Open Access
Add to BibSonomy
Abstract
With combined lifetime and intensity spectra in time-resolved photoluminescence (PL) spectroscopy, the origins of a green PL band consisting of a 2.39eV-band with lifetime ~10μs and a 2.25eV-band with lifetime ~50μs, of high-purity fused silica under ArF excimer laser (6.4eV) excitation were revealed. The 2.39eV PL defects from the surface were annealed due to the deoxidation of the dioxasilyrane group ( = SiO2). The 2.25eV PL defects from bulk showed a UV induced growth. Mechanical and laser damage induced growth of the 2.25eV PL band confirmed that it was due to physical disorder from bending of Si-O-Si bonds. Theoretical calculations further assigned the 2.25eV PL band to silanone groups ( = Si = O), which can be created by relaxation process of strained Si-O bonds in SiO2 network.
© 2017 Optical Society of America
1. Introduction
Due to excellent laser damage resistance capability and low deep ultraviolet (DUV) absorption, high-purity fused silica is widely used in UV spectral region, such as in high-energy laser systems like national ignition facility (NIF) [1] and today’s most advanced DUV lithography systems [2]. However, the optical performances of fused silica optics are strongly limited by defects mainly induced due to manufacturing processes. In high-energy laser systems, photo-active impurities [3] and surface/sub-surface fracture-induced defects [4] strongly limit the laser-induced damage thresholds of fused silica optics. Damage growth of fused silica windows under UV laser irradiation is the limiting factor for long-term operation of high-energy laser systems [5]. In DUV lithography systems, stationary absorption coefficients of type III fused silica at 6.4eV are between 0.0045 and 0.0094 cm−1 [6], which are over 4 times higher than the theoretical limit ~0.0012cm−1 caused by Rayleigh scattering and Urbach tail absorption [7]. As a result, absorption due to color centers, such as E’-centers and non-bridging oxygen hole centers (NBOHCs) [8], is non-negligible. In addition, highly strained physical disorders, which could be decomposed more easily under UV irradiation and form various dangling defects [9], speed up the deterioration of UV transmittance of fused silica [10]. Although great effort has been paid to investigate the defects of fused silica in order to improve the laser damage threshold and UV robustness of fused silica optics, the defect formation mechanisms of fused silica under UV laser irradiation are still poorly understood.
Photoluminescence (PL) is a sensitive method for defect detection and can provide real-time information on microscopic structure changes in SiO2 network during UV irradiation of fused silica. Two luminescent defects, NBOHC with ~1.90eV PL band (0.16~0.18eV full width at half maximum (FHWM), 15~20μs lifetime) [11], and oxygen vacancies center II (ODC II) with ~4.44eV (~0.44eV FHWM, 3.9~4.1ns lifetime) [12,13] and ~2.74eV (~0.30eV FHWM, 6-10ms lifetime) [13,14] PL bands, which play an important role in optical performance of UV fused silica optics, have been comprehensively investigated. It was found that NBOHC has a wide UV absorption band, which consist of 4.7eV (1.0eV FHWM), 5.5eV (0.9eV FHWM), 6.1eV (0.8eV FHWM), 6.6eV (0.8eV FHWM), and 7.3eV (1.4eV FHWM) bands [8], and ODC II has 6.8~7.0eV (0.1~0.3eV FHWM) and 4.9~5.0eV (0.15eV FHWM) absorption bands [13]. A green PL band, ranging from 2.2 to 2.4eV and commonly reported at laser-induced or mechanically damaged sites [15–17], is not firmly attributed to any known luminescent defects. This 2.3eV green PL band was firstly reported in high power 355nm laser induced damage sites and assigned to self-trapped exciton (STE) [15]. However, further investigation on the emitter of this green PL peak shows behavior different from STE in different temperatures [18], making STE explanation unconvincing. On the other hand, UV excited green PL bands with lifetime ranging from nanoseconds to microseconds has been reported in various bulk fused silica samples and assigned to peroxide radical (POR) (~2.25eV, ~0.2eV FWHM, 300ns lifetime) [19], E’ δ-center (~2.25eV, ~0.4eV FWHM, 25ns lifetime) [20], H-related defects (~2.35eV, ~12μs lifetime) [21] and copper (Cu) impurity (~2.25eV, ~0.44eV FWHM, ~40μs lifetime) [22]. However, none of these assignments could be confirmed to be the origins of this damage-induced green PL band.
In this paper, we investigate in details the formation/annealing processes of overlapped green PL bands in high-purity fused silica using time-resolved PL spectroscopy. A lifetime spectrum method is introduced to efficiently resolve severely overlapped PL bands. Furthermore, possible correlations between green PL bands and laser/mechanical induced damage have been investigated with combination of time-resolved PL and damage experiments of high purity SiO2 samples.
2. Experimental arrangement
2.1 Lifetime spectrum method
In a PL spectroscopy consisting of n PL bands with initial intensity Ik(λ) at wavelength λ and single-exponential decay lifetime τk for kth PL band, the PL signal S(λ, td) at delay time td can be described as
As a result, initial intensity, band width, and lifetime of each PL band can be resolved by multi-parameter fitting time-resolved PL signals with different delay time to Eq. (1). The number of overlapped PL bands, which is usually unknown, is the key to obtained accurately fitting parameters since it directly determines the number of fitting parameters in the multi-parameter fitting. Normally the number of overlapped bands can be estimated from the peaks of PL intensity spectroscopy if the bands are not severely overlapped, and/or by multi-exponential fitting at a single wavelength if the lifetimes of overlapped PL bands are significantly different (such as in different orders of magnitude). However, when PL bands are severely overlapped and the lifetimes of severely overlapped bands are comparable, resolving severely overlapped PL bands becomes extremely difficult in the traditional way.In this case, a lifetime spectrum method is introduced to determine the number of severely overlapped PL bands with comparable lifetime. Taking into consideration that the lifetimes of overlapped PL bands are comparable, a single-exponential decay function is used to fit the time-resolved PL signal to obtain a lifetime at each wavelength. The lifetime spectrum is the dependence of the fitted lifetime on wavelength. In a mathematical view, the profile of the lifetime spectrum depends on both relative intensity and lifetime of each PL band. As a result, the number of overlapped PL bands can be determined from the profile of the lifetime spectrum. For example, flat regions in lifetime spectrum normally correspond to peaks of PL bands. Once the number of the PL bands is determined, the initial intensity Ik(λ) and lifetime τk of each PL band can be resolved by multi-parameter fitting the measured PL signals to Eq. (1), regardless of the line shape of each PL band.
2.2 Fused silica samples
Seven high purity SiO2 samples (with Cu content < 3 ppb), including 6 fused silica samples with different H and OH contents (Corning 7980, Corning 8655 from Corning, and Spectrosil 2000, Suprasil 711, Suprasil 501, Suprasil 401 from Heraeus) and one pure quartz sample (Clear Quartz from Heraeus) with low H and OH contents (both not detectable) were tested in the experiments. PL signals from two fused silica samples (a Suprasil 401 and a Corning 7980 sample, respectively) with thickness of 8mm were firstly measured. Damage experiments were performed on these two samples and the quartz sample with thickness of 2mm. Besides, PL signals of all samples with 2mm thickness were measured to find possible correlation between green band and other known PL bands. The OH contents of all samples are measured with a FTIR spectrophotometer. The measurement uncertainty for the OH content is approximately 5ppm.
2.3 Time-resolved photoluminescence measurements
The time-resolved PL signals from fused silica samples excited by 6.4eV excimer laser pulses (pulse width: ~8ns) are collected at room temperature and N2 atmosphere with a high-resolution spectrometer (iHR320, JOBIN YVON, grating grooves: 300 line/mm, spectral resolution: 0.79nm) equipped with an ICCD (intensified charge coupled device, rise time: ~135ns). The spectral response of the spectrometer is calibrated by a tungsten lamp. The PL signals are detected from the front surfaces of fused silica samples with a 45° detection angle from the incident laser beam. Long wave pass filters are employed to eliminate interruption from high-order diffraction of 6.4eV laser beam and from short-wavelength PL signals. Lifetime spectrum is obtained from a group of PL signals detected with delay time ranging from 2.0 to 200.0μs and delay gate 2.0μs. Before lifetime spectrum measurements, fused silica samples are pre-irradiated until relative intensity change becomes less than 5% to ensure stability of PL signals. In addition, PL signals are recorded with different detection angles to separate the PL contributions from surface and bulk of fused silica samples.
2.4 Damage experiments
Mechanical and laser-induced damage experiments are performed on fused silica samples. Damages are performed at room temperature in air. Mechanically induced damage sites are created by a diamond triangular pyramid tip with a 0.75N load and 0.048N∙s momentum. Laser induced damage sites are created by focused 1064nm laser pulses (repetition rate 1Hz; fluence16.3J/cm2; pulse width: ~25ns). Both mechanical and laser-induced damage experiments are repeated three times for each sample and it is found the results are repeatable. PL signals at different positions of damage sites excited by focused 6.4eV excimer laser spot (repetition rate 400Hz; fluence 57.8mJ/cm2; spot diameter: 60μm) are measured at 5μs delay time and 100μs gate. Damage sites are also observed by a differential interference microscope.
3. Results
3.1 PL band separation for fused silica samples
Figure 1(a) shows typical fused silica’s PL spectra at 6.4eV irradiation from Suprasil 401 sample measured at different delay time. From intensity changes in PL spectra obtained without/with 2.0μs delay time, three PL peaks (at 1.90eV, 2.25eV, and 3.25eV, respectively) with lifetime in microsecond range can be found. Besides, short-lived (lifetime<<700ns) PL signals in the wavelength from 2.5 to 3.7eV can also be observed. However, as the focus of this paper is the green PL bands, these PL signals are not discussed here. Since only microsecond-lifetime green bands are detected, for time-resolved PL analysis the delay time is set to ≥ 2.0μs in order to eliminate short-lived PL bands in the PL measurements. In Fig. 1b, Decay curves at three PL peaks can be well fitted to a single-exponential decay function with lifetime ~14μs, ~53μs and ~67μs for the 1.90eV, 2.25eV and 3.29eV bands respectively. On the other hand, decay curves at two valleys can be well fitted to a double-exponential decay function with lifetimes of adjacent PL bands.
Fig. 1 a) PL intensity spectra from a 8mm-thick Suprasil 401 sample excited by 6.4eV excimer laser at 400Hz repetition rate and 7.2mJ/cm2 fluence without and with 2μs delay time and 100μs gate time after pre-irradiation treatment; b) Normalized delay curves at different wavelengths from Suprasil 401 sample. Symbols are experimental data measured with 2μs delay gate. Dotted lines are single-exponential fits for peaks (1.9eV, 2.25eV and 3.25eV) and double-exponential fits for valleys (2.00eV and 2.75eV).
Figure 2 shows the lifetime spectra and corresponding PL signals for Suprasil 401 and Corning 7980 samples at 45°detection angle. For Suprasil 401 sample (Fig. 2(a)), from the lifetime spectrum four PL bands are resolved: a 3.09eV PL band (~0.45eV FWHM, 102.2 ± 2.3μs lifetime), which is due to Ge-ODC [13]; a 1.90eV PL band (~0.14eV FWHM, 14.8 ± 0.6μs lifetime), which is confirmed to be the well-known NBOHC [11], a 2.25eV (~0.34eV FWHM, 53.3 ± 1.5μs lifetime) and a 2.39eV PL band (~0.45eV FWHM, 10.0 ± 2.8μs lifetime), which are rarely reported in bulk fused silica. From Fig. 2e, the “2.25eV green band” in intensity spectrum is irreversibly red-shifted after 6.4eV irradiation. This also indicates this green band actually consists of two PL bands from different emitters and further confirms the resolved 2.39eV PL band is reliable. From the separated PL bands this red-shift results from the laser irradiation induced intensity increase of the 2.25eV PL band as well as the intensity decrease of the 2.39eV PL band.
Fig. 2 Lifetime spectra (a, b), corresponding normalized RMSE (root-mean-square error) (c, d), and intensity spectra (e, f) from Suprasil 401 (a, c, and e) and Corning 7980 (b, d and f) samples with 8mm thickness excited by 6.4eV excimer laser at 400Hz repetition rate and 7.2mJ/cm2 fluence. In lifetime spectra (a, b), resolved lifetimes of PL bands are shown in shadow areas. In intensity spectra (e, f), integrated PL signals (full lines) and resolved PL bands (dotted lines) measured with 5μs delay time and 100μs gate time before pre-irradiation (red lines) and after 1.5kJ/cm2-dose irradiation (black lines) are shown respectively. Arrows in PL intensity spectra show dominant intensity changes of PL bands due to pre-irradiation process.
For Corning 7980 sample (as shown in Fig. 2(b)), only three PL bands are observed: a wide 2.80~3.00eV PL band (~2eV FWHM, 8.2 ± 1.8μs lifetime), which is similar to the reported STE PL band [21], a 1.90eV PL band (~0.14eV FWHM, 12.3 ± 0.7μs lifetime) and a 2.25eV PL band (~0.31eV FWHM, 58.6 ± 1.9μs lifetime). The similar line shapes and lifetimes of the 1.90eV and 2.25eV PL bands observed in both samples indicate that 1.90eV and 2.25eV PL defects are common in these fused silica samples. However, excitation processes of these two PL bands are quite different. As shown in Fig. 3, the peak intensities of green bands are linearly dependent on the laser intensity, corresponding to a one-photon excitation process at 6.4eV irradiation, while the 1.90eV PL band is quadratically dependent on the laser intensity, corresponding to a two-photon excitation process.
Fig. 3 Fluence dependence of normalized peak intensity of 2.39eV (cyan), 2.25eV (green) and 1.90eV (red) PL bands for Suprasil 401 (triangular dots) and Corning 7980 (round dots) at 400Hz repetition. The dotted lines are linear fits for green bands and quadratic fits for 1.9eV band.
3.2 Surface and bulk PL signals
Figure 4 shows the normalized surface and bulk PL signals separated by the dependence of PL signal on detection angle along the normal direction. Assuming the defect distributions in bulk and surface are different and homogeneous, the ratio of detected surface to bulk PL signals can be simplified to be a sine function of detection angle mainly due to refraction and change of bulk PL signal’s detection area under large detection angle. As a result, relative intensities of PL bands from surface or bulk can be easily obtained from PL signals at different detection angles. The 2.25eV PL band is dominating in bulk PL signal, while the 2.39eV PL band is dominating in surface PL signal, indicating the intensity changes of these PL bands are independent.
Fig. 4 Normalized surface (red lines) and bulk (black lines) PL signals (full lines), and corresponding PL bands (dotted lines) resolved by PL signals measured at different detection angles for 8mm-thick Suprasil 401 (a) and Corning 7980 (b) samples excited by 6.4eV excimer laser at 400Hz repetition rate and 5.6mJ/cm2 fluence.
3.3 Damage sites and corresponding PL signals
Figure 5 shows typical mechanical and laser induced damage sites and corresponding PL signals. Before damage process, the initial PL signal from quartz is quite different from that from the fused silica samples. The PL signal from quartz consists of a 1.90eV PL band and a wide UV PL bands with ~5μs lifetime which is partly filtered out by the long-wave pass filter and not discussed in this paper. It is noticed that no green PL band is observed in the initial PL signal of quartz sample, indicating the 2.25eV PL defects do not exist in quartz network with a periodic structure.
Fig. 5 Mechanical damage sites (a, b and c) and laser induced damage sites (g, h and i) from Quartz (a, g), Suprasil 401 (b, h), and Corning 7980 (c, i) with 2mm thickness observed by a differential interference microscope (red-bar scale: 200μm). PL signals before and after mechanical (d, e and f) and laser induced damages (j, k and l) at different positions of these damage sites from Quartz (d, j), Suprasil 401 (e, k), and Corning 7980 (f, l) excited by focused 6.4eV laser spot (repetition rate 400Hz; fluence 57.8mJ/cm2; spot diameter: 60μm) are measured at 5μs delay time and 100μs gate time.
Despite the significant differences in initial PL signals, the damage induced intensity growths of the PL bands, especially the 2.25eV PL band with ~50μs lifetime, are quite similar in both mechanical and laser damage processes in all these samples, indicating a similar damage induced PL defects’ formation process in different samples. In mechanical damage sites, the damage induced growths of the 1.90eV and 2.25eV PL bands are dominating. On the other hand, in laser induced damage sites, except the 1.90eV and 2.25eV PL bands, a strong damage-induced growth of a 2.70eV PL band, which is due to the ODC II groups [13], is observed. In addition, the comparison of the PL signals from the center (Point A) and edge (Point B) of these damage sites show that the 2.25eV PL bands have a much wider distribution than the PL bands of NBOHC that are corresponding to Si-O bonds breakage in SiO2 network in the center of damage sites.
3.4 2.25eV PL intensities in seven SiO2 samples
Figure 6a shows the peak intensities of 2.25eV from fused silica samples with different OH contents. An approximately linear relationship between the intensities of 2.25eV and 1.90eV PL bands is found, indicating possible correlation between these PL bands. As no clear correlation between 1.90eV PL intensity and OH content is presented, as shown in Fig. 6b, this 2.25eV PL band is unlikely due to SiOH groups in SiO2 network.
Fig. 6 a) Relation of peak intensities of 1.90eV and 2.25eV PL bands for 7 fused silica samples with different OH contents with 2mm thickness (1. Clear Quartz, 2. Corning 8655, 3. Suprasil 711, 4. Suprasil 501, 5. Suprasil 401, 6. Corning 7980, 7. Spectrosil 2000) excited by 6.4eV at 200Hz repetition rate and 3.2mJ/cm2 fluence. Dotted line shows a linear fit. b) Peak intensity of 1.90 PL band versus OH content.
4. Discussion
Two different green PL bands are resolved from lifetime spectra. The behaviors of 2.25eV and 2.39eV PL bands resolved from 6.4eV excited PL signals are quite different from previously reported green PL defects, especially the 2.25eV PL band created by damage processes. The decays of 2.25eV and 2.39eV PL bands nearly perfectly followed single-exponential function (R-square >0.99), which are quite different from the decay of STE following a stretched-exponential law [23]. This rules out the excitations of these PL bands are due to migration and trapping of excitons in fused silica network [24]. The microsecond lifetimes of these PL bands indicate these green PL defects cannot be due to POR and E’ centers either, which have lifetime in nanoseconds [19,20]. The PL intensities of these green bands are linearly proportional to the laser intensity, that is, a one-photon excitation process. As the one-photon excitation energy of 6.4eV is lower than the excitation energy of Si-H bonds (~8.2eV) [25], these PL bands also cannot be due to Si-H bonds in SiO2 network. In addition, Cu impurity is also excluded due to high purity of samples we used and the generation/annealing behaviors of these green PL defects under 6.4eV irradiation, while Cu impurity should be stable under 6.4eV irradiation. The emitters of these PL bands are discussed according to our experiment results.
4.1 The 2.39eV PL band
From the experimental result of 6.4eV irradiation, the annealing of 2.39eV PL band is due to the decomposition of surface defects. In silica nanoparticle samples with a high surface-to-volume ratio, a similar 2.35eV PL band was reported with ~12μs lifetime and 630 cm−1 spacing [26]. In addition, a 2.41eV green PL band with 7.5μs lifetime in air atmosphere observed by oxygen chemisorption at oxygen deficient center sites [27] was assigned to oxidation of dioxasilyrane group ( = SiO2). Since the calculated transition energy from ground state to second excited state (S0→S2) of = SiO2 group was reported to be ~5.0eV with a corresponding ~2.3eV PL band (S2→S0; oscillation strength: ~0.0001; O–O stretching vibrations: 610~630cm−1) [28], the S2→S0 transition of = SiO2 is in agreement with the 2.39eV PL band observed in our experiment as one-photon process at 6.4eV excitation as well as those green bands reported in silica nanoparticles. As a result, this surface 2.39eV PL band observed in Suprasil 401 fused silica sample is related to surface = SiO2 group, while the annealing behavior of this PL band during 6.4eV irradiation at N2 atmosphere is due to deoxidation of = SiO2 group.
4.2 The 2.25eV PL band
The similar UV induced growths of 2.25eV and 1.90eV PL bands and approximately linear relationship in intensities of these PL bands indicate the formation process of 2.25eV PL defect is similar to that of NBOHC. In our experiment, the non-linear dependence of the 1.90eV PL intensity on laser intensity shows the 1.90eV PL band is excited via a two-photon process. Since the excitation energy of NBOHC (2.2eV and 4.8eV [13]) is lower than 6.4eV, the 1.90eV PL band is actually due to the breakage of strained Si-O bonds as a two-photon absorption process under 6.4eV irradiation, other than direct excitation of NBOHC defects, which should be a one-photon absorption process. As a result, the growth of 2.25eV band is due to 6.4eV laser induced defects from physical disorder in SiO2 network.
The growths of 2.25eV PL band observed in all damage sites, even in quartz sample with low H and OH content, further prove that this PL defect is associated with physical disorder induced by pressure or laser-induced shock waves [29] other than melting on the sample surface [30] or H related groups. Compared with 6.4eV excited PL signal from stishovite [31], which mainly consist of 3.0eV and 4.7eV PL bands, the PL bands created in the damage experiments cannot be associated to highly densified disorders like 5 or 6-fold Si groups [32]. The bending of Si-O-Si angles [33] is the major structure changes in the damage experiments. From the differences in PL distribution between 2.25eV and 1.90eV PL band and the irreversible intensity growth of 2.25eV PL band during 6.4eV pre-irradiation treatment (different from NBOHC which can be annealed during darkness), we conclude that the formation of the 2.25eV PL defects is associated with irreversible rearrangement of these strained bonds without formation of dangling bonds. No 2.25eV band in the PL spectrum of un-damaged (original) quartz is experimentally observed due to an arrow Si-O-Si angle distribution of quartz network.
It was predicted that the biaxial compressive strain lead to the irreversible formation of edge-sharing tetrahedron and silanone groups ( = Si = O) to reduce the strain in thin amorphous silica layer [34]. To investigate this possibility we theoretically estimated the excitation energies of edge-sharing tetrahedron and = Si = O groups at B3LYP/6-311 + (d,p) level by ORCA (An ab initio, DFT and semiempirical SCF-MO package) [35]. The clusters are built using SiO4-tetrahedrons with fixed 0.80 Å O-H bonds as terminators. In edge-sharing tetrahedron cluster ((HO)2SiO2Si(OH)2), the vertical excitation energy of the lowest excited singlet state is ~7.8 eV with ~0.0114 oscillation strength (f), indicating the edge-sharing tetrahedron cannot be the origin of the 2.25eV PL band which is experimentally observed as a one-photon excitation process under 6.4eV irradiation. In = Si = O clusters (((HO)3Si)2Si = O), two lowest excited singlet states are found with vertical excitation energy lower than 6.4eV (S0→S1:~5.82eV, f~0.0554; S0→S2:~6.09eV, f~0.0190). The corresponding PL energy of S1→S0 and S2→S0 transitions are 1.55eV (f: 0.0025) and 2.24eV (f: 0.0034), respectively, in agreement with absorption band of silanone groups with 5.65eV peak and 0.9eV FWHM [8], indicating the calculation results can qualitatively represent PL properties of = Si = O groups. Due to comparable excitation energy and oscillator strengths for S0→S1 and S0→S2 excitations and less probable for S2→S1 emission due to the low oscillator strength and insignificant radiation-less internal conversion between S2 and S1 states [28], both 1.55eV and 2.25eV bands are considered possible in 6.4eV irradiation. Even though 1.55eV band is not observed in our experiment due to the spectral range limitation of the spectrometer used, 1.5~1.7eV red PL band was reported in silanone-based silicon oxyhydrides, sol–gel derived SiO layers doped with porous silicon and bulk SiO sample and thin SiO layers [36–38]. Therefore the = Si = O group could also well explain the experimentally observed one-photon excitation process at 6.4eV irradiation and the strain-induced growth of this 2.25eV PL band. In addition, the = Si = O group is reported to be very unstable on the surface of SiO2 network and highly reactive with O2 [27] and H2O [39], while recent research predicts that this defect can be very stable in bulk SiO2 network until 900K [40]. This prediction is supported by our experimental observation that the 2.25eV PL band is dominated in bulk PL signals but not presented in surface PL signals. Overall, from the above analysis we conclude that the = Si = O group is the possible emitter of this strain-induced PL band. However, as the excitation state of = Si = O can be easily influenced by local environment [41], green PL bands with typical ~1300cm−1 and 300cm−1 stretching vibrations modes from = Si = O group [28] in silanone-based silicon oxyhydrides, and silica nanoparticles [26,27,36] have been reported with lifetime ranging from nanosecond to microseconds, while the PL properties of = Si = O groups is still unknown. Taking into consideration the complex structure of fused silica network, other possible origins of this 2.25eV PL band may exist and more comprehensive investigation on this 2.25eV green PL band is still needed to form a more convincing explanation for this strain-induced PL band.
5. Conclusion
To summarize, two green bands, a 2.25eV green PL with ~50μs lifetime and a 2.39eV PL band with ~10μs lifetime, from fused silica have been resolved by a lifetime spectrum method. From the experimental observations of PL spectrum changes caused by UV laser irradiation as well as mechanical and laser damages, the 2.39eV PL band emitted from surface is from defects unstable under 6.4eV irradiation, while the 2.25eV PL band commonly exists in SiO2 network is from irreversible strain-induced structure disorder in SiO2 network. Together with previous publications on microsecond-lifetime green bands, we can conclude that = SiO2 groups that can be deoxided at N2 atmosphere and = Si = O groups that can be induced by strain relaxation process in SiO2 are the possible emitters of 2.39eV and 2.25eV PL bands, respectively. These results are helpful to further understanding of defect generation /annealing in SiO2 network. With current trend to control strained Si-O-Si angles in SiO2 network to improve UV transparency [7], this strain-induced 2.25eV PL band as a signature of residual stress provide a convenient tool to characterize and quantify strain-induced structure disorder in fused silica during laser damage, which is difficult to measure directly.
Acknowledgments
The authors thank Bodo Kuehn of Heraeus Quarzglas GmbH, Germany for providing those fused silica and quartz samples from Heraeus.
References and links
1. R. Betti and O. A. Hurricane, “Inertial-confinement fusion with lasers,” Nat. Phys. 12, 435–448 (2016). [CrossRef]
2. C. Wagner and N. Harned, “EUV lithography: Lithography gets extreme,” Nat. Photonics 4(1), 24–26 (2010). [CrossRef]
3. P. E. Miller, J. D. Bude, T. I. Suratwala, N. Shen, T. A. Laurence, W. A. Steele, J. Menapace, M. D. Feit, and L. L. Wong, “Fracture-induced subbandgap absorption as a precursor to optical damage on fused silica surfaces,” Opt. Lett. 35(16), 2702–2704 (2010). [CrossRef] [PubMed]
4. L. Hongjie, H. Jin, W. Fengrui, Z. Xinda, Y. Xin, Z. Xiaoyan, S. Laixi, J. Xiaodong, S. Zhan, and Z. Wanguo, “Subsurface defects of fused silica optics and laser induced damage at 351 nm,” Opt. Express 21(10), 12204–12217 (2013). [CrossRef] [PubMed]
5. R. A. Negres, M. A. Norton, D. A. Cross, and C. W. Carr, “Growth behavior of laser-induced damage on fused silica optics under UV, ns laser irradiation,” Opt. Express 18(19), 19966–19976 (2010). [CrossRef] [PubMed]
6. C. Mühlig, H. Stafast, and W. Triebel, “Generation and annealing of defects in virgin fused silica (type III) upon ArF laser irradiation: Transmission measurements,” J. Non-Cryst. Solids 354(1), 25–31 (2008). [CrossRef]
7. L. Skuja, H. Hosono, M. Hirano, and K. Kajihara, “Advances in silica-based glasses for UV and vacuum-UV laser optics,” Proc. SPIE 5122, 1–14 (2003).
8. L. Skuja, K. Kajihara, M. Hirano, and H. Hosono, “Visible to vacuum-UV range optical absorption of oxygen dangling bonds in amorphous SiO2,” Phys. Rev. B 84(20), 205206 (2011). [CrossRef]
9. U. Natura, O. Sohr, R. Martin, M. Kahlke, and G. Fasold, “Mechanisms of radiation induced defect generation in fused silica,” Proc. SPIE 5273, 155–164 (2004). [CrossRef]
10. H. Hosono, Y. Ikuta, T. Kinoshita, K. Kajihara, and M. Hirano, “Physical disorder and optical properties in the vacuum ultraviolet region of amorphous SiO2.,” Phys. Rev. Lett. 87(17), 175501 (2001). [CrossRef] [PubMed]
11. M. Cannas and F. M. Gelardi, “Vacuum ultraviolet excitation of the 1.9-eV emission band related to nonbridging oxygen hole centers in silica,” Phys. Rev. B 69(15), 153201 (2004). [CrossRef]
12. L. Nuccio, S. Agnello, R. Boscaino, B. Boizot, and A. Parlato, “Generation of oxygen deficient point defects in silica by γ and β irradiation,” J. Non-Cryst. Solids 353(5-7), 581–585 (2007). [CrossRef]
13. L. Skuja, “Optically active oxygen-deficiency-related centers in amorphous silicon dioxide,” J. Non-Cryst. Solids 239(1–3), 16–48 (1998). [CrossRef]
14. M. A. S. Kalceff, “Cathodoluminescence microcharacterization of the defect structure of irradiated hydrated and anhydrous fused silicon dioxide,” Phys. Rev. B 57(10), 5674–5683 (1998). [CrossRef]
15. M. A. S. Kalceff, A. Stesmans, and J. Wong, “Defects induced in fused silica by high fluence ultraviolet laser pulses at 355 nm,” Appl. Phys. Lett. 80(5), 758–760 (2002). [CrossRef]
16. 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]
17. L. Sun, H. Liu, J. Huang, X. Ye, H. Xia, Q. Li, X. Jiang, W. Wu, L. Yang, and W. Zheng, “Reaction ion etching process for improving laser damage resistance of fused silica optical surface,” Opt. Express 24(1), 199–211 (2016). [CrossRef] [PubMed]
18. J. Fournier, P. Grua, J. Néauport, E. Fargin, V. Jubera, D. Talaga, A. D. Guerzo, G. Raffy, and S. Jouannigot, “Temperature dependence of luminescence for different surface flaws in high purity silica glass,” Opt. Mater. Express 3(1), 1–10 (2013). [CrossRef]
19. Y. Sakurai and K. Nagasawa, “Green photoluminescence band in γ-irradiated oxygen-surplus silics glass,” J. Appl. Phys. 86(3), 1377–1381 (1999). [CrossRef]
20. Y. Sakurai, “Photoluminescence band near 2.2eV in γ-irradiated oxygen-deficient silica glass,” J. Non-Cryst. Solids 342(1–3), 54–58 (2004). [CrossRef]
21. Y. D. Glinka, S. H. Lin, and Y. T. Chen, “Time-resolved photoluminescence study of silica nanoparticles as compared to bulk type-III fused silica,” Phys. Rev. B 66(3), 035404 (2002). [CrossRef]
22. H. Ikeda, T. Murata, and S. Fujino, “Photoluminescence characteristics of sintered silica glass doped with Cu ions using mesoporous SiO2-PVA nanocomposite,” Mater. Chem. Phys. 162, 431–435 (2015). [CrossRef]
23. A. N. Trukhin, “Luminescence of localized states in silicon dioxide glass. A short review,” J. Non-Cryst. Solids 357(8-9), 1931–1940 (2011). [CrossRef]
24. J. Linnros, N. Lalic, A. Galeckas, and V. Grivickas, “Analysis of the stretched exponential photoluminescence decay from nanometer-sized silicon crystals in SiO2,” J. Appl. Phys. 86(11), 6128–6134 (1999). [CrossRef]
25. G. Pacchioni and G. Ieranò, “Ab initio theory of optical transitions of point defects in SiO2,” Phys. Rev. B 57(2), 818–832 (1998). [CrossRef]
26. Y. D. Glinka, S. H. Lin, and Y. T. Chen, “The photoluminescence from hydrogen-related species in composites of SiO2 nanoparticles,” Appl. Phys. Lett. 75(6), 778–780 (1999). [CrossRef]
27. L. Vaccaro, A. Morana, V. Radzig, and M. Cannas, “Bright Visible Luminescence in Silica Nanoparticles,” J. Phys. Chem. C 115(40), 19476–19481 (2011). [CrossRef]
28. A. S. Zyubin, A. M. Mebel, S. H. Lin, and Y. D. Glinka, “Photoluminescence of silanone and dioxasilyrane groups in silicon oxides: A theoretical study,” J. Chem. Phys. 116(22), 9889–9896 (2002). [CrossRef]
29. M. Chambonneau, R. Diaz, P. Grua, J. L. Rullier, G. Duchateau, J. Y. Natoli, and L. Lamaigne’re, “Origin of the damage ring pattern in fused silica induced by multiple longitudinal modes laser pulses,” Appl. Phys. Lett. 104(2), 021121 (2014). [CrossRef]
30. C. W. Carr, J. D. Bude, and P. DeMange, “Laser-supported solid-state absorption fronts in silica,” Phys. Rev. B 82(18), 184304 (2010). [CrossRef]
31. A. N. Trukhin, K. Smits, A. Sharakosky, G. Chikvaidze, T. I. Dyuzheva, and L. M. Lityagina, “Luminescence of dense, octahedral structured crystalline silicon dioxide (stishovite),” J. Lumin. 131(11), 2273–2278 (2011). [CrossRef]
32. F. Yuan and L. Huang, “Brittle to ductile transition in densified silica glass,” Sci. Rep. 4(1), 5035 (2015). [CrossRef] [PubMed]
33. D. Wakabayashi, N. Funamori, and T. Sato, “Enhanced plasticity of silica glass at high pressure,” Phys. Rev. B 91(1), 014106 (2015). [CrossRef]
34. C. L. Kuo, S. Lee, and G. S. Hwang, “Strain-induced formation of surface defects in amorphous silica: a theoretical prediction,” Phys. Rev. Lett. 100(7), 076104 (2008). [CrossRef] [PubMed]
35. F. Neese, “The ORCA program system,” Comput. Mol. Sci. 2(1), 73–78 (2012). [CrossRef]
36. J. L. Gole and D. A. Dixon, “Transformation, green to orange-red, of a porous silicon photoluminescent surface in solution,” J. Phys. Chem. B 102(1), 33–39 (1998). [CrossRef]
37. V. Švrček, I. Pelant, J. L. Rehspringer, P. Gilliot, D. Ohlmann, O. Crégut, B. Hönerlage, T. Chvojka, J. Valenta, and J. Dian, “Photoluminescence properties of sol–gel derived SiO2 layers doped with porous silicon,” Mater. Sci. Eng. C 19(1-2), 233–236 (2002). [CrossRef]
38. L. X. Yi, J. Heitmann, R. Scholz, and M. Zacharias, “Phase separation of thin SiO layers in amorphous SiO/SiO2 superlattices during annealing,” J. Phys. Condens. Matter 15(39), S2887–S2895 (2003). [CrossRef]
39. F. Zhou and J. D. Head, “Role of Si=O in the photoluminescence of porous silicon,” J. Phys. Chem. B 104(43), 9981–9986 (2000). [CrossRef]
40. X. Chen, Y. W. Wang, X. Liu, X. B. Wang, and Y. Q. Zhao, “Study of structural and electronic properties of the silanone group as bulk defect in amorphous SiO2,” J. Non-Cryst. Solids 414, 1–6 (2015). [CrossRef]
41. M. A. Zwijnenburg, A. A. Sokol, C. Sousa, and S. T. Bromley, “The effect of local environment on photoluminescence: a time-dependent density functional theory study of silanone groups on the surface of silica nanostructures,” J. Chem. Phys. 131(3), 034705 (2009). [CrossRef] [PubMed]
