Broadband laser spectroscopy has been proved to be a powerful analytical tool [1]. A broadband laser spectrometer is important for detecting the absorption lines of organic molecules in the 2–20 µm mid-IR range. Among other things, it can be used for breath analysis [2], gas analysis [3], ecological monitoring, and early diagnosis of diseases [4]. As it is beneficial to cover the entire mid-IR range with a single laser source, nonlinear parametric generation is used in broadband laser spectroscopy. Nonlinear crystals used for parametric generation need to provide phase matching of the conversion and transparency over the entire mid-IR range as well as have a high damage threshold and high nonlinear conversion coefficient. GaSe crystals demonstrate an optimal combination of parameters for efficient conversion in the mid- and far-IR ranges as well as in the THz range [5,6]. However, their high refractive index (average $n = 2.63$ in the 2–16 µm range) leads to significant reflection losses.

The conventional method for increasing the surface transmittance involves the use of single-layer or multi-layer antireflection coatings (ARCs). However, ARCs cannot be reliably applied to GaSe surfaces because of the adhesion problems. Owing to the highly uneven surface, applied coatings tend to delaminate and fail to provide antireflection properties. An alternative and comparatively novel approach is to fabricate antireflection microstructures (ARMs) on the crystal surface [7]. ARM is a system of micro-sized features spaced on the sample surface. The principle of ARM is described by the effective medium theory [8]. For wavelengths larger than $\lambda = n \cdot p$ (where $n$ is the refractive index and $p$ is the period of the microstructure), ARM acts as a layer with a refractive index gradient, which leads to reduced reflection [9]. Several methods have been developed for ARM fabrication [10,11]. Herein, we demonstrate the possibility of applying ARMs to a GaSe surface using the single-pulse direct femtosecond laser ablation method. The technique requires fewer steps in comparison to other techniques [7], and it provides good-quality ARMs and high throughput [12]. This method has already been tested on ZnSe crystals [13], which can be used as a gain medium for mid-IR CW tunable solid-state lasers after being doped with ${{\rm Cr}^{{2} +}}$ or ${{\rm Fe}^{{2} +}}$ ions [14]. For the first time, we demonstrate a single-pulse direct laser ablation method for ARM fabrication on highly uneven surfaces of GaSe crystals in which second-harmonic generation for 10.6 µm ${{\rm CO}_{2}}$ laser radiation could be obtained.

GaSe crystals (space group P-62m) have an optimal combination of parameters for effective conversion of laser radiation in the mid- and far-IR ranges: a wide range of transparency (0.62–20 µm) and phase synchronism of the conversion, high damage threshold (0.03 GW/cm² for 10.6 µm ${125}\;{\rm ns}\;{{\rm pulses}}$), and high nonlinear conversion coefficient (54 pm/V for 10.6 µm) [5]. However, these crystals are also characterized by high refractive index values over a wide range of wavelengths (average $n = 2.63$ in the 2–16 µm range) [5], which leads to significant reflection losses (above 20% on the surface). To increase the efficiency of GaSe-based nonlinear converters, it is necessary to use the methods that increase surface transmittance. The special features of this crystal are its layered structure, high plasticity, and a perfect cleavage only along the (001) plane, which makes it difficult to obtain thin optical plates at the phase matching angle.

To date, the acquirement of GaSe single crystals with high optical quality remains a difficult task [15]. Research in this area is currently focused on the optimal method for growing and using crystals [16–18]. To reduce strong layering and plasticity, some authors have proposed methods for doping these crystals. However, none of these methods can solve the problem of optical element manufacturing at the phase matching angle [19,20].

In practice, the cleavage method along a plane perpendicular to the $z$ axis is used to obtain optical elements. However, this method leads to an uneven surface of the chips, and applying conventional ARCs to the rough surface of a thin plate is usually impossible. Alternatively, instead of applying ARCs, ARMs can be fabricated on the optical element surface by femtosecond laser ablation. This will increase the transmittance of the crystal in the mid-IR range [21] and should still provide a high damage threshold in a widely tunable laser [22].

The single-pulse laser ablation method is based on local material removal due to the high energy density of a single femtosecond pulse. The energy from the pulse is transferred to the super-heated electron plasma induced in the volume of the substrate and then to the atomic lattice of the material. This process takes place over an extremely short period of time (on the sub-picosecond scale), which is not sufficient for long-range heat diffusion in the crystal lattice [23]. Therefore, a large amount of accumulated energy in the microscopic volume causes ablation—local material removal in the zone where a laser beam intensity exceeds the ablation threshold of the material.

A Pharos Yb:KGW laser (Light Conversion, Lithuania) was used as the femtosecond pulse source for ARM fabrication. Both the second- and third-harmonic generation systems are implemented, thereby providing a choice of three operating wavelengths: 1026 nm, 513 nm, and 342 nm. The laser system can produce pulses of 200 to 1000 fs at repetition rates of 1 kHz to 200 kHz. A 513 nm wavelength was used, and the maximum available average power was 1.5 W. To accelerate the production rate, only single pulse was used to form single micro-cavity.

For sample positioning, Aerotech ANT-90 (Aerotech Incorporated, USA) three-axis nanopositioners were used to provide an in-position stability of 2 nm and repeatability of 75 nm at velocities of up to 200 mm/s. The movement velocity of stages was adjusted to produce ARMs with a period of 3.4 µm. This should provide reasonably good antireflection properties at 10.6 µm as well as for the second harmonic at 5.3 µm. A $100^{\times}$ objective lens (Mitutoyo Corporation, Japan) was used to focus the laser beam slightly underneath the sample surface. The exact shift from the surface was determined experimentally.

The ARM transmittance was measured using two different spectrometers. For the first measurement, a Fourier-transform infrared spectrometer (FTIR) spectrometer Bruker Vertex 70v with an additional aperture of 1.8 mm diameter was used to reduce the beam size in accordance with available ARM aperture ($2 \times 2\;{{\rm mm}^2}$). The aperture was placed immediately before the sample. The chamber with the sample inside was evacuated during the measurements. The spectral resolution was $4\;{{\rm cm}^{- 1}}$, and the internal aperture was 6 mm.

The second measurement was performed using a different Bruker Vertex 70 FTIR spectrometer combined with a microscope Hyperion 2000. This device does not allow the chamber to be evacuated. Thus, the transmittance measurement was less precise due to air absorption. The aperture on the surface sample was $100 \times 100\;{{\unicode{x00B5}{\rm m}}^2}$, while the resolution was $4\;{{\rm cm}^{- 1}}$. The focal length of the objective lens was approximately 50 mm.

Two ARM samples were fabricated on a $10 \times 10\;{{\rm mm}^2}$ GaSe crystal, with each sample of size $2 \times 2\;{{\rm mm}^2}$. For transportation and storage, the crystal was embedded in a round frame with a free aperture of approximately 9 mm. The size of the ARM samples is optimal for balancing the fabrication time and measurement efficiency.

The best ARM sample was fabricated with the following parameters: wavelength of 512 nm, pulse duration of 200 fs, repetition rate of 200 kHz, and average laser power of 450 mW. These values provide the maximum energy density at the surface sufficient to ablate GaSe and form cavities with a diameter of approximately 2.6 µm and a period of approximately 3.4 µm, as shown in Figs. 1(a) and 1(b). The width of the walls between the cavities is approximately 0.8 µm, thus providing a fill factor of 0.76. Furthermore, the structure is persistent and periodical, except for some residual material that remained on the protruding parts of the ARM, which was probably caused by the solidification of the ablated material.

 

Fig. 1. SEM image of the ARM sample: (a) top-down overview; (b) magnified; (c) cross-section profile.

Download Full Size | PDF

The cross-sectional profile of the ARM is shown in Fig. 1(c). The depth of the microstructure is approximately 2.5 µm (note the sample’s 52° tilt when observed by scanning electron microscopy (SEM)) which provides an aspect ratio of approximately 0.96.

Figure 2 shows the theoretical values of the untreated GaSe transmittance for ordinary and extraordinary waves in comparison with the measured transmittance values for the flat surface. The values of the theoretical refractive indices were calculated using Sellmeier equations [24]. The optical element was obtained by cleaving along a plane perpendicular to the $z$ axis, and unpolarized light was used for transmittance measurements. Therefore, the experimental value of the untreated GaSe transmittance is closer to the transmittance value for an ordinary wave.

 

Fig. 2. Comparison of single-surface transmittance of untreated GaSe and ARM.

Download Full Size | PDF

The transmittance spectra of the fabricated ARM samples are shown in Fig. 2. The first measurement (red line) was made with a 1.8 mm diameter aperture and provides the information of overall transmittance of the ARM: an average ARM single surface transmittance of 94% in the 7–11 µm range and a maximum single surface transmittance of 94.6% at approximately 8.5 µm.

The second (green line) measurement is an averaged transmittance obtained in three points across the ARM sample with a $100 \times 100\;{{\unicode{x00B5}{\rm m}}^2}$ aperture. It demonstrates an average single surface transmittance of 94% in the 6–13 µm range and a maximum single surface transmittance of 97.8% at 8.5 µm in comparison with an average transmittance of 77% for GaSe single flat surface (recalculated from crystal transmittance measurement). The small aperture allowed for the analysis of ARM zones with less residual material and with higher uniformity; therefore, the results of the measurements with a smaller aperture are slightly better than those with larger aperture. Figure 2 shows that the absorption lines are related to ${{\rm H}_{2}}{\rm O}$ at 11–11.5 µm and ${{\rm CO}_{2}}$ at 14–15 µm.

Some areas on the sample decrease the overall transmittance and lead to discrepancies in measurements made with small and large apertures. Transmittance in one of these areas was measured (black line) to quantitatively estimate the non-uniformity of the structure. Despite the significantly lower transmittance in such areas, their overall presence should be negligible, as measurements with large aperture mostly coincide with better-case measurements (green line) performed with smaller aperture. We consider that the emergence of decreased transmittance areas is mostly connected with imperfections of the GaSe crystal surface, and not with the fabrication process malfunctions.

In Fig. 2, the transmittance increase is observed only at wavelengths larger than 5 µm, because the diffraction on structure features starts at 7 µm. The range of increased transmittance can be blueshifted or redshifted by adjusting the ARM period, if necessary [25]. At longer wavelengths, a gradual decrease in transmittance is observed because longer wavelengths fail to interact with the structure, and the spectral behavior of ARMs asymptotically tends toward Fresnel reflection of flat, untreated surfaces. This phenomenon is briefly explained by the electromagnetic roughness theory (so-called Rayleigh and Fraunhofer criteria). The idea is that for the normally incident light of wavelengths larger than some thresholds, the surface can be treated as specularly reflective. The threshold depends on the characteristic dimensions of the surface microstructure. For the case of ARMs, this can be interpreted as a gradual decrease in effective depth leading to gradual decrease of transmittance at longer wavelengths. The maximum transmittance of the ARM is more than 20% higher than that of an untreated GaSe surface. However, it is possible to achieve a maximum transmittance of 99% with a higher aspect ratio of ARM, requiring appropriate laser beam shaping.

A straightforward method for ARM fabrication was proposed to increase the transmittance of GaSe. The transmittance of the fabricated ARMs was measured using two different spectrometers. The results show a significant increase in transmittance. The best fabricated ARM demonstrated at least 94% average single-surface transmittance (recalculated from sample transmittance measurement) in the 7 to 11 µm range compared to 77% of untreated GaSe surface transmittance. Furthermore, the maximum transmittance obtained was 97.8% at 8.5 µm, which is almost 21% higher than that of an untreated GaSe surface.

In our future work, we will optimize the ARM fabrication process in order to increase the maximum transmittance values in the 7–11 µm range to 99%, and increase the average transmittance in a wider spectral range for both sides of a crystal. In addition, the techniques for residual material removal will be tested.