High second-order susceptibility has been created in a chalcogenide glass from Ge-Sb-S system. A thermal poling process was used to produce this non-linear effect and a second harmonic generation experiment allowed characterizing the phenomenon. A maximum χ(2) value of 8.0±0.5 pm/V was measured for the first time to our best knowledge in sulfide glasses.
©2006 Optical Society of America
Chalcogenide glasses are photosensitive and they have high linear and non-linear indices [1–4]. Their spectral range of transparency extends until 10-20 μm, that is why they are called “infra-red” glasses. For these reasons, they are potential candidates for optical components achievements: waveguides and electro-optic modulators for a use in the infra-red spectral region.
In amorphous materials such as glasses, second-order non-linear (NL) properties are naturally forbidden. However the thermal poling turns out to be a successful technique for creating such nonlinearities . It consists in producing and freezing an electrical polarization inside the glass. This polarization is at the origin of a second-order non-linearity (SON) which then allows optical properties such as second harmonic generation (SHG) or three waves mixing. The induced electric field inside the sample is due either to charges motions and/or atomic bonds hyperpolarizability . Generally, the SON is mostly localized under the anodic side of the poled sample.
In the present work, we mention the results of thermal poling treatments applied to a chalcogenide glass with Ge25Sb10S65 (2SG) nominal composition. This composition choice follows previous poling experiments performed on various batches containing the gallium element such as Ge20Ga5Sb10S65 (2S2G) . We report in this article that this 2SG composition exhibits a much more efficient SHG and a higher SON than chalcogenide glasses containing gallium.
2. Experimental conditions
2.1. Glass synthesis and characterization
High purity elements were used for the glass preparation: germanium, antimony and sulfur 5N. Despite the good purity of the commercial materials, sulfur can be polluted by water and carbon. Water was eliminated by heating sulfur under dynamic vacuum at 125°C and carbon by sulfur distillation at 350°C. After purification the raw materials were placed in a silica tube sealed under vacuum (10-5 mbar). After sealing, the batch was slowly heated to 850°C and homogenized in a rocking furnace at this temperature for 12 h. A glass rod was obtained by cooling the silica tube in air. It was then annealed near the glass transition temperature for 30 min before slowly cooling down to room temperature. Several glass plates about 1 mm in thickness and 20 mm in diameter were obtained from a same batch. For optical characterizations, the samples were optically polished to get sides as plan and parallel as possible.
Thermal analyses were carried out on single glass chips, about 50 mg, in sealed aluminum pans in the temperature range of 25-500°C. The measurements were performed at a heating rate of 10°C/min by means of a differential scanning calorimeter with an accuracy of ± 2°C. Analysis using Energy Dispersive X-Ray Spectroscopy (EDS) allowed the identification of the elemental composition of the studied samples imaged in a Scanning Electron Microscope (SEM) and permitted to control the stoechiometry of the obtained glasses.
2.2 Thermal poling conditions
Thermal poling process was carried out in the temperature range of 100°C to 310°C limited by the glass transition temperature Tg (Tg = 355°C ± 2°C). Thermal equilibrium duration before applying the voltage was fixed to 90 min. The electric field (4-5 kV) was then applied by means of two silicon wafer electrodes for a duration in the range of 5 to 60 min. The samples were removed from the furnace with the voltage still applied. This procedure allows cooling down the glass rapidly to room temperature As a consequence, the SON is supposed to be frozen inside the glass. Various set of poling temperature, voltage and time were tested to optimize the SON. We note that all poling treatments were realized under Ar saturated atmosphere to avoid the surface oxidation of the sulfide glass samples.
2.3. Second-order non-linearity characterization
For NL measurements, all samples were first characterized using a classical Maker fringes (MF) experiment . The optical gap for this glass being about 525 nm, the wavelength of the pump beam was chosen in the near infrared region (~ 1.9 μm) in order to avoid any absorption of the SH signal. This pump beam was focused on the sample which can rotate from -85° to 85°.
A precise determination of the NL profile was obtained by using the method described in Ref. . This technique consists in recording the SH signal for a fixed angle (30° with respect to the pump beam) while the anodic surface of the sample is etched. The chemical etching is performed by using a sodium hydroxide (NaOH) solution (0.2 mol/L). While this setup is destructive for the sample, it provides the shape of the NL susceptibility as a function of the position across the sample. The total removed thickness was measured at the end of the etching process (i.e. after the SH power has dropped to zero) using a profilometer. We assume a constant etching rate (0.028μm/s) during the experiment to deduce the removed thickness: this assumption relies on the fact that the removed thickness of poled region doesn’t not differ more than 2-3μm compared to unpoled region. In all cases, only a part of the poled region was etched, the non-etched part was used to control the stability of the SH signal. Following these experiments, the χ(2) profile was reconstructed with help of the “layer peeling” method . Indeed, providing the knowledge of the SH power as a function of the removed thickness during the etching, the SON spatial distribution can be obtained by an iterative algorithm, as extensively presented in Ref. .
The two methods (namely Maker fringe setup and layer peeling method) are based on the fact that the NL susceptibility is related to the SH peak power through the following equation:
where K(θi ) is a factor that takes into account the optical characteristics of the pump beam (peak power, wavelength and beam waist radius), of the sample indices (at both pump and SH wavelengths), and of the Fresnel transmission coefficients at the interface between air and glass. The integral contains a shape of χ(2) along the glass thickness z, which is linked to the inner field EDC via the relation : χ(2) = 3χ(3) EDC .
The refractive indexes of the studied sulfide glass are estimated to be nω = 2.193 and n2ω = 2.223, the associated coherence length Lc defined by Lc =π/∆k = ∆/4(n2ω -nω ) is about 11.9 μm.
3.1 Influence of the poling temperature
The influence of the temperature T of the poling process was studied between 100°C and 310°C. In order to well understand the temperature effect, we fixed the poling time (30 min) and the intensity of applied voltage (4-5 kV).
On Fig. 1 is plotted the maximum intensity peak of the MF curves versus the temperature used for the experiments. An optimal poling temperature is observed around 170°C. For decreasing temperatures, the poling effectiveness becomes weaker and SON can not be created for a poling temperature equal to 100°C. Above this optimal temperature, the SH signal is a decreasing function of the temperature.
On Fig. 2. are represented the full MF patterns recorded for temperatures equal to 170°C [Fig. 2(a)], 230°C [Fig. 2(b), continuous line] and 310°C [Fig. 2(b), dashed lines]. The shapes of the recorded curves differ according to the temperature. Indeed, the curve recorded for the sample poled at 170°C does not exhibit any oscillations, while they can be distinctly observed for a temperature of 310°C. Above 230°C, clear over modulations on the MF pattern are noticed.
The NaOH etching was performed for two poling temperatures: 170°C and 230°C. The width of the nonlinear (NL) layer was estimated at approximately 30 μm in the case of poling temperature at about 170°C, and around 210 μm in the case of 230°C poling temperature. In this sulfide glass, large NL width is expected to be responsible of the appearance of overmodulations for poling temperature above 200°C.
3.2 Influence of the poling duration
Considering the optimal poling temperature of 170°C., we have studied the influence of the poling duration on the SON distribution. The voltage was fixed to 4kV and three poling durations were tested: 5, 30 and 60 min. In Fig. 3 are plotted the MF curves obtained for all these three cases. No visible overmodulation is recorded at this temperature whatever the duration of the poling process. The maximum SH intensity decreases for longer poling durations and seems to reach saturation. Using the NaOH etching technique, the thickness of the NL layers of glasses poled during 5, 30 and 60 min were estimated respectively to 13, 30 and 37 μm.
Since the higher SH signal was measured for the sample poled for 5 min., its anodic side has been submitted to a NaOH etching, in order to get the precise spatial distribution of the SON. To this end, the technique described in the above section has been used. When the SH signal has dropped to zero, the chemical etching has been stopped and a profilometer was used to measure the total removed thickness with a ±0.5μm error. These experimental measurements allowed the knowledge of the SH power as a function of the depth under the anodic side, represented in the Fig. 4. One can observe a quasi smooth decreasing SH signal versus the etched depth. This curve exhibits a fast decrease until 15 μm under the surface, where the SH power drops to zero.
To get the χ(2) susceptibility spatial distribution, we proceeded as in Ref.  using the “layer peeling” method: the result is displayed on Fig. 5. The χ(2) coefficient is approximately constant from the surface until 10 μm deeper inside the sample. It then decreases abruptly until 15 μm, where it vanishes.
The real value of the SON (shown in Fig. 5) has been deduced from the Maker fringe setup which allows an absolute measurement of the SH power generated inside the poled sample. The χ(2) magnitude is obtained by comparing the experimental SH peak power recorded with the Maker fringe setup to theoretical one, calculated using the profile of Fig. 5. The SON value was thus obtained by scaling the magnitude of the reconstructed χ(2) to get the best fit of the experimental MF data as shown in the insert of Fig. 5. The magnitude of the χ(2) susceptibility was finally estimated to be around (8.0±0.5) pm/V.
As shown in Fig. 1, the optimal poling temperature is located around 170 °C for a poling duration of 30 min. Below this optimum, the temperature is too weak to provide a sufficient energy to pass the barrier of activation allowing the induction of the charge migration phenomena and/or orientation of dipoles. The appearance of overmodulation on Maker fringes curves for temperature higher than 200°C, not present in the case of glasses of the systems Ge-Ga-Sb-S , seems to be a particular behavior for glasses of the system Ge-Sb-S. It is noticeable at this point that the overmodulations are known to be present on MF in case of two differentiated NL layers (as a result of constructive and destructive interferences) or a volumic field distribution (at least larger than 2LC). The contrast (and visibility) of the overmodulation depends on the ratio between the layers thickness among others. The number of overmodulations is related to the total NL thickness. The appearance of overmodulations on the recording of the Maker fringes for temperature higher than 170 °C reflects that the thickness of the NL layer grows when the temperature of polarization increases. This assumption is confirmed by measurements of the NL thicknesses for two different temperatures (170° C and 230°C). For a given temperature, a thickness increase of the nonlinear layer is also observed when the poling time increase associated with a decrease in SH signal magnitude.
This behavior can be understood if we assume that migration of positive mobile ions are at the origin of the NL layer creation in thermally poled Ge25Sb10S65 glass. The migration of cations, like Na+, towards the cathodic side to the detriment of the anodic side probably induces a depletion layer negatively charged within a few micrometers near the anode. This accumulation of negative charges leads to the creation of a high electric field EDC near the anodic side of the sample as it has been demonstrated in poled silica glasses . The increase of the NL width for a fixed temperature can be explained in this context by the presence of a second charge carrier (for example H+) that can be injected through the anodic surface of the sample during poling. As a consequence, charge migration models predict that the magnitude of the nonlinear coefficient decreases. This feature has been experimentally observed and theoretically demonstrated in silica glasses . In particular, the extension of the NL layer that we observed with both an increasing temperature and duration of poling is in good agreement with the two charge carriers model developed by Kudlinski et al. .
In chalcogenide glasses, this mechanism can be followed by the reorientation of the lone pairs of sulfur and antimony under the induced electic field EDC, which can also induces nonlinearities via their second-order hyperpolarizability. MF curves in pp and sp polarization configurations, where pp (respectively sp) means p-polarized (respectively s-polarized) pump beam and p-polarized SH transmitted beam were recorded in order to calculate / ratio. The χ333 (2)/ ratio is estimated to be around 3 so it supposes that both charges migration and weak dipoles reorientation mechanisms could occur without the possibility to discriminate one of them.
We have presented the best level of second harmonic generation ever obtained in a bulk chalcogenide glass. A thermal poling realized under optimized conditions allows to reach a second-order susceptibility χ(2) of (8.0±0.5)pm/V. Increasing the time duration of the poling increases the nonlinear thickness and consequently decreases the SH signal peak. In the same way for higher temperature, a larger nonlinear thickness is obtained associated with an increase of the charge carriers mobility.
On an other hand, SIMS preliminary experiments display ions motions inside the sample, like the Na element occuring towards the cathod where an accumulation is measured: additive measurements are necessary to conclude on the efficiency of these charge motions on the NL thickness.
References and links
01. M. Asobe, K. Suzuki, T. Kanamori, and K. Kubodera, “Nonlinear refractive index measurement in chalcogenide-glass fibers by self-phase modulation,” Appl. Phys. Lett 60, 1153–1154 (1992). [CrossRef]
02. H. Kobayashi, H. Kanbara, M. Koga, and K. Kubodera, “Third-order nonlinear optical properties of As2S3 chalcogenide glass,” J. Appl. Phys. 74, 3683–3687 (1993). [CrossRef]
03. F. Smektala, C. Quemard, L. Leneindre, J. Lucas, A. Barthélémy, and C. De Angelis, “Chalcogenide glasses with large non-linear refractive indices,” J. Non-Cryst. Solids 239, 139–142 (1998). [CrossRef]
04. T. Cardinal, K. A. Richardson, H. Shim, A. Schulte, R. Beatty, K. Le Foulgoc, C. Meneghini, J. F. Viens, and A. Villeneuve, “Non-linear optical properties of chalcogenide glasses in the system As-S-Se,” J. Non-Cryst. Solids 256 & 257, 353–360 (1999). [CrossRef]
06. N. Mukherjee, R. A. Myers, and S. R. J. Brueck, “Dynamics of second-harmonic generation in fused silica,” J. Opt. Soc. Am. B 11, 665–669 (1994). [CrossRef]
07. M. Guignard, V. Nazabal, J. Troles, F. Smektala, H. Zeghlache, Y. Quiquempois, A. Kudlinski, and G. Martinelli, “Second-harmonic generation of thermally poled chalcogenide glass,” Opt. Express 13, 789–795 (2005), [CrossRef] [PubMed]
08. P. D. Maker, R. W. Terhune, M. Nisenoff, and C. M. Savage, “Effects of dispersion and focusing on the production of optical harmonics,” Phys. Rev. Lett. 8, 21–23 (1962). [CrossRef]
09. A. Kudlinski, Y. Quiquempois, M. Lelek, H. Zeghlache, and G. Martinelli, “Complete characterization of the nonlinear spatial distribution induced in poled silica glass with a submicron resolution,” Appl. Phys. Lett. 83, 3623–3625 (2003). [CrossRef]
10. T. G. Alley, S. R. J. Brueck, and M. Wiedenbeck, “Secondary ion mass spectrometry study of space-charge formation in thermally poled fused silica,” J. App. Phys. 86, 6634–6640 (1999). [CrossRef]
11. A. Kudlinski, G. Martinelli, and Y. Quiquempois, “Time evolution of second-order nonlinear profiles induced within thermally poled silica samples,” Opt. Lett. 30, 1039–1041 (2005). [CrossRef] [PubMed]