Open Access
Add to BibSonomyAbstract
Quest for photo-stable amorphous thin films in ternary GexAsySe100-x-y chalcogenide system is reported. Studied layers were fabricated using pulsed laser deposition technique. Scanning electron microscope with energy dispersive X-ray analyzer, Raman scattering spectroscopy, transmittance measurements, variable angle spectroscopic ellipsometry, and non-linear imaging technique with phase object inside the 4f imaging system were employed to characterize prepared thin films. Their photo-stability/photo-induced phenomena in as-deposited and relaxed states were also investigated, respectively. In linear regime, we found intrinsically photo-stable relaxed layers within Ge20As20Se60 composition. This composition presents also the highest optical damage threshold under non-linear optical conditions.
©2010 Optical Society of America
1. Introduction
Chalcogenide glasses and amorphous thin films, based on S, Se or Te elements in combination with suitable element(s) from IV. or V. group of the periodical system (typically Ge, As, etc.), are distinguished from other materials by their photosensitivity to light exposure, which presents several types of photo-induced changes in the structure and properties. Photo-induced processes are typically linked with the changes of optical parameters, mainly the band gap energy (photo-darkening or photo-bleaching) and the refractive index (photo-refraction) in linear optical regime [1,2]. Surprisingly, photo-induced phenomena are not observed in crystalline materials nor in amorphous semiconductors from IV or V group of the periodical system. They are characteristic for amorphous chalcogens and chalcogenides thanks to their structural flexibility and electronic lone-pair p states forming the top part of their valence band [2,3]. Basically, two kinds of photo-induced effects are observed: (i) irreversible effects occurring in as-deposited thin films, whereas (ii) reversible ones proceed in well-annealed (below but near by the glass transition temperature) layers and bulk glasses. The reversibility of the process means the ability of the annealed films to restore their initial properties modified by irradiation by a re-annealing [4].
On one hand, photo-induced phenomena in amorphous chalcogenides are of great interest for applications covering optical writing, photolithography, light trimming bandpass filters, etc [2,4–6]. On the other hand, for many applications in the field of infrared optics, based on linear or non-linear optical properties, amorphous chalcogenides insensitive to light exposure are required [7,8]. This is why some potential applications of amorphous chalcogenides have been restricted due to instability of their optical parameters during light exposure.
Studies on photosensitivity of amorphous chalcogenide thin film are numerous while the reports on their photosensitivity involving two photon absorption process are infrequent [7]. By an exposure at high intensity laser peak power (>1 GW/cm2), limited by ablation threshold, this specific photosensitivity occurs even for low repetition rate and with negligible linear and low non-linear absorption coefficient at used wavelength. As far as our knowledge, an investigation which studies in parallel both photo-induced effects induced in linear and non-linear optical regime is not available in the literature.
It was already shown that the photo-induced phenomena in amorphous chalcogenides might be depressed by doping with metallic elements such as Cu, Sn or Pb [9–11]. However, attempts to optimize intrinsic chemical composition of amorphous chalcogenides in order to prevent undesired photo-induced effects are rare [12]. (Ir)reversible photo-darkening is observed typically for arsenic-based sulfide/selenide glasses and their thin films [4,13]. In contrast, irreversible photo-bleaching was found in case of germanium-based amorphous chalcogenides [14,15]. It is noticeable that reversible photo-bleaching is reported rarely [4]. One can suppose that in ternary Ge-As-S(Se) materials, photo-darkening and photo-bleaching might be compensated by an appropriate choice of composition leading to photo-stability effect [12,16,17]. Such hypothesis could be interesting for the concept of intrinsically photo-stable amorphous chalcogenides. The attention of this work is thus focused on Ge-As-Se thin films optical properties and their stability/changes induced by annealing and light exposure.
As illustrated by the Fig. 1 , the glass-forming region in Ge-As-Se system is pretty large enabling relatively easy preparation of bulk glasses in wide range of chemical compositions [18]; as a result, some compositions – Ge33As12Se55 (AMTIR-1 or IG2), Ge22As20Se58 (GASIR1) or Ge10As40Se50 (IG4) – are actually commercially available.
Fig. 1 Ge-As-Se ternary diagram: glass-forming region and compositions studied in this work (circle - Ge10As30Se60, square - Ge10As35Se55, triangles (from top) - Ge10As40Se50, Ge15As30Se55, and Ge20As20Se60). MCN corresponds to the mean coordination numbers calculated from nominal composition.
Many properties of Ge-As-Se bulk glasses were already reported by several groups. The structure of Ge-As-Se glass family was studied via Raman scattering spectroscopy and x-ray photoelectron spectroscopy [19,20]. References [19,21] are focused on linear optical properties; special attention is given to non-linear optical properties [22,23]. It is worthy to note that the non-linear refractive index n2 of Ge-As-Se glasses is at least 200 times higher than that of silica glass (at 1.54 µm) with non-linear absorption β around 0.5 cm/GW [22]. The non -linear figure of merit n2/βλ is certainly promising in this system. The success of ultra-fast all optical applications based on these glasses depends on the optical losses [21] and photosensitivity control [7].
Vacuum thermal evaporation [12,24,25], radio-frequency magnetron sputtering [26] or pulsed laser deposition (PLD) [24,27,28] were shown to be useful for the preparation of amorphous Ge-As-Se thin films; each of deposition technique possessing its own advantages and disadvantages. PLD technique seems to be favorable according to its simplicity, easy control of the process, often stoichiometric transfer of target material to the films, and possibility to fabricate multilayered structures or films of unusual compositions [29]. Up to now, reported Ge-As-Se films showed both, photo-darkening or photo-bleaching, depending on fabrication method and chemical composition, respectively. To date, the lowest Ge-As-Se waveguide propagation losses reported reach value of ~0.3 dB/cm at 1.55 µm [28].
On the basis of the facts mentioned above, the aim of this work is to find photo-stable thin films in as-deposited but preferably in annealed state focusing on Ge-As-Se system. The term photo-stability can be defined as insensitivity of the material to light exposure in terms of constant values of optical parameters, both the refractive index and optical band gap. For this work, PLD was selected as an appropriate thin films growth technique. Different methods of characterization such as scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), Raman scattering spectroscopy, transmittance measurements, variable angle spectroscopic ellipsometry (VASE), and non-linear imaging technique with phase object (NIT-PO) inside the 4f imaging system were employed in this study to have complete picture of fabricated layers.
2. Experimental details
The targets used for PLD were bulk samples of chalcogenide glasses from GexAsySe100-x-y system with nominal compositions Ge10As30Se60, Ge10As35Se55, Ge10As40Se50, (i.e. x = 10, y = 30, 35, 40), Ge15As30Se55 (i.e. x = 15, y = 30), and Ge20As20Se60 (i.e. x = 20, y = 20). For reader’s convenience, individual chemical compositions are shown in ternary Ge-As-Se diagram (Fig. 1).
Targets (25 mm diameter) were prepared by weighting high purity elements (5–6 N) in a dry glove box. Then, the mixture was placed in a fused silica ampoule and pumped down for a few hours. At this stage, the tube was sealed and then heated in a rocking furnace at 850 °C for 12 h. After the quenching, each glass rod was annealed at the temperature 10 °C below its glass transition temperature for 5 h.
The targets were ablated using a KrF excimer laser, operating at the wavelength of 248 nm, with constant output energy of 300 ± 3 mJ per pulse, and with pulse duration of 30 ns. The laser energy fluency was set at ~2.6 J.cm2. Laser pulses were directed on the target at the repetition rate of 20 Hz for 480-900 seconds with a background pressure in the vacuum chamber ~2-3x10−4 Pa. In order to obtain as uniform as possible irradiation conditions, off-axis PLD technique with rotating target and substrates was used. The ablated material was collected on chemically cleaned microscope glass, SiO2 glass or silicon substrates at room temperature. The substrates were positioned parallel to the target surface at target to substrate distance of 5 cm.
Study of irreversible photo-induced phenomena in linear regime was carried out with as-deposited amorphous thin films. On the other hand, reversible ones were performed with well annealed layers. The annealing was realized in inert atmosphere (pure argon) below but near their glass transition temperature (20 °C below the respective glass transition temperature of the corresponding target glass; the values of glass transition temperatures are listed in Table 1 .) for 120 min; the samples were consequently slowly cooled down to room temperature at 1 °C.min−1.
Table 1. Nominal/real chemical composition ( ± 0.5 at. %), and thicknesses of GexAsySe100-x-y bulk glasses and PLD thin films. MCN and MCN* stand for the mean coordination numbers calculated from nominal composition and EDS results, respectively. The glass transition temperatures (Tg, ± 2 °C) shown here are for bulk glasses. Note that thicknesses of the thin films are determined from VASE data analysis ( ± 1 nm)
Either irreversible or reversible linear photo-induced effects in GexAsySe1-x-y layers were studied via exposure of the thin films by a laser diode operating at 660 nm (1.88 eV) with intensity of ~160 mW.cm−2 for exposure time of 120 min being satisfactorily long for the saturation of the photo-induced phenomena, if any. To avoid oxidation of the films during the exposure experiments, the layers under study were irradiated under inert atmosphere (nitrogen). Layers grown on glass substrates were used to eliminate possible intensity variations of exposing light inside the films.
A scanning electron microscope with an energy-dispersive x-ray analyzer (JSM 6400-OXFORD Link INCA) was used for the determination of GexAsySe1-x-y bulk glasses and thin films chemical composition. SEM technique was also applied to observe the morphology of the thin films using a field-emission gun SEM (JSM 6301F).
Raman scattering spectra were measured at room-temperature by HR800 micro-raman spectrometer (Horiba Scientific/Jobin-Yvon) with 785 nm laser diode as an excitation source. To avoid sample heating and photo-induced phenomena during measurements, optical density filters have been used to reduce effectively laser power focused on the sample.
Optical transmittance spectra both for glasses and films were recorded with VIS-NIR spectrophotometer (PerkinElmer) in the range of 450-2500 nm. In order to extract the values of optical band gap energy of the GexAsySe1-x-y thin films, Tauc’s approach was employed [30].
Optical functions and thicknesses of PLD GexAsySe1-x-y layers were obtained from the analysis of spectroscopic ellipsometry data measured using an ellipsometer with automatic rotating analyzer (VASE, J. A. Woollam Co., Inc.). The measurement parameters are as follows: spectral region 300-2300 nm (i.e. 4.13-0.54 eV) with wavelengths steps of 10 nm (20 nm in case of target glasses), angles of incidence 50°, 60°, and 70°.
In order to obtain correct near-gap and above-gap optical responses, the imaginary part of the dielectric function, ε2(E) = 2n(E)k(E), is calculated as the product of variable band edge function and Lorentz oscillator function. Two variable band edge functions – Tauc-Lorentz and Cody-Lorentz – were shown to be suitable for the description of optical responses in amorphous semiconductors. However, Cody-Lorentz (CL) model, which includes both the correct band edge function and weak Urbach absorption tail [Eq. (1)], was already shown to be more appropriate for the description of amorphous chalcogenide thin films optical functions and their photo-induced changes [13]. Following CL model, imaginary part of dielectric function is described as
where G(E) and L(E) stand for variable band edge function and Lorentz oscillator function, respectively. In CL model, G(E) function is given by empirical expression: G(E)∝[(E-Egopt)2/(E-Egopt)2 + Ep2] derived on the assumption of parabolic bands and a constant dipole matrix element. In Eq. (1), Et represents the demarcation energy between the Urbach tail transitions and the band-to-band transitions. For energies lower than Et (0<E≤ Et), Eq. (1) leads to the Urbach formula for absorption coefficient, α(E)∝exp(E/Eu); Eu is then corresponding Urbach energy. Further, in Eq. (1) E1 is defined to hold the ε2 continuity at E = Et: E1 = EtL(Et)G(Et). For E>Et, ε2CL is given by the product of G(E) and L(E) functions. Here Egopt represents the optical band gap energy. The meaning of Ep is transition energy (given by the Ep + Egopt sum) separating the absorption onset (for E<(Ep + Egopt)) from the Lorentz oscillator behavior (for E>(Ep + Egopt)). The set of three parameters describing the Lorentz oscillator, namely A, E0, and Γ, stand for Lorentz oscillator amplitude (oscillator strength), resonance (peak transition) energy, and oscillator width (broadening), respectively.The real part of dielectric function ε1(E) = n2(E)-k2(E) is obtained as a correct Kramers-Kronig transformation of imaginary counterpart of dielectric function ε2(E).
After the determination of ε1(E) and ε2(E) by fitting experimental data (evaluated from spectroscopic ellipsometry measurements) with a set of parameters (A, E0, Γ, Egopt, Et, Eu, and Ep), refractive index n(E) and extinction coefficient k(E) dispersion can be calculated as
We note that in all the cases, thicknesses of the thin films, thickness uniformity, and thicknesses of the surface layer (assuming to be composed of thin film material and voids) were fitting parameters as well. The regression analysis was performed by the mean square error (MSE) values minimization procedure; MSE is given via following expression:
where N is the number of measured pairs of ellipsometric parameters Ψ and Δ; M represents the total number of fitted parameters. and are then estimated experimental errors of Ψ and Δ, respectively.To investigate non-linear photo-induced effects in as-deposited chalcogenide thin films, the NIT-PO approach inside the 4f imaging system was used [31]. It was already shown that this technique allows to study the time evolution of non-linear photo-induced effects [32].
The excitation is provided by a Nd:YAG laser delivering 17 ps single pulses at λ = 1.064 μm with 10 Hz repetition rate. The input intensity is fixed by means of a half-wave plate and a Glan prism maintaining a linear polarization. A beam splitter at the entry of the setup permits to monitor fluctuations occurring in the incident laser beam. The focal lengths of lenses f = 20 cm. The beam waist of the incident Gaussian laser beam at the entry of the system ω0 = 1.8 mm. The phase object is placed in front focal plane of the first lens. This object is composed of circular glass plate on which a transparent dielectric disk (of radius Lp = 0.5 mm) has been deposited. The disk has a thickness and index of refraction such that it retards the phase of the incident light by π/2 radians relative to the phase outside the disk. The sample placed in the back focal plane of first lens is illuminated by the incident laser beam during 250 seconds ( = 2500 laser shots). The photoreceptor is a 1000 × 1018 pixels cooled (−30 °C) CCD camera with fixed linear gain. The camera pixels have 4095 gray levels and each pixel is 12 × 12 μm2. In the presence of non-linear photo-induced effects (modifications inside the material), the filtering process of spatial frequencies composing the spectrum of the object contribute to energy deflection or absorption inside the image of the phase plate acquired by the CCD camera. Thus the image may change with time and consequently the phase contrast variations may be observed in the acquired images. To investigate non-linear photo-induced effects quantitatively, we defined the diffraction efficiency η as the signal to measure: , where εphl and εphnl are the diffracted energies inside the phase plate in the low and high intensity regimes, respectively. Finally, εtotal denotes the total energy detected in the image plane in the high intensity regime corresponding to the first image (t = 0 s) when there is no photo-induced effect inside the films. For all the samples, we have checked that after the break position the contrast of the phase object is permanent. The absolute value of the diffraction efficiency remains relatively high memorizing the photo-induced effects even in the low intensity regime.
3. Results and discussion
As-deposited GexAsySe1-x-y thin films were homogeneous according to SEM investigations (Fig. 2 ); SEM also showed a good planarity with smooth surface of the deposited films, without cracks or corrugations, and only very limited number of sub-micrometer sized droplets. Table 1 contains chemical composition data of bulk materials and PLD films measured by EDS; all the compositions both for bulks and films are within ~3 at. % variation of the nominal composition based on atomic quantities of the elements used to prepare the glasses. Taking into account error limits of the EDS technique ( ± 0.5 at. %), we can conclude that the material transfer during PLD is almost stoichiometric and the composition of the thin films is in good accordance with used bulk glasses (Table 1).
The VASE experimental data were analyzed using three layer model of optical functions: (i) the substrate (glass slides), (ii) amorphous chalcogenide thin film, and (iii) the surface layer. The optical functions of thin films in whole measured spectral region including near- and above-band gap energies were calculated using CL model. The surface layer was defined by effective medium approximation (thin film material mixed with 50% voids).
The thicknesses of the as-deposited films varied from ~610 to ~1000 nm (Table 1), as determined by VASE data analysis, being close to the penetration depth of the light used for the exposure experiments.
A typical Tauc plot of an as-deposited film, i.e. (αhν)1/2 vs. hν dependence where α and ν stand for the absorption coefficient and the frequency of the light, respectively, is presented in Fig. 3 . For illustration, a transmission spectrum is also shown as inset in Fig. 3.
Fig. 3 As-deposited Ge10As35Se55 thin film (αhν)1/2 spectral dependence (line shows determination of Egopt value via extrapolation). Inset shows corresponding transmission spectrum.
The Egopt values obtained from Tauc plots are summarized in Table 2 . Optical band gap values calculated by CL model from VASE data are included in Table 2 for a comparison. We note that the agreement between both methods of Egopt calculation is satisfactory.
Table 2. Ge-As-Se thin films optical parameters in different stages (as-deposited, exposed, annealed, and exposed after annealing): band gap values (estimated by VASE data analysis and from Tauc plots), refractive indices at 1.54 µm (determined via VASE). For a comparison, bulk data for refractive indices are also shown
Finally, linear refractive indices values at 1.5 µm determined via CL model are given in Table 2 for both the thin films and corresponding glass targets. Examples of refractive indices spectral dependences are illustrated within Fig. 4 . The applicability of CL model for the analysis of VASE data is confirmed by low values of MSE, typically MSE<5.
Fig. 4 Refractive index ( ± 0.01) spectral dependencies of Ge10As30Se60 (a), Ge10As35Se55 (b), and Ge20As20Se60 (c) PLD thin films determined by VASE data analysis.
From the presented results one can conclude that the exposure of as-deposited GexAsySe100-x-y layers with 1.88 eV laser diode (i.e. linear regime) leads to an increase of Egopt values: Egopt (exposed) - Egopt (as-deposited) ~0.04-0.14 eV, thus clear photo-bleaching effect is observed. The exposure of relaxed (annealed) thin films results in photo-darkening effect, which is documented by a decrease of Egopt values: Egopt (exposed) - Egopt (annealed) ~-(0.01-0.09) eV. From the photo-stability point of view, the best results were found in the case of Ge20As20Se60 layers having mean coordination number (MCN) 2.6 (defined as the sum of the respective elemental concentrations times their covalent coordination numbers; these are 4, 3, and 2 for germanium, arsenic, and selenium, respectively) which are, in fact, insensitive to irradiation in relaxed state. This is in pretty good agreement with insensitivity of bulk glass of identical composition, reported by Calvez, for which a tunable laser was used [17]. Additionally, the layers with identical mean coordination number (2.6), Ge10As40Se50 and Ge15As30Se55 compositions, show very different behavior involving irreversible photo-bleaching and reversible photo-darkening what remains after all quite logical since the ratio As/Ge is not constant for the two films (Table 2). Refractive indices dispersion curves (Fig. 4) and table data extracted for telecommunication wavelength of 1.5 µm (Table 2) support the idea of photo-stable composition finding in Ge-As-Se ternary. Relaxed Ge20As20Se60 layers were evaluated as almost insensitive against exposure for their refractive index in transparent region of the spectrum. In terms of photo-stability of refractive index, we found similar result also for Ge10As30Se60 composition having mean coordination number 2.5.
In order to confirm photo-stability phenomenon both in terms of band-gap but also refractive index observed for Ge20As20Se60 layers, the experiments were successfully reproduced with samples of approximately half thickness (~500 nm). Other compositions present reversible photo-refraction (photo-induced increase of refractive index values for annealed layers around 0.02 at 1.5 µm), which is, in terms of Moss rule [33], qualitatively in agreement with observed photo-darkening. We note that in as-deposited state, all the films exhibit photo-induced refractive index decrease (Fig. 4, Table 2) ranging from 0.02 to 0.06 at 1.5 µm. The refractive indices of relaxed thin films were compared with data from corresponding bulk targets; slightly lower values for annealed layers were found but without clear correlation in respect with chemical composition of the samples under study.
Because as-deposited films show photo-induced changes of both optical band-gap and refractive index values, any of these PLD films can’t be classified as photo-stable. This fact is motivating for future work focusing on other compositions and evaluating their photo-stability. In an earlier attempt to find photo-stable composition in Ge-As-Se ternary system [12], Ge10As35Se55 thin films were identified as photo-stable in as-prepared state. The discrepancy could originate in different deposition technique (PLD vs. thermal evaporation). Other source of difference could be the real chemical composition of evaporated thin films, which can be dramatically changed from target when using thermal evaporation with high deposition rates (60 nm/min) [12]. Indeed, the optimum growth rate was found to be 3-4 nm/min and for higher deposition rates (38 ± 10 nm/min), a decrease of germanium content (−4.5 at.% when compared with bulk composition) and an increase of arsenic content ( + 6.3 at.%) were reported [25].
As mentioned above, photo-stability phenomenon is studied with constant energy (1.88 eV) of irradiating light in this work. Generally, the character as well as the amplitude of photo-induced changes of properties could be energy dependent. Due to this fact, our future work will be focused on the study of photo-stability vs. irradiation wavelength relations in Ge-As-Se thin films employing tunable laser.
Raman scattering spectra of parent GexAsySe1-x-y glasses are shown in Fig. 5 . The dominant feature of all the spectra is a broad band ranging from ~170 to 320 cm−1; this band is composed of several contributions, the most important are peaking at ~193-198, ~215, ~224-229, and ~238 cm−1. Raman spectra of relaxed layers are given in Fig. 5 as well; they show similar features as the bulk glasses do. The changes of thin film Raman spectra (documented in Fig. 6 ) due to the annealing and/or exposure can only be classified as marginal, for all the compositions excluding Ge20As20Se60 layers; this particular composition presents an increase of Raman band amplitude at ~193-198 cm−1 due to the annealing, which is connected with a decrease of Raman signal intensity in the range of ~170-185 cm−1.
Fig. 6 Raman scattering spectra of Ge10As35Se55 (a) and Ge20As20Se60 (b) PLD films (as-deposited, exposed, annealed and exposed after annealing).
The Raman band peaking at ~193-198 cm−1 can be attributed to the A1(ν1) symmetric stretching mode of GeSe4 corner-sharing tetrahedra. The band peaking at 215 cm−1, usually called “companion” mode, is due to the vibrations of GeSe4 tetrahedra sharing their edges [34,35]. In glasses containing lower content of germanium (10 at. %), this band is overlapped by stronger signal coming from other structural motifs as discussed below. Other broad Raman band with maximum at ~224-229 cm−1 is connected with the stretching vibration modes of AsSe3 pyramids [36]. Band (shoulder) at 238 cm−1 can be assigned to A1(ν 2) modes of As4Se3 cage-like molecules present in the structure. Other Raman-active strong vibrations of As4Se3 cages have frequencies at 196 cm−1 – E(ν 8) and 256 cm−1 – E(ν 7) [37]. First one (196 cm−1) is probably contributing to the amplitude of Raman band already ascribed to GeSe4 tetrahedra; second one (256 cm−1) is hidden due to overlap with AsSe3 pyramids vibrations. We cannot exclude the presence of other cage-like molecules – As4Se4 – having maxima of their most important Raman-active bands at 248 and 190 cm−1 [38]. We note that a tribute to the overall intensity of Raman signal can also originate in Ge-Ge bonds vibration modes (Ge-GemSe4-m, m = 1,2,3,4 structural units), already observed in amorphous germanium showing a wide band centered at ~270 cm−1 with a shoulder at ~170 cm−1 [39]. Finally, a shoulder in Raman spectra at ~300 cm−1 is related to F2 asymmetric vibration modes of GeSe4 tetrahedra [40]. We note that deeper analysis of Raman spectra showed by bulk Ge-As-Se glasses under study is impossible due to the proximity of the frequencies of individual vibrations and their strong overlapping.
The shape of Raman spectra of relaxed Ge-As-Se thin films is following the Raman spectra of corresponding bulk glasses (Fig. 5). However, it should be pointed out that the widths of individual bands are generally larger for the thin films when compared with bulk glasses, which is an indication of a larger degree of disorder (distortion of bond angles and bond lengths) in the films. Above-mentioned changes of Raman spectra of the Ge20As20Se60 films due to the annealing (Fig. 6) can be attributed to the structural changes between Ge-GemSe4-m (m = 1,2,3,4) or Ge2Se6 (ethan-like) structural units with Ge-Ge bonds and corner-sharing GeSe4 tetrahedra. Based on presented Raman scattering spectra, we point out that exposure of Ge-As-Se layers has no influence on the structure of the material.
The Table 3 presents I0T values, the non-linear photo-induced effects threshold, and the εpulse, the corresponding laser pulse energy measured by a joulemeter, for all the compositions under study.
Table 3. Selected characteristics of as-deposited Ge-As-Se layers: optical band gap values (Egopt) from Tauc plots, refractive indices at 1064 nm and thicknesses (determined via VASE), non-linear photo-induced threshold values I0T ( ± 20%), and corresponding laser pulse energies εpulse ( ± 10%)
Examples of the evolution of the signal (η) versus time for intensities equal to and above the threshold value are given in Fig. 7 for Ge10As40Se50 and Ge20As20Se60 PLD films. The η is constant (~0) during the time characterizing undamaged samples. For intensities above the threshold, the breakpoint position for each film can vary from 10 to 110 s depending on the composition but it can be also related to the value of the non-linear photo-induced effects threshold. Higher the power density, faster the kinetic to reach the optical threshold leading to a damage of the thin film (Fig. 7).
Fig. 7 The time evolution of the η signal for the intensities equal to the threshold value (full circles) and above the threshold (open circles): Ge10As40Se50 (right), Ge20As20Se60 (left).
The threshold values could be expected to follow the band-gap variation as a direct effect on the two-photon absorption which should increase with the decreasing of the band-gap energy. However, the process seems to be more complex and it is therefore difficult from only the linear or non-linear absorptions to know the photosensitivity effects observed for energy becoming critical or predict an optical damage threshold. One should take into account that probably other parameters come into play which can be related to the optical quality of the film, the hyperpolarizability of the elements constituting the layers, etc. However, we stress that the Ge20As20Se60 composition with the largest value of the band-gap and the best photo-stability after annealing in linear regime presents also the highest optical damage threshold in non-linear regime.
4. Conclusions
An attempt to find intrinsically photo-stable amorphous chalcogenide thin films in Ge-As-Se system is presented in this paper. The chemical composition of the pulsed laser deposited Ge-As-Se layers is in good accordance with used glassy targets; good planarity with smooth surface of the deposited films, without cracks or corrugations was found by morphology study. The structure of the films, as revealed by Raman scattering spectroscopy, is similar to structure of parent glasses and is formed mainly by GeSe4 corner- and edge-sharing tetrahedra, AsSe3 pyramids, and As4Se3(4) cage-like molecules; Ge-Ge bonds (Ge-GemSe4-m, m = 1,2,3,4 structural units) can be also presented. From compositions under study, the best photo-stability is shown by well-relaxed Ge20As20Se60 thin films: these do not change optical band gap value neither refractive index due to exposure in linear regime. Moreover, in non-linear regime, this particular composition shows the highest optical damage threshold (~35 GW/cm2) at corresponding laser pulse energy ~13 µJ.
It is worth mentioning a difficulty to envisage that exactly a single composition is photo-stable in studied system. In future work, we will continue to explore in detail Ge-As-Se ternary system in order to find other possible photo-stable compositions, stable not only in relaxed state but also in as-deposited one. It is also necessary to realize photo-stability vs. irradiation wavelength measurements to shed light on wavelength dependent photo-stable behavior of studied materials.
To summarize, an assumption of suitability of Ge-As-Se ternary system for the search of photo-stable material is confirmed as concluded from the results of several optical characterization techniques covering linear- and non-linear regime. In both regimes, Ge20As20Se60 relaxed layers were identified as the most promising composition among the studied compositions.
Acknowledgements
This work was supported by the Ministry of Education, Youth, and Sports of the Czech Republic (Project No. MSM 0021627501), the Czech Science Foundation (Project No. 104/08/0229), and via MRCT of CNRS. Thanks to ONIS platform (University of Rennes 1) for the use of the HR800 micro-raman spectrometer.
References and links
1. K. Shimakawa, A. Kolobov, and S. R. Elliott, “Photoinduced effects and metastability in amorphous semiconductors and insulators,” Adv. Phys. 44(6), 475–588 (1995). [CrossRef]
2. A. Zakery and S. R. Elliott, “Optical properties and applications of chalcogenide glasses: a review,” J. Non-Cryst. Solids 330(1-3), 1–12 (2003). [CrossRef]
3. A. Ganjoo, K. Shimakawa, K. Kitano, and E. A. Davis, “Transient photodarkening in amorphous chalcogenides,” J. Non-Cryst. Solids 299–302, 917–923 (2002). [CrossRef]
4. M. Frumar, B. Frumarova, T. Wagner, and P. Nemec, in Photo-Induced Metastability in Amorphous Semiconductors, A. V. Kolobov, ed. (Wiley-WCH, Weinheim, Germany, 2003), pp. 23–44.
5. C. Meneghini and A. Villeneuve, “As2S3 photosensitivity by two-photon absorption: holographic gratings and self-written channel waveguides,” J. Opt. Soc. Am. B 15(12), 2946–2950 (1998). [CrossRef]
6. W. D. Shen, M. Cathelinaud, M. D. Lequime, F. Charpentier, and V. Nazabal, “Light trimming of a narrow bandpass filter based on a photosensitive chalcogenide spacer,” Opt. Express 16(1), 373–383 (2008). [CrossRef] [PubMed]
7. M. Chauvet, G. Fanjoux, K. P. Huy, V. Nazabal, F. Charpentier, T. Billeton, G. Boudebs, M. Cathelinaud, and S. P. Gorza, “Kerr spatial solitons in chalcogenide waveguides,” Opt. Lett. 34(12), 1804–1806 (2009). [CrossRef] [PubMed]
8. M. Guignard, V. Nazabal, F. Smektala, H. Zeghlache, A. Kudlinski, Y. Quiquempois, and G. Martinelli, “High second-order nonlinear susceptibility induced in chalcogenide glasses by thermal poling,” Opt. Express 14(4), 1524–1532 (2006). [CrossRef] [PubMed]
9. J. Z. Liu and P. C. Taylor, “Absence of photodarkening in bulk, glassy As2S3 and As2Se3 alloyed with copper,” Phys. Rev. Lett. 59(17), 1938–1941 (1987). [CrossRef] [PubMed]
10. M. S. Iovu, S. D. Shutov, P. Boolchand, D. G. Georgiev, and E. P. Colomeico, “The relaxation of photodarkening in Sn doped amorphous As2Se3 films,” J. Optoelectron. Adv. Mater. 5, 389–395 (2003).
11. A. Masuda, Y. Yonezawa, A. Morimoto, M. Kumeda, and T. Shimizu, “Influence of Pb incorporation on light-induced phenomena in amorphous Ge100-x-yPbxSy thin films,” J. Non-Cryst. Solids 217(2-3), 121–135 (1997). [CrossRef]
12. G. Yang, H. S. Jain, A. T. Ganjoo, D. H. Zhao, Y. S. Xu, H. D. Zeng, and G. R. Chen, “A photo-stable chalcogenide glass,” Opt. Express 16(14), 10565–10571 (2008). [CrossRef] [PubMed]
13. P. Němec, V. Nazabal, and M. Frumar, “Photoinduced phenomena in amorphous As4Se3 pulsed laser deposited thin films studied by spectroscopic ellipsometry,” J. Appl. Phys. 106(2), 023509 (2009). [CrossRef]
14. E. Sleeckx, L. Tichy, P. Nagels, and R. Callaerts, “Thermally and photo-induced irreversible changes in the optical properties of amorphous GexSe100-x films,” J. Non-Cryst. Solids 198–200, 723–727 (1996). [CrossRef]
15. E. Vateva, “Giant photo- and thermo-induced effects in chalcogenides,” J. Optoelectron. Adv. Mater. 9, 3108–3114 (2007).
16. L. Calvez, Z. Yang, and P. Lucas, “Reversible giant photocontraction in chalcogenide glass,” Opt. Express 17(21), 18581–18589 (2009). [CrossRef]
17. L. Calvez, Z. Y. Yang, and P. Lucas, “Light-induced matrix softening of Ge-As-Se network glasses,” Phys. Rev. Lett. 101(17), 177402 (2008). [CrossRef] [PubMed]
18. Z. U. Borisova, Glassy Semiconductors (Plenum, New York, 1981).
19. C. J. Zha, R. P. Wang, A. Smith, A. Prasad, R. A. Jarvis, and B. Luther-Davies, “Optical properties and structural correlations of GeAsSe chalcogenide glasses,” J. Mater. Sci. Mater. Electron. 18(S1), S389–S392 (2007). [CrossRef]
20. R. P. Wang, A. V. Rode, D. Y. Choi, and B. Luther-Davies, “Investigation of the structure of GexAsySe1-x-y glasses by x-ray photoelectron spectroscopy,” J. Appl. Phys. 103(8), 083537 (2008). [CrossRef]
21. A. Prasad, C. J. Zha, R. P. Wang, A. Smith, S. Madden, and B. Luther-Davies, “Properties of GexAsySe1-x-y glasses for all-optical signal processing,” Opt. Express 16(4), 2804–2815 (2008). [CrossRef] [PubMed]
22. J. T. Gopinath, M. Soljacic, E. P. Ippen, V. N. Fuflyigin, W. A. King, and M. Shurgalin, “Third order nonlinearities in Ge-As-Se-based glasses for telecommunications applications,” J. Appl. Phys. 96(11), 6931–6933 (2004). [CrossRef]
23. C. Quémard, F. Smektala, V. Couderc, A. Barthelemy, and J. Lucas, “Chalcogenide glasses with high non linear optical properties for telecommunications,” J. Phys. Chem. Solids 62(8), 1435–1440 (2001). [CrossRef]
24. R. A. Jarvis, R. P. Wang, A. V. Rode, C. Zha, and B. Luther-Davies, “Thin film deposition of Ge33As12Se55 by pulsed laser deposition and thermal evaporation: Comparison of properties,” J. Non-Cryst. Solids 353(8-10), 947–949 (2007). [CrossRef]
25. D. A. P. Bulla, R. P. Wang, A. Prasad, A. V. Rode, S. J. Madden, and B. Luther-Davies, “On the properties and stability of thermally evaporated Ge-As-Se thin films,” Appl. Phys., A Mater. Sci. Process. 96(3), 615–625 (2009). [CrossRef]
26. D. A. Turnbull, J. S. Sanghera, V. Nguyen, and I. D. Aggarwal, “Fabrication of waveguides in sputtered films of GeAsSe glass via photodarkening with above bandgap light,” Mater. Lett. 58(1-2), 51–54 (2004). [CrossRef]
27. A. Zakery and M. Hatami, “Nonlinear optical properties of pulsed-laser-deposited GeAsSe films and simulation of a nonlinear directional coupler switch,” J. Opt. Soc. Am. B 22(3), 591–597 (2005). [CrossRef]
28. D. Y. Choi, S. Madden, A. Rode, R. Wang, and B. Luther-Davies, “Fabrication of low loss Ge33As12Se55 (AMTIR-1) planar waveguides,” Appl. Phys. Lett. 91(1), 011115 (2007). [CrossRef]
29. M. Frumar, B. Frumarova, P. Nemec, T. Wagner, J. Jedelsky, and M. Hrdlicka, “Thin chalcogenide films prepared by pulsed laser deposition - new amorphous materials applicable in optoelectronics and chemical sensors,” J. Non-Cryst. Solids 352(6-7), 544–561 (2006). [CrossRef]
30. J. Tauc, “Absorption edge and internal electric fields in amorphous semiconductors,” Mater. Res. Bull. 5(8), 721–729 (1970). [CrossRef]
31. G. Boudebs and S. Cherukulappurath, “Nonlinear optical measurements using a 4f coherent imaging system with phase objects,” Phys. Rev. A 69(5), 053813–053818 (2004). [CrossRef]
32. G. Boudebs and C. B. de Araujo, “Characterization of light-induced modification of the nonlinear refractive index using a one-laser-shot nonlinear imaging technique,” Appl. Phys. Lett. 85(17), 3740–3742 (2004). [CrossRef]
33. T. S. Moss, “A relationship between the refractive index and the infra-red threshold of sensitivity for photoconductors,” Proc. Phys. Soc. London, Sect. B 63(3), 167–176 (1950). [CrossRef]
34. S. Sugai, “Stochastic random network model in Ge and Si chalcogenide glasses,” Phys. Rev. B Condens. Matter 35(3), 1345–1361 (1987). [CrossRef] [PubMed]
35. K. Murase, T. Fukunaga, K. Yakushiji, T. Yoshimi, and I. Yunoki, “Investigation of stability of (Ge,Sn)-(S, or Se)4/2 cluster by vibrational-spectra,” J. Non-Cryst. Solids 59–6, 883–886 (1983). [CrossRef]
36. T. Mori, S. Onari, and T. Arai, “Raman-scattering in amorphous As-Se system,” Jpn. J. Appl. Phys. 19(6), 1027–1031 (1980). [CrossRef]
37. M. Ystenes, W. Brockner, and F. Menzel, “Scaled quantum-mechanical (Sqm) calculations and vibrational analyses of the cage-like molecules P4S3, As4Se3, P4Se3, As4S3, and PAs3S3,” Vib. Spectrosc. 5(2), 195–204 (1993). [CrossRef]
38. M. Ystenes, F. Menzel, and W. Brockner, “Ab-initio quantum-mechanical calculations of energy, geometry, vibrational frequencies and IR intensities of tetraphosphorus tetrasulfide, alpha-P4S4(D2d), and vibrational analysis of As4S4 and As4Se4,” Spectrochim. Acta A Mol. Biomol. Spectrosc. 50(2), 225–231 (1994). [CrossRef]
39. M. Wihl, M. Cardona, and J. Tauc, “Raman scattering in amorphous Ge and III-V compounds,” J. Non-Cryst. Solids 8–10, 172–178 (1972). [CrossRef]
40. L. Petit, N. Carlie, R. Villeneuve, J. Massera, M. Couzi, A. Humeau, G. Boudebs, and K. Richardson, “Effect of the substitution of S for Se on the structure and non-linear optical properties of the glasses in the system Ge0.18Ga0.05Sb0.07S0.70-xSex,” J. Non-Cryst. Solids 352(50-51), 5413–5420 (2006). [CrossRef]
