Open Access
Add to BibSonomy
Abstract
Here, we report on the fabrication of cm-long microchannels in LiNbO3 by selective etching of femtosecond laser inscribed tracks using hydrofluoric acid. We achieved a 1 cm long microchannel after 300 h of etching a single track inscribed into the volume along the optical axis of LiNbO3. Furthermore, we investigated the dependence of the etching behavior on various writing parameters. Highly selective etching with a selectivity up to 104 was achieved and a functional relationship between the etched depth and time was found. Thus, our results set the first milestone for future fabrication of 3D-hollow microstructures in the volume of LiNbO3 combining its outstanding physical properties such as the strong nonlinearity as well as the acousto- and electrooptic properties with both microfluidic and photonic structures in a monolithic setup.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In femtosecond (fs) laser structuring of materials, ultrashort laser pulses are strongly focused into the volume of a transparent dielectric medium. Intensity-dependent nonlinear absorption limits the energy deposition and thus the material modification to the focal volume, so that the surrounding material is not modified [1]. Femtosecond laser structuring has become a well-known and versatile technique used for material processing in many fields, ranging from integrated photonics over optofluidics to lab-on-a-chip [2]. Various optical elements such as waveguide lasers and amplifiers, couplers, splitters, optical sensors, and data memories have been fabricated in the volume of samples by femtosecond laser structuring [3].
On the other hand, selective etching of fs laser inscribed structures is a unique technique to fabricate hollow structures inside the volume of transparent dielectrics. If the fs laser modified material shows a much higher etching rate than the unmodified one, it can be selectively removed with a suitable etchant [4,5]. In this way, 3D structures can be fabricated entirely inside the sample volume without the need for sandwiching, i.e., a cover substrate on top of the structured bottom part. Moreover, in contrast to, e.g., precision diamond dicing, this method can also be applied to fabricate curved structures. Laser assisted selective etching is well established and even commercialized in glass [6,7], but it has also been demonstrated in a few crystals like sapphire [8], YAG [9], and CaF2 [10]. So far, applications such as photonic crystal fiber-like waveguides, microchannels combined with optical waveguides, microcavities, and many other microstructures of different styles and shapes have been realized [3].
Lab-on-a-chip is a research field of high interest, especially for applications in medicine and biology, since miniaturized devices offer lots of benefits such as accessibility, cost efficiency and sustainability. Polymers like polydimethylsiloxane (PDMS) have been widely used as they are inexpensive and easy to process via lithographic methods [11,12]. However, multi-functionality is usually required to realize complex analysis on a single chip. A suitable material for such multifunctional applications is lithium niobate (LiNbO3, LN). It has, for example, a high nonlinearity, high electrooptic and piezoelectric coefficients allowing for integrating waveguides, frequency converters, modulators, and resonators on a single chip [13]. To combine such elements with microfluidic channels for lab-on-a-chip devices, techniques such as ablation, precision dicing, and drilling have been applied to fabricate microfluidic channels in LN [14–17]. All these methods require sandwiching which would not be needed in the case of fs laser assisted selective etching.
Femtosecond laser structuring of LN has been investigated for many years. Waveguides, splitters, arrays, resonators, and gratings, as well as periodically-poled structures for nonlinear frequency conversion have been demonstrated [18]. Several elements have been combined effectively in LN in a monolithic and hybrid integrated approach [19]. Selective etching of fs laser-modified regions of up to a few micrometers depth has already been observed in the past [20–22]. However, selective etching of microchannels has not been studied in detail in LN until now.
In this work, we present, to the best of our knowledge, the first selectively etched cm-long microchannels in LN. Furthermore, we systematically studied the influence of the inscription parameters on the selective etching process concerning etching speed and selectivity as defined in [23]. The results of this systematic investigation will be crucial for the subsequent fabrication of complex selectively etched hollow microstructures in LN.
2. Experimental methods
2.1 fs laser structuring
A setup similar to the one described in [24] was used for fs laser structuring. However, a femtosecond laser oscillator (Pharos PH1-15) with a 1 MHz repetition rate, a pulse duration of 260 fs, a central wavelength of 1030 nm and a beam quality M2 < 1.1 was used as the laser source. The laser beam was focused into the sample using an aspheric lens with a focal length of 4.5 mm and a NA of 0.55. The lens was fully illuminated by the laser beam and was focused 100 µm deep into the sample, resulting in modifications located ∼200 µm below the surface due to the (ordinary) refractive index (2.23) of LN at 1030 nm. Instead of air-bearing stages, mechanical stages (Newport XMS50-S) were used with a translation precision of ± 0.3 µm, an acceleration of 5000 mm/s2, and a maximum scan speed of 300 mm/s. The distance for acceleration (∼9 mm), required for the stage to attain a constant speed, was considered during the experiment.
LN shows a strong nonlinear absorption compared to other materials used for fs laser structuring, e.g., glass. Hence, the threshold for nonlinear processes is lower and can be easily overcome at pulse durations below and in the range of a few 100 fs. Therefore, it can be challenging to limit the material modifications to the focal volume [20]. Because the laser pulse duration of the used system was a fixed parameter, the influence of longer pulse duration was not investigated here, though. Since selective etching of MHz repetition rate fs laser inscribed tracks in YAG showed a faster etching behavior than tracks written with 1 kHz, and high repetition rates also provide enough spatial overlap of the pulses even at high writing velocities of 100 mm/s [23], we decided to limit the current experiments to a repetition rate of 1 MHz. A drawback of the high repetition rate is heat accumulation, leading to a broadened and asymmetrical shape of the material modification [25]. Nano-gratings oriented along the writing direction have been observed in LN and glass when the inscribing laser beam is polarized perpendicularly to the writing direction [26,27]. Since the nano-gratings caused faster selective etching in glass, we also investigated the influence of the beam polarization parallel (π) and perpendicular (σ) to the writing direction.
As LN has a trigonal crystal structure with a single optical axis, it was expected that both inscription as well as etching could be different for polarizations parallel and perpendicular to the optical axis. Three samples (A, B, and C, cf. Fig. 1), all x-cut LN with dimensions of 11 mm × 20 mm and 10 mm × 10 mm, were diced from a 1 mm thick wafer. Sample A was used to investigate the influence of inscription pulse energy, translation speed, and the laser beam's polarization relative to the translation direction. Therefore, tracks separated by 100 µm in the y-direction were inscribed using pulse energies from 100 nJ - 700 nJ with a σ-polarized beam and a scan speed of 10 mm/s along the crystal's optical c-axis (z-axis). The same was repeated with a scan speed of 20, 30, 40, 50, and 60 mm/s with a 300 µm distance in the y-direction between each group of tracks written at the same speed. Furthermore, the whole protocol was repeated with a π-polarized beam. In sample B, tracks were inscribed parallel and perpendicular to the crystals’ z-axis, using an inscription energy of 150 nJ, a speed of 10 mm/s, and a σ-polarized beam to investigate the influence of the crystal orientation on the etch process. Sample C was inscribed with several tracks along the crystal’s z-axis with a pulse energy of 150 nJ and a scan speed of 10 mm/s. The tracks were stacked next to each other with a 2 µm distance in the y-direction to form a rectangular cross section in the xy-plane, i.e., to show the feasibility of producing rectangularly shaped hollow structures. After inscription the laser-induced material modifications were characterized using a digital microscope (Keyence VHX 7000) in transmitted light full coaxial mode to obtain their dimensions (width w and height h, cf. Fig. 2(a)) and morphology. Parameters leading to cracks were avoided during further investigations.
Fig. 1. Sketch of all three inscribed samples showing the orientation of the crystal with respect to the writing direction, the inscription parameters, and the position of the tracks in the sample.
Fig. 2. a) Cross section of a track inscribed with a pulse energy of 200 nJ and a scan speed of 10 mm/s, along the optical c-axis, b) Cross section of a) after 177 hours of etching, c) Top view of partly etched microchannels inscribed with 400 nJ, 300 nJ and 200 nJ (top to bottom).
2.2 Selective etching
During inscription, the laser focus was distorted at the end facet of the sample such that the modified tracks did not reach the end facet. Therefore, about 250 µm were removed from the end facet by polishing to expose the modified region before the chemical etching. For selective etching the samples were put in a teflon crucible, closed with a lid, containing 40% hydrofluoric acid (HF) at room temperature. They were removed at increasing intervals to record the channel's cross-section in the xy-plane and the etched depth per time as well as its morphology (cf. Fig. 2(b) and (c)), using the microscope described above.
The etching selectivity S is referred to in the literature as the ratio between the etching speed of laser-modified and pristine LN. Due to the limited selectivity, the channel width increases beginning at the channel entrance over time, forming a taper (cf. Fig. 2(c)). The taper directly corresponds to the selectivity and thus, S can be calculated from the change in width (Δw) over channel length (d), measured on the channel's cross-section [23]
In addition, the dependence of the etching process selectivity S on the inscription energy, writing speed, and polarization was investigated.
In order to verify that the etched structure is a hollow channel, it was diced open with a precision diamond blade dicing saw (Disco DAD322). As shown in Fig. 3(a), a cut along the z-axis was made using an 80 µm thick Z09-SD2000-Y1-60 blade, about 50 µm next and parallel to the selectively etched channel intended to be diced open. Afterward, the sample was turned 90° clockwise in the xy-plane (cf. Fig. 3(b)). Using a 200 µm thick G1A853-SD6000R21B01 blade, the material above the channel was cut down in steps of a few micrometers until the channel was revealed. Finally, the opened channel was imaged using a Keyence VK-X200 laser scanning microscope (LSM). A LSM image of a cut open channel is shown in Fig. 7 as an example.
3. Results and discussion
3.1 fs-laser induced material modifications
As an example, three different kinds of tracks inscribed are shown in Fig. 4. The tracks were inscribed with a speed of 10 mm/s and pulse energies of 200 nJ, 300 nJ, and 400 nJ. All tracks showed an elongated shape resulting from spherical aberration due to the refractive index change from air to lithium niobate as it is known for fs-laser structuring of other materials, too [28]. Spherical aberration could be compensated by various methods such as using a slit or by spatiotemporal focusing [29,30].
As can be seen, the material modification is broader and stronger in the upper part and thinner below. The critical power for self-focusing can be expressed as [31]
where n0 and n2 (2.3 and 1.5 × 10−19 m2/W for LN) are linear and nonlinear refractive indices, respectively, and λ is the laser wavelength. The critical power Pcr is 0.461 MW at 1030 nm, which corresponds to a pulse energy of 119 nJ for a pulse duration of 260 fs. Hence, the thinner and less intense part of the modification on its bottom should result from self-focusing-induced filamentation.Heat accumulation, on the other hand, results in a broadened modification because of a synergistic influence of parameters described as laser fluence [25]
where $2{\omega _o}$ is the beam diameter (2.62 µm) at the focus, and v, R, and Ep are the scan speed, repetition rate, and laser pulse energy, respectively [25]. The heat diffusion time can be calculated as tc = $(2{\omega _o})$2/κT, according to [8], with a temperature diffusivity of κT = κ /(cpρ) = 0.02 cm2s–1, obtained with the thermal conductivity κ = 5.234 Wm–1K–1, specific heat capacity cp = 628 Jkg–1K–1, and density ρ = 4.64 g cm–3 of LN [32]. Thus, the thermal diffusion time tc in LN is 3.82 µs. At a 1 MHz repetition rate, the time interval between two pulses is 1 µs. Because this is shorter than the time required for thermal diffusion, heat accumulation is expected to occur, and the broadened and segmented upper part of the tracks can be attributed to this most probably.At a constant writing speed of 10 mm/s and energies between 200 nJ and 300 nJ, the width and height of the tracks increased from w = 1.47 µm to 5.86 µm and h = 21.04 µm to 28.74 µm, respectively. This means that the tracks were up to ∼ 2 times broader than the focal diameter and less than 10 µm shorter than the focal displacement of 37 µm, resulting from spherical aberration at the air-lithium niobate interface, calculated according to [33]. With a pulse energy of 300 nJ and a scan speed between 10 and 60 mm/s, the height of the modification decreased linearly from 28.74 µm to 24.73 µm with increasing speed. All scan speeds (10 mm/s - 60 mm/s) investigated correspond to a pulse overlap of ∼ 99.9% calculated according to [34].
For energies higher than 300 nJ, the height and width of the tracks increase strongly, and cracks appear (cf. in Fig. 4, where the track inscribed with Ep = 400 nJ, v = 10 mm/s shows cracks). For further investigation of the selective etching process in LN, modifications with cracks were avoided. Pulse energies and writing velocities in the gray area of Fig. 5 are most suitable for inscription to achieve a preferably homogeneous and well-confined material modification within the investigated parameter range (cf. Fig. 4, the track inscribed with 200 nJ).
Fig. 5. 2D color map showing the suitable inscription parameter range in light grey for a: a) σ-polarized beam, and b) π-polarized beam based on inscription pulse energy and scanning speed.
3.2 Selectively etched microchannels
As already shown in Fig. 2, strongly selective etching of the laser modified material was observed after immersing the samples in HF. The etched part of a fs laser inscribed track darkens where it is etched like in fs laser modified selectively etched glass [27].
For tracks inscribed with a pulse energy of 150 nJ along the z-axis the cross-section of the etched channel is rhombically shaped (cf. Fig. 6(a)) with a height h of 3.7 µm, a width w of 2.6 µm, an acute angle of 50° and an obtuse angle of 130°. However, tracks inscribed with pulse energies of 200 nJ and above will result in channels with vertical sidewalls between the two acute angles instead of the rhombical shape (cf. Fig. 6(b) and (d)). A channel etched along the y-axis is shown in Fig. 6(c). Its cross-section is shaped like an isosceles triangle (h = 8.8 µm, w = 10.2 µm) with vertex and base angles of 55° and 70°, respectively. Since, LN has a trigonal crystal structure, and can be described by a hexagonal unit cell [35], the cross-sectional shape of the etched channel is likely related to the crystal structure. This means in certain directions the etch rate of the pristine material might be higher than in other directions due to the orientation of the bonds in the crystal. For example, it is known for many years that there is a significant etch rate along the + z-direction in LN, which H. Hu et al. reported to be 0.8 µm/h in a mixture of hydrofluoric and nitric acid and ethanol [36]. This rate would lead to an etched length of 56 µm in the + z-direction after 70 hours of etching. Here, however, a total width of 10.2 µm was observed, mostly to be attributed to the etching of laser modified material, as the width and height of the triangle differ by only 25%.
Fig. 6. Microscope image showing the top view and cross-section of channels etched from single tracks inscribed along the a) z-axis of x-cut LiNbO3 with a pulse energy of 150 nJ (after selective etching for 70 hours), b) z-axis of x-cut LiNbO3 using a pulse energy of 300 nJ (after 60 hours of selective etching), c) y-axis of x-cut LiNbO3 with a pulse energy of 150 nJ after selective etching for 70 hours, and d) z-axis of x-cut LiNbO3 with a pulse energy of 200 nJ after selective etching for 177 hours.
Fig. 7. LSM image of the opened micro-channel showing the residual groove of the channel after polishing.
As can also be seen on the strong taper of the channel shown in Fig. 6(c), the selectivity is significantly higher along the crystal's optical axis than along the y-axis. The selectivity along the optical axis using (σ, v = 10 mm/s, and Ep = 150 nJ) is 6000 ± 1200, while the selectivity along the y-axis for the same inscription conditions is 180 ± 10. This difference in the etching selectivity is attributed to the etching of the pristine material along the z-axis. However, the influence of the crystal structure, or rather the etch rate of pristine material along certain directions in the crystal is not yet completely understood and subject to further investigations. Despite the different cross-sectional shapes and selectivity, the etch direction did not influence the etch speed of the modified material significantly. After 70 hours of etching the etched depth is the same and about (1.71 ± 0.25) mm for both etching along y- and z-directions.
To confirm that the etched structures are microchannels, indeed, a channel etched from a single track, inscribed using a scan speed of 10 mm/s and inscription energy of 200 nJ, was diced open, using the technique described above. Figure 7 shows a laser scanning microscope image of the diced-open channel. The remaining groove has a width of 13.1 µm and a height of 3.2 µm since the channel was turned 90° around the z-axis. The mean surface roughness of the channel’s surface is Sa = 40 nm. This roughness is better than that observed in micro-fluidic channels fabricated by laser ablation (Sa = 300 nm) [14], however, it is rather high when compared to that reported for waveguides fabricated by chemo-mechanical polishing of LNOI (Sa = 0.452 nm) [37]. That means, in order to further utilize this technique for photonic applications, the roughness needs to be further reduced, e.g., by annealing methods as described in [38] or using other etching agents as done by H. Hu in [36].
Selective etching of sample C led to a 10 mm × 100 µm × 20 µm microchannel, with a roughly rectangular cross-section and a selectivity of 4500 ${\pm} $ 1200 (Fig. 8(a) and (c)). After etching, the sample was cleaned with water and heated to dry up all liquid in the channel. During imaging, water droplets were placed at the sample's edge which were sucked into the channel by the capillary force (cf. Fig. 8(b)).
Fig. 8. a) Side view of a 1 cm long 100 µm × 20 µm selectively etched microfluidic channel in the LN, b) left side of a) magnified showing droplet flowing, and c) Microscope image of cross section of channel.
In Fig. 9, the results of the investigation of the dependence on writing parameters of the etching selectivity are summarized. The highest selectivity achieved for σ- and π-polarization are 9000 $\pm $ 2000 and 3300 $\pm $ 600, respectively. The selectivity was measured on the taper of the channels imaged with a magnification of 300×. This was the highest magnification showing the whole length of the channels in one image without stitching images, which was avoided to prevent image stitch errors. At this magnification it was not possible to obtain length and width measurements with a precision higher than ${\pm} $ 0.6 µm, leading to a high uncertainty of up to 26% for the selectivity values. However, this work here shows that it is possible to obtain a selectivity as high as 104, which is in the range of the etching selectivity observed for sapphire (Smax = 104) [8] and YAG (Smax = 104) [23] and significantly higher than in glass (Smax = 1400) [39]. A possible reason for the polarization dependent selectivity might be the orientation of nano gratings formed in the modified region after fs laser structuring, as observed in both glass [40] and LN [26]. Though, the existence of nano gratings has not been proved in the present work. The scanning speed had no significant influence on the selectivity, but the selectivity decreases with an increase in inscription pulse energy only for σ-polarization, which has also been observed in other materials [7].
Fig. 9. Etching selectivity vs. pulse energy for writing velocities of 10, 30, and 50 mm/s for polarization of the laser beam a) parallel and b) perpendicular to the writing direction.
The etched channels were always broader than fs laser inscribed modifications. For tracks inscribed along the z-axis with (σ, v = 10 mm/s, and Ep = 100-300 nJ), the etched channels are up to 3 µm wider after 177 hours of selective etching than the inscribed modification. The height of the etched channel is the same as the height of the inscribed track for Ep > 300 nJ (cf. Fig. 10(a) and (b)) and less than the height of the inscribed track for Ep = 100 to 200 nJ since some filamented part is not etched (cf. Fig. 2(a) and (b)).
Fig. 10. Cross section of a) fs laser writing track writing with σ, v = 10 mm/s, and Ep = 300 nJ, along the optical c-axis before etching, b) Cross section of a) after 177 hours of etching.
On the other hand, channels etched along the y-axis from a track inscribed with (σ, v = 10 mm/s, and Ep = 150 nJ) also have a height less than the as the inscribed modification, however with a width about 8× the width of the inscribed track because of the higher etch rate of pristine material along the z-axis, as already explained above.
3.3 Temporal evolution of the etch speed
To investigate the etching progress quantitatively, the measured etched depth d was plotted against the etching time t (cf. Fig. 11). The etched depth showed a similar dependence on time as observed for YAG [23] and can be fitted accordingly with
Here ${D_\mathrm{\gamma }}$ is the diffusion coefficient and the parameter $\mathrm{\gamma }$ ≥ 1. In the case of $\mathrm{\gamma } = 1,$ this temporal dependence can be attributed to an etching process determined by normal diffusion as observed in YAG for most etching parameters. With higher values of $\mathrm{\gamma }$ (for LN mostly between 1.5 - 1.6), the equation describes super diffusion. Generally, super diffusion determined etching behavior could result from either the etching of porous media or the viscoelasticity of the diffusing medium. The actual physical mechanism for LN is not yet fully understood.
Fig. 11. Etched depth d vs. etching time t for tracks written with 100 nJ and 200 nJ pulse energies. The lines show the fit functions according to Eq. (4), with the fit parameters given in the legend.
At 100 nJ pulse energy and 10 mm/s writing velocity, the threshold fluence required to modify the material is attained, leading to a weak material modification and thus a slow etch speed with Dγ = 75.47 µm2/h1.6.
Between 200 nJ and 300 nJ, the influence of the pulse energy on the etching speed is negligible, as can be seen in Fig. 12, and the etching is significantly faster, compared to the case of a pulse energy of 100 nJ. Etching parameters of Dγ = 3034.64 µm2/h1.5 and Dγ = 3610.27 µm2/h1.6 have been observed for 200 nJ and 300 nJ, respectively. Thus, for pulse energies of 200 nJ – 300 nJ, a 1 cm long track can be completely etched within one week by etching from both sides (compared to 23 days in YAG [23] and ∼3 days in glass [41]). Since many samples could be etched in the same etching bath, and the process could be controlled automatically and runs at room temperature, it is also not consuming a lot of resources. Therefore, the etching time would not be a limiting factor for industrial applications.
Fig. 12. Etched depth d after 177 h of etching vs. inscription speed v for two different pulse energies and both beam polarizations.
Also, the results plotted in Fig. 12 show that there is no significant influence of beam polarization and writing speed on the etch speed for the range of parameters investigated. This is different from what is observed in glass, which is possible, since glass is an amorphous material, while lithium niobate is a crystal and the physical mechanism leading to an increase of the etching speed would be different [41].
4. Conclusion and outlook
To the best of our knowledge, we presented highly selective etching of fs laser inscribed microstructures in LN using HF acid, leading to cm-long microchannels for the first time. A fs laser oscillator was used to inscribe single tracks into the volume of the LN crystal. The modified sample was then subjected to chemical etching in hydrofluoric acid to fabricate 3D-hollow structures. Investigating the etched depth vs. time, a functional dependence for the etching progress on time was found. Using appropriate fitting parameters, the influence of the inscription parameters on the etching process can be evaluated quantitatively. A very high selectivity of up to 104 was observed. By cutting open the etched channels with a precision diamond saw, a sidewall roughness of the channels in the range of 40 nm was measured.
Our results presented here pave the way for future application of LN for integrated sensing and even lab-on-a-chip applications, combining microfluidic channels with optical elements based on the outstanding optical properties of LN such as the high nonlinearity, the high electrooptic, and acousto-optic coefficients. Nevertheless, there is still a need for further systematic parameter studies of the etching process, e.g., concerning the influence of the crystal structure, and optimization of both etching selectivity and speed using other etching agent mixtures. Furthermore, the sidewall roughness of the etched structures needs to be investigated closer, and a procedure to reduce it further needs to be found to allow for low loss selectively etched optical elements.
Disclosures
The authors declare no conflicts 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. C. B. Schaffer, A. Brodeur, and E. Mazur, “Laser-induced breakdown and damage in bulk transparent materials induced by tightly focused femtosecond laser pulses,” Meas. Sci. and Technol. 12(11), 1784–1794 (2001). [CrossRef]
2. F. Sima, K. Sugioka, R. M. Vázquez, et al., “Three-dimensional femtosecond laser processing for lab-on-a-chip applications,” Nanophotonics 7(3), 613–634 (2018). [CrossRef]
3. S. Gross and M. J. Withford, “Ultrafast-laser-inscribed 3D integrated photonics: Challenges and emerging applications,” Nanophotonics 4(3), 332–352 (2015). [CrossRef]
4. C. A. Ross, D. G. MacLachlan, D. Choudhury, et al., “Optimisation of ultrafast laser assisted etching in fused silica,” Opt. Express 26(19), 24343–24356 (2018). [CrossRef]
5. R. Drevinskas, M. Gecevičius, M. Beresna, et al., “Tailored surface birefringence by femtosecond laser assisted wet etching,” Opt. Express 23(2), 1428–1437 (2015). [CrossRef]
6. Y. Kondo, T. Suzuki, H. Inouye, et al., “Three-dimensional microscopic crystallization in photosensitive glass by femtosecond laser pulses at nonresonant wavelength,” Jpn. J. Appl. Phys. 37(1A), L94 (1998). [CrossRef]
7. J. Gottmann, M. Hermans, N. Repiev, et al., “Selective laser-induced etching of 3D precision quartz glass components for microfluidic applications-up-scaling of complexity and speed,” Micromachines 8(4), 110 (2017). [CrossRef]
8. S. Juodkazis, K. Nishimura, H. Misawa, et al., “Control over the crystalline state of sapphire,” Adv. Mater. 18(11), 1361–1364 (2006). [CrossRef]
9. J. Siebenmorgen, K. Petermann, G. Huber, et al., “Femtosecond laser written stress-induced Nd:Y3Al5O12 (Nd:YAG) channel waveguide laser,” Appl. Phys. B: Lasers Opt. 97(2), 251–255 (2009). [CrossRef]
10. S. Matsuo, K. Iwasa, T. Tomita, et al., “Femtosecond laser-assisted etching of fluoride crystals,” J. Laser Micro/Nanoeng. 6(3), 245–248 (2011). [CrossRef]
11. P. Lakhera, V. Chaudhary, B. Bhardwaj, et al., “Development and recent advancement in microfluidics for point of care biosensor applications: A review,” Biosens. Bioelectron.: X 11, 100218 (2022). [CrossRef]
12. F. Bahavarnia, M. Hasanzadeh, D. Sadighbayan, et al., “Recent progress and challenges on the microfluidic assay of pathogenic bacteria using biosensor technology,” Biomimetics 7(4), 175 (2022). [CrossRef]
13. M. G. Vazimali and S. Fathpour, “Applications of thin-film lithium niobate in nonlinear integrated photonics,” Adv. Photonics 4(03), 1 (2022). [CrossRef]
14. G. Pozza, S. Kroesen, G. Bettella, et al., “T-junction droplet generator realised in lithium niobate crystals by laser ablation,” Optofluid. Microfluid. Nanofluid. 1(1), 34–42 (2014). [CrossRef]
15. D. Huo, Z. J. Choong, Y. Shi, et al., “Diamond micro-milling of lithium niobate for sensing applications,” J. Micromech. Microeng. 26(9), 095005 (2016). [CrossRef]
16. G. Bettella, G. Pozza, S. Kroesen, et al., “Lithium niobate micromachining for the fabrication of microfluidic droplet generators,” Micromachines 8(6), 185 (2017). [CrossRef]
17. M. Chauvet, L. Al Fares, and F. Devaux, “Self-trapped beams for fabrication of optofluidic chips,” Proc. SPIE in Nonlinear Optics and Applications VI 8434, 84340Q (2012). [CrossRef]
18. B. Zhang, L. Wang, and F. Chen, “Recent advances in femtosecond laser processing of LiNbO3 crystals for photonic applications,” Laser Photonics Rev. 14(8), 1 (2020). [CrossRef]
19. J. Thomas, M. Heinrich, P. Zeil, et al., “Laser direct writing: Enabling monolithic and hybrid integrated solutions on the lithium niobate platform,” Phys. Stat. Sol. A 208(2), 276–283 (2011). [CrossRef]
20. J. Burghoff, H. Hartung, S. Nolte, et al., “Structural properties of femtosecond laser-induced modifications in LiNbO3,” Appl. Phys. A: Mater. Sci. Process. 86(2), 165–170 (2006). [CrossRef]
21. T. Wang, X. Xu, L. Yang, et al., “Fabrication of lithium niobate fork grating by laser-writing-induced selective chemical etching,” Nanophotonics 11(4), 829–834 (2022). [CrossRef]
22. T. Wang, X. Cheng, X. Li, et al., “Femtosecond-laser-assisted high-aspect-ratio nanolithography in lithium niobate,” Nanoscale 15(37), 15298–15303 (2023). [CrossRef]
23. K. Hasse, D. Kip, and C. Kränkel, “Influence of diluted acid mixtures on selective etching of MHz- and kHz-fs-laser inscribed structures in YAG,” Opt. Mater. Express 11(5), 1546–1554 (2021). [CrossRef]
24. K. Hasse, G. Huber, and C. Kränkel, “Selective etching of fs-laser inscribed high aspect ratio microstructures in YAG,” Opt. Mater. Express 9(9), 3627–3637 (2019). [CrossRef]
25. D. Tan, B. Zhang, and J. Qiu, “Ultrafast laser direct writing in glass: Thermal accumulation engineering and applications,” Laser Photonics Rev. 15(9), 2000455 (2021). [CrossRef]
26. P. Karpinski, V. Shvedov, W. Krolikowski, et al., “Femtosecond laser writing in lithium niobate,” in Conference on Lasers and Electro-Optics, OSA Technical Digest (2016) (Optica Publishing Group, 2016) (OSA, 2016), p. paper STh1Q.2. [CrossRef]
27. C. Hnatovsky, R. S. Taylor, E. Simova, et al., “Polarization-selective etching in femtosecond laser-assisted microfluidic channel fabrication in fused silica,” Opt. Lett. 30(14), 1867–1869 (2005). [CrossRef]
28. L. Huang, P. S. Salter, F. Payne, et al., “Aberration correction for direct laser written waveguides in a transverse geometry,” Opt. Express 24(10), 10565–10574 (2016). [CrossRef]
29. M. Ams, G. D. Marshall, D. J. Spence, et al., “Slit beam shaping method for femtosecond laser direct-write fabrication of symmetric waveguides in bulk glasses,” Opt. Express 13(15), 5676–5681 (2005). [CrossRef]
30. F. He, H. Xu, Y. Cheng, et al., “Fabrication of microfluidic channels with a circular cross section using spatiotemporally focused femtosecond laser pulses,” Opt. Lett. 35(7), 1106–1108 (2010). [CrossRef]
31. G. Fibich and A. L. Gaeta, “Critical power for self-focusing in bulk media and in hollow waveguides,” Opt. Lett. 25(5), 335–337 (2000). [CrossRef]
32. G. Chen, N. Li, J. Da Ng, et al., “Advances in lithium niobate photonics: development status and perspectives,” Adv. Photonics 4(03), 034003 (2022). [CrossRef]
33. Q. Sun, H. Jiang, Y. Liu, et al., “Effect of spherical aberration on the propagation of a tightly focused femtosecond laser pulse inside fused silica,” J. Opt. A: Pure Appl. Opt. 7(11), 655–659 (2005). [CrossRef]
34. G. Schnell, U. Duenow, and H. Seitz, “Effect of laser pulse overlap and scanning line overlap on femtosecond laser-structured Ti6Al4V surfaces,” Materials 13(4), 969 (2020). [CrossRef]
35. R. S. Weis and T. K. Gaylord, “Lithium niobate: Summary of physical properties and crystal structure,” Appl. Phys. A 37(4), 191–203 (1985). [CrossRef]
36. H. Hu, R. Ricken, W. Sohler, et al., “Lithium niobate ridge waveguides fabricated by wet etching,” IEEE Photonics Technol. Lett. 19(6), 417–419 (2007). [CrossRef]
37. R. Wu, M. Wang, J. Xu, et al., “Long low-loss-lithium niobate on insulator waveguides with sub-nanometer surface roughness,” Nanomaterials 8(11), 910 (2018). [CrossRef]
38. S. Suntsov, C. E. Rüter, and D. Kip, “Er:Ti:LiNbO3 ridge waveguide optical amplifiers by optical grade dicing and three-side Er and Ti in-diffusion,” Appl. Phys. B: Lasers Opt. 123(4), 118 (2017). [CrossRef]
39. M. Hermans, J. Gottmann, and F. Riedel, “Selective, laser-induced etching of fused silica at high scan-speeds using KOH,” J. Laser Micro/Nanoeng. 9(2), 126–131 (2014). [CrossRef]
40. C. Hnatovsky, R. S. Taylor, E. Simova, et al., “Fabrication of microchannels in glass using focused femtosecond laser radiation and selective chemical etching,” Appl. Phys. A 84(1-2), 47–61 (2006). [CrossRef]
41. S. Kiyama, S. Matsuo, S. Hashimoto, et al., “Examination of etching agent and etching mechanism on femotosecond laser microfabrication of channels inside vitreous silica substrates,” J. Phys. Chem. C 113(27), 11560–11566 (2009). [CrossRef]
