- Email: [email protected]
Removal of antireflection sol-gel SiO2 coating based on Ar ion beam etching
Fusion Engineering and Design 156 (2020) 111578
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Removal of antireflection sol-gel SiO2 coating based on Ar ion beam etching a
a
b
b
a
T
a
Xiaolong Jiang , Wei Liao , Bo Li , Xia Xiang , Xiaodong Yuan , Chuanchao Zhang , Xiaoyu Luana, Xiaodong Jianga,* a b
Laser Fusion Research Center, China Academy of Engineering Physics, Mianyang 621900, China School of Physics, University of Electronic Science and Technology of China, Chengdu 610054, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Inertial confinement fusion Sol-gel antireflection coating Tritium removal Coating stripping Ion beam etching
We proposed a new stripping method of sol-gel SiO2 coating widely used in inertial confinement fusion facilities based on Ar ion beam etching, during which the coating is removed in the form of gaseous and solid particle. This dry approach is more environment-friendly than the conventional wet cleaning process for saving the trouble of treating tons of polluted water, especially highly toxic tritiated water. In the IBE process, porous coating is sputtered away at a constant etching rate just like other homogeneous materials. IBE parameters are systematically optimized to obtain a selective and gentle etching. The weakly bonded coating is etched 6 times faster than the fused silica substrate, which ensures the selective sputtering of coating over the underlying fine optics. Experiments on different off-line samples as well as the actual recycled optics from high power laser system show that this method is very effective. Moreover, this stripping process does not degrade the optic’s performance. Surface shape and roughness barely change after stripping. The laser damage threshold increases slightly, probably due to the removal of re-deposition layer introduced during polishing process. This method has been successfully adopted to process the optics of ICF facility in China.
1. Introduction It is well known that in the high-power systems like National Ignition Facility (NIF) [1], Laser MégaJoule (LMJ) [2] and Shenguang (SG) series facilities [3], almost all UV optics are coated with sol-gel silica to reduce interfacial reflection [4–6]. After serving for a period of time, the optics would inevitably get laser damaged [7–9]. When it is determined that a particular optic needs to get damage mitigated, it is removed from the laser and transported to mitigation facilities. There, the optic is surveyed for tritium and other hazardous contaminations that it may have been exposed to in the target chamber and, if necessary, decontaminated [10]. Next, stripping of sol-gel coating is conducted, followed by recyclability assess (in term of surface quality, surface damage, and deeper bulk damage) and CO2 laser mitigation (if necessary) [11–14]. Conventionally, sol-gel coating is stripped using an ultra-sonic wet cleaning system, as shown in Fig. 1. Even though this method has proved to be effective in practice, its weaknesses come out too. First, it produces tons of waste water. The wet cleaning process consists of four procedures: deionized water rinse, ultrasonic cleaning using alkaline solution, ultra-sonic cleaning using deionized water, and slow lifts dry. Counting the extra water needed to flush the evacuated sinks, pipe and
⁎
percolator, one wet cleaning process produces at least 440 liters wastewater. More importantly, when the coating is contaminated with tritium, effluent disposal becomes tricky. The tritium absorbed within the porous coating would enter into the cleaning solution and make it tritium contaminated. Tritiated water, however, is the least preferred form of tritium [15]. Due to the body’s ready adsorption of tritium in the form of tritiated water, exposure to tritiated water is on the order of 10,000 times more hazardous than exposure to gaseous tritium. Tritium in the form of T2O is difficult to store for long periods due to its corrosive properties. Tritiated water is also much more difficult to dispose compared to gaseous or solid tritium. De-tritium before stripping does not make so much difference, either. In order to avoid any modification of the fine optics, gently heating the optics is one of the few remaining decontamination methods. Our experiments on de-tritiated Vacuum Isolator (VI, an UV optics, details about VI can been found in section 5.2) from SG series facilities show that tritium would accumulate in the solution, and its concentration would exceed regulatory limits after several strips. Gaseous or solid tritium, on the other hand, is easily handled and routinely disposed: their use, transfer, storage, and shipment have a long history and have been safely used for decades [16–18]. Therefore, exploring for a dry approach in which tritium is removed in the form of gas or solid is necessary.
Corresponding author. E-mail address: [email protected] (X. Jiang).
https://doi.org/10.1016/j.fusengdes.2020.111578 Received 5 November 2019; Received in revised form 1 February 2020; Accepted 20 February 2020 Available online 27 February 2020 0920-3796/ © 2020 Published by Elsevier B.V.
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 1. The ultra-sonic cleaning system used in conventional wet stripping process. (a) The overall appearance. (b) Inner details.
has a diameter of ∼1.1 m and a height of 1.1 m (inner size), with three molecular pumps (FF-250 from KYKY Technology Co.) to pump. The ion gun used is a self-developed 6 cm × 66 cm radio frequency (RF) ion beam source shooting flat-roofed ion beam. Inert gas Ar is used as the etching gas. With the long ion beam source and substrate carrier that moves back and forth along the widthwise direction, as illustrated in Fig. 2(b), this plant is capable of uniformly processing substrate of aperture up to semi-meter. For 430 mm × 430 mm aperture substrate, as in this case, only the middle 430 mm part of the ion beam is used. A beam slot, i.e. a fused silica plate with a 430 mm × 60 mm rectangle window located in the middle, is employed to ensure that only the substrate area is reached by ion beam. This is helpful for avoiding unfavorable sputtering of the substrate carrier or other chamber parts. The uniformity (along longwise direction) of beam current intensity of the unshielded part is rms ∼ 3% and PV∼ ± 6%, measured by Faraday cup. In this study, incidence angle is fixed 0°, i.e., perpendicular to substrate surface. Fused silica plates (Corning 7980 and Heraeus 312) were ground and polished using conventional ceria polishing techniques. Corning 7980 plates (50 mm × 50 mm × 4 mm) are used in most of the experiments. VI from the ICF facility of China (Heraeus 312, 430 mm × 430 mm × 10 mm) is used for comparison tests of surface shape and more importantly testifying the applicability of the proposed approach on actual optics. Smaller Heraeus 312 plates (100 mm × 100 mm × 10 mm) are used for comparison study of laser damage threshold. Substrates were dip coated with 73 nm thick sol-gel coating. Before that, all substrates were ultrasonically cleaned, unless otherwise specified. Etching depth is measured by stylus surface profilometer (NanoMap-500LS, AEP technology Inc.). Spectrophotometer (Lambda 950, PerkinElmer Inc.) was utilized to characterize the transmittance before and after stripping. Transmittance of VI is measured using large aperture spectrophotometer. Surface roughness is measured using Atomic Force Microscopy (AFM, Dimension Icon, Bruker Inc.) and optical surface profiler (FPM-1000, ZYGO Inc.). Surface shape is characterized by phase shift laser interferometer (VerifireTM MST, ZYGO Inc.). Laser damage thresholds were tested using A Q-switched Nd: YAG laser (SAGA, Thales Laser Inc.). It provides 355 nm, 6.8 ns laser pulse (temporal and spatial profile Gaussian, beam waist 1/e2 diameter 0.47 mm, modulation depth 2.2).
In this work, we proposed a new sol-gel coating striping method based on Ar ion beam etching. Ion beam etching (IBE) is a dry etching technique in which energetic ions are generated and then impinge a target to remove surface material [19,20]. Due to its atomic-scale removal capacity, non-contacting mode and good repeatability, IBE finds well-established uses in fields like ultra-smooth polishing [21,22], highprecision figuring [23,24] and micro/nano-patterning [25–27]. As to coating stripping, IBE is promising for two reasons. First, physical ion bombardment with energy of hundreds of eV could easily sputter away the coating as well as the absorbing contaminations. The etching products generally take the form of volatile plasma, or less-volatile cluster, or most likely both. In this case, the etching products theoretically consist of gaseous O2, T2 and solid Si, Si-O, Si-T. This means no liquid waste is generated, as opposed to the wet approach. The volatile (gas and small particles) could be pumped out of the chamber by vacuum pump and the non-volatile cluster would deposit on the chamber floor. If etching product gather/disposal is required, the volatiles can be collected through off-gas filtration and non-volatiles through regular cleaning of the chamber. Second, the non-contact IBE process barely modify the fine optics. Since sol-gel coating is practically a bunch of weakly bonded nano-size particles, it would theoretically have a much higher sputtering rate than the fused silica substrate. This etching selectivity is beneficial for avoiding over-etching of the substrate. The main contributions of this work are the proposal, technical design and feasibility assess of the IBE based sol-gel coating stripping method. Feasibility is systematically studied from the perspectives of effectiveness, applicability and potential side-effects. And the overall outcomes are inspiring. This paper is organized as follow: first, the stripping process of solgel silica coating under ion beam etching is studied. Then, processing parameters such as bombardment energy and ion current are optimized to obtain a selective (selectively etching the coating over substrate), slow (to control etching depth), and damage free (small implant depth of Ar + into substrate matrix) etching. After the optimum parameters are determined, stripping experiments are conducted on both off-line and on-line samples. Effectiveness of this method is experimentally confirmed based on transmission rate tests and optical microscope characterizations. The possible negative modifications this method might introduce to the fine optic are also investigated. Comparison tests on surface roughness, surface shape and laser damage resistance show that no performance degradation occurs. So far, this technique has been successfully adopted to strip the recycling optics of SG series facilities in China.
3. Stripping process of sol-gel coating Stripping experiments are conducted to understand the stripping process of sol-gel coating under ion beam etching. Dipped coated substrates are one-side etched for different durations, followed by transmittance tests and morphology characterizations using AFM and optical surface profiler. Fig. 3(a) shows the experimental results of transmittance spectra as function of IBE durations. As etching proceeds, the
2. Experimental equipment and method A large aperture IBE system with rectangle ion beam source (see Fig. 2(a)) has been developed. The main vacuum chamber of the plant 2
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 2. (a) The large aperture ion beam etching plant used in this research; (b) schematic diagram of the IBE configuration.
substrate. On the other hand, since the particle is only 20 nm in diameter, i.e., much smaller than the wavelength (351 nm), these detailed structures within one wavelength (note that the scan size of the AFM image is only 500 nm × 500 nm) would be harmless, if not beneficial [29]. Given above facts, roughness evolution is characterized using optical surface profiler (the scan size is 0.7 mm × 0.5 mm) instead of AFM. In fact, optical surface profiler is used as the standard, certificated roughness measurer of all UV optics in the SG series facilities. Fig. 5 shows the optically obtained surface morphologies of samples corresponding to Fig. 4. It is obvious that no roughening occurs after stripping.
curve peak shifts meanwhile the overall transmittance gradually drops. These apparently result from the thinning of sol-gel layer. After 5 min, transmittance curve remains unchanged and overlaps with that of one side coated substrate, indicating a clean removal. Etching rates of solgel coating under such etching recipe is calculated from the depth of steps ion-etched through polyimide mask, assuming that the coating has a constant IBE etching rate just like conventional homogeneous materials such as silica and silicon. The obtained etching rate is 16 nm/min, accordingly, the theoretical stripping time needed is 4.56 min (73 nm / (16 nm/min) =4.56 min), which coincides with the 4−5 min given by transmittance spectra. We further theoretically simulate the time evolution of transmittance spectra under 16 nm/min etching using software TFCal [28], as shown in Fig. 3(b), and compare it with the experimental results. It is clearly seen that the simulation results exhibit the same pattern as experimental observations. This confirms the assumption that porous coating is etched at a constant rate, even though it is not homogeneous matrix but porous cluster of weakly bonded nano-size particles. Fig. 4 shows the AFM images for different etching durations. For unetched surface, shown in Fig. 4(a), nano-particles of silica can be clearly seen. As etching proceeds, the contour of nano-particle gradually fades away. Especially, according to Fig. 4(f), after the coating has been removed, the porous morphology of the coating layer is partly transferred into the substrate, making the surface slightly rougher than the original bare substrate. This roughness depends on the integral thickness fluctuation of the coating layer, which is related to the size of the nanoparticles, as well as etching selectivity of coating over fused silica
4. Parameters optimization 4.1. Ion energy In order to remove the coating without hurting the underlying substrate, over etching of the underlying substrate should be restrained. One might imagine simply turning off the ion beam after the coating has been etched through. This requires a good endpoint detection system, for instance an in-situ transmittance inspection configuration. Another approach is improving the coating to substrate selectivity of the ion beam process, i.e. the ratio of sputtering yield of coating to that of substrate. Sputtering yield is defined as the number of target atoms ejected per incoming ion. In IBE, the near-surface particles are removed from the target by cascaded bombardment. When a cascade gives a
Fig. 3. The transmittances curves of coated substrate IBE processed for different durations (a) experimental results (b) simulation results. 3
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 4. AFM images of coated surface stripped for different durations: (a) 0 min; (b) 1 min; (c) 2 min; (d) 3 min; (e) 4 min; (f) 5 min. The scan size is 500 nm × 500 nm.
through polyimide and coating masks, as shown in Fig. 6(a). The lower part of fused silica substrate is dipped coated and the left part is covered with polyimide type. Polyimide is known for thermal stability, good chemical resistance and excellent mechanical properties, and can protect the substrate from being etched. Apparently, step 1 is formed between etched and un-etched areas and can be used to calculate the etching rate of fused silica. The sol-gel coating, on the other hand, can only temporally shield the substrate; therefore, once the coating is completely removed, the depth of step 2 would remain unchanged. This
target particle an energy greater than the "surface binding energy" of that target, the particle may be sputtered. Accordingly, one would expect that sputtering yield is subject to the following factors: ion type, target material, incidence angle and ion energy [20]. The first three factors, in our case, are constants; therefore, optimization of ion energy is necessary. Fig. 6 shows the experimental configuration and obtained etching selectivity dependence on Ar+ energy. Ion current is fixed at 300 mA. Etching rates were calculated from the depth of steps ion-etched
Fig. 5. Optical surface profiler images of coated surface stripped for different durations: (a) 0 min; (b) 1 min; (c) 2 min; (d) 3 min; (e) 4 min; The scan size is 0.7 mm × 0.5 mm. 4
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 6. (a) Experimental configuration used for etching rate determination (b) Experimental results of etching rates and selectivity as function of ion energy.
constant depth represents the equivalent “fused silica” thickness of solgel coating under ion beam etching. Etching selectivity is actually the ratio of coating thickness (73 nm) to the depth of step 2. Based on the obtained etching rate of fused silica and etching selectivity, etching rate of coating is further calculated. Experimental results are presented in Fig. 6(b). It is clearly seen that the coating consistently has a higher etching rate than substrate. Etching rate of coating increases rapidly from 15.2 nm/min at 200 eV to 145.6 nm/min at 1000 eV, while that of substrate increases more slowly, only from 4 nm/min to 22 nm/min. Consequently, etching selectivity gradually increases from 3.6 to 6.5. Since the density of sol-gel coating is about half of that of fused silica, the high etching selectivity should be partly attributed to the weakly bonded nature of the coating. Similarly, one would also expect a selective sputtering of contaminations over substrate because contaminations are generally also weakly bonded with each other or with the coating. It should be mentioned that ion energy beyond 1000 eV is not taken in consideration here because at that energy region, ion implant is no longer negligible. Ion implant, after certain degree, may damage the target lattice and impact the laser damage resistance. In order to study this issue, target damage is simulated using software SRIM [30,31]. Fig. 7(a) illustrates the Ar + distribution in fused silica at different ion energies. The projected range (depth of the peak concentration), as summarized in Fig. 5(b), increases with ion energy, whereas is only in
the order of several nanometers. Given these facts, ion energy is chosen to be 800 eV, at which the etching selectivity is 6.0, very close to the maximum 6.5 at 1000 eV. The projected depth is 3.7 nm, which does not impact the laser damage resistance as would be shown later. 4.2. Ion current Ion current is another important parameter. With ion energy given, ion current directly determines the etching rate, i.e., the time efficiency and controlling precision of the stripping process. The relationships between ion current and etching rate of coating and substrate are experimentally studied, as shown in Fig. 8(a). Ion energy is fixed at 800 eV. It is clearly seen that both etching rates increase linearly with ion current. This coincides with the theory that sputtering amount is proportion to the time integral of ion current. Considering that sol-gel coating is only 73 nm thick, ion current of 100 mA is feasible. The etching rates of coating and substrate at such ion current are 34.5 nm/ min and 5.8 nm/min, respectively. Hence, processing duration of no less than 2.2 min is required. Given that the contaminations absorbed on or within the coating might need extra time, 3 min in practice is feasible. As shown in Fig. 8(b), the scan length for 430 mm × 430 mm aperture optics is 540 mm: 430 mm for covering the substrate, 60 mm for the width of ion beam and extra 2 × 25 mm for alignment tolerance. That is 9 times of the width of ion beam (60 mm), hence the stripping
Fig. 7. (a) Ar+ distributions in fused silica as function of ion energy. (b) Projected depth of Ar+ in fused silica as function of ion energy. 5
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 8. (a). Experimental results of etching rates of sol-gel coating and fused silica substrate as function of ion current; (b) schematic of scanning etching.
scratches and pits on the coating. Sample #3 was poorly pre-cleaned, consequentially after coating, numerous dust particles are enclosed. After IBE stripping, however, all three samples exhibit clean and smooth surface, as in Fig. 9(b), (d) and (f). Results of transmittance tests are shown in Fig. 10. As we can see, transmittance curves of the three stripped samples overlap with that of bare fused silica, suggesting a clean removal. According to surface roughness tests, roughness (Rq) for sample #1 - #3 are 0.737 nm, 0.654 nm and 0.826 nm respectively, barely differing from the original 0.795 nm. These suggest that the proposed IBE stripping approach is broadly suitable for different kinds of sol-gel coatings.
time needed is 27 min. In circumstance higher depth-control precision is required, one can slow down the process by simply reducing the ion current. Theoretically, the control precision of IBE process is in atomic scale. 5. Effectiveness tests in practice 5.1. Applicability for defective coating In previous experiments, all the coatings are meticulously prepared and carefully handled, i.e. almost perfect. For the actual optics to be recycled, however, that is normally not the case. In this section, we further investigate the applicability of proposed approach on imperfect coatings. In Fig. 9, three typically defective coatings are shown. In sample #1, a CO2 laser mitigated crater causes nonuniform distribution of coating surrounding the crater, as indicated by the nonuniform brightness caused by nonuniform interfacial reflection of the lamplight. For sample #2, careless storage and transportation creates many
5.2. Application for actual recycled optic in ICF facility In order to further test the applicability of the proposed approach on actual UV optics, stripping experiments are performed on a VI from SG series facilities. Vacuum Isolator (also called Vacuum Window) is an UV optics locating between the Final Optics Assembly (FOA, consists of
Fig. 9. Optical microscope pictures of defectively coated sample #1, #2 and #3 taken before and after IBE stripping. 6
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 12. Transmittance spectra of samples of different treatments.
Fig. 10. Transmittance spectra of three stripped samples and bare fused silica. The overlapping of transmittance curves indicates a clean removal.
five tested sites. It is clearly seen that parameters of surface shape and roughness hardly change, which is reasonable since the coating is only etched for dozens of nanometers and over-etching is even less.
several UV optics) and the target chamber [32]. It has the dimensions of 430 mm × 430 mm × 10 mm. In SG facility, the FOA general operates at 1000 Pa while the target chamber at 10−3 Pa. VI is used to isolate the target chamber from the low-vacuum FOA, hence the name Vacuum Isolator. It has been previously CO2 laser mitigated so several mitigation craters with diameters ranging from 0.52 mm to 2 mm are present. For both incident and exit surfaces, five randomly selected sites are tracked by coordinates. Surface quality and roughness of each site are characterized before and after stripping. Surface shape and transmittance spectra are also tested. Fig. 11 shows some of the comparisons of surface images before and after stripping obtained using optical microscopy, thereinto, Fig. 11(a) is incident surface and Fig. 11(b) the exit surface. Note that after serving in the harsh laser environment for several months, the coating becomes rough and a few particles can be observed, as in Fig. 11(a). For the exit surface, some parts of the coating are even wrinkled, as in Fig. 11(b). After IBE sputtering, the rough coating and the contamination are completely removed. As shown in Fig. 12, transmittance curves of the five tracked sites overlap with that of bare fused silica, just like previous observations. Surface shape and roughness before and after stripping are summarized in Table 1. Each value of roughness and PSD2 is averaged over
5.3. Impact on laser damage resistance The potential impact the stripping approach may have on UV laser damage resistance is also investigated. Fig. 13 presents the results of 355 nm 1 on 1 laser damage thresholds (LDT) test on bare fused silica and IBE stripped substrate. For each laser fluence, 20 spots are shot to calculate the damage possibility. The distance of the adjacent spots is 4 mm. In order to minimize the test error caused by the difference of the samples, the substrates we used are cut from the same glass block, and ground and polished in the same batch. It is clearly that laser damage resistance improves slightly after stripping, with LDT increasing from 10 J/cm2 to 11 J/cm2. This is probably due to the over-etching of redeposition layer introduced during polishing, which is generally dozens of nanometers thick and identified as one of the most important laser damage precursors.
Fig. 11. Comparisons of optical microscopically obtained surface pictures of a VI from SG series facility before and after stripping: (a) Incident surface; (b) exit surface. 7
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Table 1 Surface shape and roughness of a VI from China ICF facility before and after IBE stripping. Treatment
Standard Before strip After strip
Roughness (nm)
PSD2 (nm)
0° transmission wave front (λ = 632.8 nm)
Incident surface
Exit surface
Incident surface
Exit surface
Defocusing PV
Astigmatism PV
GRMS (λ/cm)
PSD1 (nm)
1.2 0.700 ± 0.043 0.810 ± 0.056
1.2 0.440 ± 0.021 0.606 ± 0.038
1.1 0.580 ± 0.105 0.626 ± 0.064
1.1 0.430 ± 0.037 0.479 ± 0.044
0.425 0.011
0.220 0.048
1/120 1/131
1.8 1.356
0.049
0.053
1/129
1.329
Conceptualization. Chuanchao Zhang: Writing - review & editing. Xiaoyu Luan: Investigation. Xiaodong Jiang: Supervision. Acknowledgments The authors wish to thank Chunyuan Hou, Chengxiang Tian for their help with IBE experiments, and Yi Yang, Bin Hen for their help with characterization of roughness and surface contour, and Zhengkun Liu, Yin Liu for their help with AFM tests. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.fusengdes.2020. 111578. References
Fig. 13. Results of 1-on-1 laser damage threshold test before and after stripping.
[1] M.L. Spaeth, K.R. Manes, M. Bowers, P. Celliers, J.M.D. Nicola, et al., National ignition facility laser system performance, Fusion Sci. Technol. 69 (2016) 366–394. [2] N. Fleurot, C. Cavailler, J.L. Bourgade, The Laser Mégajoule (LMJ) project dedicated to inertial confinement fusion: development and construction status, Fusion Eng. Des. 74 (2005) 147–154. [3] W.G. Zheng, X.F. Wei, Q.H. Zhu, F. Jing, D.X. Hu, J.Q. Su, K.X. Zheng, X.D. Yuan, H. Zhou, W.J. Dai, W. Zhou, F. Wang, D.P. Xu, X.D. Xie, B. Feng, Z.T. Peng, L.F. Guo, Y.B. Chen, X.J. Zhang, L.Q. Liu, D.H. Lin, Z. Dang, Y. Xiang, X.W. Deng, Laser performance of the SG-xx laser facility, High Power Laser Sci. Eng. 4 (2016) e21. [4] S.I. Huang, Y.J. Shen, H. Chen, Study on the hydrophobic surfaces prepared by twostep sol–gel process, Appl. Surf. Sci. 255 (2009) 7040–7046. [5] X.S. Zou, C.Y. Tao, K. Yang, F. Yang, H.B. Lv, L.H. Yan, H.W. Yan, Y.Y. Xie, X.D. Yuan, L. Zhang, Rational design and fabrication of highly transparent, flexible, and thermally stable superhydrophobic coatings from raspberry-like hollow silica nanoparticles, Appl. Surf. Sci. 440 (2018) 700–711. [6] I.M. Thomas, Method for the preparation of porous silica antireflection coatings varying in refractive index from 1.22 to 1.44, Appl. Opt. 31 (1992) 6145–6149. [7] M. Feit, F. Genin, A. Rubenchik, L. Sheehan, S. Schwartz, M. Kozlowski, J. Dijon, P. Garrec, J. Hue, Statistical description of laser damage initiation on NIF and LMJ optics at 355 nm, Proc. SPIE 3492 (1998). [8] M. Norton, L. Hrubesh, Z. Wu, E. Donohue, M. Feit, M. Kozlowski, D. Milan, K. Neeb, W. Molander, A. Rubenchik, W. Sell, P. Wegner, Growth of laser-initiated damage in fused silica at 351 nm, Proc. SPIE 4347 (2000) 468. [9] F.T. Chi, N. Pan, C.C. Ding, X.Q. Wang, F.C. Yi, X.F. Li, J.H. Lei, Ultraviolet laserinduced damage of freestanding silica nanoparticle films, Appl. Surf. Sci. 255 (2019) 566–572. [10] M.L. Spaeth, P.J. Wegner, T.I. Suratwala, M.C. Nostrand, J.D. Bude, A.D. Conder, J.A. Folta, J.E. Heebner, L.M. Kegelmeyer, B.J. MacGowan, D.C. Mason, M.J. Matthews, P.K. Whitman, Optics recycle loop strategy for NIF operations above UV laser-induced damage threshold, Fusion Sci. Technol. 69 (2003) 265–294. [11] I.L. Bass, G.M. Guss, R.P. Hackel, Mitigation of laser damage growth in fused silica with a galvanometer scanned CO2 laser, Proc. SPIE 5991 (2005) 59910C. [12] I.L. Bass, G.M. Guss, M.J. Nostrand, P.J. Wegner, An improved method of mitigating laser induced surface damage growth in fused silica using a rastered, pulsed CO2 laser, Proc. SPIE 7842 (2010) 784220. [13] P. Cormont, L. Gallais, L. Lamaignère, J.L. Rullier, P. Combis, D. Hebert, Impact of two CO2 laser heatings for damage repairing on fused silica surface, Opt. Express 18 (2010) 26068–26076. [14] C.C. Zhang, L.J. Zhang, W. Liao, Z.H. Yan, J. Chen, Y.L. Jiang, H.J. Wang, X.Y. Luan, Y.Y. Ye, W.G. Zheng, X.D. Yuan, ATR-FTIR spectroscopic studies on density changes of fused silica induced by localized CO2 laser treatment, Chin. Phys. B 24 (2015) 024220. [15] Dept. of Energy, DOE Handbook: Tritium Handling and Safe Storage, Washington, DC (United States), (2008). [16] A. Nobile, H. Reichert, Design of the omega laser target chamber tritium removal system, Fusion Sci. Technol. 43 (2016) 522–539. [17] T. Masahiro, Control of dehumidification using polymer permeable membrane and
6. Summary We proposed an environment friendly dry stripping method of solgel coating striping based on Ar+ ion beam etching. IBE parameters like ion energy and ion beam current are systematically optimized to obtain a gentle and selective sputtering of the coating over the substrate. The effectiveness is repeatedly tested on three typically defective coatings and the actual recycled VI from SG series high power laser systems. Microscope images and transmittance test show that after stripping substrates exhibit cleaning and smooth surfaces. Moreover, no performance degradation is observed after stripping. According to comparisons tests, surface roughness and contour barely change. The laser damage resistance even increases slightly, probably due to the removal of re-deposition layer introduced during polishing process. This method has been successfully adopted to process the recycling Vacuum Isolator and becomes a favorable alternative in the design of next generation of ICF facility of China. Declaration of Competing Interest The authors declare that there are no conflicts of interest. Funding The research was supported by the National Key Research and Development Program of China [Grant No. 2018YFB1107605], the National Natural Science Foundation of China [Grant No. 61804145] and Laser Fusion Research Center Funds for Young Talents [Grant No. LFRC - CZ011]. CRediT authorship contribution statement Xiaolong Jiang: Conceptualization, Methodology, Investigation, Writing - original draft. Wei Liao: Project administration. Bo Li: Investigation. Xia Xiang: Investigation. Xiaodong Yuan: 8
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
[18] [19] [20] [21]
[22]
[23] [24]
[25]
its application to tritium removal system design, Fusion Eng. Des 36 (Part A) (2018) 141–145. J.L. Maienschein, S.W. Wilson, F. Garcia, Design and operational experience with a portable tritium cleanup system, Fusion Sci. Technol. 21 (1992) 691–695. I.S.T. Tsong, D.J. Barber, Review: Sputtering mechanisms for amorphous and polycrystalline solids, J. Mater. Sci. 8 (1973) 123–135. V.V. Manukhin, Sputtering of solids under light-ion bombardment, Nucl. Instrum. Methods Phys. Res. Sect. B 72 (1992) 45–50. M. Ando, M. Negishi, M. Takimoto, A. Deguchi, N. Nakamura, Super-smooth polishing on aspherical surfaces (I): high-precision coordinate measuring and polishing systems, Proc. SPIE 2576 (1995) e215612. W.Q. Peng, C.L. Guan, S.Y. Li, Efficient fabrication of ultrasmooth and defect-free quartz glass surface by hydrodynamic effect polishing combined with ion beam figuring, Opt. Express 22 (2014) 13951–13961. L. Zhou, Y.F. Dai, X.H. Xie, W.L. Liao, S.Y. Li, Translation-reduced ion beam figuring, Opt. Express 23 (2015) 7094–7100. T.Y. Wang, L. Huang, M. Vescovi, D. Kuhne, K. Tayabaly, N. Bouet, M. Idir, Study on an effective one-dimensional ion-beam figuring method, Opt. Express 27 (2019) 15368–15381. A. Perentos, A. Mitchell, A. Holland, Ion beam etching of high resolution structures
[26]
[27] [28] [29]
[30] [31] [32]
9
in Ta2O5 for grating-assisted directional coupler applications, Appl. Surf. Sci. 252 (2005) 1006–1012. A. Barlow, N. Sano, B.J. Murdoch, J. Portoles, P. Pigram, P.J. Cumpson, Observing the evolution of regular nanostructured indium phosphide after gas cluster ion beam etching, Appl. Surf. Sci. 459 (2018) 678–685. H.W. Dinges, B. Kempf, H. Burkhard, R. Gobel, Determination of ion beam etching damage on InP by spectroscopic ellipsometry, Appl. Surf. Sci. 459 (1991) 359–363. Http://www.sspectra.com, 2019 (accessed 29 October 2019). X. Ye, J. Huang, B. Li, F. Geng, H.J. Liu, F.R. Wang, X.D. Jiang, W.D. Wu, W.G. Zheng, Antireflection subwavelength structures on fused silica fabricated by colloidal microsphere lithography, J. Optoelectron. Adv. M. 17 (2015) 1689–1695. Http://www.srim.org, 2019 (accessed 29 October 2019). J.F. Ziegler, M.D. Ziegler, J.P. Biersack, “SRIM – the stopping and range of ions in matter (2010)”, Nucl. Instrum. Methods Phys. Res. Sect. B 268 (2010) 1818–1823. W.G. Zheng, X.F. Wei a, Q.H. Zhu, F. Jing, D.X. Hu, X.D. Yuan, W.J. Dai, W. Zhou, F. Wang, D.P. Xu, X.D. Xie, B. Feng, Z.T. Peng, L.F. Guo, Y.B. Chen, X.J. Zhang, L.Q. Liu, D.H. Lin, Z. Dang, Y. Xiang, R. Zhang, F. Wang, H.T. Jia, X.W. Deng, Laser performance upgrade for precise ICF experiment in SG-xx laser facility, Matt. Radiat. Extremes 2 (2017) 243–255.
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Removal of antireflection sol-gel SiO2 coating based on Ar ion beam etching a
a
b
b
a
T
a
Xiaolong Jiang , Wei Liao , Bo Li , Xia Xiang , Xiaodong Yuan , Chuanchao Zhang , Xiaoyu Luana, Xiaodong Jianga,* a b
Laser Fusion Research Center, China Academy of Engineering Physics, Mianyang 621900, China School of Physics, University of Electronic Science and Technology of China, Chengdu 610054, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Inertial confinement fusion Sol-gel antireflection coating Tritium removal Coating stripping Ion beam etching
We proposed a new stripping method of sol-gel SiO2 coating widely used in inertial confinement fusion facilities based on Ar ion beam etching, during which the coating is removed in the form of gaseous and solid particle. This dry approach is more environment-friendly than the conventional wet cleaning process for saving the trouble of treating tons of polluted water, especially highly toxic tritiated water. In the IBE process, porous coating is sputtered away at a constant etching rate just like other homogeneous materials. IBE parameters are systematically optimized to obtain a selective and gentle etching. The weakly bonded coating is etched 6 times faster than the fused silica substrate, which ensures the selective sputtering of coating over the underlying fine optics. Experiments on different off-line samples as well as the actual recycled optics from high power laser system show that this method is very effective. Moreover, this stripping process does not degrade the optic’s performance. Surface shape and roughness barely change after stripping. The laser damage threshold increases slightly, probably due to the removal of re-deposition layer introduced during polishing process. This method has been successfully adopted to process the optics of ICF facility in China.
1. Introduction It is well known that in the high-power systems like National Ignition Facility (NIF) [1], Laser MégaJoule (LMJ) [2] and Shenguang (SG) series facilities [3], almost all UV optics are coated with sol-gel silica to reduce interfacial reflection [4–6]. After serving for a period of time, the optics would inevitably get laser damaged [7–9]. When it is determined that a particular optic needs to get damage mitigated, it is removed from the laser and transported to mitigation facilities. There, the optic is surveyed for tritium and other hazardous contaminations that it may have been exposed to in the target chamber and, if necessary, decontaminated [10]. Next, stripping of sol-gel coating is conducted, followed by recyclability assess (in term of surface quality, surface damage, and deeper bulk damage) and CO2 laser mitigation (if necessary) [11–14]. Conventionally, sol-gel coating is stripped using an ultra-sonic wet cleaning system, as shown in Fig. 1. Even though this method has proved to be effective in practice, its weaknesses come out too. First, it produces tons of waste water. The wet cleaning process consists of four procedures: deionized water rinse, ultrasonic cleaning using alkaline solution, ultra-sonic cleaning using deionized water, and slow lifts dry. Counting the extra water needed to flush the evacuated sinks, pipe and
⁎
percolator, one wet cleaning process produces at least 440 liters wastewater. More importantly, when the coating is contaminated with tritium, effluent disposal becomes tricky. The tritium absorbed within the porous coating would enter into the cleaning solution and make it tritium contaminated. Tritiated water, however, is the least preferred form of tritium [15]. Due to the body’s ready adsorption of tritium in the form of tritiated water, exposure to tritiated water is on the order of 10,000 times more hazardous than exposure to gaseous tritium. Tritium in the form of T2O is difficult to store for long periods due to its corrosive properties. Tritiated water is also much more difficult to dispose compared to gaseous or solid tritium. De-tritium before stripping does not make so much difference, either. In order to avoid any modification of the fine optics, gently heating the optics is one of the few remaining decontamination methods. Our experiments on de-tritiated Vacuum Isolator (VI, an UV optics, details about VI can been found in section 5.2) from SG series facilities show that tritium would accumulate in the solution, and its concentration would exceed regulatory limits after several strips. Gaseous or solid tritium, on the other hand, is easily handled and routinely disposed: their use, transfer, storage, and shipment have a long history and have been safely used for decades [16–18]. Therefore, exploring for a dry approach in which tritium is removed in the form of gas or solid is necessary.
Corresponding author. E-mail address: [email protected] (X. Jiang).
https://doi.org/10.1016/j.fusengdes.2020.111578 Received 5 November 2019; Received in revised form 1 February 2020; Accepted 20 February 2020 Available online 27 February 2020 0920-3796/ © 2020 Published by Elsevier B.V.
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 1. The ultra-sonic cleaning system used in conventional wet stripping process. (a) The overall appearance. (b) Inner details.
has a diameter of ∼1.1 m and a height of 1.1 m (inner size), with three molecular pumps (FF-250 from KYKY Technology Co.) to pump. The ion gun used is a self-developed 6 cm × 66 cm radio frequency (RF) ion beam source shooting flat-roofed ion beam. Inert gas Ar is used as the etching gas. With the long ion beam source and substrate carrier that moves back and forth along the widthwise direction, as illustrated in Fig. 2(b), this plant is capable of uniformly processing substrate of aperture up to semi-meter. For 430 mm × 430 mm aperture substrate, as in this case, only the middle 430 mm part of the ion beam is used. A beam slot, i.e. a fused silica plate with a 430 mm × 60 mm rectangle window located in the middle, is employed to ensure that only the substrate area is reached by ion beam. This is helpful for avoiding unfavorable sputtering of the substrate carrier or other chamber parts. The uniformity (along longwise direction) of beam current intensity of the unshielded part is rms ∼ 3% and PV∼ ± 6%, measured by Faraday cup. In this study, incidence angle is fixed 0°, i.e., perpendicular to substrate surface. Fused silica plates (Corning 7980 and Heraeus 312) were ground and polished using conventional ceria polishing techniques. Corning 7980 plates (50 mm × 50 mm × 4 mm) are used in most of the experiments. VI from the ICF facility of China (Heraeus 312, 430 mm × 430 mm × 10 mm) is used for comparison tests of surface shape and more importantly testifying the applicability of the proposed approach on actual optics. Smaller Heraeus 312 plates (100 mm × 100 mm × 10 mm) are used for comparison study of laser damage threshold. Substrates were dip coated with 73 nm thick sol-gel coating. Before that, all substrates were ultrasonically cleaned, unless otherwise specified. Etching depth is measured by stylus surface profilometer (NanoMap-500LS, AEP technology Inc.). Spectrophotometer (Lambda 950, PerkinElmer Inc.) was utilized to characterize the transmittance before and after stripping. Transmittance of VI is measured using large aperture spectrophotometer. Surface roughness is measured using Atomic Force Microscopy (AFM, Dimension Icon, Bruker Inc.) and optical surface profiler (FPM-1000, ZYGO Inc.). Surface shape is characterized by phase shift laser interferometer (VerifireTM MST, ZYGO Inc.). Laser damage thresholds were tested using A Q-switched Nd: YAG laser (SAGA, Thales Laser Inc.). It provides 355 nm, 6.8 ns laser pulse (temporal and spatial profile Gaussian, beam waist 1/e2 diameter 0.47 mm, modulation depth 2.2).
In this work, we proposed a new sol-gel coating striping method based on Ar ion beam etching. Ion beam etching (IBE) is a dry etching technique in which energetic ions are generated and then impinge a target to remove surface material [19,20]. Due to its atomic-scale removal capacity, non-contacting mode and good repeatability, IBE finds well-established uses in fields like ultra-smooth polishing [21,22], highprecision figuring [23,24] and micro/nano-patterning [25–27]. As to coating stripping, IBE is promising for two reasons. First, physical ion bombardment with energy of hundreds of eV could easily sputter away the coating as well as the absorbing contaminations. The etching products generally take the form of volatile plasma, or less-volatile cluster, or most likely both. In this case, the etching products theoretically consist of gaseous O2, T2 and solid Si, Si-O, Si-T. This means no liquid waste is generated, as opposed to the wet approach. The volatile (gas and small particles) could be pumped out of the chamber by vacuum pump and the non-volatile cluster would deposit on the chamber floor. If etching product gather/disposal is required, the volatiles can be collected through off-gas filtration and non-volatiles through regular cleaning of the chamber. Second, the non-contact IBE process barely modify the fine optics. Since sol-gel coating is practically a bunch of weakly bonded nano-size particles, it would theoretically have a much higher sputtering rate than the fused silica substrate. This etching selectivity is beneficial for avoiding over-etching of the substrate. The main contributions of this work are the proposal, technical design and feasibility assess of the IBE based sol-gel coating stripping method. Feasibility is systematically studied from the perspectives of effectiveness, applicability and potential side-effects. And the overall outcomes are inspiring. This paper is organized as follow: first, the stripping process of solgel silica coating under ion beam etching is studied. Then, processing parameters such as bombardment energy and ion current are optimized to obtain a selective (selectively etching the coating over substrate), slow (to control etching depth), and damage free (small implant depth of Ar + into substrate matrix) etching. After the optimum parameters are determined, stripping experiments are conducted on both off-line and on-line samples. Effectiveness of this method is experimentally confirmed based on transmission rate tests and optical microscope characterizations. The possible negative modifications this method might introduce to the fine optic are also investigated. Comparison tests on surface roughness, surface shape and laser damage resistance show that no performance degradation occurs. So far, this technique has been successfully adopted to strip the recycling optics of SG series facilities in China.
3. Stripping process of sol-gel coating Stripping experiments are conducted to understand the stripping process of sol-gel coating under ion beam etching. Dipped coated substrates are one-side etched for different durations, followed by transmittance tests and morphology characterizations using AFM and optical surface profiler. Fig. 3(a) shows the experimental results of transmittance spectra as function of IBE durations. As etching proceeds, the
2. Experimental equipment and method A large aperture IBE system with rectangle ion beam source (see Fig. 2(a)) has been developed. The main vacuum chamber of the plant 2
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 2. (a) The large aperture ion beam etching plant used in this research; (b) schematic diagram of the IBE configuration.
substrate. On the other hand, since the particle is only 20 nm in diameter, i.e., much smaller than the wavelength (351 nm), these detailed structures within one wavelength (note that the scan size of the AFM image is only 500 nm × 500 nm) would be harmless, if not beneficial [29]. Given above facts, roughness evolution is characterized using optical surface profiler (the scan size is 0.7 mm × 0.5 mm) instead of AFM. In fact, optical surface profiler is used as the standard, certificated roughness measurer of all UV optics in the SG series facilities. Fig. 5 shows the optically obtained surface morphologies of samples corresponding to Fig. 4. It is obvious that no roughening occurs after stripping.
curve peak shifts meanwhile the overall transmittance gradually drops. These apparently result from the thinning of sol-gel layer. After 5 min, transmittance curve remains unchanged and overlaps with that of one side coated substrate, indicating a clean removal. Etching rates of solgel coating under such etching recipe is calculated from the depth of steps ion-etched through polyimide mask, assuming that the coating has a constant IBE etching rate just like conventional homogeneous materials such as silica and silicon. The obtained etching rate is 16 nm/min, accordingly, the theoretical stripping time needed is 4.56 min (73 nm / (16 nm/min) =4.56 min), which coincides with the 4−5 min given by transmittance spectra. We further theoretically simulate the time evolution of transmittance spectra under 16 nm/min etching using software TFCal [28], as shown in Fig. 3(b), and compare it with the experimental results. It is clearly seen that the simulation results exhibit the same pattern as experimental observations. This confirms the assumption that porous coating is etched at a constant rate, even though it is not homogeneous matrix but porous cluster of weakly bonded nano-size particles. Fig. 4 shows the AFM images for different etching durations. For unetched surface, shown in Fig. 4(a), nano-particles of silica can be clearly seen. As etching proceeds, the contour of nano-particle gradually fades away. Especially, according to Fig. 4(f), after the coating has been removed, the porous morphology of the coating layer is partly transferred into the substrate, making the surface slightly rougher than the original bare substrate. This roughness depends on the integral thickness fluctuation of the coating layer, which is related to the size of the nanoparticles, as well as etching selectivity of coating over fused silica
4. Parameters optimization 4.1. Ion energy In order to remove the coating without hurting the underlying substrate, over etching of the underlying substrate should be restrained. One might imagine simply turning off the ion beam after the coating has been etched through. This requires a good endpoint detection system, for instance an in-situ transmittance inspection configuration. Another approach is improving the coating to substrate selectivity of the ion beam process, i.e. the ratio of sputtering yield of coating to that of substrate. Sputtering yield is defined as the number of target atoms ejected per incoming ion. In IBE, the near-surface particles are removed from the target by cascaded bombardment. When a cascade gives a
Fig. 3. The transmittances curves of coated substrate IBE processed for different durations (a) experimental results (b) simulation results. 3
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 4. AFM images of coated surface stripped for different durations: (a) 0 min; (b) 1 min; (c) 2 min; (d) 3 min; (e) 4 min; (f) 5 min. The scan size is 500 nm × 500 nm.
through polyimide and coating masks, as shown in Fig. 6(a). The lower part of fused silica substrate is dipped coated and the left part is covered with polyimide type. Polyimide is known for thermal stability, good chemical resistance and excellent mechanical properties, and can protect the substrate from being etched. Apparently, step 1 is formed between etched and un-etched areas and can be used to calculate the etching rate of fused silica. The sol-gel coating, on the other hand, can only temporally shield the substrate; therefore, once the coating is completely removed, the depth of step 2 would remain unchanged. This
target particle an energy greater than the "surface binding energy" of that target, the particle may be sputtered. Accordingly, one would expect that sputtering yield is subject to the following factors: ion type, target material, incidence angle and ion energy [20]. The first three factors, in our case, are constants; therefore, optimization of ion energy is necessary. Fig. 6 shows the experimental configuration and obtained etching selectivity dependence on Ar+ energy. Ion current is fixed at 300 mA. Etching rates were calculated from the depth of steps ion-etched
Fig. 5. Optical surface profiler images of coated surface stripped for different durations: (a) 0 min; (b) 1 min; (c) 2 min; (d) 3 min; (e) 4 min; The scan size is 0.7 mm × 0.5 mm. 4
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 6. (a) Experimental configuration used for etching rate determination (b) Experimental results of etching rates and selectivity as function of ion energy.
constant depth represents the equivalent “fused silica” thickness of solgel coating under ion beam etching. Etching selectivity is actually the ratio of coating thickness (73 nm) to the depth of step 2. Based on the obtained etching rate of fused silica and etching selectivity, etching rate of coating is further calculated. Experimental results are presented in Fig. 6(b). It is clearly seen that the coating consistently has a higher etching rate than substrate. Etching rate of coating increases rapidly from 15.2 nm/min at 200 eV to 145.6 nm/min at 1000 eV, while that of substrate increases more slowly, only from 4 nm/min to 22 nm/min. Consequently, etching selectivity gradually increases from 3.6 to 6.5. Since the density of sol-gel coating is about half of that of fused silica, the high etching selectivity should be partly attributed to the weakly bonded nature of the coating. Similarly, one would also expect a selective sputtering of contaminations over substrate because contaminations are generally also weakly bonded with each other or with the coating. It should be mentioned that ion energy beyond 1000 eV is not taken in consideration here because at that energy region, ion implant is no longer negligible. Ion implant, after certain degree, may damage the target lattice and impact the laser damage resistance. In order to study this issue, target damage is simulated using software SRIM [30,31]. Fig. 7(a) illustrates the Ar + distribution in fused silica at different ion energies. The projected range (depth of the peak concentration), as summarized in Fig. 5(b), increases with ion energy, whereas is only in
the order of several nanometers. Given these facts, ion energy is chosen to be 800 eV, at which the etching selectivity is 6.0, very close to the maximum 6.5 at 1000 eV. The projected depth is 3.7 nm, which does not impact the laser damage resistance as would be shown later. 4.2. Ion current Ion current is another important parameter. With ion energy given, ion current directly determines the etching rate, i.e., the time efficiency and controlling precision of the stripping process. The relationships between ion current and etching rate of coating and substrate are experimentally studied, as shown in Fig. 8(a). Ion energy is fixed at 800 eV. It is clearly seen that both etching rates increase linearly with ion current. This coincides with the theory that sputtering amount is proportion to the time integral of ion current. Considering that sol-gel coating is only 73 nm thick, ion current of 100 mA is feasible. The etching rates of coating and substrate at such ion current are 34.5 nm/ min and 5.8 nm/min, respectively. Hence, processing duration of no less than 2.2 min is required. Given that the contaminations absorbed on or within the coating might need extra time, 3 min in practice is feasible. As shown in Fig. 8(b), the scan length for 430 mm × 430 mm aperture optics is 540 mm: 430 mm for covering the substrate, 60 mm for the width of ion beam and extra 2 × 25 mm for alignment tolerance. That is 9 times of the width of ion beam (60 mm), hence the stripping
Fig. 7. (a) Ar+ distributions in fused silica as function of ion energy. (b) Projected depth of Ar+ in fused silica as function of ion energy. 5
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 8. (a). Experimental results of etching rates of sol-gel coating and fused silica substrate as function of ion current; (b) schematic of scanning etching.
scratches and pits on the coating. Sample #3 was poorly pre-cleaned, consequentially after coating, numerous dust particles are enclosed. After IBE stripping, however, all three samples exhibit clean and smooth surface, as in Fig. 9(b), (d) and (f). Results of transmittance tests are shown in Fig. 10. As we can see, transmittance curves of the three stripped samples overlap with that of bare fused silica, suggesting a clean removal. According to surface roughness tests, roughness (Rq) for sample #1 - #3 are 0.737 nm, 0.654 nm and 0.826 nm respectively, barely differing from the original 0.795 nm. These suggest that the proposed IBE stripping approach is broadly suitable for different kinds of sol-gel coatings.
time needed is 27 min. In circumstance higher depth-control precision is required, one can slow down the process by simply reducing the ion current. Theoretically, the control precision of IBE process is in atomic scale. 5. Effectiveness tests in practice 5.1. Applicability for defective coating In previous experiments, all the coatings are meticulously prepared and carefully handled, i.e. almost perfect. For the actual optics to be recycled, however, that is normally not the case. In this section, we further investigate the applicability of proposed approach on imperfect coatings. In Fig. 9, three typically defective coatings are shown. In sample #1, a CO2 laser mitigated crater causes nonuniform distribution of coating surrounding the crater, as indicated by the nonuniform brightness caused by nonuniform interfacial reflection of the lamplight. For sample #2, careless storage and transportation creates many
5.2. Application for actual recycled optic in ICF facility In order to further test the applicability of the proposed approach on actual UV optics, stripping experiments are performed on a VI from SG series facilities. Vacuum Isolator (also called Vacuum Window) is an UV optics locating between the Final Optics Assembly (FOA, consists of
Fig. 9. Optical microscope pictures of defectively coated sample #1, #2 and #3 taken before and after IBE stripping. 6
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Fig. 12. Transmittance spectra of samples of different treatments.
Fig. 10. Transmittance spectra of three stripped samples and bare fused silica. The overlapping of transmittance curves indicates a clean removal.
five tested sites. It is clearly seen that parameters of surface shape and roughness hardly change, which is reasonable since the coating is only etched for dozens of nanometers and over-etching is even less.
several UV optics) and the target chamber [32]. It has the dimensions of 430 mm × 430 mm × 10 mm. In SG facility, the FOA general operates at 1000 Pa while the target chamber at 10−3 Pa. VI is used to isolate the target chamber from the low-vacuum FOA, hence the name Vacuum Isolator. It has been previously CO2 laser mitigated so several mitigation craters with diameters ranging from 0.52 mm to 2 mm are present. For both incident and exit surfaces, five randomly selected sites are tracked by coordinates. Surface quality and roughness of each site are characterized before and after stripping. Surface shape and transmittance spectra are also tested. Fig. 11 shows some of the comparisons of surface images before and after stripping obtained using optical microscopy, thereinto, Fig. 11(a) is incident surface and Fig. 11(b) the exit surface. Note that after serving in the harsh laser environment for several months, the coating becomes rough and a few particles can be observed, as in Fig. 11(a). For the exit surface, some parts of the coating are even wrinkled, as in Fig. 11(b). After IBE sputtering, the rough coating and the contamination are completely removed. As shown in Fig. 12, transmittance curves of the five tracked sites overlap with that of bare fused silica, just like previous observations. Surface shape and roughness before and after stripping are summarized in Table 1. Each value of roughness and PSD2 is averaged over
5.3. Impact on laser damage resistance The potential impact the stripping approach may have on UV laser damage resistance is also investigated. Fig. 13 presents the results of 355 nm 1 on 1 laser damage thresholds (LDT) test on bare fused silica and IBE stripped substrate. For each laser fluence, 20 spots are shot to calculate the damage possibility. The distance of the adjacent spots is 4 mm. In order to minimize the test error caused by the difference of the samples, the substrates we used are cut from the same glass block, and ground and polished in the same batch. It is clearly that laser damage resistance improves slightly after stripping, with LDT increasing from 10 J/cm2 to 11 J/cm2. This is probably due to the over-etching of redeposition layer introduced during polishing, which is generally dozens of nanometers thick and identified as one of the most important laser damage precursors.
Fig. 11. Comparisons of optical microscopically obtained surface pictures of a VI from SG series facility before and after stripping: (a) Incident surface; (b) exit surface. 7
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
Table 1 Surface shape and roughness of a VI from China ICF facility before and after IBE stripping. Treatment
Standard Before strip After strip
Roughness (nm)
PSD2 (nm)
0° transmission wave front (λ = 632.8 nm)
Incident surface
Exit surface
Incident surface
Exit surface
Defocusing PV
Astigmatism PV
GRMS (λ/cm)
PSD1 (nm)
1.2 0.700 ± 0.043 0.810 ± 0.056
1.2 0.440 ± 0.021 0.606 ± 0.038
1.1 0.580 ± 0.105 0.626 ± 0.064
1.1 0.430 ± 0.037 0.479 ± 0.044
0.425 0.011
0.220 0.048
1/120 1/131
1.8 1.356
0.049
0.053
1/129
1.329
Conceptualization. Chuanchao Zhang: Writing - review & editing. Xiaoyu Luan: Investigation. Xiaodong Jiang: Supervision. Acknowledgments The authors wish to thank Chunyuan Hou, Chengxiang Tian for their help with IBE experiments, and Yi Yang, Bin Hen for their help with characterization of roughness and surface contour, and Zhengkun Liu, Yin Liu for their help with AFM tests. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.fusengdes.2020. 111578. References
Fig. 13. Results of 1-on-1 laser damage threshold test before and after stripping.
[1] M.L. Spaeth, K.R. Manes, M. Bowers, P. Celliers, J.M.D. Nicola, et al., National ignition facility laser system performance, Fusion Sci. Technol. 69 (2016) 366–394. [2] N. Fleurot, C. Cavailler, J.L. Bourgade, The Laser Mégajoule (LMJ) project dedicated to inertial confinement fusion: development and construction status, Fusion Eng. Des. 74 (2005) 147–154. [3] W.G. Zheng, X.F. Wei, Q.H. Zhu, F. Jing, D.X. Hu, J.Q. Su, K.X. Zheng, X.D. Yuan, H. Zhou, W.J. Dai, W. Zhou, F. Wang, D.P. Xu, X.D. Xie, B. Feng, Z.T. Peng, L.F. Guo, Y.B. Chen, X.J. Zhang, L.Q. Liu, D.H. Lin, Z. Dang, Y. Xiang, X.W. Deng, Laser performance of the SG-xx laser facility, High Power Laser Sci. Eng. 4 (2016) e21. [4] S.I. Huang, Y.J. Shen, H. Chen, Study on the hydrophobic surfaces prepared by twostep sol–gel process, Appl. Surf. Sci. 255 (2009) 7040–7046. [5] X.S. Zou, C.Y. Tao, K. Yang, F. Yang, H.B. Lv, L.H. Yan, H.W. Yan, Y.Y. Xie, X.D. Yuan, L. Zhang, Rational design and fabrication of highly transparent, flexible, and thermally stable superhydrophobic coatings from raspberry-like hollow silica nanoparticles, Appl. Surf. Sci. 440 (2018) 700–711. [6] I.M. Thomas, Method for the preparation of porous silica antireflection coatings varying in refractive index from 1.22 to 1.44, Appl. Opt. 31 (1992) 6145–6149. [7] M. Feit, F. Genin, A. Rubenchik, L. Sheehan, S. Schwartz, M. Kozlowski, J. Dijon, P. Garrec, J. Hue, Statistical description of laser damage initiation on NIF and LMJ optics at 355 nm, Proc. SPIE 3492 (1998). [8] M. Norton, L. Hrubesh, Z. Wu, E. Donohue, M. Feit, M. Kozlowski, D. Milan, K. Neeb, W. Molander, A. Rubenchik, W. Sell, P. Wegner, Growth of laser-initiated damage in fused silica at 351 nm, Proc. SPIE 4347 (2000) 468. [9] F.T. Chi, N. Pan, C.C. Ding, X.Q. Wang, F.C. Yi, X.F. Li, J.H. Lei, Ultraviolet laserinduced damage of freestanding silica nanoparticle films, Appl. Surf. Sci. 255 (2019) 566–572. [10] M.L. Spaeth, P.J. Wegner, T.I. Suratwala, M.C. Nostrand, J.D. Bude, A.D. Conder, J.A. Folta, J.E. Heebner, L.M. Kegelmeyer, B.J. MacGowan, D.C. Mason, M.J. Matthews, P.K. Whitman, Optics recycle loop strategy for NIF operations above UV laser-induced damage threshold, Fusion Sci. Technol. 69 (2003) 265–294. [11] I.L. Bass, G.M. Guss, R.P. Hackel, Mitigation of laser damage growth in fused silica with a galvanometer scanned CO2 laser, Proc. SPIE 5991 (2005) 59910C. [12] I.L. Bass, G.M. Guss, M.J. Nostrand, P.J. Wegner, An improved method of mitigating laser induced surface damage growth in fused silica using a rastered, pulsed CO2 laser, Proc. SPIE 7842 (2010) 784220. [13] P. Cormont, L. Gallais, L. Lamaignère, J.L. Rullier, P. Combis, D. Hebert, Impact of two CO2 laser heatings for damage repairing on fused silica surface, Opt. Express 18 (2010) 26068–26076. [14] C.C. Zhang, L.J. Zhang, W. Liao, Z.H. Yan, J. Chen, Y.L. Jiang, H.J. Wang, X.Y. Luan, Y.Y. Ye, W.G. Zheng, X.D. Yuan, ATR-FTIR spectroscopic studies on density changes of fused silica induced by localized CO2 laser treatment, Chin. Phys. B 24 (2015) 024220. [15] Dept. of Energy, DOE Handbook: Tritium Handling and Safe Storage, Washington, DC (United States), (2008). [16] A. Nobile, H. Reichert, Design of the omega laser target chamber tritium removal system, Fusion Sci. Technol. 43 (2016) 522–539. [17] T. Masahiro, Control of dehumidification using polymer permeable membrane and
6. Summary We proposed an environment friendly dry stripping method of solgel coating striping based on Ar+ ion beam etching. IBE parameters like ion energy and ion beam current are systematically optimized to obtain a gentle and selective sputtering of the coating over the substrate. The effectiveness is repeatedly tested on three typically defective coatings and the actual recycled VI from SG series high power laser systems. Microscope images and transmittance test show that after stripping substrates exhibit cleaning and smooth surfaces. Moreover, no performance degradation is observed after stripping. According to comparisons tests, surface roughness and contour barely change. The laser damage resistance even increases slightly, probably due to the removal of re-deposition layer introduced during polishing process. This method has been successfully adopted to process the recycling Vacuum Isolator and becomes a favorable alternative in the design of next generation of ICF facility of China. Declaration of Competing Interest The authors declare that there are no conflicts of interest. Funding The research was supported by the National Key Research and Development Program of China [Grant No. 2018YFB1107605], the National Natural Science Foundation of China [Grant No. 61804145] and Laser Fusion Research Center Funds for Young Talents [Grant No. LFRC - CZ011]. CRediT authorship contribution statement Xiaolong Jiang: Conceptualization, Methodology, Investigation, Writing - original draft. Wei Liao: Project administration. Bo Li: Investigation. Xia Xiang: Investigation. Xiaodong Yuan: 8
Fusion Engineering and Design 156 (2020) 111578
X. Jiang, et al.
[18] [19] [20] [21]
[22]
[23] [24]
[25]
its application to tritium removal system design, Fusion Eng. Des 36 (Part A) (2018) 141–145. J.L. Maienschein, S.W. Wilson, F. Garcia, Design and operational experience with a portable tritium cleanup system, Fusion Sci. Technol. 21 (1992) 691–695. I.S.T. Tsong, D.J. Barber, Review: Sputtering mechanisms for amorphous and polycrystalline solids, J. Mater. Sci. 8 (1973) 123–135. V.V. Manukhin, Sputtering of solids under light-ion bombardment, Nucl. Instrum. Methods Phys. Res. Sect. B 72 (1992) 45–50. M. Ando, M. Negishi, M. Takimoto, A. Deguchi, N. Nakamura, Super-smooth polishing on aspherical surfaces (I): high-precision coordinate measuring and polishing systems, Proc. SPIE 2576 (1995) e215612. W.Q. Peng, C.L. Guan, S.Y. Li, Efficient fabrication of ultrasmooth and defect-free quartz glass surface by hydrodynamic effect polishing combined with ion beam figuring, Opt. Express 22 (2014) 13951–13961. L. Zhou, Y.F. Dai, X.H. Xie, W.L. Liao, S.Y. Li, Translation-reduced ion beam figuring, Opt. Express 23 (2015) 7094–7100. T.Y. Wang, L. Huang, M. Vescovi, D. Kuhne, K. Tayabaly, N. Bouet, M. Idir, Study on an effective one-dimensional ion-beam figuring method, Opt. Express 27 (2019) 15368–15381. A. Perentos, A. Mitchell, A. Holland, Ion beam etching of high resolution structures
[26]
[27] [28] [29]
[30] [31] [32]
9
in Ta2O5 for grating-assisted directional coupler applications, Appl. Surf. Sci. 252 (2005) 1006–1012. A. Barlow, N. Sano, B.J. Murdoch, J. Portoles, P. Pigram, P.J. Cumpson, Observing the evolution of regular nanostructured indium phosphide after gas cluster ion beam etching, Appl. Surf. Sci. 459 (2018) 678–685. H.W. Dinges, B. Kempf, H. Burkhard, R. Gobel, Determination of ion beam etching damage on InP by spectroscopic ellipsometry, Appl. Surf. Sci. 459 (1991) 359–363. Http://www.sspectra.com, 2019 (accessed 29 October 2019). X. Ye, J. Huang, B. Li, F. Geng, H.J. Liu, F.R. Wang, X.D. Jiang, W.D. Wu, W.G. Zheng, Antireflection subwavelength structures on fused silica fabricated by colloidal microsphere lithography, J. Optoelectron. Adv. M. 17 (2015) 1689–1695. Http://www.srim.org, 2019 (accessed 29 October 2019). J.F. Ziegler, M.D. Ziegler, J.P. Biersack, “SRIM – the stopping and range of ions in matter (2010)”, Nucl. Instrum. Methods Phys. Res. Sect. B 268 (2010) 1818–1823. W.G. Zheng, X.F. Wei a, Q.H. Zhu, F. Jing, D.X. Hu, X.D. Yuan, W.J. Dai, W. Zhou, F. Wang, D.P. Xu, X.D. Xie, B. Feng, Z.T. Peng, L.F. Guo, Y.B. Chen, X.J. Zhang, L.Q. Liu, D.H. Lin, Z. Dang, Y. Xiang, R. Zhang, F. Wang, H.T. Jia, X.W. Deng, Laser performance upgrade for precise ICF experiment in SG-xx laser facility, Matt. Radiat. Extremes 2 (2017) 243–255.