Open Access
Add to BibSonomy
Abstract
We investigate the role of each step in the combined treatment of reactive ion etching (RIE) and dynamic chemical etching (DCE) for improving the laser-induced damage resistance of fused silica optics. We employ various surface analytical methods to identify the possible damage precursors on fused silica surfaces treated with different processes (RIE, DCE, and their combination). The results show that RIE-induced defects, including F contamination, broken Si-O bonds, luminescence defects (i.e., NBOHCs and ODCs), and material densification, are potential factors that limit the improvement of laser-induced damage resistance of the optics. Although being capable of eliminating the above factors, the DCE treatment can achieve rough optical surface with masses of exposed scratches and pits which might serve as reservoirs of the deposits such as inorganic salts, thus limiting the further improvement in damage resistance of fused silica. The study guides us to a deep understanding of the laser-induced damage process in achieving fused silica optics with enhanced resistance to laser-induced damage by the combined treatment of RIE and DCE.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
High-peak-power lasers are important in developing high-energy-density science such as inertial confinement fusion (ICF) [1]. However, such systems are generally limited in output ability due to laser-induced damage (LID) in optical components [2]. Fused silica optics, which are widely used in the ICF systems, also face a serious risk of LID. For decades, extensive experimental and theoretical studies have been carried out to understand the damage initiation and growth behavior of fused silica for nanosecond pulses, and various surface defects that readily absorb sub-band-gap light have been determined to be the damage precursors of fused silica [3–6]. These defects are mainly associated with subsurface mechanical damage (SSD [7–10], typically including scratches, micro-cracks, and silica defects distributing on their surfaces) as well as the embedded absorbing impurities (e.g., Ce, Fe, etc.) introduced by grinding and polishing procedures. Besides, chemical structure defects such as oxygen-deficient center (ODC) and non-bridging oxygen hole center (NBOHC) are also possible indicators of subsurface fracture, whose luminescence properties have been widely reported [11,12].
Some success in increasing the laser-induced damage threshold (LIDT) of the fused silica has been achieved using various advanced techniques. For example, optical fabrication processes have been developed to minimize and potentially eliminate the SSD [13,14]. However, there is still significant room for damage performance improvement to approach its intrinsic threshold. Although the SSD layer of fused silica can be removed by using magnetorheological fluid finishing (MRF) technique [15], the iron in the magnetic fluid used in MRF process can contaminate the surface and decrease the surface LIDT of the optics. Moreover, due to a relatively small polishing area, multiple passes are required to polish large-aperture optics using MRF [16]. Dry etching techniques such as reactive ion etching (RIE), ion beam etching, reactive ion beam etching, and so forth have been used to physically remove the SSD of fused silica optics [17,18]. Unfortunately, bonding damage is often created in these processes due to the presence of ion bombardment. Furthermore, residue layers composed of reactant species, reaction products, and impurities which may permeate the etched silica material are also inherent during these processes [19]. Chemical wet etching routes have been previously shown to be very effective in increasing the LIDT of fused silica optics. The etching process named Advanced Mitigation Process (AMP) or Dynamic Chemical Etching (DCE) involves treating fused silica surface in the mixed aqueous solution of HF and NH4F under multi-frequency ultrasonic or megasonic agitation followed by a series of deionized water rinses [20,21]. However, this occurs at the expense of leaving behind etching traces (e.g., exposed scratches and pits) that seriously degrade the surface quality of the optic [22]. In our previous study, we attempted to address this issue by combining a RIE pretreatment with DCE to tracelessly remove the fractured defects in subsurface layer of fused silica. The combined process of RIE and DCE makes it possible to significantly increase the LIDT without surface quality degradation [21,23]. Since it is a complex set of interactions occurring at fused silica surface during the combined treatment, globally understanding the role of the defects introduced by each step is very important. For example, for each step of the combined treatment, which kinds of defects are removed or left? How will the performance change if the RIE treatment acts as the last step during the combined treatment? Although either RIE or DCE of fused silica has been separately studied before to explore the mechanism of laser damage induced by surface defects based on limited characterizations [18,20], the combination of them is still lack of deep investigation (at least from the perspective of chemical-structure defects).
In the following study, we investigate the laser-induced damage performance and surface morphology of fused silica with different treatment processes (RIE, DCE, and their combinations). We employ various surface analytical methods (including infrared spectroscopy, Raman spectroscopy, X-ray photoelectron spectroscopy, secondary ion mass spectrometer, and atomic force microscopy) to identify the possible damage precursors induced by these treatments. Samples preparation is detailed in Section 2.1. Laser-induced damage testing, surface morphology, and other characterizations are described in Section 2.2. These obtained results are presented in Section 3. Section 4 is presented for discussing the damage behavior of the treated samples concerning the analytical results. The conclusion of the work is given in Section 5.
2. Experimental
2.1 Sample preparation
Five square fused silica samples (Corning 7980) with the size of 50 mm × 50 mm × 5 mm were manufactured by the same vendor with conventional grinding and polishing processes. They were all cleaned with alkalescent detergent (10 vol% Micro 90 diluted with deionized water) and deionized water (18.2 MΩ) in an ultrasonic water bath as received, followed by surface treatment in different ways respectively, as shown in Table 1. Sample A0 was served as a virgin sample which experienced no any consequent treatments. Sample B1 was treated only by a RIE process with an etching depth of 1 μm. Sample B2 was first treated with a 1-μm RIE process and then by a 3-μm DCE process. Sample C1 was treated only by a 3-μm DCE process, while sample C2 was first treated with a 3-μm DCE process and then by a 1-μm RIE process. All the RIE processes, which were performed in a parallel-plate plasma discharge system with CHF3 and Ar as the raw etching gases, were the same to each other. The samples were placed at the center of a rotating graphite plate in the plasma chamber to realize uniform etching. The detailed etching parameters are shown in Table 2. The etching rate of the RIE process was calibrated prior to the sample preparation procedure by measuring the etching depth within a certain etching duration. The etching depth was obtained by masking a part of the sample surface during the etching process and measuring the step height with a surface profile meter after the mask was removed. The etching uniformity was also examined in the calibration. With the calibrated etching rates, the etching depth of the formal samples was controlled by adjusting the etching time. The samples were carefully cleaned and air-dried after the RIE treatment for subsequent characterizations. The DCE processes were conducted in a buffered HF/NH4F solution (where the volumetric ratio of HF to NH4F was 0.2) under ultrasonic conditions using a Teflon-lined, multi-frequency ultrasonic transducer (Blackstone multiSONIK 40, 80, 120, 140, 170, 220, and 270 kHz). The etching rate of the DCE process was calibrated in a similar way to that of the RIE process, which was measured to be about 200 nm/min. The etching depths for the RIE and DCE processes were determined to be 1 μm and 3 μm respectively based on our previous work [21]. More detailed information of this process is available in our previous publication [18]. All the cleaning, etching, and drying processes were implemented in a Class 100 clean room.
Table 1. Summary of sample treatments.
Table 2. Detailed parameters of RIE process.
2.2 Damage testing and characterization
2.2.1 Laser-induced damage testing
The laser damage resistance of the samples was assessed using a 1-on-1 laser damage testing strategy [24]. A Q-switched Nd: YAG laser generating pulses with the wavelength of 355 nm and the repetition rate of 1 Hz was used in the test. The pulse is single-longitudinal mode with the FWHM (full width half maximum) pulse duration of about 5 ns. During the test, the laser was focused on the exit surface (i.e., the treated surface) of the sample to achieve high fluence. The spatial profile and pulse energy of the beam were monitored by picking off a fraction of the beam. The beam had a near-Gaussian distribution with an efficient diameter of about 1.2 mm. The damage performance of the samples under laser irradiation was observed using a long-focal-distance microscopic CCD camera in situ. For the 1-on-1 laser damage test, each testing site was irradiated by one laser shoot with a certain fluence and twenty testing sites were chosen on the sample surface with the distance between adjacent two sites being 3 mm. The damage initiation probability versus laser fluence was obtained by calculating the number of damaged sites at each laser fluence.
2.2.2 Surface morphology characterization
An atomic force microscopy (AFM) (Dimension Icon, Bruker, Germany) was used to measure the surface roughness (RMS, root mean square) and morphologies of the samples. For each sample, the measurements were carried out with scanning sizes of 10 μm × 10 μm and 1 μm × 1 μm respectively to compare the surface morphologies at different scales. Three positions were selected randomly on each sample surface to calculate the average roughness.
2.2.3 Infrared spectroscopy
Infrared (IR) spectra were obtained using a Fourier transform infrared (FTIR) spectrometer (Nicolet iN10, Thermo Scientific, USA) in attenuated total reflection (ATR) mode with Ge as the ATR crystal and MCT/A (mercury cadmium tellurium) as the detector. This spectroscopy had an accuracy of 0.008 cm-1. During the measurements, the incident angle was fixed at 45o. The pressure loaded on the sample that was exerted by the ATR crystal was controlled by an electric torque knob to make sure that the pressure was the same in each measurement. The contact area of the ATR crystal and the sample was a circle with the diameter of 400 μm. The measurements were carried out in the range of 675 cm-1-4000 cm-1 with a resolution of 4 cm-1. During the measurement, five sites were randomly selected on each sample surface to provide an averaged spectrum.
2.2.4 Raman scattering measurement
Raman scattering measurements were carried out on a high-resolution confocal Raman spectroscopy (LabRAM HR Evolution, Jobin Yvon, France). A diode-pumped laser operating at 532 nm was used as the source light. A 100× objective lens was used as the focus and collection optic. The scattering light from the near-surface area of the samples was collected using the confocal system in the measurement. The Raman light was detected through an 1800 grooves/mm, 500 nm blazed holographic grating, with the Raman shift at the range of 200∼1750cm-1 and the resolution of ∼2 cm-1. The integration time for each spectrum was 10 s.
2.2.5 X-ray photoelectron spectroscopy
An X-ray photoelectron spectroscopy (XPS) (Quantera II, ULVAC PHI) with monochromatic Al Kα as the X-ray source was used to investigate the surface chemistries of the samples. The X-ray beam had a diameter of 100 μm in the measurement. The Ar ions were used as the sputtering source to expose the areas beneath the sample surfaces. Each sample was analyzed at different depths with an interval of 5.45 nm until the XPS spectra had no change. The binding energies were referenced to the C 1s photoelectron line at 284.8 eV arising from the atmospheric carbon adsorbed at the top surfaces of the samples.
2.2.6 O/Si stoichiometry and contamination analysis
A time-of-flight secondary ion mass spectrometer (ToF-SIMS, TOF.SIMS 5-100, ION-TOF GmbH, Germany) was used to investigate the stoichiometric ratio of O to Si and the distribution of impurity atoms as a function of depth beneath the sample surfaces. Since the depth profiles of metal elements on fused silica surfaces treated with different processes have been studied extensively and systematically in our previous work [18], only the probable Fluorine (F) contamination that might be introduced during the RIE or DCE process was investigated in the present study. The acceleration voltage of the primary ions was 30 keV with Bi as the ion beam source. The area of the sputtered crater was 200 μm × 200 μm. Cs+ was used as the secondary sputtering source with the acceleration voltage of 1 keV and incident angle of 45o, producing a sputtering rate of 0.087 nm/s for the fused silica. The samples were sputtered until the intensities of the interested elements did not change with depth. The relative sensitivity factor of O to Si was calibrated by assuming that the stoichiometric ratio of O/Si was 2 on the substrate of the material. The intensity of F was normalized with the silicon particle number as a standard.
3. Results and discussion
3.1 Laser-induced damage testing
The laser-induced damage probability versus laser fluence of the samples is shown in Fig. 1. The laser damage resistance is dramatically improved with all the etching processes (RIE, DCE, or their combination). Compared with the virgin sample A0, the 0% probability damage threshold of the 1-μm-RIE etched sample B1 increases more than 60% (from ∼9 to ∼15 J/cm2) and the damage threshold of 100% probability increases even more significantly (from ∼11.5 to ∼23 J/cm2). In comparison, the 3-μm DCE treatment (sample C1) can improve the damage performance more dramatically, with the damage threshold of 0% probability increased from ∼9 J/cm2 to nearly 18 J/cm2. However, as the DCE-treated sample is further treated with RIE (sample C2), the damage resistance turns back to a level that is even lower than that of the RIE-treated sample. Their damage thresholds of 0% probability are close while the damage threshold of 100% probability of sample C2 is much lower than that of sample B1. Amongst the four etching processes, the combined process of RIE and DCE exhibits the most excellent performance in improving the laser damage resistance of the fused silica sample. The sample treated with this process (sample B2) has the highest damage threshold (∼21.5 J/cm2 for 0% probability).
Fig. 1. Damage initiation probability versus laser fluence at 355 nm, 5 ns for the virgin sample and the samples treated with different processes. The quality factor (peak energy / average energy) of the laser beam spot was around 2.6.
3.2 Surface morphology
The typical surface morphologies and roughness of the samples treated with different processes in a 10 μm × 10 μm scanning region are shown in Fig. 2. As can be seen from Figs. 2(a)-(c), the virgin sample A0, the RIE-treated sample B1, and the combined-treated sample B2 (first RIE and then DCE) all have relatively smooth surfaces with the RMS roughness of 0.259 nm, 0.43 nm, and 0.47 nm respectively. No obvious defects such as cracks and scratches can be observed on the surfaces of samples A0 and B1 except for slim and shallow polishing traces. For sample B2, even the polishing traces can be hardly observed. However, as the sample is directly treated with the DCE process (sample C1), the surface quality deteriorates seriously with the RMS value increased dramatically from 0.26 nm to 1.66 nm. Deep and large physical-structure defects, i.e., scratches and pits, distribute discretely on the sample surface, as can be seen in Fig. 2(d). After the DCE-treated surface is retreated with RIE (sample C2, first DCE and then RIE), the surface quality of the sample is considerably improved with the RMS value decreasing to 0.844 nm. The result suggests that the RIE treatment can smooth the optical surface of fused silica, which has been extensively reported before [25]. It can also be noted from Fig. 2(e) that there are still large scratches on the sample surface, suggesting that the RIE treatment is incompetent to remove the as-exposed physical-structure defects. The difference in surface quality between samples B2 and C1 also implies that the RIE process can effectively remove the embedded cracks and scratches in the subsurface layer of polished fused silica.
Fig. 2. (a)-(e) Surface morphologies of the samples measured by AFM in a 10 μm × 10 μm scanning region. (f) Corresponding RMS roughness for all the samples.
The surface morphologies and roughness of the samples in a much smaller scanning region (1 μm × 1 μm) were subsequently studied, as shown in Fig. 3. The morphological results show that all the sample surfaces have similar high-frequency fluctuations at the scale of sub-nanometers. The RMS roughness values are all quite low (less than 0.4 nm). Compared to the virgin sample A0 and the samples treated with RIE as the last step (B1 and C2), the samples treated with DCE as the last step (B2 and C1) have a relatively high surface roughness. Much more tiny peaks with different heights can be observed on the surfaces of samples B2 and C1. Nevertheless, the difference in surface roughness at this scale might have no substantial influence on the laser-induced damage resistance of the samples, which will be shown below.
Fig. 3. (a)-(e) Surface morphologies of the samples measured by AFM in 1 μm × 1 μm scanning region. (f) Corresponding RMS roughness for all the samples.
The AFM results indicate that the RIE treatment can tracelessly remove the SSD of the conventionally polished fused silica without exposing or extending the topological structures of the scratches and cracks embedded in it. In contrast, the DCE treatment would expose and enlarge the fractured defects in the subsurface layer of the samples. As a result, the RIE and the combined RIE and DCE processes can both smooth the optical surface of fused silica, while the DCE and the combined DCE and RIE processes will leave a mass of open fractures such as scratches and cracks, causing degradation of the surface quality of the optics.
3.3 Infrared spectroscopy
The IR signals can provide rich information regarding defects and microstructures in the near-surface of fused silica, which mainly locate below the wavenumbers of 1500 cm-1. In our previous study, nearly no obvious absorption peaks can be found in the range of 1400∼4000 cm-1. We thus mainly focus on the FTIR spectra between 675 cm-1 and 1425 cm-1 in the present study. In this range, the sampling depth was calculated to be less than 1 μm with an incident angle of 45o and Ge as the ATR crystal, which can reflect the vibration modes in the near-surface region. Typical FTIR reflection spectra taken of the virgin sample and the samples treated with different processes are shown in Fig. 4(a). It can be noted from this figure that all the sample surfaces have the same absorption peak profile but the intensities are different. The corresponding Gaussian fit results for all the samples are shown in Figs. 4(b)-(f). Three peaks which are associated with the fundamental Si-O vibrational modes of tetrahedral SiO2 can be recognized from the Gaussian fit curves: peak centered at ∼800 cm-1 is attributed to the bending vibration (TO2) of Si-O-Si bond; peaks centered at ∼1120 cm-1 and ∼1200 cm-1 can be ascribed to the transverse optical (TO3) and longitudinal optical (LO3) modes respectively of the Si-O-Si asymmetric stretching vibration [26]. The TO3 peak can be decomposed into two subpeaks, named TO3-1 and TO3-2. In addition, the peak centered at ∼950 cm-1 is associated with the non-bridging (NB) Si-O or Si-OH stretching mode [27]. The Gaussian band parameters (peak position and height) obtained from FTIR spectra of the samples are summarized in Table 3.
Fig. 4. FTIR reflection spectra of the virgin sample and the samples treated with different processes. The black dash-dotted lines in panel (a) mark the locations of TO2, TO3, and LO3 vibration frequencies of the samples. The inset in panel (a) shows the spectra region of the non-bridging mode of the samples. The blue solid and dashed curves in panels (b)-(f) denote the multi-component Gaussian fit to the measured FTIR data of the samples.
Table 3. Gaussian band parameters obtained from FTIR spectra of the samples: peak positions (cm-1) and heights of the absorbance.
To compare the NB peaks of the samples, their heights were normalized by the TO2 peak heights respectively for each sample, as shown in Fig. 5(a). It can be learned from this figure that, samples A0, B1, and C2 have stronger absorption than samples B2 and C1 as for the NB stretching mode. Amongst these samples, the virgin one has the highest peak height, indicating that the polishing process might produce NBOHC defects in the near-surface region. As the polished fused silica samples undergo RIE or/and DCE treatments, the intensity of the non-bridging mode decreases. Especially, when DCE is used as the last step to treat the fused silica surface, the absorption level of the corresponding peak decreases, suggesting that DCE treatment can eliminate the NBOHC defects on fused silica surface. However, as the DCE-treated sample is further treated with RIE, the corresponding absorption rises again to the level of the sample only treated with RIE. This indicates that NBOHC might be introduced on fused silica surface during the RIE process. We also investigate the relative change of the TO3-1 and TO3-2 subpeaks, as shown in Fig. 5(b). It can be found from this figure that the samples treated with DCE as the last step had relatively high TO3-1/TO3-2 peak height ratios while the samples treated with RIE as the last step had relatively low ratios. The results also seem to tell us that chemical-structure defects introduced by RIE have a correlation with the ratio of TO3-1 to TO3-2.
Fig. 5. (a) NB peak heights normalized by TO2 peak heights and (b) ratio of TO3-1 peak height to TO3-2 peak height for all the samples.
3.4 Raman spectroscopy
The Raman spectra of the virgin sample and the samples treated with different processes are shown in Fig. 6, where the spectra are offset for clarity. The relatively broad bands in the spectra, including the peaks centered at ∼423 cm-1 and ∼800 cm-1 respectively, are originated from the coupled vibrational modes of the silica random network [28]. Besides, two relatively sharp bands centered at ∼496 cm-1 (D1) and ∼606 cm-1 (D2) respectively can be clarified from this figure. They are the typical bands of fused silica which stem from the in-phase breathing motions of oxygen atoms in puckered four- and planar three-membered ring structures respectively [12,28]. It is obvious from this figure that the spectra can be classified into three sets with the spectra in each set coincident with each other respectively: the first set is the black curve which is the spectrum of the virgin sample; the second set is the red and pink curves which are the spectra of the sample treated with RIE and the sample treated with DCE followed by RIE respectively; the last set is the green and blue curves which are the spectra of the sample treated with DCE and the sample treated with RIE followed by DCE respectively. Compared with the virgin sample, the relative intensities of D1 and D2 significantly increase for the sample treated with RIE but decrease for the sample treated with DCE. After the RIE-treated sample is retreated with DCE, the relative intensities of D1 and D2 decrease significantly to a level that is nearly the same as that of the DCE-treated sample. In contrast, after the DCE-treated sample is retreated with RIE, the relative intensities of D1 and D2 increase significantly to a certain level which is nearly the same as that of the RIE-treated sample. Since the increase in the number of small-membered rings indicates material densification, the results shown in Fig. 6 indicate that the DCE treatment can remove the densified layer of the conventionally polished or RIE-treated samples while the RIE treatment would induce even more intense material densification compared to the virgin sample. This is related to the disruption of the continuous random network of Si-O tetrahedron, which is attributed to the change in ring statistics where sixfold rings transform to threefold and fourfold rings upon the ion bombardment during the RIE procedure [29]. Similar results have also been suggested in the Refs. [25,30], which convince us that the RIE process would induce breakage of the Si-O bonds and material compression of the near-surface region of fused silica.
Fig. 6. Raman spectra of the virgin sample and the samples treated with different processes. The spectra are offset for clarity.
3.5 X-ray photoelectron spectroscopy
XPS was used to investigate stoichiometry and electronic states of Si and O for the samples. The first issue needed to be addressed is whether the intensity of the spectrum for each sample is dependent on the detecting depth (especially for the RIE-treated sample). We thus measured the XPS spectra of the samples with different detecting depths by sputtering the fused silica surface with Ar ions. However, the obtained results show that there is no obvious variation in spectrum intensity as the sputtering begins and continues to remove the material for several nanometers (e. g., 5.45 nm and 10.9 nm). The difference in spectrum intensity is observed only between the sputtered surfaces and the un-sputtered surfaces, as shown in Fig. 7. The XPS spectra of the five samples coincide with each other for the same sputtering depth (5.45 nm). A little C 1s peak can be detected only on the surfaces without Ar-ion sputtering, which is due to contamination from the environment. No other elements except for O and Si are detected either on the sputtered surfaces and the un-sputtered surfaces of the samples, which indicates that the DCE and/or the RIE treatments introduce no obvious contaminations in the present study. Table 4 summarizes the XPS compositional analysis in atomic percentage of the sputtered fused silica surfaces. We can conclude from this table that neither RIE nor DCE treatments will change the stoichiometric ratios of O to Si in the subsurface regions of the samples. Nevertheless, the detection of F contamination and O/Si stoichiometry with much higher sensitivity and accuracy was conducted in our work using ToF-SIMS, with the results being presented in the following section.
Fig. 7. XPS spectra of the surfaces with or without Ar-ion sputtering for all the samples. (a) Un-sputtered surfaces; (b) sputtered surfaces (at 5.45 nm detecting depth).
Table 4. Elemental analysis from XPS measurements.
Figure 8 compares the high-resolution O 1s and Si 2p spectra measured in the subsurface regions of the virgin sample and the samples treated with different processes. Single bands at 532.7 eV and 103.4 eV can be observed respectively in Figs. 8(a) and (b), which correspond to O2- reduction state and Si4+ state respectively associated with SiO2 [31]. It is clear from Fig. 8(a) that the DCE treatment would cause an obvious decrease in the binding energy of O 1s both for the conventionally polished sample and RIE-pretreated sample, which indicates an increase in the amounts of accepted electrons for the oxygen atoms in SiO2 [32]. However, as the DCE-pretreated sample is retreated with RIE, the binding energy increases instead, indicating a decrease in the amounts of accepted electrons. Since the stoichiometric ratio of O to Si does not change for all the samples as Table 4 shows, the decrease in amounts of accepted electrons for O implies an increase in non-bridging oxygen atoms. The XPS results shown in Fig. 8(a) are in good agreement with the FTIR results shown in Figs. 4 and 5. It thus makes us believe that there are a certain amount of non-bridging oxygen atoms in the subsurface region of the conventionally polished fused silica. It probably originates from the intrinsic silica defects distributed on the fractured surfaces in the SSD layer [29]. The DCE treatment can effectively reduce the concentration of the non-bridging oxygen atoms since it can remove the surface layers of the cracks and fractures without introducing new silica defects. In contrast, the RIE treatment might induce new non-bridging Si-O on the sample surface with the SSD layer removed, which would happen both for the originally polished sample and the DCE-pretreated sample.
Fig. 8. High-resolution spectra of (a) O 1s and (b) Si 2p for the virgin sample and the samples treated with different processes.
3.6 O/Si stoichiometry and F contamination analysis
Figure 9 shows the stoichiometric ratio of O to Si as a function of depth beneath the surface for the virgin sample and the samples treated with different processes. For all the samples, the peak O/Si ratios are as high as 4 which locate on the top surfaces and decrease quickly to a constant value of 2 at the depth of ∼2 nm. This is probably because some gaseous oxygen from the atmosphere is adsorbed on the surface of the samples. The O/Si-ratio profiles of all the samples nearly coincide with each other which indicates that the RIE or DCE treatments induce no change in the O/Si stoichiometry. Nevertheless, this does not imply that the RIE or DCE processes induce no change in the O-Si networks or chemical states of the atoms at the near-surface region of fused silica. From our previous work, the RIE treatment would increase the photoluminescence (PL) signals that are related to the ODC [23,33,34]. The results in Fig. 9 suggest that the lost oxygen atoms accompanied by the generation of ODCs did not escape from the material surfaces in the present case. We thus believe that the lost oxygen atoms remain in the Si-O networks in the forms of interstitial oxygen or peroxy radials or bridges.
Fig. 9. The O/Si stoichiometry versus depth detected by ToF-SIMS for the virgin sample and the samples treated with different processes.
The distribution of F as a function of detecting depth is shown in Fig. 10. The samples B2 and C1 treated with DCE as the last step have relatively low F concentration, which is nearly the same to the virgin sample A0. Compared to these three samples, the samples B1 and C2 treated with RIE as the last step have much higher F concentration along the measured depth. In addition, these two samples have nearly the same depth profiles of F with each other. The results indicate that the RIE treatment would introduce F contamination in the surface region of the fused silica with a penetration depth of at least 20 nm. In contrast, the 3-μm DCE treatment can effectively remove the F-contaminated layer induced by the 1-μm RIE treatment while not introducing additional F on the sample surface. The results also indicate that the penetration depth of the F contamination introduced by the 1-μm RIE is lower than 3 μm.
Fig. 10. Depth profiles of F in the near-surface region detected by ToF-SIMS for the virgin sample and the samples treated with different processes.
4. Discussion
Laser-induced damage precursors that may be present in the near-surface region of a conventionally polished fused silica optics include extrinsic precipitates or organic contamination that adhere tightly on the surface, extrinsic impurities (mainly metallic elements originated from the polishing powder) that embedded in the Beilby layer or SSD layer, and intrinsic silica defects that locate on the fractured surfaces [35]. From the previous study, both the RIE and DCE processes can remove these precursors effectively. However, there is a substantial difference in the laser-induced damage resistance when the fused silica surface is treated with either of the two processes or combined [21,23]. In our present study, the investigations of the surface morphology, chemical states, and possible contamination in the near-surface region of the samples treated with different processes guide us to distinguish the factors that play roles in determining the UV laser damage performance of fused silica. Moreover, they help to improve our understanding on the underlying mechanism of the excellent damage performance of fused silica when its surface is treated with the combined process of RIE and DCE, which has been proposed by our group before [21]. In this section, the following two aspects will be discussed concerning the laser-induced damage resistance with respect to the characterization results obtained by using the various analytical techniques: first, what exactly happens in the near-surface of fused silica during each treatment process (either of the two single processes or their combined processes)? Second, why the surfaces treated with RIE as the last step do not show the same excellent damage performance as the treated ones with DCE as the last step and what kind of damage precursors are responsible for that with different etching processes?
For all the processes (single or combined) mentioned above, the SSD layer on fused silica surface can be considerably removed but the resulting surfaces are different. It can be noted from Figs. 4, 5, 6, and 8 that the chemical and mechanical states of the fused silica surface are largely determined by the final step of the treatment that the sample undergoes. For a conventionally polished fused silica surface, there are a large number of metallic elements and broken Si-O bonds which contribute to the chemical structural defects (i.e., the NBOHCs) and material densification in the near-surface region. They can be thoroughly removed along with the SSD layer by the 1-μm RIE process, which leads to an obvious increase in the LIDT (see Fig. 1). However, the RIE process itself would induce new chemical-structure defects either for an originally polished surface or a DCE-pretreated surface. These defects include broken Si-O bonds (see Fig. 8), NBOHCs (see Figs. 4 and 5), material densification (see Fig. 6) as well as ODCs (Refs. [23,33,34]). The XPS and the ToF-SIMS results shown in Figs. 8 and 9 make us believe that the RIE treatment will not change the O/Si ratio of the material. Besides the chemical-structure defects, trace amounts of F contamination are introduced in the RIE process with the penetration depth of more than 20 nm (see Fig. 10).
Different from the RIE treatment, the DCE treatment would not induce new chemical defects, material densification, or F contamination as it removes the pre-existing defects on the sample surface (see Figs. 4, 6, 8, and 10). Therefore, compared to the samples treated with RIE as the final step, the samples treated with DCE as the final step present more significant enhancement in LIDT. That is, samples B2 (first RIE and then DCE) and C1 (DCE alone) have higher LIDTs than samples B1 (RIE alone) and C2 (first DCE and then RIE), as Fig. 1 indicates. For the former two samples, the sample treated with the combined process of RIE and DCE has even much higher LIDT than the sample treated only with DCE process. There are two factors that may play roles in leading to this result. For one hand, it has been proved that relatively large etched amount (> 20 μm) was needed for a single DCE treatment to maximize the LIDT of the optics [20]. Apparently, 3-μm DCE treatment might not be sufficient to thoroughly remove the damage precursors in the subsurface layer of the conventionally polished fused silica, which restricts the further enhancement in LIDT. For another, as shown in Fig. 2, the sample treated only by the DCE process has a quite rough surface with many enlarged scratches and pits exposed on it. These opened scratches and pits would serve as reservoirs of the redeposited silica compounds during the DCE process or deposited inorganic salts during the rinsing and drying process, which would lower the LIDTs of the samples [20,36]. In contrast, sample C1 which is pretreated with RIE process has a relatively smooth surface with no evident scratches and pits. The smooth surface can decrease the deposition probability of compounds and inorganic salts, leading to a more significant enhancement in LIDTs. We thus speculate surface morphology is another important factor limiting the damage resistance improvement of the DCE-treated sample B2. This speculation can be verified by the discrepancies in the damage performances between samples B1 and C2. Both of these two sample surfaces present a similar level of broken Si-O bonds, F contamination, and material densification in the near-surface regions (see the FTIR, XPS, Raman scattering, and ToF-SIMS results). The only obvious difference between them is that sample C2 has evident scratches and pits on its surface while sample B1 doesn’t, which makes the former sample has a little lower LIDTs than the latter one (see Fig. 1). The surface morphology might play a key role in affecting the high-fluence damage performance of the optics since the damage thresholds of 0% probability of the two samples are close.
5. Conclusion
We investigated the damage performance and surface quality of the fused silica with different treatment processes (RIE, DCE, or their combination). Various surface analytical methods were used to identify the potential damage precursors introduced by these treatments. The results demonstrated that potential factors limiting the improvement of laser-induced damage resistance of the RIE-treated fused silica include F contamination, broken Si-O bonds and luminescence defects (i.e., NBOHCs and ODCs), and material densification. Factors limiting the improvement in damage resistance of DCE-treated fused silica are exposed scratches and pits which can serve as reservoirs of the deposits including inorganic salts. For this reason, orderly coupling RIE with DCE is critical for achieving high-quality fused silica optics with excellent damage performance and surface quality.
To further extend this study, deeper investigations will be performed to accurately determine the influencing depth of each step during the combined process of RIE and DCE. Physical models of fused silica with different potential damage precursors (broken Si-O bonds or F contamination) encountered in this study will be also established and studied. Such studies could help us better improve the post-treatment process and gain insight into the fundamental damage mechanisms of fused silica used in high-fluence laser systems.
Funding
National Natural Science Foundation of China (61705204, 61705206, 61805221, 62005258); Laser Fusion Research Center Funds for Young Talents (LFRC-CZ028, RCFPD3-2019-2); Science and Technology on Plasma Physics Laboratory, China Academy of Engineering Physics (2018).
Acknowledgments
The authors wish to acknowledge the Raman scattering measurement performed by Wei Le.
Disclosures
The authors declare no conflicts of interest.
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. H. B. Radousky, A. M. Rubenchik, M. D. Feit, and S. G. Demos, “Localized dynamics during laser-induced damage in optical materials,” Phys. Rev. Lett. 92(8), 087401 (2004). [CrossRef]
3. S. G. Demos, B. N. Hoffman, C. W. Carr, D. A. Cross, R. A. Negres, and J. D. Bude, “Mechanisms of laser-induced damage in absorbing glasses with nanosecond pulses,” Opt. Express 27(7), 9975–9986 (2019). [CrossRef]
4. N. Shen, J. D. Bude, and C. W. Carr, “Model laser damage precursors for high quality optical materials,” Opt. Express 22(3), 3393–3404 (2014). [CrossRef]
5. R. A. Negres, D. A. Cross, Z. M. Liao, M. J. Matthews, and C. W. Carr, “Growth model for laser-induced damage on the exit surface of fused silica under UV, ns laser irradiation,” Opt. Express 22(4), 3824–3844 (2014). [CrossRef]
6. A. A. Manenkov, “Fundamental mechanisms of laser-induced damage in optical materials: today’s state of understanding and problems,” Opt. Eng. 53(1), 010901 (2014). [CrossRef]
7. L. Hongjie, “Subsurface defects of fused silica optics and laser induced damage at 351 nm,” Opt. Express 21(10), 12204–12217 (2013). [CrossRef]
8. T. Suratwala, L. Wong, P. Miller, M. Feit, J. Menapace, R. Steele, P. Davis, and D. Walmer, “Sub-surface mechanical damage distributions during grinding of fused silica,” J. Non-Cryst. Solids 352(52-54), 5601–5617 (2006). [CrossRef]
9. P. E. Miller, T. I. Suratwala, L. L. Wong, M. D. Feit, J. A. Menapace, P. J. Davis, and R. A. Steele, “The distribution of subsurface damage in fused silica,” Proc. of SPIE 5991, 599101 (2005). [CrossRef]
10. W. Liao, Y. Dai, Z. Liu, X. Xie, X. Nie, and M. Xu, “Detailed subsurface damage measurement and efficient damage-free fabrication of fused silica optics assisted by ion beam sputtering,” Opt. Express 24(4), 4247–4257 (2016). [CrossRef]
11. J. Fournier, J. Neauport, P. Grua, E. Fargin, V. Jubera, D. Talaga, and S. Jouannigot, “Green luminescence in silica glass: A possible indicator of subsurface fracture,” Appl. Phys. Lett. 100(11), 114103 (2012). [CrossRef]
12. S. Kucheyev and S. Demos, “Optical defects produced in fused silica during laser-induced breakdown,” Appl. Phys. Lett. 82(19), 3230–3232 (2003). [CrossRef]
13. T. Suratwala, R. Steele, M. Feit, L. Wong, P. Miller, J. Menapace, and P. Davis, “Effect of rogue particles on the sub-surface damage of fused silica during grinding/polishing,” J. Non-Cryst. Solids 354(18), 2023–2037 (2008). [CrossRef]
14. 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 351 nm 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), 71321I-71321I-71327.
15. R. Catrin, J. Neauport, D. Taroux, P. Cormont, C. Maunier, and S. Lambert, “Magnetorheological finishing for removing surface and subsurface defects of fused silica optics,” Opt. Eng. 53(9), 092010 (2014). [CrossRef]
16. R. Catrin, J. Neauport, D. Taroux, C. Maunier, S. Lambert, and O. Rondeau, “Contamination of fused silica optics surface after magnetorheological polishing,” in Optical Fabrication and Testing(Optical Society of America, 2014), p. OM4B. 2.
17. M. Xu, F. Shi, L. Zhou, Y. Dai, X. Peng, and W. Liao, “Investigation of laser-induced damage threshold improvement mechanism during ion beam sputtering of fused silica,” Opt. Express 25(23), 29260 (2017). [CrossRef]
18. 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]
19. S. J. Fonash, “An Overview of Dry Etching Damage and Contamination Effects,” J. Electrochem. Soc. 137(12), 3885–3892 (1990). [CrossRef]
20. 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]
21. 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]
22. M. Feit, T. Suratwala, L. Wong, W. Steele, P. Miller, and J. Bude, “Modeling wet chemical etching of surface flaws on fused silica,” in Laser Damage Symposium XLI: Annual Symposium on Optical Materials for High Power Lasers(International Society for Optics and Photonics, 2009), 75040L-75040L-75013.
23. 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]
24. L. M. Sheehan, M. R. Kozlowski, C. J. Stolz, F. Y. Genin, M. J. Runkel, S. Schwartz, and J. Hue, “Large-area damage testing of optics,” in Optical Instrumentation & Systems Design, 2775357–369 (International Society for Optics and Photonics, 1996). [CrossRef]
25. F. Frost, R. Fechner, D. Flamm, B. Ziberi, W. Frank, and A. Schindler, “Ion beam assisted smoothing of optical surfaces,” Appl. Phys. A 78(5), 651–654 (2004). [CrossRef]
26. M. Tomozawa, J.-W. Hong, and S.-R. Ryu, “Infrared (IR) investigation of the structural changes of silica glasses with fictive temperature,” J. Non-Cryst. Solids 351(12-13), 1054–1060 (2005). [CrossRef]
27. R. M. Almeida, T. A. Guiton, and C. G. Pantano, “Characterization of silica gels by infrared reflection spectroscopy,” J. Non-Cryst. Solids 121(1-3), 193–197 (1990). [CrossRef]
28. A. Pasquarello and R. Car, “Identification of Raman Defect Lines as Signatures of Ring Structures in Vitreous Silica,” Phys. Rev. Lett. 82(19), 3230–3232 (1998). [CrossRef]
29. H. Sugiura, R. Ikeda, K. Kondo, and T. Yamadaya, “Densified silica glass after shock compression,” J. Appl. Phys. 81(4), 1651–1655 (1997). [CrossRef]
30. B. Li, C. Hou, C. Tian, J. Guo, X. Xiang, X. Jiang, H. Wang, W. Liao, X. Yuan, X. Jiang, and X. Zu, “Layer by layer exposure of subsurface defects and laser-induced damage mechanism of fused silica,” Appl. Surf. Sci. 508, 145186 (2020). [CrossRef]
31. F. J. Himpsel, F. R. McFeely, A. Talebibrahimi, J. A. Yarmoff, and G. Hollinger, “Microscopic structure of The SiO2/Si interface,” Phys. Rev. B 38(9), 6084–6096 (1988). [CrossRef]
32. J. F. Moulder, J. Chastain, and R. C. King, Handbook of X-ray photoelectron spectroscopy: a reference book of standard spectra for identification and interpretation of XPS data, (Physical Electronics, Eden Prairie, MN, 1992).
33. L. Sun, T. Shao, J. Xu, X. Zhou, X. Ye, J. Huang, J. Bai, X. Jiang, W. Zheng, and L. Yang, “Traceless mitigation of laser damage precursors on a fused silica surface by combining reactive ion beam etching with dynamic chemical etching,” RSC Adv. 8(57), 32417–32422 (2018). [CrossRef]
34. T. Shao, L. Sun, W. Li, X. Zhou, F. Wang, J. Huang, X. Ye, L. Yang, and W. Zheng, “Understanding the role of fluorine-containing plasma on optical properties of fused silica optics during the combined process of RIE and DCE,” Opt. Express 27(16), 23307–23320 (2019). [CrossRef]
35. P. Miller, T. Suratwala, J. Bude, T. Laurence, N. Shen, W. Steele, M. Feit, J. Menapace, and L. Wong, “Laser damage precursors in fused silica,” in Laser Damage Symposium XLI: Annual Symposium on Optical Materials for High Power Lasers (International Society for Optics and Photonics, 2009), 75040X-75040X-75014.
36. J. Bude, P. Miller, S. Baxamusa, N. Shen, T. Laurence, W. Steele, T. Suratwala, L. Wong, W. Carr, and D. Cross, “High fluence laser damage precursors and their mitigation in fused silica,” Opt. Express 22(5), 5839–5851 (2014). [CrossRef]
