Open Access
Add to BibSonomy
Abstract
Formation of relief structures on the surface of thermally poled silicate glasses with reactive ion etching (RIE) is performed for the first time. RIE of a poled soda-lime glass is compared with acidic chemical etching of the glass by a polishing etchant. The chemical etching provides surface relief ∼20 times deeper than RIE, because the RIE selectively etches “silica-like” poled layer of the glass contrary to the acidic etching in which the poled glass layer behaves as a mask for the etching. However, the quality of the transfer of the anodic electrode relief pattern to glass surface by RIE is higher because the latter allows smoothing of peculiar stronger poled regions near the edges of the anodic electrode used in the glass poling. In manufacturing structures, which elements are hundreds of nanometers in size, RIE is preferable over the chemical etching because of its directivity. Additional thermal treatment of the poled glasses allowed increasing of the surface relief formed by RIE ∼ 230% and by chemical etching ∼ 140% due to the relaxation of the poled region of the glass.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The opportunity to change the rate of the etching of silicate glasses with hydrofluoric acid (HF) by means of thermal poling [1] has opened a new way towards the formation of a micron-scale relief on the surface of the poled glasses with chemical etching [2–5]. Specifically, chemical etching allows one to reveal poling-induced hidden variations of the glass properties [6–8] and to increase tens of nanometers scale surface relief formed by the poling [9–10]. This latter relief is due to decreasing the volume of the poled region of the glass because of relaxation [11]. In multicomponent silicate glasses, thermal poling results in the depletion of their subsurface region with mobile alkali and alkali earth ions and, if open anode poling is used, in the enrichment of the subsurface layer with hydrogenated species generated from atmospheric water vapors [12]. In both cases, the poled regions have a “silica-like” structure beneath the glass surface, which, however, contains much more defects than a bare fused silica. As a result, the poled regions possess a higher chemical resistance to acidic etchants [1] and lower chemical resistance to alkaline etchants [2].
Poling of glasses with patterned anodic electrodes pressed to the glass surface is often referred to as thermal-electric field imprinting (TEFI) because it enables multiple “imprinting” of the hidden image of anodic electrode surface pattern in the subsurface layer of a glass [13]. The latter makes TEFI procedure similar to lithographic processes. It has been recently demonstrated [3,4] that both acidic (F- ions based) and alkaline (OH- ions based) etching can be employed to form complicated relief structures on the surface of TEFI-processed glasses. The former results in the etching of unpoled glass regions, while the latter preferably etches poled regions of the glass. The TEFI can be described in terms of the formation of a mask reproducing the profile of a patterned anodic electrode in a glass poled with this electrode. Although this mask is unable to provide the contrast of the chemical etching rates comparable with the contrast provided by polymer or metal masking films, its formation does not require photo- or e-beam lithography. It is worth noting that a TEFI mask was also used in plasma-chemical etching of glasses [14], however reported spatial resolution of that process was rather low.
In this paper, we present the first experiments on the application of a “dry” etching technique that is widely used in semiconductor technology reactive ion etching (RIE) [15,16] to thermally poled soda-lime glass. As reported in [17], the RIE rate of fused silica is higher than that of soda-lime glass, and we attempted to use the difference of the RIE rates for the formation of relief structures on the surface of a soda-lime glass poled with profiled anodic electrode. We also compare RIE of the soda-lime glass with acidic chemical etching. Additionally, we consider the influence of thermal treatment of poled glasses on their both RIE and chemical etching and demonstrate the formation of sub-micron relief structures by RIE of a TEFI-processed soda-lime glass.
2. Experimental
In the experiments, we used soda-lime glass slides [18] thermally poled with glassy carbon electrodes in air atmosphere. The composition of the slides is presented in Table 1.
Table 1. Chemical composition of used glass [18]
The poling was performed at 300°C under DC voltage of 300-700 V. Both flat and patterned electrodes were used in the experiments. The patterns presented several sets of periodic grooves with the periods from 200 nm to 80 µm. The patterned electrodes were manufactured using electron beam lithography (Ebeam EBPG 500, Raith Lithography, USA) and the positive resists AR-P 6200 (CSAR 62, Allresist GmbH), then reactive ion etching (Plasmalab System 100, Oxford Instruments, UK) for the chromium mask and finally Plasmalab System 80 for the glassy carbon etching. The relief height of the glassy carbon electrodes varied from 300-400 nm for the patterns with the smallest structures to ∼700 nm for the biggest ones. A SEM image of the glassy carbon electrodes is presented in Fig. 1.
Fig. 1. SEM images (45-degree view) of glassy carbon electrodes with period of 200 nm (a), 600 nm (b), and 2000nm (c). Insert: SEM image shows the relief height in the edge of the pattern. The SEM images were taken after several tens of TEFI processes.
To characterize the anodic electrodes and patterns etched on the surface of the poled glasses we used atomic force microscopy (AFM, Dimension-3100, Veeco, USA), scanning electron microscopy (SEM, Leo 1550 Gemini, Zeiss, Germany) and stylus (Dektak 150, Veeco, USA) and optical (Wyko NT9300, Veeco, USA) profilometers. Several poled glass samples were additionally annealed to understand the influence of this thermal processing on the behavior of the poled glass regions. RIE of the poled and TEFI processed glasses was performed in a plasma etching system (Plasmalab System 80) using CHF3 and Ar with the gas flows of 12 and 38 sccm, respectively, the base pressure of 43 mTorr, and the RF power was 220 W. It should be emphasized that the gas composition we used is a typical one for etching of silica [19]. Correspondingly, the RIE resulted in preferable etching of the poling-modified “silica-like” glass regions, in which the concentration of alkali and alkali earth ions have been greatly reduced by the poling. These are regions of the glass, which were in the direct contact with the flat glassy carbon electrode or with the electrode ridges (see Fig. 2). In contrary, performed at room temperature wet etching of poled and TEFI processed glass samples in acidic polishing agents 100NH4F:800Н2О:7HF49% (in weight) and NH4F:8H2O (in weight) resulted in the preferable etching of the unpoled glass regions, which did not contact with the anodic electrode, in particular, corresponding to the grooves in the electrode. We used NH4F:8H2O for TEFI processed glasses because of slower and, respectively, more accurate etching of the pattern through the mask formed by the poled regions of the glass. The difference between chemical and reactive ion etching of a TEFI processed glass is schematically illustrated in Fig. 2.
Fig. 2. Schematics of thermal-electric field imprinting, “wet” chemical and “dry” reactive ion etching of a glass. Thermal poling: applying 300-700 V DC with profiled glassy carbon anodic electrode to the glass heated to 300°C results in the formation of a hidden structure via deepening of alkali and alkali earth ions. Chemical etching: faster etching of virgin (poling-untouched) regions of the glass slide in acidic polishing agent results in the formation of the reversed electrode profile. Reactive ion etching: faster etching of poling-modified regions of the glass results in the formation of the direct electrode profile.
3. Results and discussion
3.1 Etching of poled glasses
In the experiments, we measured the height of the step between the glass regions under and out of the anodic electrode for poled glass (red dots), for poled and annealed glass (blue triangles) and the height of the step between protected and etched regions of the virgin glass (black squares) as the dependence of the RIE and chemical etching time – see Fig. 3. When performing the measurements under the electrode, we took the region distant by ∼ 300-500 microns from the electrode edge. These dependences allowed evaluating the difference in the etching rate of the poled and unpoled glasses (slope of the curves for the poled glasses), the durations of the etching, which corresponded to the complete removal of “silica-like” subsurface glass layer, and the thickness of this layer. The removal of the poling-modified layer corresponds to the kinks in the curves: 5 min and 13 min for the RIE of the poled and poled and annealed glass, respectively as seen in Fig. 3(a), and 46 min for the chemical etching of the both poled and poled and annealed glass as shown in Fig. 3(b). As awaited according to the data on silica glass and soda-lime glass RIE [17], in our experiments the RIE rate of the poled (modifiers-depleted) glass region exceeded the rate of the unpoled soda-lime glass. The thicknesses of the preferably etched in RIE (“SiO2-like”) layers can be found as the sum of the step heights in the poled and virgin glasses in Fig. 3(a) taken at the kink moment with the account for the initial (after thermal poling) step height. This gives 110 nm and 220 nm for the poled and poled and annealed glasses, respectively. Thus the annealing has increased the thickness of the preferably etched layer twice. Supposedly, this is due to structural relaxation of the poled subsurface region, which makes the glass depleted with metal ions closer to fused silica. This structural relaxation is followed by the relaxation of the poling-induced stress, which can also influence the etching rate of this region. The influence of the structural relaxation (possibly annealing of “defects”) and stress relaxations is seen in the difference in the slopes of the curves for the poled and poled and annealed glasses. The RIE etching rate has become higher after the annealing, for the relaxed poled region has become more close to fused silica. It is worth to note that the increase in the height of the poling-induced step from 75 to 200 nm after annealing shown in Fig. 3(a) is also due to the thermal relaxation of the poled glass volume [11,20], however this relaxation should also include deeper regions of the glass, which contain both glass formers and glass modifiers. In the case of the chemical etching illustrated with Fig. 3(b), the thicknesses of the ether-resistant layers correspond to the differences of step heights in the virgin and poled glasses taken at the kink moment with the account for the initial (after thermal poling) step height: 1960nm and 660 nm for the poled and poled and annealed glasses, respectively. The difference in the slopes of the curves for the poled and poled and annealed samples reflects the difference in the glass etching rate out of the electrode region. This difference can be due to a partial relaxation of the stress out of the electrode region, which is induced by the compressed poled region under the electrode. After the annealing, the compression stress decreases, which results in the increase in the etching rate of the glass near (not under) the anodic electrode region. This is reflected in the slope of the corresponding dependence. The annealing made that rate closer to the etching rate of the virgin glass, however it remains lower. This could be due to incomplete relaxation of stressed regions of the non-poled glass, which depth essentially exceeds the thickness of the poled region [5]. Comparing the data of RIE and chemical etching one can conclude that the thicknesses of formed in thermal poling less resistant layers to RIE and the layers, which are more resistant to acidic chemical etching, differ. This should be due to the properties of the glass regions laying near the lower border of the “silica-like” glass layer where essential fluctuations of elemental composition take place in poled glasses [20].
Fig. 3. Time dependence of the drop at poled/unpoled glass interface before and after additional thermal processing and the height the step at the interface of protected and etched regions of the virgin glass slides resulting from RIE (a) and chemical (b) etching. Poling conditions: 300°C, 700 V, 30 min; thermal processing conditions: 400°C, 50 min; RIE in 12CHF3:38Ar gas composition, chemical etching in 100NH4F:800Н2О:7HF49% (in weight) at room temperature.
One may observe from Fig. 3(a) that the annealing of the poled glass decreases its RIE rate from 22 to 17 nm/min (possibly because of a decrease in the concentration of defects in the subsurface “silica-like” layer of the poled glass), the RIE rate of the virgin glass being 4.9 nm/min. The difference in the etching rates of the virgin, poled, and poled and annealed glass indicates that the surface relief achievable with RIE can be increased by ∼2.3 times by additional annealing of the poled glass. Supposedly, the increase in the thickness of RIE-resistant layer is due to the relaxation of the elastic stress and defects which makes this layer closer to fused silica which durability in RIE is less than the glass durability. Besides, the stress relaxation in glass can also fasten the glass modifiers diffusion at lower interface of the poling modified layer.
To evaluate possible diffusion of the glass modifiers we used the data on the glass modifiers concentration profiles measured with secondary ion mass spectrometry after poling and after subsequent annealing [21]. According to our evaluations, possible shift of the mobile cations cloud towards anodic surface of the glass after the annealing hardly exceeds 50 nm and can be neglected in this consideration. The relation of the wet etching rates of the poled, poled and annealed, and unpoled glasses (43 nm/min, 14 nm/min and 114 nm/min, respectively) in Fig. 3(b) allows achieving by the wet chemical etching the maximum relief height about 20 times deeper than by RIE. The annealing of the poled glasses increases the relief, which can be formed by the chemical etching by ∼1.4 times, that is ∼1.2 µm. Supposedly, this is due to the growth of structural perfection and, respectively, chemical durability of the poled silica-like subsurface glass layer because of the annealing.
3.2 Etching of TEFI glasses
In order to compare performance of the RIE and chemical etching in the processing of TEFI-formed structures in glass, we used glass slides poled using anodic electrode with periodic patterns (see Fig. 1). Characterizing the RIE we measured the height ${h_s}$ of the surface relief in the formed grating after each step of the etching as indicated by the markers in Fig. 4a. One can see that the relief height is about 100-110 nm for grating period from 6 to 80 µm, and that the temporal dependence of the height is almost the same for all structures. Decreasing the structure periodicity results in minor reducing the relief height. Surprisingly the temporal dependences of the etching height are non-monotonous despite the fact that RIE rate for poled regions, which corresponds to the ridges of the profiled electrode, is higher than that for unpoled ones (see Fig. 2). We suppose that the initial decrease of the grating height (the dip at 2.5 min of etching in all the curves in Fig. 4(a)) is due to the presence of a thin layer of strongly hydrogenated glass under the electrode hollows where atmospheric gas discharge takes place [5]. The hydrogen originates from atmospheric water vapors [12], and its higher concentration and, respectively, lower concentration of the glass modifiers, are expected to make this thin layer less durable in RIE then the regions under the electrode ribs, which contain less hydrogen [5]. This statement is based on the data of secondary ion mass spectroscopy, which indicated a thin, about 40 nm half-width, strongly hydrogenated layer beneath the soda-lime poled glass surface. Supposedly, the appearance of this layer is connected with the existence of a leached layer beneath the initial microscope slides surface, which could be due to their interaction with atmosphere during storage time after manufacturing. After etching off that layer, the dip disappears and the relief height monotonically increases until saturation. Besides, penetration of hydrogen in glasses induced by atmospheric poling is not properly studied yet as well as the influence of the penetrated hydroxyl groups on the poled glass properties. Our preliminary experiments allow yet unproved supposing that the essential volume relaxation of the poled glasses can be related with hydrogen release. The presence of hydrogen should also influence glass etching, thermally stimulated depolarization current (charge relaxation) behavior and the second optical harmonic generation. The latter has been reported recently: it was demonstrated that the second harmonic signal from the glasses thermally poled in close and in open anode configuration differed by about 50 times [6].
Fig. 4. Height of the gratings relief during the RIE (a) and the lower bound of the grating height during the chemical etching (b). Periods of the gratings are marked near corresponding curves. TEFI is performed with pressed glassy carbon electrode at 300°C under 300 V, 30 min (a), and at 325°C under 1000 V, 2 min (b), RIE was performed in 12CHF3:38Ar gas composition, chemical etching in NH4F:8H2O (in weight) at room temperature.
Figure 4(b) shows the dependence of the lower bound the surface relief height estimation as a function of chemical etching duration for anodic electrodes with pattern periods of 2 and 20 µm. Specifically, we measured the diffraction efficiency of the gratings using He-Ne laser beam (wavelength λ=632.8 nm) in the course of the etching as described elsewhere [4]. In order to estimate the surface relief we used the following equation for the diffraction efficiency of the thin sine grating [22]:
where ${J_1}$ is the Bessel function, $\Delta \varphi = \frac{\pi }{\lambda }{h_S}\Delta n$, $\Delta n = {n_{glass}} - {n_{etchant}}$ is the difference between indices of the glass ${n_{glass}}$ and of the etchant ${n_{etchant}}$, and ${h_s}$ is the height of the sine grating. Fitting the experimental data allowed us to obtain dependence of the ${h_s}$ on etching time shown in Fig. 4(b). This is lower bound because the height is underestimated: we did not account for the poled soda-lime glass index drop, ∼0.03 [23]. This index decrease and the surface relief affect the grating efficiency the opposite way. At initial stage of the etching, when the relief is low, we evaluate maximal possible underestimation as ∼30 nm, this underestimation gradually decreases with the relief growth and goes to zero when the relief height saturates. Besides, in the case of other glasses, which provide higher index decrease [24] the underestimation can be several times higher.In the calculations, we used ${n_{glass}} = 1.5$ and ${n_{etchant}} = 1.33$, i.e. we assumed that the refractive index of the etchant is close to that of water.
One can observe that ${h_s}$ decreases in the initial stage of the wet etching. This is the manifestation of the glass relaxation in the process of poling. After the thermal poling the poled glass regions relax, and this results in their going down relatively to the unpoled regions. Thus the unpoled regions in the periodically poled glass form hillocks on the glass surface, which diffract the light together with the regions of unpoled glass which index exceeds the index of the poled glass regions. The poled regions are similar to fused silica, and in acidic chemical etching they are more durable than the unpoled ones. Thus, the chemical etching of the unpoled regions is faster. This results first in flattening the glass surface (etching of the hillocks) in the beginning of the etching and the corresponding decrease of the diffraction efficiency. When the glass surface is flat, the light is still diffracted by the refractive index grating formed by the periodicity of poled and unpoled glass regions which indices differ. Zero diffraction efficiency corresponds to the situation when the phase shift by the poled glass region index decrease is equal to the phase shift by the surface relief (grooves) of the unpoled glass, the grooves being placed at the positions where the hillocks were initially placed. Further etching of the unpoled regions leads to the formation of deeper grooves in the glass. Deepening the grooves corresponds to the increase of the grating height and, respectively, the diffraction efficiency. One can see from Fig. 4 (b) that the longer the grating period, the longer time is needed to achieve zero diffraction efficiency. This is because the height of the poling-formed relief of 2-µm period grating is less than that of 20-µm grating and the glass index between the electrode ribs is less than one in 20-µm grating due to additional poling of the glass between the electrode ribs in gratings of higher periodicity [25]. Respectively, the contrast in the glass relaxation of regions below and between the electrode grooves is lower. Comparison of RIE and chemically etched periodic structures evidences almost order of magnitude relief height difference in 20-µm structure and about half of order difference in 2-µm structure. Further decreasing the grating periodicity results in reducing of the relief height in chemically etched structures because of both weaker poling contrast and non-directivity of the chemical etching (etching sidewalls of the structure). For smaller structures, which size is less or comparable with the thickness of the poled “silica-like” layer beneath the glass surface, chemically etched structures have no preferences compared to RIE structures.
In the case of the RIE of sub-micron scale gratings the relief height decreases almost linearly from 110 nm for 2 µm in period grating to ∼5 nm for ∼200 nm in period grating as shown in Fig. 5(a).
Fig. 5. Sub-micron gratings height (a) and dynamics of the glass pattern formation (b) in RIE. TEFI is performed with pressed glassy carbon electrode at 300°C under 300 V, 30 min, RIE in 12CHF3:38Ar gas composition.
An essential difference between the RIE and chemical etching relates to the quality of glass profiling after TEFI. In chemical etching, peculiar stronger and deeper poled regions near the electrode edges result in the formation of “whiskers” near the edges [5] because of their etching-resistance. In RIE, which preferably etches “silica-like” poled glass regions, in the beginning of the process the presence of the stronger poled regions near the anodic electrode edges because of electric field peculiarities and non-relaxed (due to below glass transition poling) elastic stress induced by the poled/unpoled regions interface result in the faster etching and the formation of V-grooves at corresponding positions - see 4 min line in Fig. 5(b). According to the performed modeling [5], this stress is the tensile stress which should fasten the etching. Further RIE (after etching off these peculiar regions) smoothens the deepenings. The smoothing starts when these peculiar regions are completely etched off and the RIE rate at these positions decreased while the etching of “normally” poled regions beneath the anodic electrode is still going on - see 8 - 25 min lines in Fig. 5(b). Thus longer RIE allows the fabrication of high quality structures - see 25 min line in Fig. 5(b). One may conclude that the RIE allows formation of precise relief structures in TEFI glasses, and the RIE smoothens defects of glass poling related to electrode edge peculiarities. Several examples of RIE processed TEFI structures are presented in Fig. 6.
Fig. 6. Images of RIE-patterned glass surfaces. AFM images of 0.6 (a), 2 (b) and 6 µm (c) gratings. Acquired with optical profiler images of 80 µm grating (d) and a piece of a pattern (filter, mixer) for microfluidics (e).
4. Conclusions
Reactive ion etching differs with acidic chemical etching of thermally poled silicate glasses and structures imprinted in the glasses in the height of surface relief, which can be formed: in our experiments chemical etching provided surface relief 20 times deeper than RIE. Additional thermal treatment of the poled glasses allowed ∼2.3 times increase in the surface relief by RIE and ∼1.4 times of the relief by chemical etching. However, the quality of the transfer of an anodic electrode relief pattern to the glasses is higher with RIE because it allows smoothing of peculiar regions near the edges of the anodic electrode pattern. In manufacturing structures with features of hundreds of nanometers size RIE prevails over the chemical etching because of its directivity. Generally, thickening of the “silica-like” layer and search for glass and gas compositions providing higher contrast of RIE rates between poled and unpoled glass region should allow higher relief. We suppose that the developed two-step “dry” process makes it possible to avoid using liquid chemicals in the formation of microrelief patterns on the glass surface.
Funding
Ministry of Education and Science of the Russian Federation (Minobrnauka) (3.2869.2017); Suomalainen Tiedeakatemia (323052).
Acknowledgments
This study was supported by the Ministry of Education and Science of the Russian Federation (project#3.2869.2017) and by Finnish Academy project #323052. AFM characterization were performed using equipment owned by the Federal Joint Research Center “Material science and characterization in advanced technologies” (Saint-Petersburg, Russia).
References
1. W. Margulis and F. Laurell, “Interferometric study of poled glass under etching,” Opt. Lett. 21(21), 1786–1788 (1996). [CrossRef]
2. N. Ikutame, K. Kawaguchi, H. Ikeda, D. Sakai, K. Harada, S. Funatsu, and J. Nishii, “Low-temperature fabrication of fine structures on glass using electrical nanoimprint and chemical etching,” J. Appl. Phys. 114(8), 083514 (2013). [CrossRef]
3. N. Kubo, N. Ikutame, M. Takei, B. Weibai, S. Ikeda, K. Yamamoto, K. Uraji, T. Misawa, M. Fujioka, H. Kaiju, G. Zhao, and J. Nishii, “Nano-imprinting of surface relief gratings on soda-aluminosilicate and soda-lime silicate glasses,” Opt. Mater. Express 7(5), 1438–1445 (2017). [CrossRef]
4. A. N. Kamenskii, I. V. Reduto, V. D. Petrikov, and A. A. Lipovskii, “Effective diffraction gratings by acidic etching of thermally poled glasses,” Opt. Mater. 62, 250–254 (2016). [CrossRef]
5. I. Reduto, A. Kamenskii, A. Redkov, and A. Lipovskii, “Mechanisms and Peculiarities of Electric Field Imprinting in Glasses,” J. Electrochem. Soc. 164(13), E385–E390 (2017). [CrossRef]
6. M. Dussauze, V. Rodriguez, A. Lipovskii, M. Petrov, C. Smith, K. Richardson, T. Cardinal, E. Fargin, and E. I. Kamitsos, “How does thermal poling affect the structure of soda-lime glass?” J. Phys. Chem. C 114(29), 12754–12759 (2010). [CrossRef]
7. S. Ikeda, K. Uraji, T. Suzuki, K. Yamamoto, and J. Nishii, “Migration behavior of alkali and alkaline-earth cations in soda-lime silicate glass surface by electrical nanoimprint,” J. Non-Cryst. Solids 453, 103–107 (2016). [CrossRef]
8. A. Lepicard, T. Cardinal, E. Fargin, F. Adamietz, V. Rodriguez, K. Richardson, and M. Dussauze, “Micro-structuring the surface reactivity of a borosilicate glass via thermal poling,” Chem. Phys. Lett. 664, 10–15 (2016). [CrossRef]
9. P. N. Brunkov, V. G. Melekhin, V. V. Goncharov, A. A. Lipovskii, and M. I. Petrov, “Submicron-resolved relief formation in poled glasses and glass-metal nanocomposites,” Tech. Phys. Lett. 34(12), 1030–1033 (2008). [CrossRef]
10. A. V. Redkov, V. G. Melehin, V. V. Statcenko, and A. A. Lipovskii, “Nanoprofiling of alkali-silicate glasses by thermal poling,” J. Non-Cryst. Solids 409, 166–169 (2015). [CrossRef]
11. I. V. Reduto, V. P. Kaasik, A. A. Lipovskii, and D. K. Tagantsev, “Volume relaxation of poled glasses: surface relief enhancement,” J. Non-Cryst. Solids 499, 360–362 (2018). [CrossRef]
12. C. M. Lepienski, J. A. Giacometti, G. F. Leal Ferreira, F. L. Freire Jr, and C. A. Achete, “Electric field distribution and near-surface modifications in soda-lime glass submitted to a dc potential,” J. Non-Cryst. Solids 159(3), 204–212 (1993). [CrossRef]
13. A. A. Lipovskii, V. G. Melehin, M. I. Petrov, and Yu. P. Svirko, “Thermal electric field imprinting lithography: fundamentals and applications,” chapter 6 in Lithography: Principles, Processes and Materials, Theodore C. Hennessy, ed. (Nova Science Publishers, 2011), pp. 149–163.
14. S. E. Alexandrov, A. A. Lipovskii, A. A. Osipov, I. V. Reduto, and D. K. Tagantsev, “Plasma-etching of 2D-poled glasses: A route to dry lithography,” Appl. Phys. Lett. 111(11), 111604 (2017). [CrossRef]
15. S. J. Pearton, “Reactive Ion Etching of III–V Semiconductors,” Int. J. Mod. Phys. B 8(14), 1781–1876 (1994). [CrossRef]
16. K. Nojiri, Dry Etching Technology for Semiconductors (Springer, 2015).
17. P. W. Leech, “Reactive ion etching of quartz and silica-based glasses in CF4/CHF3 plasmas,” Vacuum 55(3-4), 191–196 (1999). [CrossRef]
18. Soda-lime glass slides composition, http://www.agarscientific.com/microscope-slides.html
19. J. Schmitt, A. Meier, U. Wallrabe, and F. Völklein, “Reactive ion etching (CF4/Ar) and ion beam etching of various glasses for diffractive optical element fabrication,” Int. J. Appl. Glass Sci. 9(4), 499–509 (2018). [CrossRef]
20. A. V. Redkov, V. G. Melehin, D. V. Raskhodchikov, I. V. Reshetov, D. K. Tagantsev, V. V. Zhurikhina, and A. A. Lipovskii, “Modifications of poled silicate glasses under heat treatment,” J. Non-Cryst. Solids 503-504, 279–283 (2019). [CrossRef]
21. D. V. Raskhodchikov, I. V. Reshetov, D. K. Tagantsev, A. A. Lipovskii, and V. P. Kaasik, “Study of charge relaxation in poled silicate glass,” in Book of Abstracts of 5th International School and Conference Saint-Petersburg OPEN 2018, Zh. I. Alferov, ed. (Academic University Publishing, 2018), pp. 400–402.
22. J.W. Goodman, Introduction to Fourier Optics (McGraw-Hill, 1968), xiv + 287.
23. R. Oven, “Measurement of the refractive index of electrically poled soda-lime glass layers using leaky modes,” Appl. Opt. 55(32), 9123–9130 (2016). [CrossRef]
24. M. Dussauze, E. I. Kamitsos, E. Fargin, and V. Rodriguez, “Refractive index distribution in the non-linear optical layer of thermally poled oxide glasses,” Chem. Phys. Lett. 470(1-3), 63–66 (2009). [CrossRef]
25. K. Sokolov, V. Melehin, M. Petrov, V. Zhurikhina, and A. Lipovskii, “On spatially periodical poling of silica glass,” J. Appl. Phys. 111(10), 104307 (2012). [CrossRef]
