Open Access
Add to BibSonomy
Abstract
Reactive ion etching (RIE) is crucial for fabricating high-quality fused silica optics since this technique can be used as a first step before dynamic chemical etching (DCE) for tracelessly removing the fractured defects in subsurface layer. The final quality of the optics is dramatically influenced by the plasma etching condition but still lacks sufficient information for practical application. In this work, combination of RIE and DCE was investigated deeply on polished fused silica surface by changing the gas type and flow rate. We show that the proper choice of fluorine-containing plasma condition during the RIE process allows the simultaneous occurrence of high surface quality and a low concentration of etching-introduced defects on fused silica. This leads to an ultrahigh laser-induced damage threshold at 355 nm while substantially keeping the surface roughness unchanged. This study paves the way for designing and developing a next-generation surface modification ability of high-quality fused silica with the great potential for high-power laser application.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Fused silica optical components have been the workhorse for high-power fusion-class laser facilities, such as the National Ignition Facility [1], the Megajoule laser [2], and so forth. For decades, the fabrication ability of fused silica optics has been continually improved [3–6]. However, there are still risks of laser-induced damage (LID) because sorts of damage precursors are readily left on the fused silica surface during the grinding/polishing processes. HF-based acid etching has been traditionally used as a final step of the finishing process to remove the damage precursors [5,7–9]. After the etching, the Beilby layer and embedded metallic impurities (Ce, Al, Cu, Fe, etc.) induced by the polishing process are effectively removed [10]. Moreover, subsurface damage (SSD) such as narrow scratches and micro-cracks are opened and enlarged [9,11,12]. The laser damage threshold of the optics is thus dramatically enhanced. However, some potential risks of this technique should not be ignored. For example, a huge amount of material removal (20 μm approximately) is needed by wet etching to allow for sufficient mass transfer of the reaction products such as SiF62- away from the silica surface [8,13]. In this case, the optical surface becomes very rough and non-uniform [12]. Moreover, such deep etching treatment is generally time-consuming, laborious and harmful to human body. At last but not least, additional burden on environment is increased since the recycling rate of the HF-acid solution is limited [14]. These risks have made its future uncertain.
In our recent studies [15], we proposed a novel surface modification method of fused silica by combining the reactive ion etching (RIE) with dynamic chemical etching (DCE) techniques. RIE pretreatment, which involves a fluorine-containing plasma dry etching, is used to tracelessly remove the fractured defects as well as the embedded metallic impurities in the surface and subsurface layer of fused silica. DCE retreatment, which involves a HF-acid shallow etching, is used to effectively remove the chemical structural defects such as oxygen deficient center (ODC), non-bridging oxygen hole center (NBOHC) or others produced by plasma etching [16]. This combined etching treatment can dramatically increase the initial damage threshold of the fused silica by 1.5~2 times compared to the traditional HF-acid etching treatment [15,17]. Although the positive impact of pronounced combined etching on fused silica damage threshold at 355 nm is unquestionable, some underlying mechanisms explaining this effect are still elusive. For example, how plasma etching influences the chemical structural defects once the gas type and flow rate are changed? Since reactive gases such as SF6, CHF3 or others have significant influence on plasma parameters, potential distribution, discharge chemistry, and proceeding instabilities [18–21], how they should be chosen to achieve a more efficient combined treatment for fused silica surface? The aim of this paper is to explore these questions.
In this work, a systematical study was implemented to investigate the effect of fluorine-containing plasma on the optical properties of fused silica optics during the combined treatment of RIE and DCE. Different gas types as well as their flow rates were used for the RIE treatment. Section 2 presents the information of experiments, including preparation of samples, characterization of defects and surface quality, and testing of laser-induced damage. Experimental results are given in Section 3, which involves two parts, optical properties of the samples treated by RIE and those treated by combined etching. Finally, the natures of fluorine-containing plasma influencing the optical properties of RIE-treated fused silica and further influencing the optical properties of combined-treated fused silica are discussed. We then bring our conclusions.
2. Experimental
2.1 Samples preparation
Four square high-purity fused silica samples (Corning 7980, 50 mm × 50 mm × 5 mm) were manufactured with conventional grinding and polishing processes using CeO2 particles as the abrasives, followed by a ultrasonic cleaning with alkalescent detergent (Micro 90) and deionized (DI) water at 45°C. After cleaning, the surfaces of these unetched samples were almost free of external contamination and exhibited a relatively high damage threshold at 355 nm. These samples were marked as A, B, C, and D respectively. Sample A was not treated by any etching process and just used as a reference. The other three samples were treated by the combined process of RIE and DCE.
The RIE treatment was performed in a parallel-plate plasma discharge system using ultrapure CHF3, SF6 and Ar (99.999%) as the reactive gases. The experimental setup of RIE is schematically shown in Fig. 1(a). In the chamber, with the aid of the electric field, collisions between electron, Ar, CHF3 and SF6 will generate neutral plasma species, reactive radicals and energetic ions. Fluorine-contained neutral species (such as CF3, CF2) and reactive radicals (such as F-) will react with the silica to form volatile products or fluorocarbon polymer. Energetic ions, which are accelerated by electric field, assist to etch the silica surface anisotropically. In addition, Ar plasma can make the etching more stable and change the electron energy distribution. More detailed information about the RIE mechanism has been provided in our previous publication [22]. The samples were put flat on the center of a round graphite stage to realize a uniform plasma etching treatment. The treated surfaces were regarded as the rear surfaces for the subsequent laser damage testing. Three protocols of RIE were conducted for these samples. All the samples were etched with a 1 μm removal amount of the material. For all the protocols of RIE, the working pressure and input power were 20 mTorr and 200W, respectively. During the etching, there were complex plasma-surface interactions involving chemical etching, physical etching, polymer deposition and enhanced chemical/physical etching. The chemical etching mainly resulted from the reactions between the fluorine-contained radicals and the silica, while the physical etching mainly resulted from the energetic ions bombardment of the substrate surface. After the etching, the reaction chamber was immediately purged with ultrapure Ar for 2 minutes to remove the residual reactive gases. The RIE rates for each protocol are given in Fig. 1(b) and the detailed preparation procedures for the samples are shown in Fig. 2. Comparison between Protocols B and C was mainly to investigate the effect of SF6 on the etching, while comparison between Protocols C and D was mainly to investigate the effect of gas flow ratio Q(Ar):Q(CHF3) on the etching.
After the RIE pretreatment, DCE retreatment was conducted in a double frequency megasonic washing machine (multiMEGt 430 kHz, 1.3 MHz). The DCE process mainly consisted of three steps: weak alkali cleaning, inorganic acid cleaning and HF-based shallow etching. The HF-based shallow etching (3 μm removal) was performed with a mixture of HF (49 wt. %) and NH4F (30 wt. %) at 45°C. After the acid etching, the samples were immediately cleaned using DI water and then exposed to air dry. The cleaning, etching, and drying processes were totally implemented in a Class 100 clean room. A more detailed description of the DCE process is available elsewhere [22].
2.2 Laser-induced fluorescence imaging
Laser-induced fluorescence (FL) imaging technique offers the ability of making high resolution spatial mapping of both FL and scattering (SC) images. FL signal can indicate the fractured defects in the subsurface layer of fused silica since the embedded polishing residues inside the scratches or micro-crack are photosensitive to laser light. SC signal can indicate the open fractured defects on the surface of fused silica since the rough structures on the surface can cause the scattering of the detected light. In this study, a self-developed laser-induced FL microscope was used to obtain the FL and SC signals on the sample surfaces. FL images were detected with the excitation wavelength of 355 nm. SC images were detected with the illumination wavelength of 532 nm. A computer-controlled CCD camera with a high-pass filter in front of it was used to record the signals of the FL and SC light with the wavelength larger than 375 nm. The size of the view field obtained by CCD camera was 0.6mm × 0.6 mm. In this study, 9 mm × 9 mm FL and SC images each comprising 225 stitching sub-images (15 × 15) were respectively obtained by using a home-made imaging platform.
2.3 Fluorescence spectra analysis
FL spectra were used for analyzing the chemical-structural defects such as ODC or NBOHC on the sample surface. FL spectra were detected by a FL spectrometer using a photomultiplier as a detector. The slit widths for excitation and emission were set to be 5 nm and 20 nm respectively. FL excitation was performed at 5.4 eV (~230 nm) using a Xe lamp as the illumination at room temperature. The placement of the samples in the measurement was strictly fixed so that the magnitudes of their FL spectra were comparable.
2.4 Surface quality characterization
Surface quality of the samples after the combined treatment of RIE and DCE was analyzed using an optical microscope and a white light interferometer. Optical microscope was used to obtain the surface morphology after etching and laser damage. White light interferometer was used to analyze the surface morphology and roughness of the samples with the measured area of 0.7 mm × 0.5 mm. For each sample, five areas were chosen randomly on the sample surface to obtain the mean roughness value. Surface errors of the samples were also measured in terms of peak-valley (P-V) values using interferometer. The test area was the full aperture of the sample in surface error measurement.
2.5 Laser-induced damage testing
The laser facility described in our previous work [17] was employed for the laser damage testing on the rear surface of the samples. Briefly, the testing facility comprised a tripled Nd: YAG laser generating laser pulses at 355 nm (3ω) (mono-wavelength configurations) with the repetition rate of 1 Hz. At the focus the laser beam was near-Gaussian shape and the beam diameter (measured at 1/e) was about 1.2 mm. Since the depth of focus was much higher than the thickness of the sample, the diameter of the beam was considered as constant along its propagation route through the sample. The maximum fluence of the laser system was about 25 J/cm2 (with the maximum peak fluence about 66 J/cm2) at 3ω, 7ns.
3. Results
3.1 Optical properties of the plasma-etched samples
3.1.1 FL and scattering microscopy
Figure 3 presents the FL images of the unetched and etched samples (1 μm removal) treated with different gas flow rates. As Fig. 3(a) shows, a large number of scratches can be observed on the unetched sample surface. However, no fractured defects appear to exist on the surfaces of the etched samples (see Figs. 3(b)-3(d)). This indicates that the SSD depth of the polished fused silica optics that we study here was nearly less than 1 μm. The embedded polishing-residues in the subsurface layer can be thoroughly removed during the 1-μm plasma shallow etching (no matter how the gas flow rates change).
Fig. 3 FL images for the unetched and etched samples (1 μm removal) treated with different gas flow rates. (a) Unetched sample, (b) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3):Q (SF6) = 1:14:2, (c) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3) = 1:14, (d) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3) = 10:5. Each image is a 9 mm × 9 mm square stitched by 255 sub-images (15 × 15) with the side length of 0.6 mm.
SC images obtained for the unetched and etched samples treated with different gas flow rates are presented in Fig. 4. From the figures, we can again observe a mass of scratches on the unetched sample surface, which are very consistent with the FL result. Moreover, strong signal of scratches can be observed on the surfaces of etched samples B and C, as Figs. 4(b) and 4(c) illustrate. The scratches on the surfaces of these two etched samples are even more evident compared to those on the unetched sample, indicating that the plasma etching can expose the SSD below the optical surfaces of the samples. However, almost no SC signal can be observed on the surface of sample D. Since the same measured areas of all these three etched samples exhibit no FL signal (as Fig. 3 shows), we believe that protocol B and C tend to expose the fractured defects in the surface of fused silica because chemical reaction dominates the etching under fluorine-enriched condition while protocol D benefits the removal of these defects because physical bombardment dominates the etching under fluorine-deficient condition.
Fig. 4 SC images for the unetched and etched samples (1 μm removal) treated with different gas flow rates. (a) Unetched sample, (b) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3):Q (SF6) = 1:14:2, (c) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3) = 1:14, (d) Etched sample with a mixed gas flow of Q (Ar):Q (CHF3) = 10:5. The measured areas are the same to those shown in Fig. 3.
3.1.2 Surface fluorescence spectra analysis
FL spectra obtained for unetched and etched samples treated with different gas flow rates are shown in Fig. 5. All the spectra have a similar profile where there are two broad FL peaks around 376nm (3.3 eV) and 671 nm (1.85 eV) respectively. Meanwhile, there is a significant increase in the FL intensity for the etched samples with respect to the unetched one. This is probably because chemical structural defects were introduced during the RIE process. It can be further noticed from Fig. 5(a) that the increasing degrees for the etched samples are different from each other. Clearly, the sample etched with protocol B has the highest intensity of FL defects. Compared to this, a little decrease in the FL intensity can be observed for the sample etched with protocol C. We believe it is probably due to the existence of SF6 gas in protocol B since high etching rate is obtained under this condition (see Fig. 1(b)). Protocols B and C provide fluorine-enriched plasma condition, which brings strong chemical etching. And in this case, the bonding structure on the fused silica surface is deteriorated, causing the increase of FL intensity. For the sample etched with protocol D, the increasing degree of the FL intensity is the lowest. That’s because the sample surface undergoes a fluorine-deficiency plasma etching since low etching rate is obtained under this condition (see Fig. 1(b)).
Fig. 5 (a) Comparison of the FL spectra under 230 nm (5.4 eV) excitation for the unetched and etched samples. The dot line presents the three Gaussian components of the spectrum deconvolution for the unetched sample; (b)-(d) Spectrum deconvolution for the etched samples.
It’s also quite obvious from Fig. 5(a) that the FL spectrum has a redshift after etching, which was also observed in our previous experiments [17]. For comparison purpose, the spectrum of each sample was deconvoluted into several Gaussian components, with the Gaussian components of the unetched one centered at 376 nm (3.3 eV), 496 nm (2.5 eV), 620 nm (2.0 eV) and 671 nm (1.85 eV) (see Table 1) respectively. The spectra analysis brings additional information. Etching does not appear to create new kinds of FL defects in the 280-900 nm (1.38-4.43 eV) range. Though the FL intensity increases, the spectra composition becomes less complex after etching. The narrow bands centered at 376 nm (3.3 eV) and 382 nm (3.25 eV) are attributed to the well-known ODC while the broad emission peak centered at about 671 nm (1.85 eV) is ascribed to the typical NBOHC [23]. Peak centered at around 689 nm is probably a result of redshift of the peak centered at 671 nm. We are not very sure what the peaks centered at 496 nm (2.5 eV), 506 nm (2.45 eV) and 620 nm (2.0 eV) attributed to, they are probably associated with the self-trapped exciton (STE) [23].
Table 1. Gaussian band parameters of FL spectra for the unetched sample A: peak position, energy, FWHM, and intensity at each position.
3.2 Optical properties of the combined-etched samples
3.2.1 Laser-induced damage testing
The laser damage probability of the samples treated with different processes is shown in Fig. 6, from which we can obtain many meaningful information. First, the unetched sample exhibits a very low laser-induced damage threshold (LIDT) since the optical surface is full of polishing-induced absorptive defects that would trigger damage. Second, for all the combined-etched samples, there is a remarkable improvement in LIDT compared to the unetched sample. Among them, samples treated with protocols C and D have the highest zero-probability damage threshold, around 25 J/cm2 (peak fluence ~66 J/cm2), which is just the maximum laser fluence of the testing facility (see the red dot line in the figure). By contrast, there is a relatively low improvement in damage threshold for the sample treated with protocol B (see the blue squares). The results seem to tell us that the SF6-containing plasma is not very helpful for the damage threshold enhancement of fused silica.
Fig. 6 Damage initiation probability versus laser fluence at 355 nm, 7 ns for the samples treated with different combined-etching processes. The quality factor (the peak energy/ the average energy) of the laser beam spot was around 2.6 and the beam diameter was about 1.2 mm in the test.
3.2.2 Optical microscope observation
Morphologies of the unetched and etched samples were obtained using optical microscope, as shown in Fig. 7. A molten damage crater surrounded by fractured periphery can be observed on the surface of the unetched sample (see Fig. 7(a)). Near the damage crater, we also notice the existence of splashed sites with very small size (approximately less than 1 μm), which are likely associated with the low-speed molten residues. The damage for the unetched sample is believed to be initiated by low-fluence damage precursors since polishing-introduced scratches and residuals are left on the polished sample surface. The damage morphology of treated sample with RIE of protocol B and DCE is quite different. It can be noticed from Fig. 7(b) that the molten damage crater locates just right on a brittle scratch with the width of several micrometers. Clearly, the splashing energy of the molten material during the damage is very high so that many radial lines around the damage crater are still left. Furthermore, it can be noticed that the material near the edge of the damage crater is plastically deformed, suggesting that the lattice skeleton of SiO2 is soften during the damage accompanied by a strong shock wave and high temperature.
Fig. 7 Surface morphologies of the samples treated with different combined-etching processes. (a) Damage region of the unetched sample, (b) Damage region of the etched sample with 1-μm RIE (Protocol B) pretreatment followed by 3-μm DCE retreatment, (c) Edge region of the etched sample with 1-μm RIE (Protocol C) pretreatment followed by 3-μm DCE retreatment, (d) Edge region of the etched sample with 1-μm RIE (Protocol D) pretreatment followed by 3-μm DCE retreatment.
To further understand the influence of plasma condition on etched surface quality, we also observed the surface morphology on the edge of the two undamaged samples. On the optical surface of the sample RIE-pretreated with protocol C and retreated with DCE (see Fig. 7(c)), there are many brittle scratches which have been passivated and enlarged by HF acid. On the contrast, the sample surface RIE-pretreated with protocol D and retreated with DCE is free of scratches (see Fig. 7(d)). The results suggest that plasma etching condition not only influences the damage performance but also influences the final surface quality of fused silica in the combined treatment of RIE and DCE.
3.2.3 Surface quality measurement
We also investigated the difference in the surface quality for the samples treated with different combined-etching processes via white-light interferometer, as shown in Figs. 8 and 9. For the unetched sample, the surface is very flat and uniform, with the roughness value of only 0.576 nm (see Fig. 8(a)). When the sample undergoes the SF6-containing plasma etching and DCE, the optical surface becomes very rough (RMS = 1.086 nm) and nonuniform. Some exposed scratched can even be identified (see Fig. 8(b)). For the etched sample treated with RIE of protocol C and DCE, periodic trips with some slight scratches can be observed on the surface, as Fig. 8(c) shows. The surface roughness of this sample is around 0.9 nm, which is also significantly higher than that of the unetched sample. The surface of the etched sample with RIE of protocol D and DCE remains flat and uniform. Its roughness value is only 0.655 nm, which is close to that of the unetched sample. These results are quite coincident with those shown in Fig. 4. From the surface errors shown in Fig. 9, the P-V values of the combined etched samples with different RIE protocols are quite close to each other, and they are almost identical to that of the unetched one. The P-V values of all these samples are below 0.4λ. These results indicates that the 1-μm RIE + 3-μm DCE combined etching processes caused no deterioration in surface flatness, and the different RIE protocols showed no difference in remaining the surface flatness.
Fig. 8 Surface morphologies of the samples treated with different combined-etching processes via white-light interferometer. (a) Unetched sample, (b) Etched sample with 1-μm RIE (Protocol B) pretreatment followed by 3-μm DCE retreatment, (c) Etched sample with 1-μm RIE (Protocol C) pretreatment followed by 3-μm DCE retreatment, (d) Etched sample with 1-μm RIE (Protocol D) pretreatment followed by 3-μm DCE retreatment. The surface roughness (RMS) is also given in the figures.
Fig. 9 Surface errors of the unetched samples and combined etched (1-μm RIE + 3-μm DCE) samples with different RIE protocols.
3.2.4 Surface fluorescence spectra analysis
Figure 10 exhibits the FL spectra of the sample surfaces treated with different combined-etching processes. It can be noticed that all the characteristic peaks observed in Fig. 5(a) become much weaker after the RIE-pretreated samples undergoes the 3-μm DCE retreatment. The FL intensities of the etched samples are even lower than that of the unetched sample. Furthermore, the redshift phenomenon shown in Fig. 5(a) does not occur after the DCE retreatment. Instead, the peak positions of the etched samples appear to be consistent with that of the unetched sample. These results make us believe that the RIE-introduced chemical-structural defects are effectively eliminated by the 3-μm DCE retreatment. Another fact Fig. 10 illustrates is that the FL intensities of all the three combined-etched samples are nearly the same to each other. This result further indicates that the 3-μm DCE treatment is enough for completely removing the deterioration layer introduced by the 1-μm RIE pretreatment, no matter what the gas flow rates of the RIE are.
Fig. 10 Comparison of the FL spectra under 5.4 eV excitation for the unetched and combined-etched samples. All the parameters used for the spectra detecting here were completely the same as those in Fig. 5.
4. Discussion
Fluorine-containing plasma etching is ideal for the removal of the damage precursors of high-quality UV fused silica optics. Especially, it can be used as a smoothing process to tracelessly remove the fractured defects in the subsurface layer of fused silica. Gas type and their flow rates are two important factors influencing the surface quality after etching. Figure 3 indicates that the SSD and residues embedded in the layer were effectively removed after the samples were RIE-treated with any protocols studied here. However, the RIE-treated surfaces were significantly different when using different protocols. From the SC images shown in Figs. 4(a) and 4(b), it can be clearly noticed that a great number of scratches were exposed on the RIE-treated surface when using protocols B and C. By contrast, the sample surface RIE-treated with protocol D was totally free of scratches. These differences are mainly due to the concentration change of F radical in the plasma. Under fluorine-enriched plasma condition such as protocols B and C, the concentration of F radicals in the plasma are considerably high, which produces a chemical-reaction-dominated etching. The embedded residues, which have higher reaction rate than the silica matrix, are released and dissolved away from the micro-cracks because of their higher specific areas for the chemical reaction. The fractured physical-structural defects are thus opened and enlarged by the isotropic etching. Under fluorine-deficient plasma condition such as protocol D, the concentration of F radicals in the plasma are relatively low, the energetic ion bombardment thus dominates the etched surface morphology. In the case, the subsurface layer and the embedded residues are removed with nearly the same rates, which generates smooth and flat surface of fused silica.
Fluorine-containing plasma etching can be adopted as a pretreatment process before HF-based acid etching to realize high laser damage threshold on fused silica surface with low roughness. However, it should be noted that not all the plasma conditions have the same effect. Compared with the pure fluorocarbon-containing plasma, SF6-containing plasma seemed not to be a superior candidate for mitigating subsurface damage on ICF silica optics since just a little addition of SF6 in CHF3 plasma caused a limited increase in LIDT (see the blue symbols in Fig. 6). We speculate that the etching of fused silica with SF6-containing plasma might introduce photoluminescence defects since we observed significant increase in FL defects intensity (compared with the counterpart for the pure CHF3 plasma etching) via the FL spectra analysis (see the red line in Fig. 5(a)). In addition, the redshift phenomenon of the spectra peak for the SF6-containing plasma etching was more obvious. Previous studies also revealed that ion bombardment silicon-based material surface during the RIE with SF6 introduced strong PL defects [24,25]. Furthermore, residual fractured scratches after the etching were evidenced to be a potential factor limiting the LIDT improvement since a large proportion of the damage craters were generated just on the brittle scratches (see Fig. 7(b)). The surface microscopy shown in Fig. 8 also indicates that RIE with protocol B has some adverse effects on SSD removal and surface smoothing for fused silica. Under the condition of RIE with protocol C, the combined treatment of 1-μm RIE and 3-μm DCE gave a promising result for strongly enhancing the damage threshold of fused silica optics. However, the surface quality of the sample is unacceptable since numerous scratches were exposed by fluorine-enriched plasma, causing the obvious increase in surface roughness and flatness, as shown in Fig. 4(c). These fracture defects were even enlarged after DCE retreatment, which can be clearly observed in Figs. 7(c) and 8(c). The rough surface may provide a trap for depositing HF-acid etching products such as SiF62- [8]. In contrast, combination of 1-μm RIE (protocol D) and 3-μm DCE can extremely improve the damage performance of the fused silica optics. Compared with the unetched sample, the zero-probability LIDT had a ~2.7 times enhancement (see Fig. 6). Moreover, there was nearly no physical-structural defects on the optical surface, even on the edge of the sample (see the obtained morphology in Fig. 7(d)). The LIDT improvement was mainly attributed to two factors: 1) the SSD was effectively removed by the RIE pretreatment of protocol D, leaving smooth optical surface of fused silica without any etching traces, and 2) the RIE-produced deterioration layer was completely removed by the subsequent 3-μm DCE retreatment. We speculate that the real value of LIDT for this combined-etched sample is even higher than the currently measured one (66 J/cm2) in Fig. 6. However, it is regret that we can’t obtain the corresponding result in this study due to the limitation of maximum fluence of the laser system.
5. Conclusions
We have investigated the important role of fluorine-containing plasma on the surface quality and damage resistance of fused silica during the combined treatment of RIE and DCE. SF6-containing plasma was evidenced to be labored for effectively mitigating the SSD of fused silica. The RIE-introduced defects also give rise to a limitation of LIDT improvement. In the case of CHF3 and Ar mixtured gas, there is a dramatic enhancement in zero-probability LIDT (approaches the maximum fluence of the laser facility) by combining a 3-μm DCE retreatment. However, it is also difficult for fluorine-enriched plasma to tracelessly remove the fractured defects in subsurface layer of fused silica. In contrast, we demonstrated that for the combination of 1-μm fluorine-deficient plasma etching and 3-μm DCE, the treated surface roughness is only 0.655 nm, which is close to that of the polished fused silica. This work therefore indicates that the plasma etching condition such as gas types and their flow rates do necessarily need to be chosen properly during the combined process of RIE and DCE.
The resulting outstanding damage performance, on par with or even above the engineering requirement for ICF, makes extreme interactions between high-power lasers and matter interesting for clearly understanding the damage nature of fused silica optics at high-power UV nanosecond lasers.
Funding
National Natural Science Foundation of China (NSFC) (61705204, 61705206, and 61805221).
Acknowledgments
The authors would like to thank Dr Shufan Chen for helpful discussions.
References
1. E. I. Moses, “Ignition on the National Ignition Facility: a path towards inertial fusion energy,” Nucl. Fusion 49(10), 104022 (2009). [CrossRef]
2. J. L. Miquel, C. Lion, and P. Vivini, “The Laser Mega-Joule: LMJ & PETAL status and program overview,” J. Phys. Conf. Ser. 688, 012067 (2016). [CrossRef]
3. F. Shi, Y. Zhong, Y. Dai, X. Peng, M. Xu, and T. Sui, “Investigation of surface damage precursor evolutions and laser-induced damage threshold improvement mechanism during Ion beam etching of fused silica,” Opt. Express 24(18), 20842–20854 (2016). [CrossRef] [PubMed]
4. J. Neauport, C. Ambard, H. Bercegol, O. Cahuc, J. Champreux, J. Charles, P. Cormont, N. Darbois, P. Darnis, and J. Destribats, “Optimizing fused silica polishing processes for 351nm high-power laser application,” in Boulder Damage Symposium XL Annual Symposium on Optical Materials for High Power Lasers (International Society for Optics and Photonics, 2008), pp. 71321I. [CrossRef]
5. S. Baxamusa, P. E. Miller, L. Wong, R. Steele, N. Shen, and J. Bude, “Mitigation of organic laser damage precursors from chemical processing of fused silica,” Opt. Express 22(24), 29568–29577 (2014). [CrossRef] [PubMed]
6. F. Shi, Y. Tian, X. Peng, and Y. Dai, “Combined technique of elastic magnetorheological finishing and HF etching for high-efficiency improving of the laser-induced damage threshold of fused silica optics,” Appl. Opt. 53(4), 598–604 (2014). [CrossRef] [PubMed]
7. X. Ye, J. Huang, H. Liu, F. Geng, L. Sun, X. Jiang, W. Wu, L. Qiao, X. Zu, and W. Zheng, “Advanced mitigation process (AMP) for improving laser damage threshold of fused silica optics,” Sci. Rep. 6(1), 31111 (2016). [CrossRef] [PubMed]
8. T. I. Suratwala, P. E. Miller, J. D. Bude, W. A. Steele, N. Shen, M. V. Monticelli, M. D. Feit, T. A. Laurence, M. A. Norton, C. W. Carr, and L. L. Wong, “HF-based etching processes for improving laser damage resistance of fused silica optical surfaces,” J. Am. Ceram. Soc. 94(2), 416–428 (2011). [CrossRef]
9. J. Neauport, C. Ambard, P. Cormont, N. Darbois, J. Destribats, C. Luitot, and O. Rondeau, “Subsurface damage measurement of ground fused silica parts by HF etching techniques,” Opt. Express 17(22), 20448–20456 (2009). [CrossRef] [PubMed]
10. J. Bude, C. W. Carr, P. E. Miller, T. Parham, P. Whitman, M. Monticelli, R. Raman, D. Cross, B. Welday, F. Ravizza, T. Suratwala, J. Davis, M. Fischer, R. Hawley, H. Lee, M. Matthews, M. Norton, M. Nostrand, D. VanBlarcom, and S. Sommer, “Particle damage sources for fused silica optics and their mitigation on high energy laser systems,” Opt. Express 25(10), 11414–11435 (2017). [CrossRef] [PubMed]
11. B. Da Costa Fernandes, M. Pfiffer, P. Cormont, M. Dussauze, B. Bousquet, E. Fargin, and J. Neauport, “Understanding the effect of wet etching on damage resistance of surface scratches,” Sci. Rep. 8(1), 1337 (2018). [CrossRef] [PubMed]
12. L. Wong, T. Suratwala, M. Feit, P. Miller, and R. Steele, “The effect of HF/NH4F etching on the morphology of surface fractures on fused silica,” J. Non-Cryst. Solids 355(13), 797–810 (2009). [CrossRef]
13. W. Wang, P. Lu, L. Han, C. Zhang, C. Yang, R. Sui, and J. Chen, “Diffusion behavior of ammonium group and its interaction mechanisms with intrinsic defects in fused silica,” Appl. Phys., A Mater. Sci. Process. 122(10), 929 (2016). [CrossRef]
14. M. Pfiffer, P. Cormont, E. Fargin, B. Bousquet, M. Dussauze, S. Lambert, and J. Néauport, “Effects of deep wet etching in HF/HNO<sub>3</sub> and KOH solutions on the laser damage resistance and surface quality of fused silica optics at 351 nm,” Opt. Express 25(5), 4607–4620 (2017). [CrossRef] [PubMed]
15. L. Sun, J. Huang, H. Liu, X. Ye, J. Wu, X. Jiang, L. Yang, W. Zheng, and W. Wu, “Combination of reaction ion etching and dynamic chemical etching for improving laser damage resistance of fused silica optical surfaces,” Opt. Lett. 41(19), 4464–4467 (2016). [CrossRef] [PubMed]
16. S. J. Fonash, “An Overview of Dry Etching Damage and Contamination Effects,” J. Electrochem. Soc. 137(12), 3885–3892 (1990). [CrossRef]
17. L. Sun, J. Huang, T. Shao, X. Ye, Q. Li, X. Jiang, W. Wu, L. Yang, and W. Zheng, “Effects of combined process of reactive ion etching and dynamic chemical etching on UV laser damage resistance and surface quality of fused silica optics,” Opt. Express 26(14), 18006–18018 (2018). [CrossRef] [PubMed]
18. A. K. Paul, A. K. Dimri, and R. P. Bajpai, “Plasma etch models based on different plasma chemistry for micro-electro-mechanical-systems application,” Vacuum 68(2), 191–196 (2002). [CrossRef]
19. L. Lallement, C. Gosse, C. Cardinaud, M.-C. Peignon-Fernandez, and A. Rhallabi, “Etching studies of silica glasses in SF6/Ar inductively coupled plasmas: Implications for microfluidic devices fabrication,” J. Vac. Sci. Technol. A 28(2), 277–286 (2010). [CrossRef]
20. P. Chabert, G. Cunge, J.-P. Booth, and J. Perrin, “Reactive ion etching of silicon carbide in SF6 gas: Detection of CF, CF2, and SiF2 etch products,” Appl. Phys. Lett. 79(7), 916–918 (2001). [CrossRef]
21. D. Bose, M. Rao, T. Govindan, and M. Meyyappan, “Uncertainty and sensitivity analysis of gas-phase chemistry in a CHF3 plasma,” Plasma Sources Sci. Technol. 12(2), 225–234 (2003). [CrossRef]
22. L. Sun, H. Liu, J. Huang, X. Ye, H. Xia, Q. Li, X. Jiang, W. Wu, L. Yang, and W. Zheng, “Reaction ion etching process for improving laser damage resistance of fused silica optical surface,” Opt. Express 24(1), 199–211 (2016). [CrossRef] [PubMed]
23. R. Salh, “Defect related luminescence in silicon dioxide network: a review” in Crystalline Silicon – Properties and uses, (INTECH Open Access Publisher, 2011).
24. I. A. Buyanova, A. Henry, B. Monemar, J. L. Lindstrom, and G. S. Oehrlein, “Photoluminescence of defects induced in silicon by SF6/O2 reactive-ion etching,” J. Appl. Phys. 78(5), 3348–3352 (1995). [CrossRef]
25. A. Belkacem, E. André, J. C. Oberlin, C. Pomot, B. Pajot, and A. Chantre, “Electronic defects induced in silicon by SF 6 plasma etching,” Mater. Sci. Eng. B 4(1–4), 451–455 (1989). [CrossRef]
