Open Access
Add to BibSonomy
Abstract
We demonstrate, for the first time to our knowledge, single-scan ultrafast laser inscription and performance of mid-infrared waveguiding in IG2 chalcogenide glass in the type-I and type-II configurations. The waveguiding properties at 4550 nm are studied as a function of pulse energy, repetition rate, and additionally separation between the two inscribed tracks for type-II waveguides. Propagation losses of ∼1.2 dB/cm in a type-II waveguide and ∼2.1 dB/cm in a type-I waveguide have been demonstrated. For the latter type, there is an inverse relation between the refractive index contrast and the deposited surface energy density. Notably, type-I and type-II waveguiding have been observed at 4550 nm within and between the tracks of two-track structures. In addition, although type-II waveguiding has been observed in the near infrared (1064 nm) and mid infrared (4550 nm) in two-track structures, type-I waveguiding within each track has only been observed in the mid infrared.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Beyond the limitation of silicon-based photonics, the chalcogenide glasses are attractive for applications in the mid-infrared (mid-IR) regions because of their transparency arising from phonon suppression by heavy elements, tunability of the optical and thermal properties via changes in composition, high photosensitivity, and high ultrafast Kerr nonlinearity, making them desirable for infrared photonics and all-optical signal processing [1–3]. Because of signature absorption peaks in molecular spectra, infrared photonics allows for the spectroscopy of chemical and organic molecules. For instance, mid-IR medical imaging has been applied for pathological diagnosis [4], and mid-IR sensing has been enabled by integrated photonics based on chalcogenide waveguides [5,6]. Recently, a programmable chalcogenide-based all-optical deep neural network has been demonstrated [7]. Another interesting application is in astronomy [8,9], for example, photonic devices for observing exo-planets which emit strongly in the mid-IR spectral region in contrast to their host stars.
One of the challenges for these applications is that chalcogenide glasses are typically fragile and brittle because of the weak covalent bonding between elements. Therefore, common manufacturing approaches such as lithography, chemical etching, and fiber drawing are limited for these materials. Additionally, these approaches cannot easily be applied to fabricate three-dimensional structures, e.g., the photonic lantern (PL) [10] targeted for improving the light collection efficiency in astrophotonics. Ultrafast-laser inscription (ULI) is a flexible and powerful tool for the micro-engineering of transparent dielectrics and is capable of producing a large range of photonic structures [11,12]. It relies on the tight focusing of an ultrashort optical pulse, typically with sub-picosecond duration, to generate peak power densities sufficient for permanent local modification of the material’s properties via enhanced nonlinear absorption of sub-bandgap photons.
IG2 (Ge33As12Se55) is a commercially available photo-stable infrared glass [13]. It transmits approximately from 1 to 12 µm [14], a range that covers the two atmospheric transmission windows, making it suitable for remote sensing and astrophotonics. The reported studies of ULI in Se-based glasses focusing on As2Se3, IG2, or Ge22As20Se58 correspond to the inscription of diffraction grating [15,16], type-I and type-II waveguides [17–20], and waveguides for directional couplers and supercontinuum generation [18,19]. (The diffraction grating inscribed in IG2 operates in mid- infrared wavelengths 2–5 µm, and the grating in the As2Se3 glass operates in the near infrared wavelength 808 nm [16]). Type-I waveguides rely on a positive change in refractive index in the modified region induced by one or several passes of the focused ultrafast laser beam. Typical type-I waveguides in chalcogenide glasses require multiple scans or spatial shaping of the focal spot for single-scan fabrication [21]. The basic single-scan method has been reported mostly in the Gallium Lanthanum Sulphide (GLS) glass operating at 1550, 2485 and 3850 nm, with the lowest propagation loss of 1.47 dB/cm at 1550 nm, however the propagation losses were not reported at longer wavelengths [21–24]. For Se-based glasses, the multi-scan type-I waveguides operating at 7.8 µm or 2.94 µm usually have propagation losses of ∼1 - 1.75 dB/cm [17–19]. The maximal refractive index change is on the order of 10−2 [16–19,25]. To date, there is no study of single-scan inscription of type-I waveguides in Se-based glass like IG2. In addition, there has been no investigation of single-scan type-I guiding optimized at wavelengths longer than 2 µm for any chalcogenide glass we know of. On the other hand, type-II waveguides rely on the stress-induced increase in refractive index between two inscribed tracks. They can usually be effectively inscribed with a broader range of ULI parameters. They are generally more thermally stable [26], and can preserve the intrinsic material’s properties as the central guiding region is unchanged. We have recently reported ULI type-II waveguides in IG2 and described their waveguiding properties in the near infrared [20]. However, waveguiding in the mid-IR, for which such glass is advantageous, has not been studied.
In this article, we demonstrate, for the first time to our knowledge, the single-scan ultrafast laser fabrication and performance of mid-IR waveguides in IG2 for both type-I and type-II configurations, exploring a range of writing parameters (energy, repetition rate, deposited energy density, and track separation). In Section 2, we describe the experimental setups used for waveguide inscription and characterization in the mid-IR. In Section 3, we present the dimensional analysis of the ultrafast-laser-inscribed tracks for waveguides inside IG2, as a function of the writing parameters. In Section 4, the mid-IR waveguiding properties are presented for both type-II and type-I waveguides, demonstrating consistent single-mode operation with propagation loss optimized at 4550 nm, as well as emergence of multimode operation.
2. Experimental setups
2.1 Waveguide writing
The waveguide-writing setup is based on a femtosecond laser (Monaco, Coherent) operating at 1030 nm, with pulse duration τp of 350 fs, and repetition rate fr adjustable between 10 and 500 kHz. A 0.55-NA 50× plan-apochromat objective (Mitutoyo) focuses the beam to a ∼1.6 µm spot size located approximately 200 µm below the substrate’s surface. A pulse energy Ep between 10 and 120 nJ is sufficient to inscribe single or pairs of tracks suitable for waveguiding at 4550 nm as well as 1064 nm. A quarter-wave plate at 1030 nm is used to induce a circular polarization state [20]. The sample is mounted on a computer-controlled air-bearing stage allowing translation in three spatial dimensions. For all the data presented in this article, the sample translation speed relative to the static focused laser beam is 10 mm/s.
The IG2 glass substrates [10 mm × 5 mm × 2.5 mm, with edges defining an (x, y, z) coordinate system] are polished on 4 faces: top/ bottom faces (y, x plane: 5 mm × 10 mm) for waveguide writing through the top surface, and input/output faces (y, z plane: 5 mm × 2.5 mm) for coupling in and out of the inscribed waveguides. Both the single tracks forming the type-I waveguides and the two tracks defining the type-II waveguides were inscribed parallel to the longest dimension, i.e., in the x direction, at a constant depth z. The waveguides were characterized without any post-inscription polishing of the input or output surface.
2.2 Waveguide characterization
The inscribed tracks are directly imaged using a Differential Interference Contrast microscope in a transmission configuration. Planes corresponding to constant longitudinal position x or a constant depth z can be imaged, allowing either to characterize a cross-section, or a top view of the inscribed tracks [20]. The illumination source is a halogen bulb with a spectrum extending into the infrared. Figure 1 illustrates the setup used for characterizing the guiding properties of the waveguides. The mid-IR laser for the waveguide characterization is a continuous-wave (cw) quantum cascade laser (QCL) operating at 4550 nm. The light polarization is parallel to the tracks, i.e. along the z direction. Throughout this investigation, we did not observe a noticeable dependence of output modes on polarization. The camera used for beam profiling (PY-IV-C-A PRO, Spiricon) features a 25-mm imager (320 × 320 pixels) and a 16-bit A/D converter. The aspheric CaF2 focusing lens has a focal length of 20 mm, and the collection lens is a ZnSe objective which has a focal length of 18 mm and a numerical aperture (NA) of 0.08. The mode field diameter (MFD) is calculated by fitting the mode measured by the beam profiler (after averaging over 20 acquisitions) with a two-dimensional Gaussian profile after background subtraction. The insertion loss is measured by the power meter. An output iris having a diameter of 3 mm is placed 500 mm away from the back end of the ZnSe objective to block the scattered light propagating towards the power meter and the beam profiler. For additional information on type-I guiding in the near IR, as has been done for type-II guiding, characterization at 1064 nm was also performed using the hardware described in [20].
The waveguide propagation loss (PL) is calculated from the measured insertion loss, the ratio of output power of the guided mode and input power of the focused beam, taking into account the coupling loss (∼ 1 - 3 dB), the inherent transmission loss, and Fresnel reflection at the input and output faces (1.8 dB). The coupling loss is determined by calculating the overlap integral of the focused input beam (waist ∼37 µm) and the mode profile. The transmission loss due to IG2 absorption at 4550 nm over 10 mm is 0.1 dB.
The wavelength-dependent Fabry-Perot resonances, i.e., the multiple internal reflections between the two coupling faces having high-index contrast across the interface, lead to uncertainty of the propagation-loss characterization [27]. The uncertainty at a given wavelength can be calculated as a function of the single-pass loss, including the 0.1-dB transmission loss. The calculated uncertainties are of the order of 1.2, 0.8, 0.4, and 0.2 dB for losses of 1.5, 3, 6, and 10 dB, respectively. We have quantified the experimental uncertainty, for a type-II and a type-I waveguide, by heating the substrate to induce a gradual change in the propagation phase and determining the resulting maximal and minimal transmission. The resulting uncertainties were found to be equal to 0.4 and 0.3 dB for propagation losses of 1.2 and 2.5 dB, respectively. These experimental uncertainties are much smaller than the calculated values and give an estimate of the lower limit of the error bars on the propagation losses reported in this article.
The refractive index change Δn of a given waveguide is estimated according to the type of waveguides. For the type-II waveguides, the step-index approximation of the Δn is estimated from the profile of the fundamental guided mode [28–31]. Following the same practice, we use the equation Δn = NA2/(2n), where n is the material’s optical index, and NA is the numerical aperture whose lower limit is given by NA = 2λ/(πMFD). For the type-I waveguides, we determine Δn based on comparison with the explicit numerical simulation considering the measured track dimensions and mode field diameters. Although both these methods are subject to large uncertainty per se, they provide general indications about the magnitude of refractive index change and their evolution with inscription parameters.
3. Ultrafast-laser-inscribed tracks characterization
3.1 Tracks cross-section
Figures 2 and 3 present microscope images of the typical type-II and type-I waveguides suitable for operation at 4550 nm. In Fig. 2, type-II waveguides defined by double tracks were inscribed inside the IG2 substrate with pulse energy from 50 to 102 nJ, and separation between two tracks from 50 to 70 µm. For these tracks, the repetition rate is equal to 500 kHz. The distance between the boundary tracks of neighboring waveguides is 120 µm. In Fig. 3, single tracks as type-I waveguides were inscribed with energy from 5 to 74 nJ, at a repetition rate equal to 10, 100, 200 and 500 kHz. The distance between the centers of neighboring tracks is 80 µm, which is large enough to avoid crosstalk between neighboring waveguides. Considering the focal spot size and spatial pulse separation induced by substrate translation, the average deposited optical energy density (Ed) ranges from 0.31 J/cm2 (5 nJ, 10 kHz) to 375 J/cm2 (120 nJ, 500 kHz). The cross-sections of the Type-II waveguide demonstrate the impact of pulse energy on the size of the tracks. The pulse energy is critical for track formation: if the energy is too low, no tracks are observed on the microscope images. At the repetition rate of 500 kHz, the threshold for track formation with a single scan is approximately 9 nJ, which increases to 12.5 nJ at the repetition rate of 10 kHz, for a reduced cumulative effect of heat. At higher energies, we have previously reported the observation of self-organized periodic structures and a transition from irregular or grain-like structures to grating-like structures [20].
Fig. 2. Cross-section images of the type-II waveguides defined by two tracks inscribed in IG2 with different pulse energies Ep = 50 - 102 nJ (Energy density, Ed = 156.3 - 318.8 J/cm2), with track separations of 70, 60, and 50 µm from left to right, respectively. The repetition rate is 500 kHz. The focusing depth in the material is 200 µm. The distance between the boundary tracks of neighboring waveguides is 120 µm.
Fig. 3. Cross-section images of single-track type-I waveguides inscribed in IG2 with Ep = 5 - 74 nJ, at the repetition rate fr = 10, 100, 200, and 500 kHz (Ed = 0.63 - 231.3 J/cm2). The distance between single tracks is 80 µm. The dashed line indicates the position of geometric focus, and the purple arrow on the rightmost image indicates the position of the tail of the tracks.
3.2 Dependence on energy density and repetition rate
The characteristics of the tracks are quantified as a function of energy density in Fig. 4 (not all energies used here were shown in Fig. 3). As the energy density increases, the track length and width increase approximately linearly at 200 and 500 kHz [Figs. 4(a) and 4(b)], indicating a linear relationship between the size of the region where the IG2 glass is being modified and the surface density of deposited laser pulse energy. At the scanning speed of 10 mm/s, the 100-kHz repetition rate is a critical point between thermal and athermal regimes [20], and the track length exhibits a saturation behavior with increased energy density. This phenomenon can also be observed at 10 kHz for both the tracks length and width. For all the tested conditions, higher energy density generally induces an upward (toward the input surface and laser source) displacement of the tracks, and in particular, of the top of the tracks (Figs. 2 and 3). The bottom of the tracks stays near the geometrical focus (within ∼5 µm), as indicated by the dashed line, where the purple arrow indicates the thin tail of the long track (which can be discerned if zoomed in but not obvious in the image). In the athermal regime with repetition rate at or below 100 kHz, track formation closer to the surface at higher energy is consistent with self-focusing occurring during the propagation of the focused pulse, as the critical power for self-focusing 3 kW has already been exceeded at the modification threshold Ep = 9 nJ [20]. In this regime, the origin of saturation behavior might be related to the nonlinear two-photon absorption process. The material's modification is dominated by pulse energy. Conversely, in the thermal regime at repetition rates of 200 and 500 kHz where the heat has less time to dissipate, both track lengths and widths increase more linearly with energy density, and in a similar fashion. In this regime, the effect of heat accumulation is more significant [32,33], the material's modification is dominated by energy density, and the movement of the track top toward the surface is approximately linear on the log-log plot (also on the linear scale).
Fig. 4. (a) Length and (b) width of the tracks inscribed in IG2, as a function of deposited energy density, for repetition rate fr equal to 500 kHz (solid, blue), 200 kHz (dotted, red), 100 kHz (dashed, black), and 10 kHz (dash-dotted, green). The two dashed frames correspond to energy density ranges with lowest PLs for type-I and type-II waveguiding: 10–18 J/cm2 and 180–300 J/cm2, respectively.
We have observed consistent waveguiding at 4550 nm either within the region located between the two tracks (type-II) or within the tracks themselves (type-I) (Section 4). This indicates that the optical index within those regions is larger than in the surrounding or contiguous areas. No type-I waveguiding has been observed at 1064 nm within the tracks. Possible physical mechanisms that explain the type-II waveguiding include a rarefaction in average density within the tracks, because of thermal expansion or formation of voids, and a stress-induced increase in the optical index of the region in between two tracks [33]. The possible mechanisms that explain type-I guiding include a chemical modification with rearrangement of the covalent bonds or bond angles, or ion migration and disorder in structure, which produces an increase of polarizability and/or density of the glass network [34–38].
Specifically, as shown in Fig. 4, for the best waveguiding, we indicated the two ranges of energy density with the lowest PLs: 10–18 J/cm2 for type-I waveguiding and 180–300 J/cm2 for type-II waveguiding. Incidentally, the PL for type-I guiding also drops significantly in the energy-density range corresponding to minimal PI for type-II guiding. More details will be discussed later.
4. Waveguiding properties
4.1 Type-II waveguiding
For all the type-II waveguides inscription, we have chosen the repetition rate of 500 kHz. No or weak guiding is observed for tracks inscribed at or below 200 kHz. Similarly to what has been observed at 1064 nm [20], better mid-IR propagation is observed for tracks inscribed at higher repetition rates, and with larger separations between two tracks. Figure 5 shows two typical examples of the guided modes measured at 4550 nm, with the corresponding images of the type-II waveguide tracks in the same imaging plane. Singlemode guiding is presented in Fig. 5(a) for the waveguide in Fig. 5(b). For this waveguide, the pulse energy is Ep = 80 nJ, the track separation is 50 µm, and the dimension of single tracks is 84 µm × 8.5 µm. The PL is estimated to be 1.3 dB/cm (while the insertion loss is 4.4 dB), the MFD averaged over two directions (41 µm and 59 µm) is 49 µm, and the lower limit of refractive index change (Δn) is estimated to be 0.7 × 10−3. For the second waveguide shown in Fig. 5(d), a higher order mode TEM01 is obtained by properly tuning the injection condition, as presented in Fig. 5(c). For this waveguide, the pulse energy is Ep = 120 nJ, the track separation is 70 µm, and the dimension of single tracks is 139 µm × 15.5 µm. This specific waveguide can support more than one mode, and excitation of the higher-order mode is possible. The size of the top and bottom lobe is 40 µm × 46 µm and 45 µm × 44 µm, respectively. For both modes, we can observe weak scattering in the background. This is partly due to crosstalk with other proximal (adjacent) guiding channels formed intrinsically around or inside the tracks.
Fig. 5. Guided modes at 4550 nm for type-II waveguides inscribed with (a) Ep = 80 nJ (Ed = 250 J/cm2), and track separation of 50 µm, and (c) Ep = 120 nJ (Ed = 375 J/cm2), and track separation of 70 µm. The corresponding cross-section images of the double tracks in the same imaging plane are shown in (b) and (d).
Figure 6 shows more examples of the guided modes at 4550 nm for different pulse energies (Ep = 29 to 120 nJ) and track separations (50 to 80 µm). The typical images of the double-track waveguides have been shown for some combinations of parameters in Fig. 2 (the cases for Ep = 29 and 35 nJ, and track separation of 80 µm were not shown). The thresholds of energy and track separation for mid-IR guiding are found experimentally to be approximately 29 nJ and 30 µm, respectively. With pulse energy equal to or greater than 29 nJ, and pulse separation equal to or larger than 40 µm, guided modes are consistently observed between the tracks. This is consistent with a relative increase in refractive index in this region. The guided modes are relatively irregular and sometimes elongated along the track direction, in particular when the track separation is small. With the pulse energy between 29 and 50 nJ, the guiding is generally poor, with more scattering from tracks or coupling to the proximal (and adjacent) guiding channels.
Fig. 6. Guided modes at 4550 nm (normalized to 1) for type-II waveguides inscribed with Ep = 29 - 120 nJ (Ed = 91 - 375 J/cm2), track separation of 50 - 80 µm, at the repetition rate fr = 500 kHz. The modes with a green circle correspond to the lowest propagation loss (within experimental uncertainty).
Figure 7 shows the propagation loss and mode field diameter for the type-II waveguides inscribed with pulse energy from 29 to 120 nJ (energy density from 91 to 375 J/cm2), with track separation from 40 to 80 µm. Propagation losses for the track separation of 40 µm is 0.5 - 3 dB/cm higher than for the other cases because the narrow guiding region leads to more scattering of the guided mode. Waveguides with track separation from 50 to 80 µm have similar propagation loss as a function of pulse energy. Significantly higher losses with PL = 2–5 dB/cm (IL = 6–10 dB) have been observed with lower pulse energy between 29 to 50 nJ. In this energy range, the guiding is comparatively poorer, with more scattering or coupling to the proximal channels. This is due to inadequate confinement and the inherently complex refractive index profiles formed (to be elaborated later). With larger energy, the loss decreases. Low propagation losses approaching 1 dB/cm within the experimental uncertainty have been measured in several type-II waveguides [Fig. 5(a), and those marked with green circles in Fig. 6], for tracks separated by 50 to 80 µm inscribed with energy between 60 to 90 nJ. The lowest propagation loss of ∼1.2 dB/cm (insertion loss of 4.4 dB) has been achieved at pulse energy of 80 nJ for a track separation of 60 µm. At even higher energy, the loss rises again. This increase in loss could be due to more defects caused by enhanced damage, melting and/or vaporization, leading to more scattering. The higher-order mode is not observable until 120 nJ with a track separation of 70 or 80 µm [see Fig. 5(d)].
Fig. 7. (a) Propagation loss, and (b) MFD of type-II waveguides inscribed in IG2, as a function of pulse energy and energy density, for track separation from 40 to 80 µm. The repetition rate is 500 kHz.
Figure 7(b) shows that the MFD generally increases with the track separation, and it is within the range of 39 to 64 µm. For track separations between 50 and 80 µm, the mode size does not dependent on the pulse energy. When the track separation is 40 µm, with Ep = 60 to 120 nJ, the MFD slightly decreases from 50 µm at 60 nJ to a minimal value of 39 µm at 120 nJ. The corresponding PL is 5.3 dB/cm (IL = 8.8 dB), along with the highest index change with lower limit Δn = 1.1 × 10−3. This suggests a trade-off between loss and confinement of the mode.
4.2 Type-I waveguiding
Type-I waveguiding in the mid-IR has been consistently observed along the directly modified region inside single tracks, besides the described type-II guiding in between the two tracks (Section 4.1). This is unlike the situation at 1064 nm for which no type-I guiding has been observed in the tracks that define the type-II waveguides or in single isolated tracks. The propagation loss and the mode field diameter have been determined for the type-I waveguides. Figures 8(a) and 8(c) show two typical examples of the guided type-I modes measured at 4550 nm, with the corresponding images of the single tracks in the same imaging plane displayed in Fig. 8(b) and 8(d). Singlemode guiding is presented in Fig. 8(a) for the waveguide displayed in Fig. 8(b). For this waveguide, the pulse energy is Ep = 20 nJ, the repetition rate is fr = 100 kHz, and the dimension of single tracks is 19 µm × 2 µm. The PL is estimated to be 2.5 dB/cm (while the insertion loss is 4.7 dB), the MFD is 44 µm, and the refractive index change is estimated to be 0.015. For the leftmost track shown in Fig. 8(d), a higher-order mode TEM01 has been observed by properly tuning the injection condition, as presented in Fig. 8(c). For this waveguide, the pulse energy is Ep = 90 nJ, the repetition rate is fr = 500 kHz, and the dimension of single tracks is 120 µm × 12 µm. The size of the top and bottom lobe is 30 µm × 29 µm and 35 µm × 31 µm, respectively. Δn is estimated to be of the order of 0.002 from the corresponding singlemode profile. For the high-order mode, weak scattering can be observed in the background. This is partly due to crosstalk with other intrinsic proximal guiding channels around or inside the track, similar to the cases of type-II guiding in section 4.1.
Fig. 8. Guided modes at 4550 nm for type-I waveguides inscribed with (a) Ep = 20 nJ, fr = 100 kHz (Ed = 12.5 J/cm2), and (c) Ep = 90 nJ, fr = 500 kHz (Ed = 281.3 J/cm2). The cross-section images (b) and (d) of single tracks are in the same imaging plane of (a) and (c). The track separation in (d) is 60 µm.
Figure 9 shows more examples of the type-I guided modes at 4550 nm, with different pulse energies Ep from 10 to 74 nJ, at various repetition rates fr = 10, 100, 200 and 500 kHz. Type-I guided modes are consistently observed inside most of the tracks shown in Fig. 3, with the color scale being normalized to the peak of the measured profile. The guided modes are less confined at low energies for the rep. rates of 10, 100 kHz, and 200 kHz. No guiding or poor guiding with scattered light in the background is evident for pulse energy up to 20 nJ at 10 kHz, and up to 15 nJ at 100 kHz, nearing the modification thresholds of materials. For the 200 kHz case, although well-confined single mode guiding is obtained at 15 nJ, the mode is less confined and larger than those obtained at higher energies. Single mode guiding with consistent mode size has been achieved for all the energies from 10 to 120 nJ at 500 kHz (Ep ≥ 80 nJ not shown).
Fig. 9. Guided modes (normalized to 1) at 4550 nm for type-I waveguides inscribed with Ep = 10 - 74 nJ, and repetition rate fr = 10 - 500 kHz (Ed = 0.9 - 231.3 J/cm2). The mode identified by a green circle has the lowest propagation loss.
Figure 10 shows the plots of measured propagation loss, measured mode field diameter, and estimated refractive index change Δn versus the energy density (more cases included here with Ep up to 120 nJ). The performance trends are separated into the athermal and thermal regimes characterized by the critical frequency 100 kHz. At repetition rates of 10, 100 kHz, and 200 kHz, PL generally increases with the energy density after reaching the minimal values at which the modes are confined (Fig. 9), as shown in Fig. 10(a). The minimal PLs are 3.3, 2.1 and 2.9 dB/cm around 1.9 J/cm2 (Ep = 30 nJ), 15.6 J/cm2 (Ep = 25 nJ), and 18.8 J/cm2 (Ep = 15 nJ), at repetition rates of 10, 100 and 200 kHz, respectively. The corresponding MFDs are 45 and 50 µm for the 100 and 200 kHz, respectively. For the repetition rate of 500 kHz, a minimal PL of 2.9 dB/cm at 250 J/cm2 (Ep = 80 nJ) was achieved. The corresponding MFD increases slightly from 35 to 40 µm when the energy density increases from 90 to 250 J/cm2. The PL first increases with the energy density between 31 to 62.5 J/cm2 (Ep = 10 to 20 nJ), reaching a maximum of 9.3 dB/cm (IL = 12.3 dB) around 47 to 62.5 J/cm2 (Ep = 15 to 20 nJ), then decreases all the way to the local minimum. After that, there is a slight increase in loss.
Fig. 10. (a) Measured propagation loss, (b) measured MFD, and (c) estimated Δn of type-I waveguides inscribed in IG2, as a function of energy density, at repetition rate of 10, 100, 200, and 500 kHz.
Figure 10(b) shows that the measured MFD generally decreases with increasing pulse energy at repetition rate from 10 to 200 kHz, with noticeable drop at the lower energy density end nearing the modification thresholds. Above the thresholds, the propagation behavior evolves from no guiding, weak confinement, and tight confinement with increasing energy density, as shown in Fig. 9. The MFD is less dependent on energy density for the repetition rate of 500 kHz, with tighter confinement. When the energy density reaches 30 J/cm2, although the MFD stays relatively stable between 35 - 40 µm for repetition rates of 100 to 500 kHz, the corresponding PL increases sharply for the 100 kHz and 200 kHz cases. This suggests that using energy density higher than 30 J/cm2 for writing at 100 kHz and 200 kHz is not beneficial.
A complex interplay between the different processes leading to the index change (magnitude and sign) within chalcogenide glasses has previously been documented [21,39]. The experimental data (microscope images of the tracks cross-section and MFD) and modeling are combined to identify the different regimes of index modification. The peak refractive index change is iteratively estimated by comparing the experimentally measured MFDs with the simulated values, using a commercial beam propagation software package (RSOFT BeamPROP). For this simulation, we assume a two-dimensional Gaussian distribution for the refractive modulation induced by the laser writing process. The dimension of the refractive index change is defined based on the spatial profile of the inscribed tracks obtained from the microscope images, assuming that the observed feature size is representative of the region where a refractive-index change has occurred. Figure 10(c) shows that the determined refractive index change decreases with increasing energy density for all repetition rates except for 10 kHz. In addition, at energy densities of 10 J/cm2 and below, the refractive index changes in the athermal regime are larger than most of those in the thermal regime, which indicates that it is pulse energy that drives the refractive index change in the athermal regime, and pulse energy density that drives in the thermal regime. Additional experimental investigation, e.g., direct measurement of the refractive index change, and modeling are required to get a more in-depth understanding of these processes.
4.3 Capability of type-I and type-II waveguiding
Type-I and type-II waveguiding has been observed at 4550 nm in two-track structures inscribed at energies beyond 29 nJ. Figure 11 shows an example of type-I and type-II guiding in two tracks inscribed with an 80-nJ energy and 60-µm separation. Figure 11(a) shows the cross-section image of the two tracks. The mid-IR source was sequentially coupled into each of the two tracks and through the area in between the two tracks by translating the waveguide substrate in the horizontal direction (y direction) while keeping the substrate at a constant height (z direction). The two type-I modes on the left and right side were guided by each of the two tracks [Figs. 11(b) and 11(d)], while the central type-II mode was guided within the area between the two tracks [Fig. 11(c)]. The MFDs are 42.2, 58.7, and 42.9 µm for these three configurations. The ellipticity for type-I and type-II modes is 0.9 and 0.57, respectively. The propagation loss of the type-I modes on the left and the right sides is 4.4 and 3.7 dB/cm, respectively, while the propagation loss of the type-II mode is 1.4 dB/cm, being approximately 2.6 dB/cm lower than that of the type-I mode. During this experiment, the individual modes were excited separately and distinctly, and we did not observe strong coupling of the modes of different types.
Fig. 11. (a) Cross-section image of the two tracks having a separation of 60 µm. (b), (c) and (d) correspond to guiding in the leftmost track (Type I), in-between the two tracks (Type II), and in the rightmost track (Type I), respectively.
Figure 12 shows the comparison of propagation loss and mode field diameter vs pulse energy (and energy density), for the type-I and type-II waveguides inscribed at 500 kHz (data extracted from Figs. 7 and 10). A track separation of 60 µm is selected for this comparison because this corresponds to the lowest PL for type-II waveguides and the PL does not vary significantly for track separations larger than 50 µm (Fig. 7). For the compared configurations, the lowest PL for type-II waveguiding is approximately 2 dB lower than for type-I waveguiding, and type-II has a broader energy range (60-80nJ) within which the best PL is achieved [Fig. 12(a)]. The loss follows a similar trend between 60 and 120 nJ for the two configurations, and the PL is the lowest at 80 nJ for the two configurations. The better guiding between the tracks relative to within the tracks is attributed to lower scattering induced by tracks defects. Figure 12(b) shows that the MFDs of type-II guiding with track separation of 60 µm are larger than those of type-I guiding (this was also observed for a 40-µm track separation). The type-II mode can be well confined only above 40 µm, and MFDs increase with track separation [Fig. 7(b)]. Therefore, the MFD of type-II guiding is in general larger than that of type-I guiding due to lower refractive index contrast and larger ellipticity.
Fig. 12. (a) Propagation loss, and (b) MFD of type-I (blue, solid) and type-II (black, dashed) waveguides inscribed in IG2, as a function of pulse energy (and energy density), at the repetition rate of 500 kHz. The track separation for the type-II waveguides is 60 µm. The vertical dashed lines indicate the lowest inscription energy allowing for type-I and type-II guiding.
4.4 Discussion
The possible mechanisms that explain type-I guiding include a chemical modification with formation of homopolar bonds after breaking of heteropolar bonds [34,35], intermolecular and intramolecular bond breaking [23], rearrangement of the covalent bond angles among As-Se and Ge-Se [36], or ion migration and disorder in structure [37,38], which produce an increase of polarizability and/or density of the glass network. With increasing deposited energy density, extended defects and broken bonds form [40], more local damages result in the glass where it is unable to rearrange the broken bonds in a fixed volume and expansion happens [36], with possible formation of voids. This leads to a rarefaction in average density within the tracks, and a stress-induced increase in the optical index outside the tracks [20]. With additional energy deposited, there is a possible reduction of defects due to uniform and sustained heating and annealing in the materials. Therefore, the propagation loss decreases at higher energy density after 90 - 100 J/cm2 [Figs. 10(a) and 11(a)]. Another possible reason for the improved guiding is that there is a continuous evolution of refractive index profiles with increasing energy density, similar to what happens in a ZBLAN-type fluoride glass [41] (though not quite similar to that in the bulk of As2S3 glass [42]). This may eventually facilitate the confinement and propagation of light. For instance, if the radial index profile of a single track forms a W-like shape [43], with a positive index bump at the track center, light can potentially be guided within each track and in the region between the tracks.
Lastly, throughout our experiments, we have not observed any type-I waveguiding at 1064 nm within the inscribed tracks. One explanation is the wavelength dependence of scattering. Light at 1064 nm would naturally suffer more scattering than light at 4550 nm (assuming that the feature size of defects is smaller than the wavelength in materials), and significant scattering could therefore hinder guided propagation in the near infrared. Another possibility is the inherent wavelength dependence of the index change, including its sign, as is the case for ZnSe, in which the sign of the index change induced by a femtosecond laser is wavelength-dependent [44].
5. Conclusions
In conclusion, single-scan type-II and type-I waveguides for propagation at the mid-IR wavelength of 4550 nm have been inscribed with an ultrafast laser in the chalcogenide glass IG2 for the first time to our knowledge. The evolution of track sizes has been analyzed as a function of the energy density, and the waveguiding properties of the waveguides have been studied for a range of track separations (40–80 µm), irradiation pulse energies (10 - 120 nJ), and repetition rates (10 - 500 kHz), with the 350 fs circularly polarized laser pulses. For type-II, better propagation is observed for tracks inscribed at higher pulse energy and higher repetition rates (i.e. higher energy density), and with larger spacing of 50 to 80 µm. For type-I, better propagation is observed for tracks inscribed either at lower repetition rates and lower pulse energy (lower energy density) or at higher repetition rates and higher pulse energy (higher energy density). A propagation loss of ∼1.2 dB/cm has been measured in a type-II waveguide with tracks separated by 60 µm, inscribed with 80-nJ pulses, at a 500-kHz repetition rate. A propagation loss of 2.1 dB/cm has been measured in a type-I waveguide with tracks inscribed with 25-nJ pulses, at a 100-kHz repetition rate. The emergence of higher-order modes only occurs at high energy density, e.g., a two-lobe mode was observed in a type-II waveguide with a 70-µm track separation written at a pulse energy of 120 nJ and a repetition rate of 500 kHz, and in a type-I waveguide written at a pulse energy of 90 nJ and repetition rate 500 kHz. The capability of two inscribed tracks to independently guide light either within each track (type-I) or within the region between each track (type-II) has been documented.
The fabrication of mid-IR waveguides of different types in a chalcogenide glass will help understand the basic mechanisms of waveguide formation and wavelength dependence. These waveguides will enable the fabrication of integrated photonic circuits in the mid-IR, including beamsplitters and photonic lanterns.
Funding
National Aeronautics and Space Administration (80NSSC20C0027, 80NSSC21C0638, 80NSSC22PA936).
Acknowledgments
Portions of this work were presented at the Advanced Solid-State Lasers Conference in 2022, paper ATh2A.4. The authors thank Dr Anthony Yu (NASA Goddard Space Flight Center) for fruitful technical discussions and the loan of hardware that supported this investigation.
Disclosures
The authors declare that there is no conflict of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. A. B. Seddon, “Chalcogenide glasses: a review of their preparation, properties and applications,” J. Non-Cryst. Solids 184, 44–50 (1995). [CrossRef]
2. M. D. Pelusi, V. G. Ta’eed, L. Fu, E. Magi, M. R. E. Lamont, S. Madden, D. Y. Choi, D. A. P. Bulla, B. Luther-Davies, and B. J. Eggleton, “Applications of Highly-Nonlinear Chalcogenide Glass Devices Tailored for High-Speed All-Optical Signal Processing,” IEEE J. Sel. Top. Quantum Electron. 14(3), 529–539 (2008). [CrossRef]
3. B. J. Eggleton, B. Luther-Davies, and K. A. Richardson, “Chalcogenide photonics,” Nat. Photonics 5(3), 141–148 (2011). [CrossRef]
4. E. Willenbacher, A. Brunner, B. Zelger, S. H. Unterberger, R. Stalder, C. W. Huck, W. Willenbacher, and J. D. Pallua, “Application of mid-infrared microscopic imaging for the diagnosis and classification of human lymphomas,” J. Biophotonics 14(9), e202100079 (2021). [CrossRef]
5. N. Carlie, J. D. Musgraves, B. Zdyrko, I. Luzinov, J. Hu, V. Singh, A. Agarwal, L. C. Kimerling, A. Canciamilla, F. Morichetti, A. Melloni, and K. A. Richardson, “Integrated chalcogenide waveguide resonators for mid-IR sensing: leveraging material properties to meet fabrication challenges,” Opt. Express 18(25), 26728–26743 (2010). [CrossRef]
6. Q. Du, Z. Luo, H. Zhong, Y. Zhang, Y. Huang, T. Du, W. Zhang, T. Gu, and J. Hu, “Chip-scale broadband spectroscopic chemical sensing using an integrated supercontinuum source in a chalcogenide glass waveguide,” Photonics Res. 6(6), 506–510 (2018). [CrossRef]
7. T. Y. Teo, X. Ma, E. Pastor, H. Wang, J. K. George, J. K. W. Yang, S. Wall, M. Miscuglio, R. E. Simpson, and V. J. Sorger, “Programmable chalcogenide-based all-optical deep neural networks,” Nanophotonics 11(17), 4073–4088 (2022). [CrossRef]
8. L. Labadie and O. Wallner, “Mid-infrared guided optics: a perspective for astronomical instruments,” Opt. Express 17(3), 1947–1962 (2009). [CrossRef]
9. A. Arriola, S. Gross, M. Ams, T. Gretzinger, D. Le Coq, R. P. Wang, H. Ebendorff-Heidepriem, J. Sanghera, S. Bayya, L. B. Shaw, M. Ireland, P. Tuthill, and M. J. Withford, “Mid-infrared astrophotonics: study of ultrafast laser induced index change in compatible materials,” Opt. Mater. Express 7(3), 698–711 (2017). [CrossRef]
10. T. A. Birks, I. Gris-Sánchez, S. Yerolatsitis, S. G. Leon-Saval, and R. R. Thomson, “The photonic lantern,” Adv. Opt. Photonics 7(2), 107–167 (2015). [CrossRef]
11. D. Choudhury, J. R. Macdonald, and A. K. Kar, “Ultrafast laser inscription: perspectives on future integrated applications,” Laser Photonics Rev. 8(6), 827–846 (2014). [CrossRef]
12. T. Dezhi, W. Zhuo, X. Beibei, and Q. Jianrong, “Photonic circuits written by femtosecond laser in glass: improved fabrication and recent progress in photonic devices,” Adv. Photonics 3(02), 1–24 (2021). [CrossRef]
13. G. Yang, H. Jain, A. Ganjoo, D. Zhao, Y. Xu, H. Zeng, and G. Chen, “A photo-stable chalcogenide glass,” Opt. Express 16(14), 10565–10571 (2008). [CrossRef]
14. https://www.vitron.de/files/VITRON_IG-2_Datenblatt_Jan_2015.pdf.
15. H. L. Butcher, D. Lee, R. Brownsword, D. G. MacLachlan, R. R. Thomson, and D. Weidmann, “Ultrafast laser-inscribed mid-infrared transmission gratings in IG2: modelling and high-resolution spectral characterization,” Opt. Express 25(26), 33617–33628 (2017). [CrossRef]
16. W. Ma, L. Wang, P. Zhang, W. Zhang, B. Song, and S. Dai, “Femtosecond laser direct writing of diffraction grating and its refractive index change in chalcogenide As2Se3 film,” Opt. Express 27(21), 30090–30101 (2019). [CrossRef]
17. H. L. Butcher, D. G. MacLachlan, D. Lee, R. R. Thomson, and D. Weidmann, “Demonstration and characterization of ultrafast laser-inscribed mid-infrared waveguides in chalcogenide glass IG2,” Opt. Express 26(8), 10930–10943 (2018). [CrossRef]
18. H. L. Butcher, D. G. MacLachlan, D. Lee, R. R. Thomson, and D. Weidmann, “Ultrafast laser-inscribed mid-infrared evanescent field directional couplers in GeAsSe chalcogenide glass,” OSA Continuum 1(1), 221–228 (2018). [CrossRef]
19. J. M. Morris, M. D. Mackenzie, C. R. Petersen, G. Demetriou, A. K. Kar, O. Bang, and H. T. Bookey, “Ge22As20Se58 glass ultrafast laser inscribed waveguides for mid-IR integrated optics,” Opt. Mater. Express 8(4), 1001–1011 (2018). [CrossRef]
20. W. Hu, M. Kilinc, W. Gebremichael, C. Dorrer, and J. Qiao, “Morphology and waveguiding properties of ultrafast-laser-inscribed type-II waveguides in IG2,” Opt. Mater. Express 12(1), 360–373 (2022). [CrossRef]
21. T. Gretzinger, S. Gross, M. Ams, A. Arriola, and M. J. Withford, “Ultrafast laser inscription in chalcogenide glass: thermal versus athermal fabrication,” Opt. Mater. Express 5(12), 2862–2877 (2015). [CrossRef]
22. M. Hughes, W. Yang, and D. Hewak, “Fabrication and characterization of femtosecond laser written waveguides in chalcogenide glass,” Appl. Phys. Lett. 90(13), 131113 (2007). [CrossRef]
23. M. A. Hughes, W. Yang, and D. W. Hewak, “Spectral broadening in femtosecond laser written waveguides in chalcogenide glass,” J. Opt. Soc. Am. B 26(7), 1370–1378 (2009). [CrossRef]
24. J. E. McCarthy, H. T. Bookey, N. D. Psaila, R. R. Thomson, and A. K. Kar, “Mid-infrared spectral broadening in an ultrafast laser inscribed gallium lanthanum sulphide waveguide,” Opt. Express 20(2), 1545–1551 (2012). [CrossRef]
25. D. Lee, D. G. MacLachlan, H. L. Butcher, R. A. Brownsword, D. Weidmann, C. R. Cunningham, H. Schnetler, and R. R. Thomson, “Mid-infrared transmission gratings in chalcogenide glass manufactured using ultrafast laser inscription,” in Proc. SPIE 9912, Advances in Optical and Mechanical Technologies for Telescopes and Instrumentation II, 99122X (2016).
26. S.-E. Wei, Y. Wang, H. Yao, M. Cavillon, B. Poumellec, G.-D. Peng, and M. Lancry, “Thermal Stability of Type II Modifications by IR Femtosecond Laser in Silica-based Glasses,” Sensors 20(3), 762 (2020). [CrossRef]
27. R. Regener and W. Sohler, “Loss in low-finesse Ti:LiNbO3 optical waveguide resonators,” Appl. Phys. B 36(3), 143–147 (1985). [CrossRef]
28. J. Siebenmorgen, T. Calmano, K. Petermann, and G. Huber, “Highly efficient Yb:YAG channel waveguide laser written with a femtosecond-laser,” Opt. Express 18(15), 16035–16041 (2010). [CrossRef]
29. T. Calmano, A.-G. Paschke, S. Müller, C. Kränkel, and G. Huber, “Curved Yb:YAG waveguide lasers, fabricated by femtosecond laser inscription,” Opt. Express 21(21), 25501–25508 (2013). [CrossRef]
30. A. Rodenas and A. K. Kar, “High-contrast step-index waveguides in borate nonlinear laser crystals by 3D laser writing,” Opt. Express 19(18), 17820–17833 (2011). [CrossRef]
31. T. Feng, P. K. Sahoo, F. R. Arteaga-Sierra, C. Dorrer, and J. Qiao, “Pulse-Propagation Modeling and Experiment for Femtosecond-Laser Writing of Waveguide in Nd:YAG,” Crystals 9(8), 434 (2019). [CrossRef]
32. I. Miyamoto, Y. Okamoto, R. Tanabe, Y. Ito, K. Cvecek, and M. Schmidt, “Mechanism of dynamic plasma motion in internal modification of glass by fs-laser pulses at high pulse repetition rate,” Opt. Express 24(22), 25718–25731 (2016). [CrossRef]
33. I. H. W. Nordin, Y. Okamoto, A. Okada, T. Takekuni, and T. Sakagawa, “Effect of focusing condition on molten area characteristics in micro-welding of borosilicate glass by picosecond pulsed laser,” Appl. Phys. A 122(5), 492 (2016). [CrossRef]
34. O. M. Efimov, L. B. Glebov, K. A. Richardson, E. Van Stryland, T. Cardinal, S. H. Park, M. Couzi, and J. L. Brunéel, “Waveguide writing in chalcogenide glasses by a train of femtosecond laser pulses,” Opt. Mater. 17(3), 379–386 (2001). [CrossRef]
35. A. Zoubir, M. Richardson, C. Rivero, A. Schulte, C. Lopez, K. Richardson, N. Hô, and R. Vallée, “Direct femtosecond laser writing of waveguides in As2S3 thin films,” Opt. Lett. 29(7), 748–750 (2004). [CrossRef]
36. M. R. Vázquez, B. Sotillo, S. Rampini, V. Bharadwaj, B. Gholipour, P. Fernández, R. Ramponi, C. Soci, and S. M. Eaton, “Femtosecond laser inscription of nonlinear photonic circuits in Gallium Lanthanum Sulphide glass,” J. Phys. Photonics 1(1), 015006 (2018). [CrossRef]
37. T. Gretzinger, T. T. Fernandez, S. Gross, A. Arriola, and M. J. Withford, “Boson band mapping: revealing ultrafast laser induced structural modifications in chalcogenide glass,” Opt. Lett. 45(13), 3369–3372 (2020). [CrossRef]
38. P. K. Sharma and S. C. Katyal, “Effect of Ge Addition on the Optical Band Gap and Refractive Index of Thermally Evaporated As2Se3 Thin Films,” Research Letters in Materials Science 2008, 1–4 (2008). [CrossRef]
39. C. D’Amico, G. Cheng, C. Mauclair, J. Troles, L. Calvez, V. Nazabal, C. Caillaud, G. Martin, B. Arezki, E. LeCoarer, P. Kern, and R. Stoian, “Large-mode-area infrared guiding in ultrafast laser written waveguides in Sulfur-based chalcogenide glasses,” Opt. Express 22(11), 13091–13101 (2014). [CrossRef]
40. B. McMillen, B. Zhang, K. P. Chen, A. Benayas, and D. Jaque, “Ultrafast laser fabrication of low-loss waveguides in chalcogenide glass with 0.65 dB/cm loss,” Opt. Lett. 37(9), 1418–1420 (2012). [CrossRef]
41. J.-P. Bérubé, M. Bernier, and R. Vallée, “Femtosecond laser-induced refractive index modifications in fluoride glass,” Opt. Mater. Express 3(5), 598–611 (2013). [CrossRef]
42. P. Masselin, D. Le Coq, and E. Bychkov, “Refractive index variations induced by femtosecond laser direct writing in the bulk of As2S3 glass at high repetition rate,” Opt. Mater. 33(6), 872–876 (2011). [CrossRef]
43. O. Caulier, D. L. Coq, E. Bychkov, and P. Masselin, “Direct laser writing of buried waveguide in As2S3 glass using a helical sample translation,” Opt. Lett. 38(20), 4212–4215 (2013). [CrossRef]
44. J. Lapointe, J.-P. Bérubé, Y. Ledemi, A. Dupont, V. Fortin, Y. Messaddeq, and R. Vallée, “Nonlinear increase, invisibility, and sign inversion of a localized fs-laser-induced refractive index change in crystals and glasses,” Light: Sci. Appl. 9(1), 64 (2020). [CrossRef]
