Photoluminescence (PL) microscopy and spectroscopy under 266 nm and 355 nm laser excitation are explored as a means of monitoring defect populations in laser-modified sites on the surface of fused silica and their subsequent response to heating to different temperatures via exposure to a CO2 laser beam. Laser-induced temperature changes were estimated using an analytic solution to the heat flow equation and compared to changes in the PL emission intensity. The results indicate that the defect concentrations decrease significantly with increasing CO2 laser exposure and are nearly eliminated when the peak surface temperature exceeds the softening point of fused silica (~1900K), suggesting that this method might be suitable for in situ monitoring of repair of defective sites in fused silica optical components.
©2010 Optical Society of America
Fused silica is commonly used for the manufacturing of optical elements for high average laser power applications such as microlithography, deep UV and excimer lasers, and large aperture laser systems designed to achieve laser-driven inertial confinement fusion (ICF). The large photon energies used in these applications result in degradation of material performance mainly via generation of atomic defects or the formation of macroscopically observed damage sites, thus limiting system performance and increasing cost of operation. Exposure of fused silica to temperatures and pressures created during a laser-induced damage event  causes material modifications which include the formation of a crater, creation of micro-fractures, and the generation of a number of defects [2–7]. Studies have shown that upon damage initiation on the surface of optical components for ICF class laser systems, subsequent laser pulses can grow these damage sites to macroscopic dimensions [8,9]. To avoid having to replace the optic each time a damage event occurs, extensive research has been directed towards increasing the lifetime of these optics by arresting damage site growth [10,11]. One of the most promising methods towards this goal involves exposure of the damage site to a CO2 laser , thereby increasing the local temperature to the point where cracks can be fused together. The development of in situ diagnostic tools [3,7,13] to guide this repair process requires a better understanding of the processes involved, including thermally-activated annihilation of the absorbing defects that have been postulated to play a key role in the re-initiation of the damage process at pre-existing damage sites leading to damage growth [3,4,7,14–16].
Photoluminescence (PL) microscopy has the potential to offer such diagnostic capabilities because it can spatially map the concentration of various defect species in laser-modified fused silica, namely oxygen deficiency centers (ODC), unknown laser-induced defects (referred to as LID), and non-bridging oxygen hole centers (NBOHC), which are known to have specific PL emission bands centered at ~470, 560, and 650 nm, respectively [2–4,17]. In this work we evaluate PL microscopy under UV excitation as a means of monitoring the distribution of defects formed following laser damage and over various CO2 laser treatments (corresponding to various surface temperatures). The objective of this work is two-fold: 1) to develop a better understanding of the annihilation efficiency of the different defect species as a function of the surface temperature and 2) to develop a tool that can be used as an in situ diagnostic during the implementation of various repair protocols.
2. Experimental arrangement
Experiments were conducted on a 5 cm-diameter, 1 cm-thick Corning 7980 fused silica sample. Fifty-eight damage sites were formed on the exit surface of the sample in a grid pattern with site nearest-neighbor separation of 3 mm. Each damage site was initiated in an ambient environment by a single 7-ns FWHM pulse from a Nd:YAG laser operating at 355 nm with a fluence at the exit surface of the sample of ~30 J/cm2. These sites were then grown in 10 torr of Ar simultaneously using a 5-shot sequence of 5-ns, flat-in-time pulses with an average fluence of ~10 J/cm2 at 351 nm delivered in a spatially uniform, large area laser beam . The initiated damage site diameter was ~40 μm, while that of the final grown damage site ranged from 140 to 240 μm.
Thirty of the 58 damage sites were subsequently exposed for 40 s to different irradiances from a CO2 laser operating at 10.6 µm wavelength and with a 1/e beam radius of a = 0.71 mm. These 30 sites were then categorized into 6 sets (5 sites per set) corresponding to 6 levels of CO2 laser peak axial irradiance: I = 0.44, 0.57, 0.63, 0.69, 0.76 and 0.88 kW/cm2. The remaining damage sites received no CO2 laser exposure, thus serving as control sites against which to compare sites exposed to the CO2 laser beam.
For the purpose of monitoring the distribution of defects in damage sites as a function of increasing CO2 laser exposure, we employed a portable PL microscopy system  depicted in Fig. 1 . This system consists of a long working-distance (34 mm), 5X/0.14 NA objective (Mitutoyo, Japan) and a 5X zoom lens followed by a 420 nm long-pass (LP) filter. To eliminate the possibility of photobleaching, the excitation was provided by two low-irradiance (3 mW/cm2) compact laser sources operating at 266 nm and 355 nm wavelengths. UV laser illumination was incident at an angle of ~60° with respect to the sample’s normal and was switched by means of electronically-controlled shutters. Images were captured with a liquid nitrogen-cooled charge-coupled device (CCD) under either excitation with integration times ranging from 30 to 180 s. Because of the reduced sensitivity of the CCD detector in the near IR, this combination of long-pass filter and CCD resulted in a spectral range of ~420-1000 nm for PL imaging. Back and ring-illuminated white light scattering (LS) images of the same sites were also recorded using a standard microscope equipped with a 10X/0.28NA objective. The spatial resolution for both the PL and LS images was on the order of 2 µm.
PL spectra were acquired from the control damage sites using a spectrometer (Jobin Yvon Horiba) and were subsequently corrected for instrument response by acquiring a spectrum of the emission from a quartz tungsten-halogen calibration lamp (Oriel) and determining the correction factor for the instrument response based on a 6th-order polynomial fit of the published spectrum. The limits of the spectral range of 400-800 nm were determined by the decreased transmission of the microscope deeper in the UV and the decreased efficiency of the spectrometer grating in the near IR. Typical spectra are shown in Fig. 2 . The labels A and B refer to two different locations, separated by several microns, within the center of the same damage site. The spectra under 266 nm excitation contain one prominent emission peak centered at ~470 nm (corresponding to ODC) and a weaker peak at ~650 nm (minor contribution from NBOHC). In contrast, the spectra under 355 nm excitation contain one emission peak centered at ~560 nm (corresponding to LID defects, see Ref. ) and another at ~650 nm (corresponding to NBOHC). The difference in the relative intensities of these peaks at locations A and B suggests a non-uniform distribution of the same defects on a local scale within a damage site. We note that as shown in Fig. 2, the 420 nm LP collection scheme used in the imaging set up of Fig. 1 sufficiently captures the PL bands in both laser excitations.
Figure 3 shows PL images under the different UV excitation sources (266 nm excitation (Fig. 3(a)) and 355 nm excitation (Fig. 3(b))) of the same control (no CO2 laser treatment) damage site as well as a back-illuminated LS image (Fig. 3(c)). We note that the spatial distribution of the 266 and 355 nm excited images are roughly inverted, and, interestingly, only vaguely correlated to what is observed under standard microscopy. Following the PL peak assignments described above, the image formed under 266 nm excitation (Fig. 3(a)) portrays the spatial distribution of mainly ODC defects while that under 355 nm excitation (Fig. 3(b)) portrays that of mainly NBOHC and LID defects.
With knowledge of this spatial distribution of defect concentrations, we now consider the LS image (Fig. 3(c)) which gives an indication of the relative morphology of the damage site. For example, the central region of the site (indicated as region 1 in Fig. 3(b)) under 355 nm excitation is darker (relative to the surrounding region) on average when compared to the PL image under 266 nm excitation of Fig. 3(a). This would indicate that these central regions, which tended to appear more ‘molten’, are comprised of mainly ODC defects. In contrast, region 2 corresponds to fractured material found around the periphery of the site and is somewhat richer in NBOHC and LID defectsThese results are consistent with the model of a surface damage site proposed by Wong et al.  describing the center of the site as a molten ‘botryoidal’-shaped core where ODC defects tend to be localized and the outer region as a fractured shell where NBOHC defects tend to be localized. The region to the far right of Fig. 3 shows a minor degree of smoothing but is dominated by the appearance of fractures.
Turning now to the CO2 laser irradiance-dependent data, a similar result is observed in the images of the control damage site of column 1 of Fig. 4 , namely that the core region when viewed under 355 nm excitation (Fig. 4(a)) appears darker on average than when viewed under 266 nm excitation (Fig. 4(f)). This figure displays PL (rows 1 and 2) and ring LS (row 3) images of selected damage sites of similar post-damage diameter (160-180 µm) after exposure to various levels of CO2 laser irradiance, as indicated at the top of each column (one site per column). In this set of comparisons, ring-illumination was favored over back- illumination due to the higher sensitivity to non-specular surface morphology (e.g. fractures). In stark contrast to the images shown in Figs. 3(a)-3(b) (and first column of Fig. 4), after exposing a site to at least I = 0.44 kW/cm2 of CO2 laser irradiance, (e.g. Figures 4(b) and 4(g)), there were virtually no differences in the intensity maps under either UV excitation source.
We now present quantitative analysis of the data. Since the resistance to damage that results from laser-induced defect annihilation is thermally-driven , it is instructive to assign temperature estimates to the laser treatments to better understand the change in defect populations and hence PL results. Although finite element-based solutions to the full non-linear heat flow equation are technically required to describe the temperature rise, simplifications can be made for relatively long exposure times under Gaussian beam heating. Specifically, for exposure times long compared to the thermal diffusion time, τth~0.5a2/D, where D~0.008cm2/s is the thermal diffusivity (yielding τth ~300 ms), and for beam radii much larger than the absorption depth, the spatial distribution of the temperature rise can be approximated as 
In Eq. (1), Ia is the absorbed irradiance (which takes into account the 15% reflectivity of silica at 10.6 µm), k = 0.020W/cm∙K is the thermal conductivity , and J0 represents the modified Bessel function of order zero. While the above expression has been shown to correlate with direct thermographic measurements of heated silica slabs to better than 5% , we acknowledge that it is likely to be a slight over-estimate in our case since scattering losses from the irregular surface are not taken into account. Nonetheless, we used this expression for the laser-induced temperature field and then evaluated the peak over the area of the damage site (~200 μm), yielding values shown in Table 1 .
The mean PL intensity of a damage site was calculated by defining a square region of interest inscribed within the region giving rise to PL in the image. The mean intensity (counts per pixel) was then calculated, and this value was normalized to the image integration time. Alternatively, a threshold above background could be set defining a site- specific region of interest, though this approach did not yield significant differences in our final conclusions. Moreover, for the purpose of facilitating fast pixel count acquisition consistent with a simple diagnostic, our inscribing approach was more suitable.
The result of these calculations is summarized in Fig. 5 , where the mean intensity of PL (Fig. 5(a)) and ring LS (Fig. 5(b) are plotted as a function of the local peak surface temperature corresponding to each CO2 laser exposure used in this study. Error bars indicate one standard deviation (n = 5 sites).
Fused silica undergoes material transformations at well-defined temperatures. The glass transition temperature (TG) is the temperature at which, upon cooling, the material transforms from a supercooled liquid to a solid. Above this temperature, the material can relax structurally without deforming. The softening temperature (TS) is above TG and is the temperature above which the glass can deform. Further heating to the boiling temperature (TB) leads to evaporation of the material. We now discuss the measured PL and LS intensities of Fig. 5 in relation to these material transformations.
While the mean PL intensity (Fig. 5(a)) from the site decreased under both excitations after I = 0.44 kW/cm2 of irradiation, the decrease under 266 nm excitation is much larger (by a factor of about 4), indicating a significant decrease in the concentration of ODC defects. It is worth noting that this exposure raises the sample surface temperature just above the glass transition temperature (TG = 1315 K) but is well below the softening point (1860 K), implying that structural relaxation over a <100 s time scale is possible but viscous flow is unlikely. The difference in the response of the PL intensity under 266 nm excitation as TG is exceeded suggests that annihilation of ODC is thermally-activated at this temperature (1500 K) for a 40 s exposure. However, though not explicitly explored here, it could be expected that lower temperatures over longer time scales would lead to similar defect annihilation levels .
Mean PL intensity further decreased by a factor of about 20, but proportionally between the two excitation wavelengths, as CO2 laser irradiance was increased to I = 0.69 kW/cm2 (corresponding to 2200 K). This result suggests that, when exceeding the softening point of fused silica (~1900 K) over a 40 s time scale, the material has sufficient thermal energy to anneal defects of all three types within the damage site through either surface diffusion or capillary action. CO2 laser heating may facilitate the reconstruction of bonds, thus restoring the original molecular network and annihilating defects initially formed during the thermal shock  induced by the damage laser pulse.
We now turn to the effect of laser heat-treatment of silica surface damage on surface morphology as characterized by edge-illuminated scattered light microscopy and compare these results to those acquired by PL microscopy. As shown in Fig. 5(b), the mean scattered intensity is relatively constant over temperatures less than the softening point (~1860 K), or equivalently, for viscosities higher than 107.6 Poise.
A recent study of surface smoothing of etched silica gratings showed that over similar time scales (10’s of seconds), the onset of capillary-driven surface relaxation takes place at temperatures of ~1800 K, with complete relaxation at ~2000 K . Because capillary forces are derived from local curvatures, the length scale over which laser smoothing is observed is dependent on both time and the temperature-dependent viscosity associated with the laser exposure. Specifically, under conditions where diffusion and evaporation are negligible , it can be shown that the relaxation time τ of a feature of length λ is approximately τ~ηλ/πσ where η and σ are the viscosity and surface tension, respectively, for type III fused silica. If we equate this relaxation time to the laser exposure time (assuming a negligible thermal diffusion time), we can estimate an upper bound on feature sizes that will be appreciably relaxed. For example, using published expressions for η(T)  and σ~0.3N/m , we find that for T = 1480 K and τ = 40 s, λ~2 nm, and no appreciable change in visible light scattering should be expected. In contrast, at T = 2000 K and τ = 40 s, λ~20 µm, and clear morphological changes should be detectable by visible light scattering or inspection. At high enough temperatures, macroscopic evolution of the entire damage site will occur such that λ>200 µm, or at temperatures greater than approximately T = 2200 K.
In contrast to the change in light scattering as a function of estimated temperatures, the PL signals from both excitation wavelengths continually change, even at temperatures below the softening point. This fact illustrates the annealing behavior of CO2 laser treatment at low temperatures, and is to be compared to PL changes in oven-annealed studies of fractured SiO2 . However, because successful treatments towards damage repair and resistance must remove both the UV light absorbing defects and surface irregularities that can cause near-field intensification , monitoring diagnostics must be sensitive to both of these facets.
These observations are consistent with what is observed from our light scattering images of Figs. 4(k)-4(o). The spatial distribution of the light scattering is quite similar to that of the 266 and 355 nm PL images at temperatures <1800 K (I < 0.57 kW/cm2) as shown in Figs. 4(k)-4(l) but deviates from the PL at higher temperatures (Figs. 4(m)-4(n)). Indeed for T ≥ 2000 K, as temperature increases, the light scattering region begins to show signs of spatial expansion, consistent with macroscopic changes that can lead to focusing effects, while the PL intensity map continually diminishes (Figs. 4(h)-4(j)). The light scattering regions are larger than the corresponding PL regions because of the different effect of viscous flow on surface roughness (where scattered light is enhanced) vs. surface defect distribution and concentration. Morphological features are extended and become less localized as a result of this flow, increasing the size of the region of scattered light, and are eventually smoothed out to a large degree at temperatures above ~2300 K or near the boiling point of SiO2 (Fig. 4(o)). The PL response, however, is indicative of local changes in the evolution of point defects which are progressively reduced in number via annihilation with higher temperatures (Figs. 4(h)-4(j)). Since the damage threshold of CO2 treated sites tends to show a monotonic increase with increasing treatment temperature , this result emphasizes the utility of measuring defect concentrations with PL over light scattering or optical inspection.
A negligible amount of PL intensity above the noise level was detected from sites exposed to I = 0.88 kW/cm2, corresponding to a surface temperature of about 2700 K which is above the fused silica boiling point (2500 K). Thus, at this temperature, the majority of defective silica is expected to have been removed through evaporation.
In this work we have presented a diagnostic system capable of monitoring the effect of CO2 laser heating on the distribution and relative concentration of different defect species within pre-existing laser-induced damage sites on the surface of fused silica. Annihilation of these species was found to occur efficiently at temperatures above the fused silica softening point. The information provided by this system can potentially be used to monitor the efficacy of protocols for damage repair of optical components.
The authors wish to thank Mary A. Norton and Gabriel M. Guss for assistance with sample preparation. This work was performed under the auspices of the U. S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
References and links
1. C. W. Carr, 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] [PubMed]
2. L. Skuja, “The nature of optically active oxygen-deficiency-related centers in amorphous silicon dioxide,” J. Non-Cryst. Solids 239(1-3), 16–48 (1998). [CrossRef]
3. S. G. Demos, M. Staggs, K. Minoshima, and J. Fujimoto, “Characterization of laser induced damage sites in optical components,” Opt. Express 10(25), 1444–1450 (2002), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-10-25-1444. [PubMed]
4. S. O. Kucheyev and S. G. Demos, “Optical defects produced in fused silica during laser-induced breakdown,” Appl. Phys. Lett. 82(19), 3230–3232 (2003). [CrossRef]
5. J. Wong, J. L. Ferriera, E. F. Lindsey, D. L. Haupt, I. D. Hutcheon, and J. H. Kinney, “Morphology and microstructure in fused silica induced by high fluence ultraviolet 3ω (355 nm) laser pulses,” J. Non-Cryst. Solids 352(3), 255–272 (2006). [CrossRef]
6. S. Xu, X. Zu, X. Jiang, X. Yuan, J. Huang, H. Wang, H. Lv, and W. Zheng, “The damage mechanisms of fused silica irradiated by 355 nm laser in vacuum,” Nucl. Instrum. Methods Phys. Res. B 266(12-13), 2936–2940 (2008). [CrossRef]
7. T. A. Laurence, J. D. Bude, N. Shen, T. Feldman, P. E. Miller, W. A. Steele, and T. Suratwala, “Metallic-like photoluminescence and absorption in fused silica surface flaws,” Appl. Phys. Lett. 94(15), 151114 (2009). [CrossRef]
8. M. A. Norton, L. W. Hrubresh, Z. L. Wu, E. E. Donohue, M. D. Feit, M. R. Kozlowski, D. Milam, K. P. Neeb, W. A. Molander, A. M. Rubenchik, W. D. Sell, and P. Wegner, “Growth of laser initiated damage in fused silica at 351 nm,” Proc. SPIE 4347, 468 (2001). [CrossRef]
9. H. Wan-Qing, H. Wei, W. Fang, X. Yong, L. Fu-Quan, F. Bin, J. Feng, W. Xiao-Feng, Z. Wan-Guo, and Z. Xiao-Min, “Laser-induced damage growth on larger-aperture fused silica optical components at 351 nm,” Chin. Phys. Lett. 26(1), 017901 (2009). [CrossRef]
11. T. Kamimura, K. Nakai, M. Yoshimura, Y. Mori, T. Sasaki, M. Tanaka, Y. Okada, H. Yoshida, M. Nakatsuka, T. Kojima, and K. Yoshida, “High damage resistivity of optical surface for UV lasers by ion beam etching,” Rev. Laser Eng. 27, 623–627 (1999).
12. R. M. Brusasco, B. M. Penetrante, J. A. Butler, and L. W. Hrubesh, “Localized CO2-laser treatment for mitigation of 351-nm damage growth in fused silica,” Proc. SPIE 4679, 40–47 (2002). [CrossRef]
13. J. Neauport, P. Cormont, P. Legros, C. Ambard, and J. Destribats, “Imaging subsurface damage of grinded fused silica optics by confocal fluorescence microscopy,” Opt. Express 17(5), 3543–3554 (2009), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-17-5-3543. [CrossRef] [PubMed]
14. V. Uhl, K. O. Greulich, and S. Thomas, ““Comparison of the influence of the fictive and the annealing temperature on the UV-transmission properties of synthetic fused silica,” Appl. Phys,” Adv. Mater. 65, 457–462 (1997).
15. C. Muhlig, W. Triebel, S. Bark-Zollmann, and D. Grebner, “In situ diagnostics of pulse laser-induced defects in DUV transparent fused silica glasses,” Nucl. Instrum. Methods Phys. Res. B 166-167, 698–703 (2000). [CrossRef]
16. R. A. Negres, M. W. Burke, P. P. DeMange, S. B. Sutton, M. D. Feit, and S. G. Demos, “Evaluation of UV absorption coefficient in laser-modified fused silica,” Appl. Phys. Lett. 90(6), 061115 (2007). [CrossRef]
17. P. Barritault, P. R. Bouchut, H. Bercegol, P. Chaton, and G. Ravel, “Fluorescence of mitigated laser damage in fused silica,” Proc. SPIE 5647, 188–196 (2005). [CrossRef]
18. R. A. Negres, M. A. Norton, Z. M. Liao, D. A. Cross, J. D. Bude, and C. W. Carr, “The effect of pulse duration on the growth rate of laser-induced damage sites at 351 nm on fused silica surfaces,” Proc. SPIE 7504, 750412 (2009). [CrossRef]
19. S. G. Demos, C. A. Lieber, B. Lin, and R. Ramsamooj, “Imaging of tissue microstructure using a multimodal microscope design,” IEEE J. Sel. Top. Quantum Electron. 11(4), 752–758 (2005). [CrossRef]
20. M. von Allmen, and A. Blatter, Laser-Beam Interactions with Materials: Physical Principles and Applications, (Springer-Verlag, Hiedelberg, Germany, 1998).
21. S. Elhadj, M. J. Matthews, S. T. Yang, D. J. Cooke, J. S. Stolken, R. M. Vignes, V. G. Draggoo, and S. E. Bisson, “Determination of the intrinsic temperature dependent thermal conductivity from analysis of surface temperature of laser irradiated materials,” Appl. Phys. Lett. 96(7), 071110 (2010). [CrossRef]
22. S. T. Yang, M. J. Matthews, S. Elhadj, V. G. Draggoo, and S. E. Bisson, “Thermal transport in CO2 laser irradiated fused silica: In situ measurements and analysis,” J. Appl. Phys. 106(10), 103106 (2009). [CrossRef]
23. F. Messina and M. Cannas, “Temperature dependence of the generation and decay of E’ centers induced in silica by 4.7 eV laser radiation,” J. Non-Cryst. Solids 355(18-21), 1038–1041 (2009). [CrossRef]
24. N. Shen, M. J. Matthews, J. E. Fair, J. A. Britten, H. T. Nguyen, D. Cooke, S. Elhadj, and S. T. Yang, “Laser smoothing of sub-micron grooves in hydroxyl-rich fused silica,” Appl. Surf. Sci. 256(12), 4031–4037 (2010). [CrossRef]
25. W. W. Mullins, “Flattening of a Nearly Plane Solid Surface Due to Capillarity,” J. Appl. Phys. 30(1), 77–83 (1959). [CrossRef]
26. V. Zandian, J. S. Florry, and D. Taylor, “Viscosity of Fused-Silica with Different Hydroxyl Contents,” Br. Ceram., Trans. J. 90, 59–60 (1991).
27. W. D. Kingery, “Surface Tension of Some Liquid Oxides and Their Temperature Coefficients,” J. Am. Ceram. Soc. 42(1), 6–10 (1959). [CrossRef]
28. N. Shen, J. D. Bude, T. A. Laurence, W. A. Steele, P. E. Miller, M. D. Feit, L. L. Wong, and T. I. Suratwala, “Isothermal anneal of damage precursors on fused silica surfaces,” submitted.
29. F. Y. Genin, A. Salleo, T. V. Pistor, and L. L. Chase, “Role of light intensification by cracks in optical breakdown on surfaces,” J. Opt. Soc. Am. A 18(10), 2607–2616 (2001). [CrossRef]