Surface damage is known to occur at fluences well below the intrinsic limit of the fused silica. A native surface precursor can absorb sub band-gap light and initiate a process which leads to catastrophic damage many micrometers deep with prominent fracture networks. Previously, the absorption front model of damage initiation has been proposed to explain how this nano-scale absorption can lead to macro-scale damage. However, model precursor systems designed to study initiation experimentally have not been able to clearly reproduce these damage events. In our study, we create artificial absorbers on fused silica substrates to investigate precursor properties critical for native surface damage initiation. Thin optically absorbing films of different materials were deposited on silica surfaces and then damage tested and characterized. We demonstrated that strong interfacial adhesion strength between absorbers and silica is crucial for the launch of an absorption front and subsequent damage initiation. Simulations using the absorption-front model are performed and agree qualitatively with experimental results.
© 2014 Optical Society of America
Surface laser damage limits the lifetime of optics in high fluence laser systems such as large scale experimental laser facilities designed to explore Inertial Confinement Fusion and High Energy Density Science as well as applications which require compact high energy light sources. While the intrinsic damage threshold of bulk optical material such as silica in the UV is greater than 100 J/cm2  damage can occur at fluences as low as a few J/cm2 with nanosecond scale pulses. This damage results from the absorption of sub band-gap light by defects near the surface of an optic. In many cases these damage precursors are unknown. In order to identify and ultimately remove damage precursors, it is desirable to understand the general optical and physical properties a precursor must have in order to initiate a damage event. Studies of optical damage with model or surrogate precursors with known properties are often used to this end [2–6]. Many of these experiments with model precursors are performed on fused silica substrates because this material can be made with very high quality and is an important component in most high fluence systems.
Metals are the materials most frequently chosen as model precursors because they can be easily deposited on surfaces, they are known absorbers for all wavelengths of interest, and they have fairly well-known physical properties. They are also a known source of damage, as metal particulates are a common source of contaminants in many laser systems [2, 4, 7–9]. Studies of contamination-induced damage have explored a variety of deposition methods and size regimes including metal particles greater than 20 μm dispersed on optical surfaces  and micron-thick photo-lithographically defined dots of Al on silica surfaces . Absorbing defects can also be trapped near the surface, typically as absorbing inclusions or absorbing particles trapped in sub-surface fractures. Model systems for these conditions include Au nano-particles deposited on silica surfaces which are then buried by a thin (100 nm thick) deposited oxide layer  and sub-micron Au particles covered by micron thick oxide over-layers .
While contaminants and inclusions are important sources of damage in many cases, surface damage on high quality optics is typically dominated by other sources, all of which lie either on or within a few hundred nanometers of the surface. To distinguish these precursors and damage sites from contaminants and inclusions, we refer to them simply as “native” surface precursors and damage sites. State-of-the-art optics can be made essentially clean of inclusions and trapped impurities . For these materials, fracture surfaces are the most important native precursors for low fluences (below 10 J/cm2, 3 ns Gaussian, 355 nm pulses). These sub-surface fractures can be effectively mitigated by carefully controlled chemical etching processes [10, 11]. However, the maximum fluence an optical surface can withstand is limited by additional unknown type(s) of native precursors – “high fluence precursors.” They are not associated with known flaws, not detectable by optical microscopy, are nano-scale and proximate to the surface. The damage density associated with high fluence precursors increases strongly with fluence [1, 12], resulting in a threshold behavior. This type of quasi-threshold behavior is seen in many types of optical materials, and for high quality fused silica surfaces, the onset of damage is observed around 15 J/cm2 for 3 ns Gaussian, 355 nm laser pulses with a spot size of ~100 μm in diameter. Damage sites associated with these sources (native surface damage sites) are typically many microns deep, open to the surface, and surrounded by extensive fracture  which suggests that although the initial absorption occurs at the surface, significant energy must be deposited deep (microns) below it . This morphology is quite important: these types of deeply fractured sites lead to damage growth when exposed to additional laser pulses, and hence ultimately limit an optic’s lifetime [15, 16].
While previous model precursor studies are well suited to study environmental contaminants and inclusions, they are not well suited to study native surface damage. Good model precursors for damage on high quality optics must lie within 100 nm of the surface and must absorb enough energy to produce many micron deep fractured damage sites. While the particle contamination work in , Honig et al, produced fractured damage sites, the particles which created damage were greater than 50 μm, much larger than the nano-scale precursors associated with surface damage sites, and therefore, not good surrogates for native surface damage. The micron thick Al layers used in the disc studies  only produce extensive fracture for high fluence (40 J/cm2 1064 nm pulses). At fluences lower than 20 J/cm2, fairly smooth pits were formed in the silica with maximum depths of 400 nm. These pits are likely formed by evaporation or shock densification from the heated expanding metal layer on the silica surface. Unlike native surface damage sites, shallow pits without fracture do not grow easily . Finally, buried Au nano-particles  produce damage typical of inclusions; when buried deep (10 μm below the surface) they produce extensive fracture, but when buried close to the surface (2 μm), the laser-heated gold particles removed the over-layer easily leaving a pit; these nano-particles do not produce much material removal or fracture below the absorbing particle itself.
Previously, we proposed a model in which nano-scale absorption can produce native damage sites . In this model, a surface absorber creates a native fractured damage site through formation of a laser-driven absorption-front (AF). When the absorber heats the underlying silica, the silica begins to absorb. This temperature-activated bulk absorption can be explained through Urbach broadening and band-gap narrowing . Temperature-activated absorption triggers a thermal run-away as more bulk material begins to absorb. The increased absorption is accompanied by temperature-activated thermal conduction, and the two processes act to generate an absorption-front which propagates into the bulk where it deposits energy deep below the surface. Energy deposited within the bulk leads to explosive ejection of material  and fracture due to thermal expansion of sub-surface material. In contrast, when a laser heats a thin deposited metal film, the film expands, ejects away and/or possibly evaporates. At sufficiently high fluences (more than a few J/cm2 at 355 nm), the metal is removed, and heat and pressure from the metal leave pits without significant fracture in the underlying silica . In the context of the absorption-front model, it is possible that the metal absorbers described above don’t create native damage sites because they lose thermal contact with the bulk material before the bulk substrate is heated sufficiently to initiate an absorption front.
In this work, we seek to create a model precursor system appropriate for initiation similar to native surface damage sites. This model system will help us determine the physical and optical properties of the precursors essential for producing typical fractured damage sites on silica surface. We will also be able to test the absorption-front model of damage initiation against experimental data. In the first part of this study, we extend thin metal film studies to include a variety of absorbing films chosen to increase residence time on the surface making generation of an absorption-front more likely. Then, thermal annealing is explored as a means to increase residence time by increasing the surface adhesion between the absorbers and the substrate. The discs were laser tested, and the results were compared to computational models developed in , Carr et al, to help clarify elements of the physics which produce the limiting damage associated with native damage on high quality optics. Furthermore, these comparisons demonstrate a temperature threshold for native damage and support the general features of the absorption-front model. While directly relevant to the study of native damage, this work also helps to clarify elements of the laser-matter interaction associated with contaminants and inclusions differentiating them from native surface damage processes.
2. As-deposited thin metal film on silica surface
Since native damage precursors absorb very near the surface, we first explored a variety of thin absorbing films. The absorption characteristics of metal films has been well-studied which allows us to better estimate and control the substrate temperature caused by the artificial absorber. Thin films of different materials with varying thicknesses (5 nm to 1 μm) were deposited on fused silica substrates previously treated using the AMP protocol . All films were deposited using electron-beam evaporation method and were then laser damage tested. In this study, we used small diameter laser beams for damage testing in order to target specific regions of the sample surface with spatial selectivity (our absorbers are highly localized: uniform thin films and later, small metal discs). The small beam laser damage test in this study was performed using a Coherent Infinity Q-switched Nd:YAG laser operating at 355 nm (Fig. 1). The temporal profile of the laser pulse is Gaussian with a full-width-half-maximum pulse duration of ~3 ns. The laser is focused onto the thin film on the exit surface of the silica part unless otherwise noted using a long focal distance lens. A single laser pulse is used for every test site. The laser pulse energy and its spatial profile are monitored by picking off a fraction of the beam and recording it using a charge coupled device (CCD) camera. The beam has a 1/e2 spot size of ~80 μm. An imaging microscope is set up to observe the sample under laser irradiation and identify laser-induced morphological changes to the film. Two types of morphological changes were identified on the sample after laser irradiation in this study. When the laser fluence is sufficiently high, the absorbing film is removed and fractures are generated in the substrate silica. We call this the damage regime. Below this laser damage fluence, the film may still detach but will leave behind a shallow pit without observable fracture. This is the cleaning regime as seen in the some of the work cited above.
For gold films 5 and 20 nm in thickness, no substrate damage was observed until ~35 J/cm2 and higher, which is close to the typical damage fluence for the pristine (cleaned and etched using the AMP protocol) fused silica after AMP cleaning. With a more careful examination, we noticed that the metal films were cleaned off using less than 1 J/cm2 of laser fluence. It is clear that although the gold film should have been heated up to at least a few thousand degrees, there was no significant energy transfer to the underlying silica – no large-scale melting, vaporization or facture. We also deposited aluminum and silicon films on silica substrate and observed a similar laser “cleaning” threshold with no reduction in the silica damage threshold. Optical microscopy images of cleaned (top) and damaged films (bottom) are shown in Fig. 2.Panels (a) and (c) were taken in transmission mode for gold and silicon films respectively. Reflected light was used for panel (b) with aluminum film. While the laser beam is Gaussian and slightly elliptical in the central region, the irregular shape of the film removal is due to fact that there is still sufficient fluence at the wings of the beam for cleaning. We believe that the film quickly lost contact with the substrate silica during laser irradiation before significant energy was thermally coupled to the glass.
In order to choose a more appropriate surrogate absorber with sufficient thermal coupling to the substrate silica, we tested a series of materials as thin films deposited on SiO2. Table 1 lists the materials used and their properties. Tungsten was chosen for its higher melting point and lower vapor pressure; thin deposited Si films were chosen as they were expected to bond better to the substrate; and capping layers of chemical vapor deposited SiO2 up to 1μm thick were used to help confine the expanding films to the surface.
The laser fluences at which the film detachment from the substrate is observed (cleaning) and the fluence at which substrate silica damage takes place are also reported in Table 1. In all cases, the thin films chosen in this study were removed from the substrate at very low laser fluences with little dependence on the film material properties. The thicker capping over-laying had little effect as well. No silica substrate damage was observed below the damage fluence of the fused silica control part. It appears that as-deposited thin films quickly lose contact with the underlying fused silica, even with the additional capping over-layer, before transferring heat to the substrate.
This film-substrate energy transfer is a crucial step in the absorption-front model and is clearly necessary to generate classical damage sites. Based on simulation (see details of the simulations in Section 5), the thin films reach temperatures above their melting points at a fluence of 1 J/cm2 but below their boiling points. Since the films are completely cleaned from the surface at these fluences, the dominant force for film removal is likely inertial ejection as the center of mass of the expanding film moves away from the surface. Hence, it is likely that the bonding forces between the films and the substrate are not strong enough to ensure thermal contact between the film and the substrate during the duration of the pulse. In this scenario, even at higher fluences where film temperatures become much higher (> 5,000 K), the films are ejected during the pulse before enough heat is transferred to the substrate. While this is a likely explanation, discussion of the details of the physical film ejection processes are beyond the scope of this work.
3. Thermally annealed particles and thin metal discs on silica
3.1 Annealed aluminum particles
It is likely that the thermal coupling between the deposited film and substrate in our study needs to increase in order to be representative of surface absorbers of laser induced damage in the absorption-front model. Thermal annealing has been widely used to improve bonding (adhesion) between films through the formation of strong chemical bonds or the inter-diffusion of atomic species [21, 22]. Melting can also increase the surface area of contact if the interface is rough. Therefore, we chose to anneal model aluminum precursors (thin films or aluminum particles as described next) on silica surfaces in an oven to attempt to improve the adhesion between to the substrate, and consequently, the thermal coupling at the interface. Because of the large difference in the thermal expansion coefficients between aluminum and silica, annealing fractures may occur in a large area film and/or glass. Vaporization of aluminum should also be minimized during annealing. Careful determination of the annealing parameters such as temperature, heating and cooling rate and time therefore is necessary.
We first tested the annealing of artificial absorbers on silica surface in order to enhance thermal coupling using aluminum particles. The particles were prepared by filing a piece of aluminum and collecting the particles onto a precision sieve. The particles 1-10μm in diameter were then deposited on a fused silica part previously processed using the AMP protocol. Lose particles were shaken off by lightly tapping the silica part. After 1-2 days of settling time in a cleanroom environment, the silica parts with aluminum particles were loaded into an oven for annealing. The heating and cooling cycle was controlled to minimize the possibility of introducing thermal stress in the glass by using a 5 K/minute heating rate and a 0.5 K/minute cooling rate.
We annealed three aluminum particle samples at 873 K, 973 K and 1023 K for 2 hours each. The samples were inspected using optical microscopy and SEM to ensure that no stress fracture was generated in the glass or the aluminum particles after annealing. At laser fluences too low to produce observable damage for pristine substrate glass, the aluminum particles in the un-annealed control sample created smooth and shallow pits upon laser irradiation. The pit depth was estimated to be roughly a few hundred nanometers deep using SEM measurements, likely due to impact plastic deformation when the particles were ejected from the glass surface. This observation is consistent with previous studies of damage initiations using gold nano-particles .
The fluence at which substrate damaged with fracture and molten cores characteristic of “native damage” was observed and is plotted in Fig. 3 as a function of annealing temperature. Note the fluence where the pristine (no particle) surface was observed to damage under similar test conditions was about 41 J/cm2, consistent with silica surface with un-annealed particles. As the annealing temperature increases, the damage threshold in the substrate decreases, indicating a gradual strengthening of the energy coupling between the Al particle and glass which suggests the activation of an absorption-front in the glass. This increase in thermal coupling appears to level off for annealing temperatures ~973 K and higher. An area of ~3 cm2 was laser tested and we found that at ~8 J/cm2 about 50% of the aluminum particles caused damage in the substrate glass.
3.1 Annealed aluminum discs
The study using aluminum particles demonstrated that artificial absorbers can be created on silica surface which mimic native precursors. Thermal annealing is a crucial step that ensures bonding hence thermal coupling between surface absorbers and substrate silica for energy transfer. For better control of the size distribution of the artificial absorbers, we created an array of 10 μm-diameter, ~350 nm and ~500 nm thick aluminum discs using electron beam deposition on silica substrates previously cleaned, followed by wet etch with a photolithographic mask. Compared to the aluminum particles, the disc also offers a more uniform and well-defined contact area with the silica and a more precise estimation of optical absorption and temperature increase during laser pulse irradiation. The aluminum discs were then tested using the small laser beam setup described in the previous section before and after thermal annealing. We characterized the different morphology changes and chemical composition of the aluminum discs and substrate silica which resulted from the laser irradiation using a number of methods including white light and confocal laser scanning microscopy, SEM, energy-dispersive X-ray spectroscopy (EDS), AFM and contact needle profilometry.
As a control, the aluminum disc samples were first laser tested without any thermal annealing. Again, the silica substrate did not show any fractured damage until ~35 J/cm2 with the 350 nm and 500 nm discs on the exit surface of the sample in the laser beam path. They start to detach from the substrate when the laser fluence exceeds ~1 J/cm2, but the underlying silica only shows cleaning behavior–shallow smooth plastic deformation pits with slight increase in depth as a function of fluence.
The samples were then annealed at 873 K for 2 hours with a controlled heating and cooling cycle and were laser tested again. The ~350 nm thick metal discs detach from silica surface at ~1 J/cm2 and create rough pits a few hundred nanometers deep without fracture in the glass and stay close to constant in depth for fluence up to ~8 J/cm2 (Fig. 4). However, beyond a certain fluence threshold, the behavior changes dramatically showing strong evidence that energy now is deposited deeper in the substrate consistent with the formation of an absorption-front.
In the ~350 nm thick disc sample at about 8 J/cm2, fractures start to appear right around the edge of the aluminum disc. With increasing fluence, these fractures become more pronounced first within the footprint of the disc. Molten silica can be seen in the SEM images, and as the fluence increases evidence of material eruption from beneath the surface appears with “petals” of material which rise from the center and fill the site. Eventually energy deposition is deep enough that subsurface stresses release lateral fractures which extend well beyond the perimeter of the patterned aluminum (see for example, Fig. 4(f)). The depth of the pits also takes a sharp turn and begins to increase almost linearly for fluences higher than ~8 J/cm2. We measured the average depth of the disc sites as a function of the incident laser fluence for both the control and thermal annealed samples using laser scanning confocal microscopy, and plotted the result in Fig. 5. Here, average depth is defined as the average depth of the line profile within the footprint of the 10 μm disc (e.g., it does not include fractures or small area pits).
Figure 5 shows that the annealed discs can be laser-cleaned but do not generate damage below 8 J/cm2. In this cleaning range between 1 and 8 J/cm2, the surface morphology of the silica substrate is different than the control: it is rougher and recessed about 300nm below the substrate surface (some substrate material left with the metal) – see Fig. 4(c). Clearly the annealing has affected the layer of glass up to ~300 nm below the interface, similar in depth to recent laser ablation studies of contaminants on silica . While the diffusion length of aluminum atoms in glass under these conditions is very short (a few nm ), it is possible that the anneal left a layer of devitrified glass stuck to the bottom of the aluminum. One question to ask is whether the substrate is modified below this 300 nm layer in such a way that it could complicate interpretation of the experiment. To test whether this happens, we first laser-cleaned a number of annealed aluminum discs using a low fluence of ~5 J/cm2 to minimize unnecessary heat deposition. These sites were then shot again at a higher fluence to determine the threshold at which damage occurs in cleaned silica beneath the discs. As shown also Fig. 5, substrate damage does not happen until ~22 J/cm2. This verifies that once annealed, the aluminum discs (and, while unlikely, possibly the material just beneath it) indeed act as the initial absorber and facilitate the damage event in the substrate silica. The annealing process does not create absorbers in the bulk of the silica far enough below the surface (i.e. deeper than this 300nm affected region) to explain the damage shown in Fig. 5 created at higher fluence. Further evidence that this is so is given in the simulation section.
We have argued above that the annealing process improved the adhesion between the aluminum disc and substrate silica allowing better thermal coupling of the absorbed laser energy to the glass and longer contact at the interface. One might also expect that given stronger adhesion, it would take more energy to clean off the metal disc after annealing. It appears initially that the discs detach at a similar fluence for both annealed and un-annealed samples when examined under optical microscope (images not shown). To show that annealing leads to stronger adhesion, we compared the morphology of the site with the aluminum discs laser cleaned off with and without annealing. We also performed energy-dispersive X-ray spectroscopy (EDS) to determine whether any aluminum remains after the laser-cleaning of the anneal samples. These experiments were performed on the sample with the 500 nm thick discs for convenience. Both the AFM and SEM images of the laser cleaned sites without annealing show complete removal of the disc at ~1.5 J/cm2 and a residual shallow pit ~100 nm in depth. After thermal annealing, although most of the aluminum discs appeared to be removed at ~1.5 J/cm2, a few patches higher than the substrate surface still remain after laser-irradiation. This is shown in Fig. 6 which compares the depth profiles for un-annealed and annealed laser-irradiated samples: removal of the aluminum disc in the un-annealed sample in (a) shows a clean pit, while the surface profile of the annealed disc site in panel (b) still appears to be above the substrate surface even after laser irradiation at ~4 J/cm2. Clearly, some aluminum remains even after 4 J/cm2 in the annealed sample. EDS on these raised patches in Fig. 6(c) confirms a clear aluminum peak. Furthermore, the level of aluminum decreases with increasing laser fluence. At ~5 J/cm2, the level of remaining aluminum on the surface is down to our detection noise. This provides evidence that thermal annealing indeed increases adhesion between the metal film disc and silica so that the film disc remains in thermal contact with the substrate long enough to heat the glass in contact with the absorbing metal and initiate a damage-producing absorption-front.
4. Simulation and modeling
In the following section, we compare measured pit depths with simulations using the absorption-front model. The purpose of these simulations is to find qualitative agreement with measurement; the details of the process including the material response which leads to metal ejection and silica pit formation are beyond the scope of this work. The AF model simulates energy absorption and energy and light transport including temperature-activated absorption and temperature-activated thermal-conductivity in silica.
Since the discs are much wider (10 μm) than they are thick (350-500 nm), the simulations were performed in one dimension. To treat this problem, several modifications were made to the computational model. To properly capture absorption and reflection in thin metal films, the simple laser propagation model used in , Carr et al, was replaced with a Helmholtz solver which captures reflection from the metal, interference effects in the silica bulk and skin-depth effects in the metal films. Simple material motion was included through an elastic response model which couples lattice temperature to thermal expansion in order to better understand film release; it was not used in computing pit depths. Values for the real and imaginary parts of the indices of refraction for aluminum at 355 nm were taken from  (at room temperature). Room temperature thermal conductivity was used for aluminum (235 W/m-K). While the thermal conductivity of aluminum is expected to drop somewhat at higher temperature, few reliable values are published. Several values were explored, but the impact on the simulated results is not large. Temperature-activated absorption in silica was taken from fits to the simulations described in , Sadigh et al; we explored several value near the best fit of α(355 nm,TL) = 4x106 exp(−37,000/TL) cm−1, where TL is the lattice temperature. In one-dimension, the discs were represented as thin metal layers on the silica; both the 350 nm and 500 nm thick dick samples were simulated. The laser excitation source was modeled to fit experimental conditions as a Gaussian in time having a full-width-at-half-max of 3 ns. The simulations assumed that the metal films (annealed) were thermally coupled to the surface until the end of the pulse. This is likely a decent approximation. In all cases at higher temperatures, the thin films detach from the substrate near the end of the pulse: solution of the heat flow equations along with the thermo-elastic equations predicted that even strongly bound films should eject from the surface sometime near the end of the pulse when they are liquid (no longer strongly bound to the surface) and as the thermal expansion in the metal driving its center-of-mass forward turns off and inertial pulls the film away.
Figure 7 shows the simulated pit depths along with measurements (data re-plotted from Fig. 5) for the 350 and 500 nm thick aluminum disc films. The experimental depths are plotted as the depth of the pit relative to the depth of the cleaned pit. As seen in Figs. 5 and 7, the depth remains at the cleaned value of about 400 nm until 8 J/cm2. The depth of the pits generated by the laser-metal interaction was taken as the depth of the absorption front at the end of the laser pulse. The best fit to the data was found using α(355 nm,TL) = 1.8x106 exp(−37,000/TL) cm−1, close to the value from . The fit is reasonably good considering the complexity of the problem and the uncertainty in when the annealed film leaves thermal contact with the surface.
Most importantly, both the measured and simulated data show two clear trends: they exhibit a clear threshold below which there is only laser cleaning and above which, deep substrate damage – native damage sites–are formed; they also indicate that at the threshold fluence for this transition is higher for the thicker aluminum disc film. The threshold behavior in the simulation is due to the formation of a laser-supported AF as described above; with no temperature-activated absorption, heat does not penetrate into the substrate (gray curve in Fig. 7). Furthermore, the simulation predicts a higher threshold for AF formation in the 500 nm thick case because it takes more energy to heat the 500nm film than it does to heat the 350 nm thick disc film due to the larger thermal mass. This is very strong evidence that energy begins to be deposited in the bulk when a precursor (extrinsic absorber) reaches a critical temperature. Simulations estimate that this occurs between 5,000 K and 6,000 K in silica depending on the driving laser fluence. This behavior suggests that the onset of thermal runaway and AF formation is strongly fluence dependent; there is no evidence of a laser intensity dependence for this precursor system and as a consequence, no strong laser dependence in the intrinsic absorption. Finally, it is also strong evidence in support of an AF and the link between AF formation and native damage site formation: both measurement and simulation show a quasi-linear increase in depth with fluence above threshold; this behavior is also expected in the AF model.
Model damage precursors on silica surface can be used to help understand the damage initiation event and infer important properties of the actual native damage precursors. Using metal films deposited on silica substrate, we have learned that adhesion between the film and silica is crucial for precursor initiated damage. Here, we have experimentally demonstrated how nano-scale absorption can lead to macro-scale damage (i.e. native damage sites) as the absorption front model predicts. The important element appears to be adhesion to the surface: the absorbing precursors need to be strongly bound to the glass surface in order to initiate temperature-activated absorption in the substrate within the laser pulse duration. These experiments also support key features of the absorption-front model: both experiment and simulation show in qualitative agreement a sharp threshold temperature (~6000 K) above which an absorption-front forms in silica, and both indicate that the absorption front grows nearly linearly back into the bulk with increasing fluence consistent with results presented in , Carr et al. With careful design of further experiments, we hope to learn more about the properties of native damage precursors on the surface of silica optics such as size, density and method of deposition which cause them to be damage-prone.
The authors wish to thank Dr. Manyalibo J. Matthews for helpful discussions. We are also grateful to Elaine Behymer, Julie Hamilton and Alford Craig for preparing the thin film samples. This work was performed under the auspices of the U. S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC5207NA27344.
References and links
1. T. A. Laurence, J. D. Bude, S. Ly, N. Shen, and M. D. Feit, “Extracting the distribution of laser damage precursors on fused silica surfaces for 351 nm, 3 ns laser pulses at high fluences (20-150 J/cm2),” Opt. Express 20(10), 11561–11573 (2012). [CrossRef] [PubMed]
2. J. Honig, M. A. Norton, W. G. Hollingsworth, E. E. Donohue, and M. A. Johnson, “Experimental study of 351-nm and 527-nm laser-initiated surface damage on fused silica surfaces due to typical contaminants,” Proc. SPIE 5647, 129–135 (2005). [CrossRef]
3. P. Jonnard, G. Dufour, J. L. Rullier, J. P. Morreeuw, and J. T. Donohue, “Surface density enhancement of gold in silica film under laser irradiation at 355 nm,” Appl. Phys. Lett. 85(4), 591–593 (2004). [CrossRef]
4. S. Palmier, J. L. Rullier, J. Capoulade, and J. Y. Natoli, “Effect of laser irradiation on silica substrate contaminated by aluminum particles,” Appl. Opt. 47(8), 1164–1170 (2008). [CrossRef] [PubMed]
5. S. Papernov and A. W. Schmid, “Correlations between embedded single gold nanoparticles in SiO2 thin film and nanoscale crater formation induced by pulsed-laser radiation,” J. Appl. Phys. 92(10), 5720–5728 (2002). [CrossRef]
6. H. Bercegol, F. Bonneau, P. Bouchut, P. Combis, J. Donohue, L. Gallais, L. Lamaignere, C. Le Diraison, M. Loiseau, J. Y. Natoli, C. Pelle, M. Perra, J. L. Rullier, J. Vierne, and H. Ward, “Laser ablation of fused silica induced by gold nano-particles comparison of simulations and experiments at lambda=351 nm,” in High-Power Laser Ablation Iv, Pts 1 and 2 (SPIE, 2002), pp. 1055–1066.
7. M. J. Matthews, N. Shen, J. Honig, J. D. Bude, and A. M. Rubenchik, “Phase modulation and morphological evolution associated with surface-bound particle ablation,” J. Opt. Soc. Am. B 30(12), 3233–3242 (2013). [CrossRef]
8. F. Y. Genin, K. Michlitsch, J. Furr, M. R. Kozlowski, and P. Krulevitch, “Laser-induced damage of fused silica at 355 and 1064 nm initiated at aluminum contamination particles on the surface,” Proc. SPIE 2966, 126 (1996).
9. F. Y. Genin, A. M. Rubenchik, A. K. Burnhan, M. D. Feit, J. Yoshiyama, A. Fornier, C. Cordillot, and D. Schirmann, “Thin film contamination effects on laser-induced damage of fused silica surfaces at 355 nm,” Proc. SPIE 3492, 212–218 (1998).
10. 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]
11. 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]
12. C. W. Carr, M. D. Feit, M. C. Nostrand, and J. J. Adams, “Techniques for qualitative and quantitative measurement of aspects of laser-induced damage important for laser beam propagation,” Meas. Sci. Technol. 17(7), 1958–1962 (2006). [CrossRef]
13. C. W. Carr, D. A. Cross, M. A. Norton, and R. A. Negres, “The effect of laser pulse shape and duration on the size at which damage sites initiate and the implications to subsequent repair,” Opt. Express 19(S4Suppl 4), A859–A864 (2011). [CrossRef] [PubMed]
14. 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]
15. M. A. Norton, E. E. Donohue, M. D. Feit, R. P. Hackel, W. G. Hollingsworth, A. M. Rubenchik, and M. L. Spaeth, “Growth of laser damage on the input surface of SiO2 at 351 nm,” Proc. SPIE 6403, 64030L (2006). [CrossRef]
16. 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]
17. R. A. Negres, G. M. Abdulla, D. A. Cross, Z. M. Liao, and C. W. Carr, “Probability of growth of small damage sites on the exit surface of fused silica optics,” Opt. Express 20(12), 13030–13039 (2012). [CrossRef] [PubMed]
18. 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]
19. B. Sadigh, P. Erhart, D. Åberg, A. Trave, E. Schwegler, and J. Bude, “First-Principles Calculations of the Urbach Tail in the Optical Absorption Spectra of Silica Glass,” Phys. Rev. Lett. 106(2), 027401 (2011). [CrossRef] [PubMed]
20. R. N. Raman, S. Elhadj, R. A. Negres, M. J. Matthews, M. D. Feit, and S. G. Demos, “Characterization of ejected fused silica particles following surface breakdown with nanosecond pulses,” Opt. Express 20(25), 27708–27724 (2012). [CrossRef] [PubMed]
21. K. Takahashi, H. Ishii, Y. Takahashi, and K. Nishiguchi, “Valence auger analysis of the annealing effect on atomic interaction at titanium sapphire, titanium silica and silver silica interfaces,” Thin Solid Films 221(1-2), 98–103 (1992). [CrossRef]
22. K. L. Mittal and A. Pizzi, Adhesion Promotion Techniques: Technological Applications (Marcel Dekker, 1999).
23. H. G. Francois-Saint-Cyr, F. A. Stevie, J. M. McKinley, K. Elshot, L. Chow, and K. A. Richardson, “Diffusion of 18 elements implanted into thermally grown SiO2,” J. Appl. Phys. 94(12), 7433–7439 (2003). [CrossRef]