaporation of SbI3 and oxidation) and volume (contraction as well as expansion) changes under irradiation with a super-bandgap CW Ar+ laser due to the thermal runaway and photoexpansion effects. The crystalline grains grow, but in a very narrow range of applied laser power density (P) from 0.25 to 0.32 mW/μm2. At P>0.4 mW/μm2 SbI3 evaporation dominates and produces strong surface erosion. For P<0.2 mW/μm2 and long exposure times (1-10 min), only laser-induced expansion is observed.
©2013 Optical Society of America
1. Introduction
Synthesis and processing with energetic photons or electrons opens up new ways to create materials and devices that are not currently possible with established techniques. An intriguing feature of this approach is manifested when the material begins in the glassy state. The pre-determined two-dimensional (2D) structures or architectures can be made or written at or near the surface by selectively heating the glass with a laser or electron beam that is absorbed by the glass matrix. Through this heating crystalline features can be fabricated at the surface of glass, and in some special cases also deep inside the bulk.
Two different laser crystallization techniques are presently recognized. CW laser assisted modification of glass surfaces has been reported in many papers [1–9] and is facilitated by doping the desired glasses with rare-earth (RE) or transition metal (TM) ions. This concept is based on the absorption by the glass matrix, which leads to heating of the laser-irradiated region. Since RE or TM ions can be introduced into most glasses, this approach is considered the main tool for surface crystallization of glasses [1–9]. The second method is more universal to material micropatterning, including crystallization of glasses, and employs femtosecond lasers [10–18]. Due to the ultrashort light-matter interaction time and the high peak power density, localized crystallization with the femtosecond laser is observed only near the focused part of the laser beam. The wavelengths of these lasers belong to the transparency region of the glasses, and localized heating is invoked by multiphoton absorption processes. It should be emphasized that the CW laser is more conventional compared to femtosecond lasers, which are usually much more expensive.
These laser methods have been tested for crystallization of oxide [1–17] and fluoride [18] glasses, but very little is known about their applicability to chalcogenide glasses. Antimony sulphoiodide (SbSI) is a chalcogenide compound which exhibits ferroelectric properties in its crystalline phase at room temperature [19]. The fabrication of ferroelectric crystal architectures in an infrared transparent chalcogenide such as SbSI glass by a laser beam leads to a wide range of potential new applications due to the combination of exceptional electrical, mechanical and optical functionalities of ferroelectric crystals with the robustness, easy formability and low cost of chalcogenide glasses.
The goal of the present study is to crystallize stoichiometric SbSI glass using a CW laser beam. This was accomplished using the 488 nm emission of CW Ar+ laser. The wavelength of the laser is well above the bandgap of SbSI and is thus strongly absorbed, eliminating the need for TM or RE dopants as mentioned earlier for the case of oxides. Consequently, the penetration depth of the laser radiation is limited to several hundred nanometers. In this paper we report progress in the laser patterning of crystalline spots on the surface of SbSI glass, as well as challenges which appear due to the strong absorption of the CW irradiation. By using different exposure times and varying power densities we found regimes for crystallization on SbSI glass surface, which are identified here.
2. Experimental procedure
SbSI glasses were produced following the ampoule quenching method previously used for the SbSI-GeS2 system [20]. To obtain fast cooling rates, the inner diameter of the ampoules was reduced from 11 mm to 6 mm. The ampoules were slowly heated at 1°C/s successively to 128°C, 250°C, 450°C and 650°C and kept at each temperature for one hour. Finally, the batch was heated to 730°C and held for 12 h in a rocking furnace to facilitate mixing. Then the ampoules containing reacted melt were slowly cooled to 650°C and quenched in cold water to form glass. The 6 mm diameter glass rods were removed after cutting the quartz ampoule. From this rod 2-mm-thick pieces of SbSI were cut and polished with grit sizes down to 0.1 μm. X-ray powder diffraction (XRD) analysis of the as-quenched glasses confirmed their amorphous state. The glass transition (Tg) and maximum (Tc) crystallization temperatures were determined with a TA Instruments Model Q2000 differential scanning calorimeter (DSC). Samples were also characterized by thermo gravimetric analysis (TGA) using TA Instruments Model Q500 at 10°C/min.
A CW Ar+ laser operating at 488nm was used for writing spots [20,21]. The laser beam was focused on the polished surface of the glass sample by a microscope objective with a numerical aperture (NA) of 0.75. A polarizer served as an attenuator, while the hand-controlled shutter was used to block the beam after a spot formed. The power density used for laser writing varied between 0.1 and 1.7 mW/μm2.
The spots formed by the laser beam were observed with Scanning Electron Microscopy (Hitachi 4300 SE). Their crystallinity was examined by electron backscatter diffraction (EBSD) technique, in which Kikuchi patterns were obtained with a Hikari detector. The accelerating voltage for the EBSD analysis was 30 kV. The patterns were analyzed and indexed using TSL orientation data collection and analysis software.
Chemical analysis of the samples was performed using an Energy Dispersive X-ray (EDX) spectroscopy device attached to a Hitachi 4300 SEM in a low vacuum environment to eliminate the charging effects usually observed with nonconductive samples. For EDS analyses, an acceleration voltage of 20 kV and water vapor pressure of 30 Pa was chosen. The spectra were collected by using the EDAX-Genesis software package. The parameters for data acquisition (time, full scale for intensity, pulse processing time) were kept the same for all the spots. For the calibration of EDS measurements we fabricated SbSI crystals using the similar procedure as for making glass, but with the difference that rather than quenching, the ampoule was slowly cooled from 730°C to room temperature. X-ray powder diffraction analysis show that the resulting needle-like druse was SbSI crystalline phase (without Sb2S3 and SbI3 phases). The EDS analysis of one such crystalline “needle” yielded chemical composition within ± 1 at.% of the stoichiometric value.
3. Results
3.1 Thermal stability and crystallization propensity
To establish thermal stability, DSC measurements were performed on powders of SbSI glasses with different particle sizes at a heating rate of 10 K/min. The DSC trace for particles >500 μm in size shows glass transition (Tg) at 130°C, crystallization temperature (Tc) peaks at 155 and 195°C, and a very weak exothermal peak at 270°C. The powders with particle sizes of 63-177 μm crystallized almost fully at temperatures close to Tg and displayed only one strong exothermal crystallization peak at 136°C (Fig. 1, inset). A broad peak of low intensity was again observed above ~270°C. XRD analysis (of samples heated to 170 and 205°C) showed that the first and second peaks correspond to crystallization into the SbSI phase. The calculations of the Avrami exponent as well as an analysis of the intensities of DSC exothermal peaks and their dependence on particle size (Fig. 1, inset) show that the first crystallization process corresponds to one-dimensional crystallization that starts from surface. We believe that the second process represents a three-dimensional bulk crystallization of the SbSI phase. Two different kinds of crystallization are probably predetermined by strong anisotropy of SbSI chain’s crystal structure. The details of these results will be reported in a future paper.
Fig. 1 TGA and DSC (inset) data (heating rate 10°/min) of the SbSI glass for >500 μm (1) and <177 μm (2) size particles.
The thermo-gravimetric analysis (TGA) shows decomposition of the glass sample above 250°C (Fig. 1). Presumably, the DSC peak at 270°C is due to exothermal decomposition of the SbSI crystalline phase and the corresponding evaporation of SbI3 [22].
3.2 Laser-induced modification of SbSI glass
DSC analysis showed that SbSI glasses devitrify via surface as well as volume crystallization mechanisms. To produce laser-fabricated architectures, seed crystals can be made by focusing the laser at a particular spot from which crystal lines can be drawn subsequently [1]. We investigated morphological, structural and chemical changes induced by a CW laser beam in a spot as a function of laser power density (P) and time of exposure.
As the 488nm light is absorbed by the surface layer of the SbSI glass, the spot temperature strongly depends on the power density, which can be manipulated by adjusting the power as well as the focus position of the laser beam. Surprisingly, by applying relatively low power densities, craters are introduced on the surface. In order to determine the laser spot size for use in calculating the nominal laser power density at the glass surface, the laser power was incrementally lowered while the focal position was varied up and down relative to the glass surface. Eventually a power was reached for which even slight adjustment of the focus prevented the modification effect. We adopted this position as the “zero” focus point of the laser and the focus position used for all induced spots was referenced relative to this “zero” position. However, it should be noted that for the data discussed in this paper, the focus position was held constant at “zero” and P was varied.
For short times of laser exposure (~5s), structural or morphological changes at the glass surface were observed above a minimum threshold value of P, in which case a crater formed. Generally, these craters were observed to consist of two distinct morphological areas including a central region, the size of which did not depend strongly on the laser power, and an outer region whose width increased with increasing laser power. Figure 2 shows the relationship between both the width of the craters and their central “remelted” regions vs. P at the “zero” focus position. Evidently, the shape and size of the central region correspond to the shape and size of the focused laser beam. This experimentally determined value of the laser beam diameter (~2 μm) was used to evaluate the nominal power density of the laser beam.
Fig. 2 Width dependence of craters (1) and their central “remelted” region (2) on laser power after 5 s exposure at the “zero” focus position.
Whereas the width of the central region of the induced craters was unaffected by the laser power, the morphology was strongly dependent on P. In the range of 0.4-0.6 mW/μm2 the central region appeared to consist of mirror-smooth re-solidified liquid (Fig. 3(a)). At slightly higher power densities this region became rough (Fig. 3(b)) and at high densities (>1.0 mW/μm2) drops of sputtered material were observed in the area surrounding the central region (Fig. 3(c)). An analysis of the exposure time dependence showed that the craters appear in the first milliseconds of the laser irradiation process, and do not change with increasing irradiation time.
Fig. 3 SEM images of laser-fabricated spots by different laser powers 0.4 (a), 1.0 (b) and 1.7 mW/μm2 (c) at 5 s exposure. Scale bar corresponds to 0.5 μm.
Long exposures (from 1 to 10 min) of the SbSI glass surface using power densities less than 0.2 mW/μm2 induce expansion (Figs. 4(a), 4(b)). At a slightly higher P of 0.21 mW/μm2 the surface morphology becomes complicated, appearing as a “hill surrounded by moat” (Fig. 4(c)). This morphology was obtained for spot created with 0.22 mW/μm2 after 1 minute of exposure. Further exposure (e.g. a total of 2 or 3 minutes) removed the hill apex and resulted in the spots having 3 distinct morphological regions best described as a crater on the apex of a hill surrounded by a moat (Fig. 5).
Fig. 4 SEM images of laser-fabricated spots by different laser power densities 0.15 (a), 0.20 (b) and 0.21mW/μm2 (c) at 10 min exposure. Scale bar corresponds to 1.0 μm.
Fig. 5 SEM images of laser-induced spots created during 1, 2 and 3 min exposures (marked on the fig. in light font) with power densities of 0.22, 0.25, 0.29 and 0.32 mW/μm2 at the glass surface. Scale bar corresponds to 2.0 μm.
The most complicated morphology was observed for spots created with power densities from 0.25 to 0.32 mW/μm2. During exposures lasting a few minutes, a “hill surrounded by a moat” which formed initially was transformed into a few grains located centrally within a larger crater (see Figs. 5 and 6). Using Electron Back Scatter Diffraction (EBSD), Kikuchi patterns (see inset on Fig. 6) were obtained for some of these separated grains, which confirmed their crystallinity. The observed weak contrast of lines on Kikuchi patterns is probably the result of surface roughness, which introduces uncertainty to the determination of correct phase and orientation of a grain. Nevertheless, when the patterns were manually indexed using crystal structure parameters of the most likely phases, viz. SbSI, Sb2S3 and Sb2O3, the best agreement between predicted and experimental pattern was obtained with SbSI phase (Fig. 6, top right). So the complicated morphology of laser-modified spot is a result of competing processes including contraction, expansion, and crystallization of SbSI.
Fig. 6 SEM image of laser-induced spots created during 2 and 3 min exposures with a power density of 0.32 mW/μm2. Inset: collected and indexed EBSD patterns received from crystalline grain in the spot indicated within the modified area. Scale bar corresponds to 1.0 μm.
3.3 Chemical composition of spots
We conducted chemical analysis of the laser-induced spots using Energy Dispersive X-ray spectroscopy (EDS). The chemical composition was calculated from the acquired spectra by the ZAF procedure for the three major elements: Sb, S and I. Chemical compositions of the central points of spots are shown on the ternary phase diagram for the Sb-S-I system (Fig. 7). The chemical composition after laser irradiation with power densities from 0.4 to 0.8 mW/μm2 during 5s exposures “shifts” parallel to the quasi-binary Sb2S3-SbI3 section. Such a variation of composition suggests that the observed contraction results from the evaporation of SbI3 from the glass surface. The chemical composition of spots fabricated with higher power densities (for example, > 0.8 mW/μm2) shows additional depletion of S and an increase of Sb (see Figs. 7 and 8). A more detailed inspection of the EDS spectra reveals the presence of an additional peak at 523 eV, which corresponds to the Kα-edge of oxygen (Fig. 8). It should be noted that this peak is located close to the Mz-peak of iodine (498 eV) and their partial overlap complicates the ZAF calculation of the correct chemical composition.
Fig. 7 Chemical compositions at the central point of the laser-fabricated spots. The initial glass composition is marked as point 0. The black series of markers (1-4) indicates the compositional changes after 5s exposures with laser power densities of 0.6, 0.8, 1.0 and 1.7 mW/μm2,respectively. The red series (1-3) shows changes after 1, 2, and 3 minute exposures at a constant power density of 0.32 mW/μm2.
Fig. 8 EDS spectra for the initial glass (curve marked by 0) and central points of laser-fabricated spots using laser power densities 0.6 (curve 1) and 1.7 mW/μm2 (curve 2) for 5 s.
The chemical composition of spots fabricated by the laser beam with medium power density (0.25 to 0.32 mW/μm2) after 1 min exposure also indicates a “shift” parallel to the Sb2S3-SbI3 pseudo-binary line that corresponds to the evaporation of SbI3 (Fig. 7). The EDS spectra of spots fabricated by the laser beam with longer exposure times (2-3 min) show the presence of oxygen, the concentration of which increases with increased time of laser irradiation (Fig. 9). As in the case of spots induced by high power density for short irradiation times, these spots “shift” on the ternary diagram into a region with a lower concentration of S and a higher concentration of Sb in comparison to chemical contents of spots created after only 1 minute of laser exposure (Fig. 7). Comparison of an SEM image with a map of the oxygen distribution shows that oxidation occurs mainly in crystalline grains around the center of the spots (Figs. 5 and 10). Additionally, SbI3 evaporation continues from the center of these spots (Figs. 7 and 10).
Fig. 9 EDS spectra in range of the Kα-edge (523 eV) for the initial glass (0) and spots after 1, 2, and 3 minute exposures (1-3) with a constant power density of 0.32 mW/μm2.
Fig. 10 SEM image (a) and corresponding maps of O (b), I (c), and S (d) elemental distributions in the area of spots created by the laser beam during 1, 2 and 3 min exposures (marked by red numbers) with power densities of 0.22, 0.25, 0.29 and 0.32 mW/μm2 (shown by black values). Scale bar corresponds to 2.0 μm.
For spots that exhibited radiation-induced photoexpansion at low power density (<0.25 mW/μm2), we did not detect any deviation (within the experimental error of 1 at. %) of the chemical composition from the stoichiometric SbSI, even after exposure for 10 min.
4. Discussion
From the EDS analysis of the laser-induced spots the observed contraction appears to be a result of SbI3 evaporation. Raman spectroscopy of Sb-S-I glasses shows that the matrix is built of mostly SbS3 and SbI3 structural groups with weak molecular interaction between them [23–25]. The quasi-eutectic structure of Sb-S-I system glasses comprising of weak molecular bonds between SbS3 and SbI3 units is amenable to decomposition and evaporation of SbI3 molecules from the surface, especially in the liquid state. Apparently, the observed thermal instability (decomposition) of Sb-S-I glasses is not special to iodine-containing compounds [26], as it was observed also for other chalcogenides, for example, Ge20Se80 [27]. In this case, the authors explained the observed sample decomposition (the ejection of excess Se from Se-rich samples) as a thermal runaway effect [28]. They observed that when the rate of laser heating was greater than the rate of heat removal, the sample reached the ablation (decomposition) temperature in the local irradiated region. The onset of decomposition (ablation) occurs, as a rule, at sharp threshold fluence.
The selective evaporation can be obtained with any strongly absorbed laser beam. Especially in our case, where the CW laser wavelength of 488 nm is above the bandgap of SbSI, the laser beam is fully absorbed within a thin surface layer less than a few hundred nanometers thick. As shown in section 3, for SbSI glass the decomposition was observed clearly for spots created using P = 0.4 mW/μm2. If we assume that all laser energy is converted to heat and neglect all sources of heat loss, we can evaluate the amount of time needed to melt a particular volume of SbSI. In order to compensate for our previous assumptions, we used a volume equal to a few times that of the irradiated area multiplied by the nominal penetration depth of photons with a wavelength above the bandgap (a few hundred nanometers). In the case where the laser power density is 0.4 mW/μm2, we determined that enough energy to melt (405°C [19,29]) this volume is deposited within just a few microseconds. At the same time, TGA measurements revealed that SbSI sample lost mass at 250°C (section 3.1), which is well below the melting point. Thus, we can expect that a power density of 0.4 mW/μm2 is more than enough to heat a thin surface layer to temperatures higher than the decomposition temperature (250°C) in a relatively short time.
At power densities ≥0.4 mW/μm2 evidently the spot surface reaches temperatures which are higher than the decomposition temperature. Then, as a result of fast decomposition and evaporation of SbI3 molecules, a crater is formed at the glass surface. As the crater formation progresses the depth increases, thereby shifting the absorbing surface into the diverging region of the laser focus. This serves to decrease P and halt the process of SbI3 evaporation. Under lower P the rate of laser heating becomes smaller than the rate of heat removal and the spot temperature decreases. Additionally, we may expect that the decomposition power density threshold would increase with the depletion of volatile SbI3. For some comparatively high concentration of Sb2S3, the decomposition temperature could be higher than the eutectic melting point (~387°C [19,29]) as we observe a “remelted” region in the central part of spots (Fig. 3).
Alternatively, at intermediate power densities which deposit heat at a rate less than the decomposition threshold but higher than heat removal at room temperature, the temperature of the spot surface reaches a steady level after some time of irradiation. In this case the laser heat deposition becomes equal to the loss of heat as the temperature rises, and thermal gradient extending radially outward is established with the maximum value in the illuminated region. It should be noted that such heating of undersurface layers must involve their thermal expansion, which would disappear with the termination of irradiation. However, we observe a permanent ‘hill’, the height of which depends on P (see Figs. 4(a), 4(b)). Similar photoexpansion effects are observed in a wide range of chalcogenide glasses [30–33] as well as in some oxide glasses [34,35].
The value 0.22 mW/μm2 evidently corresponds to the decomposition power density threshold for SbSI glass. Around this power density we obtained three distinct spot morphological regions (Fig. 5). At the beginning the spot surface reaches a temperature which is close to decomposition temperature. As a result there is very small evaporation of SbI3 (less than the sensitivity of EDS method – see Fig. 10), which appears as a low depth crater at the glass surface. The radiation-modified surface shifts down slightly relative to laser focus and the power density becomes less than the decomposition threshold. Nevertheless, the laser heat deposition is enough for installation of a steady temperature distribution around spot. Certainly, continued heating of the surrounding bulk regions involves their thermal expansion, which was observed after 1 min exposure (Fig. 5). Longer laser irradiation (2 and 3 min) causes further expansion and so the glass surface begins to rise until it reaches a beam cross-section where value of P is higher than decomposition threshold. Correspondingly, SbI3 again evaporates from the top of this “hill”.
At some laser power density and time conditions (for example, at P ~0.25-0.32 mW/μm2), surface crystallization of the SbSI phase becomes possible (Figs. 5 and 6). This change appears around the center of the spot. Crystallization does not occur in the center of the spot due to the evaporation of SbI3 (see Figs. 5 and 10). Here the laser power density exceeds the decomposition threshold owing to non-uniform Gaussian distribution of power across the laser beam. The chemical composition is enriched with Sb2S3 and this prevents the process of SbSI surface crystallization there.
On further increasing the power density to 0.35 mW/μm2, the SbI3 evaporation dominates in the whole illuminated region and thus only the photocontraction of the crater is observed. Thus, surface SbSI crystallization can occur only in a comparatively narrow range of power density from 0.25 to 0.32 mW/μm2 at long exposure times (>1 min). Our experimental data show that at these conditions the rate of surface crystallization is low and decomposition threshold is close to the temperature range of surface crystallization (120-160°C, see section 3.2). Possibly for the illuminated viscous supercooled liquid state the decomposition begins at temperature lower than 250°C noted from for the slowly heated (10K/min) and fully crystallized sample in TGA measurements (Fig. 1). The proximity of the temperatures of decomposition and surface crystallization complicates the crystallization of stoichiometric SbSI glass by CW laser irradiation significantly.
From the EDS map of the oxygen distribution (Fig. 10) we can conclude that surface oxidation is observed at low power densities and long exposure times simultaneously with the crystallization. At present it is difficult to determine the role of oxygen in crystallization process, but we must take into account oxidation especially when using long periods of laser irradiation. The rate of oxidation increases with rising temperature (proportional to P) and becomes detectable in spot-craters fabricated by the laser beam even for short exposure times (5 s) for power densities higher than 0.8 mW/μm2.
5. Summary
To understand the complexities of laser interactions with technologically important chalcogenide glasses, samples of stoichiometric SbSI model composition were irradiated with a super-bandgap CW laser. In contrast to slow thermal heating (typically <15K/min), the laser causes chemical changes (damages) due to a thermal runaway effect. At low laser power density (P<0.2 mW/μm2) and long exposure times (0.5-10 min) laser-induced expansion is observed. At higher P (>0.4 mW/μm2) SbI3 evaporation dominates and produces strong surface erosion. The surface crystallization to the SbSI phase occurs in a narrow range of P from 0.25 mW/μm2 to 0.32 mW/μm2. A smaller P is not enough to heat the surface region to crystallization, and at greater P the decomposition takes over with the preferential evaporation of SbI3. For P between these limits the decomposition is still observed but at the center of spots. The evaporation of the SbI3 enriches the remaining supercooled liquid with Sb2S3, which complicates surface crystallization of the SbSI phase. In contrast, the changes of chemical composition are lower around the central region and this allows the formation of SbSI crystal grains there. However, the growth of SbSI crystalline grains in regions enriched with Sb2S3 allows us to propose SbI3-poor compositions as a more promising precursor for the laser fabrication of SbSI crystalline architectures.
Laser writing in chalcogenide glasses such as the present Sb-S-I system is found to be inherently more complicated than in oxide glasses due to preferential evaporation and oxidation (if performed in air). Notwithstanding, the present results have demonstrated that crystalline architecture can be created with optimal combination of laser power density, focus position, duration of irradiation and chemical composition of precursor glass. Examples of such architectures are presented elsewhere [27]. Moreover, the laser-induced shrinkage and expansion in combination with SbSI crystallization opens a way to fabricate and integrate different types of passive (e.g. optical waveguides, Bragg gratings, microlenses, lens arrays) as well as active (electro-optic modulators, wavelength convertors) micro-optical elements in the same glass substrate using different regimes of CW laser irradiation.
Acknowledgments
This work is supported by the Basic Energy Sciences Division, Department of Energy (project DE-FG02-10ER46698). BK, who contributed to the laser irradiation part of the work, is supported by NSF (DMR-0906763).
References and links
1. T. Komatsu, R. Ihara, T. Honma, Y. Benino, R. Sato, H. G. Kim, and T. Fujiwara, “Patterning of non-linear optical crystals in glass by laser-induced crystallization,” J. Am. Ceram. Soc. 90(3), 699–705 (2007). [CrossRef]
2. S. Mizuno, Y. Benino, T. Fujiwara, and T. Komatsu, “Novel technique for fabrication of nanoparticle structures in KNbO3-TeO2 glass for photonic integrated circuit,” Jpn. J. Appl. Phys. 45(8A), 6121–6125 (2006). [CrossRef]
3. T. Fujiwara, T. Honma, S. Mizuno, N. Iwafuchi, Y. Benino, and T. Komatsu, “Order/disorder hybrid structures in photonic glass materials,” Adv. Mater. Res. 11–12, 53–56 (2006). [CrossRef]
4. R. Ihara, Y. Benino, T. Komatsu, and T. Fujiwara, “Fabrication of optical waveguide in glass by laser-induced crystallization,” Adv. Mater. Res. 11–12, 197–200 (2006). [CrossRef]
5. N. Iwafuchi, S. Mizuno, Y. Benino, T. Fujiwara, T. Komatsu, M. Koide, and K. Matsushita, “Nano-crystallized glass fibers with second-order optical nonlinearity,” Adv. Mater. Res. 11–12, 209–212 (2006). [CrossRef]
6. T. Honma, “Laser-induced crystal growth of nonlinear optical crystal on glass surface,” J. Ceram. Soc. Jpn. 118(1374), 71–76 (2010). [CrossRef]
7. T. Honma, Y. Benino, T. Fujiwara, T. Komatsu, and R. Sato, “Technique for writing of nonlinear optical single-crystal lines in glass,” Appl. Phys. Lett. 83(14), 2796–2798 (2003). [CrossRef]
8. T. Honma, Y. Benino, T. Fujiwara, and T. Komatsu, “Transition metal atom heat processing for writing of crystal lines in glass,” Appl. Phys. Lett. 88(23), 231105 (2006). [CrossRef]
9. H. Masai, T. Fujiwara, Y. Benino, T. Komatsu, and H. Mori, “Selective surface crystallization of nonstoichiometric 30BaO-15TiO2-55GeO2 glass,” J. Appl. Phys. 101(3), 033530 (2007). [CrossRef]
10. Y. Yonesaki, K. Miura, R. Araki, K. Fujita, and K. Hirao, “Space-selective precipitation of non-linear optical crystals inside silicate glasses using near-infrared femtosecond laser,” J. Non-Cryst. Solids 351(10-11), 885–892 (2005). [CrossRef]
11. Y. Dai, B. Zhu, J. Qiu, H. Ma, B. Lu, S. Cao, and B. Yu, “Direct writing three-dimensional Ba2TiSi2O8 crystalline pattern in glass with ultrashort pulse laser,” Appl. Phys. Lett. 90(18), 181109 (2007). [CrossRef]
12. Y. Dai, B. Zhu, J. Qiu, H. Ma, B. Lu, and B. Yu, “Space-selective precipitation of functional crystals in glass by using a high repetition rate femtosecond laser,” Chem. Phys. Lett. 443(4-6), 253–257 (2007). [CrossRef]
13. B. Zhu, Y. Dai, H. Ma, S. Zhang, G. Lin, and J. Qiu, “Femtosecond laser induced space-selective precipitation of nonlinear optical crystals in rare-earth-doped glasses,” Opt. Express 15(10), 6069–6074 (2007). [CrossRef] [PubMed]
14. K. Miura, J. Qiu, T. Mitsuyu, and K. Hirao, “Space-selective growth of frequency-conversion crystals in glasses with ultrashort infrared laser pulses,” Opt. Lett. 25(6), 408–410 (2000). [CrossRef] [PubMed]
15. Y. Dai, H. Ma, B. Lu, B. Yu, B. Zhu, and J. Qiu, “Femtosecond laser-induced oriented precipitation of Ba2TiGe2O8 crystals in glass,” Opt. Express 16(6), 3912–3917 (2008). [CrossRef] [PubMed]
16. A. Stone, M. Sakakura, Y. Shimotsuma, G. Stone, P. Gupta, K. Miura, K. Hirao, V. Dierolf, and H. Jain, “Directionally controlled 3D ferroelectric single crystal growth in LaBGeO5 glass by femtosecond laser irradiation,” Opt. Express 17(25), 23284–23289 (2009). [CrossRef] [PubMed]
17. Y. Teng, B. Qian, N. Jiang, Y. Liu, F. Luo, S. Ye, J. Zhou, B. Zhu, H. Zeng, and J. Qiu, “Light and heat driven precipitation of copper nanoparticles inside Cu2+-doped borate glasses,” Chem. Phys. Lett. 485(1-3), 91–94 (2010). [CrossRef]
18. Y. Liu, B. Zhu, Y. Dai, X. Qiao, S. Ye, Y. Teng, Q. Guo, H. Ma, X. Fan, and J. Qiu, “Femtosecond laser writing of Er3+-doped CaF2 crystalline patterns in glass,” Opt. Lett. 34(21), 3433–3435 (2009). [CrossRef] [PubMed]
19. E. I. Gerzanich, V. A. Lyakhovitskaya, V. M. Fridkin, and B. A. Popovkin, in Current Topics in Materials Science, E. Kaldis, ed. (North-Holland, Amsterdam, 1982), Vol. 10, pp. 55–190.
20. P. Gupta, A. Stone, N. Woodward, V. Dierolf, and H. Jain, “Laser fabrication of semiconducting ferroelectric single crystal SbSI features on chalcohalide glass,” Opt. Mater. Express 1(4), 652–657 (2011). [CrossRef]
21. DB. Savytskii, V. Knorr, Dierolf, and H. Jain, “Formation of laser-induced SbSI single crystal architecture in Sb–S–I glasses,” J. Non-Cryst. Solids (to be published) http://www.sciencedirect.com/science/article/pii/S0022309313000264#.
22. http://www.antimony.atomistry.com/antimony_thioiodide.html.
23. I. D. Turyanitsa, L. K. Vodopyanov, V. M. Rubish, L. Y. Kengermensky, and V. M. Dobosh, “Raman spectra and dielectric properties of glasses in Sb–S–I system,” Zh. Prikl. Spektrosk. 44, 798–804 (1986) (in Russian).
24. I. D. Turyanitsa, T. N. Melnichenko, P. P. Shtets, and V. M. Rubish, “Glass preparation, structure and properties of alloy in Sb–S–I system,” Izv. Akad. Nauk SSSR, Neorg. Mater. 22, 2047–2050 (1986) (in Russian).
25. V. M. Rubish, “Thermostimulated relaxation of SbSI glass structure,” J. Optoelectron. Adv. Mater. 3, 941–944 (2001).
26. S. R. Lukic, D. M. Petrovic, I. O. Get, and M. I. Avramov, “Complex non-crystalline chalcogenides: technology of preparation and spectral characteristics,” J. Phys. Res. 30, 111–130 (2006).
27. G. Beadie, W. S. Rabinovich, J. Sanghera, and I. Aggarwal, “Fabrication of microlenses in bulk chalcogenide glass,” Opt. Commun. 152(4-6), 215–220 (1998). [CrossRef]
28. G. Beadie and N. M. Lawandy, “Single-step laser fabrication of refractive microlenses in semiconductor-doped glasses,” Opt. Lett. 20(21), 2153–2155 (1995). [CrossRef] [PubMed]
29. T. Mori and H. Tamura, “Preparation of ferroelectric SbSI single crystals,” J. Phys. Soc. Jpn. 19(7), 1247 (1964). [CrossRef]
30. H. Hisakuni and K. Tanaka, “Optical microfabrication of chalcogenide glasses,” Science 270(5238), 974–975 (1995). [CrossRef]
31. H. Hisakuni and K. Tanaka, “Giant photoexpansion in As2S3 glass,” Appl. Phys. Lett. 65(23), 2925–2927 (1994). [CrossRef]
32. L. Calvez, Z. Yang, and P. Lucas, “Light-induced matrix softening of Ge-As-Se network glasses,” Phys. Rev. Lett. 101(17), 177402 (2008). [CrossRef] [PubMed]
33. P. Knotek and L. Tichy, “On photo-expansion and microlens formation in (GeS2)0.74(Sb2S3)0.26 chalcogenide glass,” Mater. Res. Bull. 47(12), 4246–4251 (2012). [CrossRef]
34. A. Streltsov, J. Dickinson, R. Grzybowski, D. Harvey, S. Logunov, A. Ozturk, J. Sutherland, and M. Potuzak, “Laser texturing of doped borosilicate glasses,” Proc. SPIE 7584, 75840S (2010). [CrossRef]
35. R. Grzybowski, S. Logunov, A. Streltsov, and J. Sutherland, “Extraordinary laser-induced swelling of oxide glasses,” Opt. Express 17(7), 5058–5068 (2009). [CrossRef] [PubMed]
