High energy laser systems are ultimately limited by laser-induced damage to their critical components. This is especially true of damage to critical fused silica optics, which grows rapidly upon exposure to additional laser pulses. Much progress has been made in eliminating damage precursors in as-processed fused silica optics (the advanced mitigation process, AMP3), and very high damage resistance has been demonstrated in laboratory studies. However, the full potential of these improvements has not yet been realized in actual laser systems. In this work, we explore the importance of additional damage sources–in particular, particle contamination–for fused silica optics fielded in a high-performance laser environment, the National Ignition Facility (NIF) laser system. We demonstrate that the most dangerous sources of particle contamination in a system-level environment are laser-driven particle sources. In the specific case of the NIF laser, we have identified the two important particle sources which account for nearly all the damage observed on AMP3 optics during full laser operation and present mitigations for these particle sources. Finally, with the elimination of these laser-driven particle sources, we demonstrate essentially damage free operation of AMP3 fused silica for ten large optics (a total of 12,000 cm2 of beam area) for shots from 8.6 J/cm2 to 9.5 J/cm2 of 351 nm light (3 ns Gaussian pulse shapes). Potentially many other pulsed high energy laser systems have similar particle sources, and given the insight provided by this study, their identification and elimination should be possible. The mitigations demonstrated here are currently being employed for all large UV silica optics on the National Ignition Facility.
© 2017 Optical Society of America
Over the past decade, significant effort has been invested in identifying and eliminating operationally limiting damage precursors from the surfaces of fused silica optics [1–11]. Two of the most important precursors, sub-surface damage and precipitates of trace impurities from processing steps, have been discovered, and approaches have been developed to reduce  and mitigate them. We refer to such processing techniques as the Advanced Mitigation Processes AMP2 [5,6] and AMP3 [7,8]. Significant improvements in damage resistance have been realized in laboratory tests of small optics treated with the AMP processes including a reduction of damage density by a factor of 10,000 for fluences up to 30 J/cm2. Similar efforts have improved the damage density of as-fabricated frequency conversion crystals and multi-layer dielectric mirrors; laser conditioning has been particularly successful for these classes of optics .
However, when fielded on actual laser systems, the full benefit of these improvements has not been fully realized. For example, a full scale AMP3 process has been developed to treat the UV (351 nm) silica optics used in the final optics assembly of the NIF laser (192 focusing lenses and sampling gratings each with 1,200 cm2 laser beam size) [14,15]. The NIF laser takes laser shots which expose these optics to UV fluences up to 8 to 9 J/cm2 (3ns Gaussian equivalent pulse lengths, 3nsG ). Based on laboratory experiments, one would expect no damage initiation on fused silica optics fielded on the NIF laser. In practice, many hundreds of damage sites can occur on optics exposed to several (five or more) shots at 8 J/cm2. Damage is managed on NIF through a robust optic recycle process, but reducing damage levels to those expected of optics processed using AMP3 would lead to a significant reduction in the cost for optic repair and replacement [17,18], an increase in the number of high fluence shots which could be taken, and significant increases in the maximum operating fluence of the laser. Over the past several years, we have investigated the source of this discrepancy between the damage rate observed on NIF and the rates of damage that would be expected from extrapolations from off-line optical tests. Several potential explanations have been proposed including: overall beam contrast, beam modulation from up-stream phase objects, underestimation of as processed damage precursors, and contamination of installed optics. While damage levels were ultimately reduced by improving the UV beam contrast to values of less than 10% root-mean-square using spatial light shapers , they remained at least ten times higher than expectations derived from small optic laboratory testing. Moreover, it was found that only a small number of sites could be traced to modulation from phase objects.
In this work, we show that as-processed AMP3 precursor densities are extremely low at fluences up to 9 J/cm2, accounting for five or fewer sites over 12,000 cm2 of beam area (the equivalent of five 0.5 m optics). Instead, we find that a particular class of contamination – contamination from laser-driven particle sources – is the dominant factor in high energy laser systems, and in this case, accounts for almost all of the observed UV fused silica damage. We demonstrate two important sources of laser-driven particle production, and describe the mechanisms which make this damage source so potent. Finally, we present mitigations for these damage sources, and demonstrate their effectiveness on a laser shot campaign involving 20 beamlines and 24,000 cm2 of optical surface on the NIF laser. Mitigation of laser-driven particle sources allows demonstration of the full potential of the AMP process improvements, eliminating more than 98% of the damage to large-scale (1,200 cm2) fused silica optics at 3 ns fluences near 9 J/cm2. These mitigations are currently being employed on all NIF’s 192 beamlines. This work shows that the elimination of laser-driven particle sources is an important consideration for any high fluence laser system [15,20–22].
2. Experimental methods and materials
2.1 Small area off-line damage testing
Off-line tests of small area (50 mm diameter) fused silica samples prepared using our laboratory AMP3 process  were performed in a controlled environment using the Optical Sciences Laser (OSL) laser . The OSL laser system and data analysis techniques used for damage testing are described in detail elsewhere . Unlike most damage test systems, the OSL laser uses a comparatively large area beam (one to ten cm2 area). Beam contrast provides damage density information for fluences ranging within 20% of the beam average by registering the location of damage sites to local fluence. OSL tests can reach an average fluence of up to ~40 J/cm2 with pulse shaping from 500ps to about 15 ns. Damage tests are performed in a vacuum chamber with a controlled environment from 10−6 Torr to about 10 Torr. Damage locations are mapped and registered to on-sample fiducials by an automated microscope which can provide individual damage site images with a resolution of 0.2 μm.
2.2 On-line experiments fielded on a full high energy laser system
Tests of optics fielded on the multi-mega Joule class NIF laser system  are referred to here as “on-line” tests. Note that since the NIF laser is a continually operating user facility, experiments dedicated to understanding damage are limited. Hence, some data presented here resulted from inspections of optics resulting from normal NIF operation. However, studies designed specially to understand optical damage were also performed and are presented here.
Here, we use the grating debris shield (GDS) as an example of a large, damage-prone fused silica optic, and the NIF final optics assembly (FOA)  as an example of a high energy laser environment (typical of high energy lasers for target shooting [15,20–22]) [see Fig. 1]. Within each NIF beamline, the GDS serves several functions. First it serves as a vacuum barrier between the hard vacuum (10−6 Torr) of the target chamber and the FOA which is maintained at 10 Torr of clean dry air (CDA). Second the GDS serves as a beam sampling grating allowing the energy and power of each beam to be monitored. Finally, the GDS serves as a mechanical barrier to protect upstream optics from damage from target debris in the event of a failure of the disposable debris shield (DDS).
As shown in Fig. 1 the GDS is located between the fused silica wedged focusing lens (WFL) and the DDS which is fabricated from a borosilicate glass (Borofloat) . These optics range in size from about 43 cm by 43 cm to 43 cm by 47 cm; the beam area is approximately 1,200 cm2, slightly smaller than the full optic aperture. All three optics see UV (351 nm) light, typically presented in a variety of complex pulse shapes. The NIF laser takes shots with UV energies up to 1.8 MJ which is equivalent to an approximate 351 nm 3 nsG fluence of 9 J/cm2 on these large silica optics.
Stray light from back reflections off optic surfaces (and for NIF, the laser targets as well) is unavoidable in any complex laser system. In the NIF laser, stray light from specular reflections from optical surfaces (ghosts) amounting to even a tiny fraction of the 1 kJ laser beam propagating through each final optics assembly can reach levels (~200 mJ/cm2) which will damage metal surfaces. The 6061-T6 Al alloy used in the construction of the beam enclosure is protected (armored) from stray light using commercial architectural (absorbing) glass (Supergrey ).
In many cases, the paths of the highest intensity stray light rays are readily predicted, and absorbing glass can be installed in places outside the beam area to intercept them before they reach dangerous fluence levels and strike metal surfaces. The general location of some key pieces of absorbing glass used in the FOA are indicated in Fig. 1, although no attempt has been made to represent the specific size or shape of each pieces. In most cases, the absorbing glass is deployed in frames inserted within the beam tube, extending a few centimeters into the beam tube aperture without intersecting the high-energy laser beam.
The WFL and the GDS are high value optics fabricated from high quality inclusion free fused silica finished with processes designed to achieve minimal sub-surface damage and minimal scatter for 351 nm wavelength light . Both surfaces of the WFL have a silanol terminated colloidal silica anti-reflection (AR) coating  that has been applied by dip coating. Historically, only the input surface of the GDS facing the WFL has been AR coated, in this case using a hexamethyldisilazane (HMDS) treated trimethylsilyl terminated colloidal silica that is applied using spin coating . The exit surface was not coated because of the difficulty in manufacturing a coated grating which met the diagnostic requirements for uniformity and efficiency.
While both the GDS and WFL damage at rates higher than expectations from laboratory scale testing, historically the GDS has experienced a much higher damage rate than the WFL. As its name suggests, the DDS is disposable – target debris limits its lifetime to less than ten shots; because the DDS is replaced frequently, it is made of lower cost, lower quality Borofloat glass .
On-line data was collected both in specifically designed on-line damage experiments and from inspections of optics from normal NIF operations. In both cases, optic inspections were performed with a commercial automated microscope (View Micro-Metrology, Summit 600) herein referred to as the “scanning optical microscope” capable of scanning an entire NIF optic and imaging individual damage sites with a resolution of 0.2 μm. These inspections were performed on optics received from the NIF laser and after cleaning and removal of the AR coatings. The optics are cleaned of particles and residue and stripped of the AR coating using an automated optic cleaning system which employs aqueous solutions of commercial detergents and basic (NaOH) solutions. For all the on-line experiments described here, only the 351 nm component of the fluence seen at the GDS is quoted, and fluences have been scaled to a 3nsG equivalent pulse length. Some unconverted light at 1053 nm is also present but is not considered to be a significant factor relative to damage initiation. Such unconverted 1053 nm light is, however, important for damage growth [31,32].
Note with respect to optic input and exit surfaces: for both small experimental optics and full-size optics tested on-line, light propagates first through the optic “input surface” and then exits the optic through the “exit surface”.
3. Sources of damage
3.1 As-processed damage precursors
Prior to the development of the AMP processes, optical damage on silica optics was dominated by innate (as processed or fabricated) damage precursors [1–4]. The AMP2 process mitigates damage precursors at the lowest fluences by etching away silica containing residual polishing compound and electronically defective surfaces associated with fractures from polishing and incidental contacts that occur during normal handling of the optics. The AMP3 process adds improvements to control the impurity levels (parts per billion) in processing liquids and control over the drying process to reduce nano-scale precipitates. Laboratory damage tests of small scale optics demonstrate the potential of enormous reductions in damage which can be expected to yield essentially damage free operation up to 10 J/cm2 for 351 nm, 3nsG light (a NIF equivalent of a 1.8 MJ ignition shot). Both AMP processes have been implemented as production processes for fused silica optics. Fig. 2(a) shows the density of damage sites initiated as a function of UV fluence on small area witness samples tested in the off-line OSL lab; here, the red line is a best fit (smooth, monotonic in fluence) to the data. This figure highlights two issues which complicate projections for large optics (~1,200 cm2 beam size in NIF). First, there is process variation associated with the ability to control impurities and drying: better results are routinely seen with the small optic laboratory process compared to the large optic production process employed in the Optics Processing Facility, OPF. There is also more run-to-run variability seen in the OPF process [process variability is shown as the shaded region in Fig. 2(a)]. The uncertainty in damage density due to process variability is amplified by the second issue, extrapolating results from small area tests with higher fluence (20 J/cm2) to the lower fluences (≤ 10 J/cm2) experienced by optics on NIF. Accurately measuring damage at low fluences requires testing much larger areas closer to the full size of the fielded optics. To better understand how a production process scales to lower fluence, we tested a 110 cm2 area from typical OPF witness samples using multiple exposures of the 3 cm diameter OSL laser beam on multiple samples. The results, shown in Fig. 2(b), produced two damage sites at roughly 15 J/cm2 mean fluence. This is consistent with the extrapolation from high fluence damage density shown as the solid blue curve in Fig. 2(b); this extrapolation is a piecewise exponential (monotonic in fluence) constrained as a best fit through the high fluence data labelled “OPF AMP” and chosen such that it yields two damage sites per 110 cm2 when integrated with the fluence histogram (labelled in the figure) of the laser shots taken on the samples in the experiment. The extrapolation gives us additional confidence that as-processed AMP3 fused silica optics (GDSs) should not damage up to 1.8 MJ operation; nonetheless, this extrapolation is still based upon a small area sampling.
These tests do however, confirm that damage sites generated from as-processed precursors appear different from those seen coming from the laser at fluences down near 10 J/cm2. Damage sites from AMP precursors appear symmetric (circular) with typical diameter to depth aspect ratio of 3:1. Damage sites imaged on GDSs fielded on NIF often appear irregular in shape with a much shallower aspect ratios. As we will demonstrate, these irregularly shaped sites are evidence of particle damage which has grown upon exposure to additional laser shots.
3.2 Environmental contaminants
Many studies have shown that the contamination of optical surfaces can dramatically reduce their damage threshold, e.g [33,35]. Whether they are a greater source of damage than the as processed precursors in high energy laser systems is unclear. Common contamination sources include organic compounds produced by out-gassing of construction materials, particularly in low pressure beam-line environments, and airborne particulates. Out-gassing of organics is minimized using strict control of using low out-gassing materials, various getters, and clean gas supply to ensure no degradation of AR performance due to pore filling of the coating. However, our studies have shown little impact of gas-phase organic compounds on the damage resistance of coated or uncoated silica optics . Although high quality optics are produced in cleanroom environments, preventing particle contamination is challenging, especially during the assembly, installation and use of optical components. Metallic and organic particles from optic frame assemblies and plastic (e.g. polyethylene terephthalate (PETG)) storage containers are examples of sources of such particles. Finally, particles on beam tube surfaces or other hardware in the beam tube enclosure can migrate to optical surfaces during laser operations including vacuum system pump-purge, exchange of optics (especially the DDSs which are routinely replaced), and small shocks to the metal beam tube walls from stray light impacts.
In addition to general environmental particle contaminants, one main source of beam-tube particle contamination comes from the DDS optics themselves. Because the DDSs are designed to intercept target debris, their exit surfaces can inject Borofloat particles into the downstream environment. As DDSs are exchanged, these particles can migrate upstream through a complicated series of steps and contaminate the GDS exit surface. Similarly, damage sites on the GDS can introduce silica glass fragments into the beamline. Absorbing glass can also introduce particle contamination in much the same way. Indeed, optics removed from the laser can have thousands of particles greater than 10 μm in size (many more less than 10 μm). Chemical analysis, of such particles, from optical components fielded on such systems, has shown that these include Borofloat glass, absorbing glass, silica glass fragments, solid organic polymers (e.g. Delrin PETG), target debris, and metals.
Any of these particle sources can potentially explain the higher than expected damage rates on GDS optics installed in an FOA. Some guidance is provided by previous laboratory studies of particle-related damage to fused silica optics [34,35]. In general, particle damage increases with particle size and laser fluence. Particle damage sites initiate as small (tens of microns) sites, but such initiated sites grow upon exposure to subsequent laser shots [37,38]. As with all laser damage, it is the growth of the initiated damage sites which limits the lifetime of the optic. This is particularly true of the exit surface damage which grows rapidly, sometimes exponentially, with additional laser shots. In contrast, input surface damage sites grow much more slowly. Hence, particles on the exit surface of an optic are much more problematic than those on input surfaces.
Semi-transparent particles (plastics and glass fragments) more often lead to growing damage than opaque metallic particles . Optical absorption occurs on the surfaces of metallic particles, generating high plasma pressures which propel them from the surface leaving small smooth pits [39,40]. Most energy deposition in semi-transparent particles occurs in the bulk resulting in the explosive destruction of the particle which is sometimes accompanied by the dispersal of material on the surface. Here, we describe the dispersed remnants of a particle following laser-induced destruction as a “splat.” Splats can be inert during further laser shots or they can grow to become laser damage sites on the silica substrate itself. Metallic particle damage has a fairly predictable behavior while damage in semi-transparent particles is largely stochastic. Glass fragment damage is described by a probability to damage, to laser-clean (removal without damage), or to remain unchanged as a function of particle size and laser shot fluence [34,35].
Based on these studies, one would expect that semi-transparent particles larger than about 10 μm are the most dangerous; however, because of the complex environment an installed optic sees and the effort required to address each potential contamination source, it is essential to investigate the impact of each using controlled on-line experiments. To that end, we designed experiments to track particles from the moment they left the optics processing facility cleanroom assembled in their mounting hardware through installation, laser operations, laser shots and optic removal. First, a GDS was processed through AMP3, assembled in its mounting frame, and then imaged in the scanning optical microscope to record the size and number of particles on the optic; about 500 particles of sizes up to about 30 μm were found. Then the optic was installed in an FOA, and a laser shot of 6 J/cm2 (approximate beam area of 1,200 cm2) was taken. The optic was removed and returned to the OPF cleanroom for a post-shot scan by optical microscopy.
Figure 3(a) shows the result of this experiment where the location of each site has been plotted as a function of their X and Y spatial coordinates (in millimeters) over the approximately 350 mm by 350 mm beam area (shown as a black square). The yellow symbols designate locations where as-installed particles were found that had been laser cleaned without resulting in damage initiation. Cyan symbols designate as-installed particles which exploded leaving a faint trace on the optic surface (a splat). The blue symbols indicate splats from particles not identified as pre-installed; such particles were most likely deposited during optic transport and installation. Green symbols indicate as-installed particles which did not move (these were outside the beam area and hence, not exposed to laser light. Finally, the red symbols indicate new particles which likely appeared either during the laser shot or during the removal and transport of the optic back to OPF. Upon reinstallation and further laser shots, none of the splats grew or damaged further. Table 1 summarizes the results of a series of controlled experiments like this that were designed to isolate various environmental, non-laser-shot particle sources. The results of these tests indicate that almost all non-shot particles (i.e. those introduced from the cleanroom, optic assembly, transport, optic installation, vacuum system venting, and the exchanges of adjacent optics) were removed from the optic by laser cleaning without damaging the underlying surface of the optic. All these particles were deposited at low velocity – either particles settling from air in the environment or from air contaminated by particles agitated by modest mechanical action or from another nearby contaminated surface. These results indicated that either a much higher density of particles or a different, more dangerous, source/type of particles is necessary to account for the hundreds of damage sites that routinely appear on-line, even with shots at fluence as low as 6 J/cm2.
3.3 Laser-driven particle contamination sources
3.3.1 General behavior
While non-shot related particle sources typically result in less than a thousand particles with little apparent damage to the fused silica GDS optic surface, laser shots introduce many new particles and many growing damage sites, possibly related to the introduction of such particles. Fig. 3(b) shows an optical microscope image of the same optic shown in Fig. 3(a) after an additional 20 shots at fluences near 6 J/cm2. More than 20,000 new particles, more than 2,000 larger than 15 μm, appear after these shots (see Table 1, laser shot row). Chemical analysis indicates that most of them are borosilicate (Borofloat) glass fragments, indicating that they originated from the DDS (Fig. 1). Many of them are low velocity, subject to settling on this upward facing optic due to gravity (gravity down arrow shown on the right). It is apparent that each additional shot introduces new particles; the particles exposed to the beam are largely cleaned with each shot, but particles accumulate outside the beam area, especially in regions below the beam. These laser shots produce more than 200 damage sites which will likely grow (damage sites are characterized by a fractured structure and are typically larger than about 20μm). Particle distributions typical of Fig. 3(b) are common in upward facing optics (the exit surface of the optic is facing up with respect to gravity).
We have made scanning optical microscopy measurements of particle number, size and distribution on optics taken from normal operations of the NIF laser, and correlated these to the orientation and laser damage found on the optics. Table 2 lists some typical findings: in most cases, the number of particles found on upward facing optics is much larger than the number found on downward facing optics, consistent with the notion that laser shots distribute many low velocity particles which subsequently settle on upward facing GDS surfaces. However, these excess particles don’t seem to be strongly associated with optical damage. Similarly, we attempted to clean the beam tube (BT) region between the GDS and the DDSs and found that while the number of particles appeared to be slightly reduced, the number of damage sites did not decrease. While the statistical variation in these measurements is large given that they were taken from optics fielded during normal operation of the NIF user facility, they do not appear to support an association between the number of particles and observed rates of optical damage. Certainly, the notion that low velocity particles of any sort, even in high numbers, produce damage is not supported by these observations.
3.3.2 Laser-driven particle sources
While several studies have proposed direct deposition of particles from nearby growing damage sites as a potential source of particle-related damage [34,35] none have directly explored deposition method as it relates to the damage process or its severity. The results of the study described above strongly suggests that if particle damage is the dominant source seen below 10 J/cm2, either particle type or deposition method must be considered. Since we suspect that the type of DDS particles seen after laser shots are of the most dangerous class (semi-transparent fragments), the deposition method is likely critical. The most natural source of laser-driven borosilicate glass is line-of-sight ejecta from the DDS input damage site [see Fig. 4]. Input surface damage in glass optics is typically small, grows very slowly (linearly with laser shots compared to exponential growth of exit surface sites in glasses) and rarely occurs. Given this most damage to the DDS would be present on the exit surface, where damage from target debris results in rapidly growing damage sites. However, if particles deposited by direct laser-driven events result in a higher probability of damage compared to particles which have settled on the GDS surfaces, even a small or rare site may be prove to be problematic.
To determine whether damage sites appear on DDS input surfaces, we developed techniques to mount and image these optics using scanning optical microscopy. A survey of a number previously used DDS optics indicates the presence of numerous locations of bulk damage. Depending on how long the DDS was installed, approximately one third to one half of used DDSs had one or more bulk damage sites which had vented to the input surface. Such erupted sites had diameters ranging from several hundred microns to one millimeter. Images of typical Input Surface Bulk Eruptions (ISBEs) are shown in Fig. 4. In two of the images (upper right and lower left), the bulk damage has not fully vented to the surface. Because the DDSs are disposable low cost optics, they are made of a commercial grade borosilicate glass. Each DDS has been observed to have up to 30 bulk inclusions per optic at NIF relevant fluences which are distributed randomly throughout the volume of the optic. As more laser shots are accumulated on a given DDS, these ISBEs continue to grow. Similarly, new sites can also vent to the surface, although this is less likely since each DDS is used for less than ten shots.
To better understand the magnitude of the particle problem from a single ISBE, we performed a series of off-line damage tests using the OSL laser. In these experiments, an ISBE approximately 650 μm in diameter was located on a used DDS, and a piece of the optic containing this site was sectioned for use in the OSL vacuum chamber. The ISBE was located on the input side facing incident laser light, and an AMP3 fused silica witness sample (called the catcher optic – Sample A) was placed upstream of it [see Fig. 5(a)]. After a single laser shot, the fused silica witness was imaged using optical microscopy. Fig. 6(a) shows the resulting particle size distribution for a 7 J/cm2 and a 9 J/cm2 shot.
After imaging and construction of the particle size histogram, the original fused silica catcher (Sample A) and its mount, was inverted (particles facing down) and dropped 10 cm onto a support designed to engage only the optical mount, arresting its fall. Loosely bound debris originally collected on the surface of the original catcher optic (Sample A) was dislodged by the mechanical shock and collected on the surface of a second optic (Sample B). Here, Sample A is designed to mimic direct laser-driven deposition of ISBE particles from the DDS input surface to the GDS exit surface (direct high energy deposition). Sample B represents the collection of slow, cool ISBE particles from the beam tube environment (indirect low energy deposition). This process was repeated, producing two samples with high velocity direct deposition [samples of type A shown as Samples 1 and 2 in Fig. 6(b)], and two samples with low velocity indirect deposition [samples of type B shown as Samples 3 and 4 in Fig. 6(b)]. All four samples were imaged by optical microscopy, and the location and size of each particle was recorded. To ensure that typical handling processes did not perturb the particles, a sample of each type was loaded into a vacuum chamber, the chamber was pumped down and subsequently vented. The substrates were then examined by optical microscopy, and it was determined that very few particles were found to have moved or to have been dislodged during handling.
All samples were then damage tested (DDS sample removed) with particles on their exit surface [see Fig. 5(c)] with a 351 nm 5ns flat-in-time (3nsG equivalent) laser shot at 9 J/cm2. The samples were imaged again to determine which particles created substrate damage (fractured damages sites which will grow), which were cleaned by the laser with little or no evidence left on the substrate, and which were unaffected by the laser. A fourth outcome was also noticed: some of these particles exploded on the surface dispersing material around the original particle site–a splat–as has been described above. A second shot at 9 J/cm2 was then taken to test whether the splats damaged during subsequent shots. The results of these damage tests are shown in Fig. 6(b) in which damage probability following two 9 J/cm2 shots is plotted as a function of particle diameter (specifically the effective circular diameter); the fractions labelled near each data point represent the number which damaged over the total number of particles in that size bin which were tested. In almost all cases tested here, the cold particles deposited at low velocity (indirect deposited Samples 3 and 4) were cleaned by the laser without producing damage on the AMP3 exit surface; only 5 out of 825 cold deposited particles (0.6%) damaged after these two shots. In contrast, the high-energy laser-driven particles which remained on the original catcher (laser-driven deposited particles on Samples 1 and 2) damaged with high probability which increased strongly with particle size – up 50% for particles larger than 20 μm.
There are two possible explanations for the high damage probability of the particles deposited directly from laser-driven ejecta from a growing ISBE damage site: adhesion to the surface and damage to the surface. Laser-driven particle ejection produces particles with a wide distribution of sizes and shapes ejected at velocities from a few meters per second (largest particles) up to 1 km/s . Many particles appear to make it to the catcher surface partially molten; occasionally, there is indication of substrate brittle fraction (chipping) nearby. These features are highlighted in the SEM images of ISBE particles on a catcher surface [Fig. 7(a)]. The most dangerous particles are those which cannot be easily dislodged by mechanical means. In fact, most of these tightly bound particles cannot be cleaned with ultra-sonic di-ionized water rinses or vigorous surfactant cleaning. Particle adhesion can in principal be enhanced when a molten particle solidifies on surface, or through the formation of a thin molten layer due to heating from high velocity impacts which solidifies between the particle and the substrate.
While not typical of the particles we inspected here, a small percentage show direct evidence of impact fracture to the substrate indicative of high particle velocity/kinetic energy (similar evidence for substrate fracture from high velocity particle impacts was also seen in ). However, it is possible the impact fracture is much more common than it would appear from these inspections. In fact, small surface fractures can close becoming invisible optically or with SEM. In past studies, we have used HF acid-based etching to enhance the visibility of surface fractures [2,3]. We have applied this technique here to determine whether small closed fractures typically occur beneath or near ISBE particle impacts. A fused silica ISBE catcher such as sample A was prepared in the configuration shown in Fig. 5(a), and particle locations were mapped using scanning optical microscopy. The sample was cleaned with an ultrasonic di-ionized water rinse and was then etched with buffered HF acid (BOE) to remove about 2.5 μm of silica from the surface.
The sample was then re-imaged with the scanning optical microscope and a subset of locations where particles had been was investigated using SEM. Figure 8 shows a sequence of images (pre-etch microscopy, post-etch microscopy, post-etch SEM) from two typical particles. In both cases, regions very near or beneath locations where the particles had been located formed etch pits [Fig. 8(c)]. While etch pits can result from regions of densified glass, those highlighted in Fig. 8 are typically associated with small sub-surface fractures proximate to a high velocity impact zone. Almost all the regions of particle impact we investigated were revealed to host impact fractures despite having no direct evidence of fracture before etching. Fracture surfaces are known to be some of the most potent damage precursors on silica surfaces [2–4]. In fact, AMP directly targets such precursors. Impact fracture can explain the high probability of laser damage for laser-driven particle sources [Fig. 6(b)]. However, adhesion to the surface may also play a role in increasing the probability that a surface-bound particle will damage by better allowing laser-absorbed energy in the particle to heat the substrate priming it for initiation of a laser-induced absorption-front . Laser-driven particles are clearly a much higher risk for laser damage than other means of particle contamination, likely due to their propensity for impact fracture.
Figure 7(b) shows SEM images of damage to the substrate at particles sites from Fig. 7(a) after a 9 J/cm2 laser shot. These damage sites initially appear as the shallow lateral fractures we refer to as “mussels” due to their resemblance to these bi-valves. Mussels like these are commonly found on GDS surfaces. Due to their fractured defect layers, mussels grow with subsequent laser shots. The probability of growth increases strongly with damage site size and laser fluence . Mussels grown by subsequent laser shots are the likely source of the many irregularly shaped damage sites seen GDSs. In these off-line laboratory experiments, we have demonstrated that a single 1-mm diameter ISBE with a volume of a fraction of 1 mm3 can produce thousands of particles [Fig. 6(a)] and a large number of damage sites (tens of sites with a single shot at 7 J/cm2). From this work and previous studies, we expect that any growing damage site in a glassy material (Borofloat, fused silica, absorbing glass) can generate highly damage-prone particles sites on fused silica surfaces which have a direct line-of-site for laser-driven deposition. Laser-driven particle sources are likely more dangerous than environmental particle contamination for generic glasses and other optical materials as well. However, the details of particle damage are specific to the substrate material. For example, while metal particles collected at low velocity from environmental sources are not problematic for transmissive fused silica optics, they are potent damage sources for mirrors with multi-layer dielectric interference coatings due to the thermal-mechanical properties of deposited coating materials .
4. Isolation and mitigation of particle damage sources
A series of on-line laser experiments were designed specifically to determine the importance of laser-driven ISBEs as a particle-damage source. The results of the first experiment (on-line damage Experiment 1) revealed the presence of a second, previously unknown laser-driven particle source with a line-of-sight to the GDS exit surface. The first experiment was designed to test the impact of ISBEs on the on-line GDS damage rate. A specially designed optical assembly was fabricated to mount an inclusion-free fused silica optic between the DDS input surface and the GDS exit surface to block the ISBEs [see Fig. (9)]. Unlike the low-cost DDS Borofloat glass, the fused silica debris screen (FSDS) uses the same high quality inclusion free-glass as the GDS and the WFL optics. It was processed with an AMP3 process in OPF to mitigate the known as-processed damage precursors – it does not produce ISBEs.
Two four beam-line quads were chosen for this experiment, one where the exit surfaces of the optic faces downward with respect to gravity (Q14T), and one with optics facing upward (Q14B). New AMP3 GDSs were fielded in this on-line experiment, and two beam-lines from each quad were protected from ISBEs by the FSDSs. Fig. 10 shows the layout of the eight beam-line experiment. The direction of gravity is shown by an arrow on the left, and the location of the beam energy sampling drive diagnostics (DrD) are indicated for reference to the orientation of the FOAs shown in Fig. 9. Ten non-target shots were taken at approximately 8.6 J/cm2 3 nsG, and GDSs, FSDSs and DDSs were removed for inspection. Scanning optical microscopy images of the GDSs prior to cleaning are shown in Fig. 10; these detections include all particles and damage sites. The total number of particles larger than 10 μm is labelled for each beam-line.
Figure 11 shows scanning optical microscope images after cleaning the optics using a process that removes almost all particles. The remaining detections were inspected at a higher resolution, and were categorized as growing damage sites based on morphology (typically sites larger than 20um, irregularly shaped with fractured surfaces). The total number of growing sites (NG) is shown in Fig. 11 (blue) together with and the total number of ISBEs and their approximate volume as detected by scanning optical microscope images of the DDSs (green). Several important conclusions can be drawn from these measurements. First, the ISBE particle screens (FSDSs) do not appear to have been effective in reducing the number of growing damage sites. Second, several areas of enhanced particle density appear. These are correlated to the concentrations of growing damage sites which gives a clear sense of an association between particle location and damage location. These areas of enhanced particle collection are in the same position within each beam-line. Specifically, they are concentrated at the location facing an energy sampling diagnostic (the drive diagnostic, DrD) 1.7 m away. Hence, there appears to be another source of laser-driven particle contamination coming from the vicinity of the DrD. It should be noted that this concentration of damage had not been evident in earlier GDS examinations from normal NIF operations since controlled experiments without recycling optics are rarely done. Although careful retrospective examination of accumulated data from previous studies did reveal evidence of such localized damage it was not as clear as the data from the two quads described above.
The location of this localized damage was eventually identified as originating from an unexpectedly energetic stray light ghost originating from the back reflection off the bare (uncoated) GDS exit surface, which propagates through the wedged focusing lens (WFL), reflects off the input surface of the THG frequency conversion crystal (AR coated for 1053 nm and 527 nm, but not 351 nm), and back through the WFL coming to focus on a piece of absorbing glass [see Fig. 12]. This piece of absorbing glass was originally placed in this location to block forward propagating diffracted light from striking the metal beam tube surface. Inspections of this armor glass clearly showed a burn mark approximately 1-cm long (the exact location depends upon the specific focus points set for each beam-line, so the severity of this source and the exact particle trajectory changes slightly with time). Laser damage on these pieces of armor glass from a focusing stray-light ghost produces laser-driven particles similar to those produced by ISBEs. They also have a direct line-of-sight to the GDS exit surface and represent a second important laser-driven particle damage sources. In the following, we refer to this second damage source as Armor Glass Damage (AGD).
A second on-line experiment (on-line damage Experiment 2) was then designed to test for the presence and mitigation of both laser-driven particle sources [Fig. 12]. Mitigation of the second source was provided by a GDS exit surface AR coating that was developed to reduce energy loss (process development details will be reported elsewhere). Both the ISBE mitigation (FSDS) and the absorbing glass ghost damage mitigations are shown in Fig. 13. This second experiment also included a test of a potential laser cleaning mitigation. Laser cleaning has been described in previous works [35,39,44]. Here, we have tested the efficacy of laser cleaning as a mitigation for laser-driven particles (ISBEs). Some off-line tests in OSL indicated that laser-driven particles (ISBEs and AGD) could be safely removed used a moderate fluence (approximately 6 J/cm2). Subsequent high fluence shots did not produce damage at previous particle sites. Perhaps a more gentle shot overcomes particle adhesion without initiating substrate damage, or the shallow fractures near a particle can be conditioned by a lower fluence shot without damage. For example, the damage threshold of a fracture produced by static indentation is higher when the fluence is ramped from low to high (R/1 tests) compared to a single shot at higher fluence (S/1). While mechanisms of laser cleaning are beyond the scope of the present work, the potential of a laser-cleaning mitigation was tested here by alternating lower fluence cleaning shots (6 J/cm2) with the higher fluence shots of interest (9.8 J/cm2).
Experimental details for the second on-line particle mitigation experiment are summarized in Table 3. Both anti-reflective coated GDSs to mitigate absorbing glass damage and the fused silica debris screen to mitigate ISBEs were tested. Ten beam-lines tested only the AR coated GDS mitigation, while ten employed both the AR-coated GDSs and the FSDS. Laser cleaning was also tested on four beam-lines. Twelve beam-lines from quad Q14T, Q14B and Q26B were tested with ten shots at a mean fluence of 8.6 J/cm2, a full NIF equivalent (FNE) energy of 1.6 MJ. Note that Q14T and Q14B were also employed in the first on-line experiment (Figs. 11 and 12) and were shot at the same fluence. The first on-line experiment serves as a control set for GDSs without an AR coating.
Eight beamlines were tested at a higher fluence of 9.8 J/cm2, a FNE of 1.8 MJ – Q16T and Q23B. These higher fluence experiments also included a test of laser cleaning. Here low fluence (6 J/cm2) cleaning shots were interleaved between high fluence (9.8 J/cm2) shots in the Q23B quad. These GDSs saw a total of five high fluence (9.8 J/cm2) shots. The beam-lines in quad Q16T served as a laser cleaning control; they used the same number of 6 J/cm2 and 9.8 J/cm2 shots as in Q23B, but they were arranged so that after an initial low fluence shot, all the high fluence shots came in a group, which were then followed by the remaining five lower fluence shots.
Table 3 also presents the results of this study by beamline reporting the number of damage sites which aren’t obviously growing but could grow and the number of growing damage sites which have been assessed as described above. This data is also summarized in Table 4. Data corresponding to no mitigations (labelled “None”) were taken from the first on-line experiment which used beam-lines from two of the same quads (Q14T and Q14B) and saw the same shots. The findings from this experiment convincingly prove the role of laser-driven particle damage sources in on-line damage. First, compare the optics with no AR coated GDS to those with the AR coated GDSs (first and second rows of Table 4). Elimination of the ghost (reduced by several orders of magnitude) and absorbing glass damage by the anti-reflective coating drops the number of growing damage sites by a factor of five and the total number of damage sites by a factor of 40. Hence, for the NIF beam-line environment, the laser-driven absorbing glass particle source accounts for 80% of the growing damage sites and more than 95% of the total number of small sites which could eventually grow. While the statistics are low, the addition of laser particle cleaning can drop the number of growing sites by another factor of ten. More work is required to determine how robust laser cleaning is in practice.
The ten optics which benefitted from elimination of both laser-driven particle sources (last row of Table 4 – AR GDS and FSDS) are nearly damage-free. In fact, all but three optics were completely free of growing damage sites including all of those exposed to the five shots at 9.8 J/cm2 (see Table 3). Several of the damage sites which did appear for this set happened on unmitigated handling scratches that had been unwittingly introduced after the AMP3 process. Consequently, we can conclude that essentially all optical damage on AMP3 processed fused silica components appearing on-line at fluences up to 9.8 J/cm2 can be attributed to laser-driven particle sources. That is, AMP3 processed optics can be made essentially damage-free to these fluence levels.
We have shown the importance (and difficulty) of identifying and eliminating laser-driven particle sources to control the damage to large fused silica optics (and potentially other optics as well) in real laser system environments. For the specific case of the GDS optic on the NIF laser, we have identified the two important particle sources which account for almost all the damage observed on-line: absorbing glass damage from an unexpected energetic focusing ghost and input surface damage due to the eruption of bulk inclusions in DDSs fabricated from commercial grade glass. Particles due to a myriad of other sources were shown to be inconsequential with respect to damage. We have proposed and demonstrated two mitigations for the relevant damage mechanisms. A production process for applying an AR coating to the GDS exit surface has been implemented, and a reasonably low cost solution for a fused silica debris screen (FSDS) has been proposed with an expected lifetime of about half a year in normal NIF laser operation.
All pulsed high energy laser systems have similar laser-driven particle sources. This work provides the insight to identify and eliminate them. While the details are system dependent, care should be taken to eliminate laser damage sources which can particulate and which have a direct line-of-sight to the exit surfaces of critical optical components. Optical materials are an important class of materials which fall into this category. These laser-driven particle sources can occur on the input surfaces of optics facing other optics up-stream in a beam-line. We have identified one example, input-surface bulk eruptions from inclusions in economical commercial grade glass; input surface scratches which have not been treated by the AMP process are another. These sources can also be absorbing glass pieces exposed to stray light which exceeds their stray light capture design fluence. They can also be found in beam dump materials. When it is not possible to eliminate laser-driven sources either by improving material quality (inclusion free glass), eliminating the ghosts, or moving the sources out of the line-of-sight, they can be blocked by fused silica protective screens treated with the AMP process.
Finally, we have demonstrated that once extrinsic damage sources (laser-driven particles) have been eliminated, AMP3 processed fused silica optics can be made essentially damage-free up to fluences approaching 10 J/cm2, 3ns Gaussian equivalent pulse lengths. Off-line testing suggests that they can be made essentially damage free to even higher levels up to full NIF equivalent energies well beyond 2 MJ. Additional on-line experiments made possible by these particle mitigations are needed to demonstrate full AMP3 potential on-line.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344.
The authors wish to acknowledge NIF facility management for making on-line optical damage experiments possible.
References and links
1. J. Bude, P. Miller, T. Suratwala, T. Laurence, W. Steele, S. Baxamusa, L. Wong, C. Carr, D. Cross, M. Monticelli, M. Feit, and G. Guss, “Silica laser damage mechanisms, precursors and their mitigation,” Proc. SPIE 9237, 92307S (2014).
2. 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]
3. P. E. Miller, T. Suratwala, J. Bude, T. A. Laurence, N. Shen, W. A. Steele, M. D. Feit, J. A. Menapace, and L. L. Wong, “Identification of laser damage precursors in fused silica,” Proc. SPIE 7504, 75040X (2009). [CrossRef]
4. 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]
5. 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]
6. P. E. Miller, T. Suratwala, J. Bude, N. Shen, W. Steele, T. Laurence, M. Feit, and L. Wong, “Methods for globally treating silica optics to reduce optical damage,” US Patent 8,313,662, 4/7/2011.
7. J. Bude, P. Miller, S. Baxamusa, N. Shen, T. Laurence, W. Steele, T. Suratwala, L. Wong, W. Carr, D. Cross, and M. Monticelli, “High fluence laser damage precursors and their mitigation in fused silica,” Opt. Express 22(5), 5839–5851 (2014). [CrossRef] [PubMed]
8. S. Baxamusa, P. E. Miller, L. Wong, R. Steele, N. Shen, and J. Bude, “Mitigation of organic laser damage precursors from chemical processing of fused silica,” Opt. Express 22(24), 29568–29577 (2014). [CrossRef] [PubMed]
9. X. Ye, J. Huang, H. Liu, F. Geng, L. Sun, X. Jiang, W. Wu, L. Qiao, X. Zu, and W. Zheng, “Advanced mitigation process (AMP) for improving laser damage threshold of fused silica optics,” Sci. Rep. 6(1), 31111 (2016). [CrossRef] [PubMed]
10. J. Neauport, L. Lamaignere, H. Bercegol, F. Pilon, and J. C. Birolleau, “Polishing-induced contamination of fused silica optics and laser induced damage density at 351 nm,” Opt. Express 13(25), 10163–10171 (2005). [CrossRef] [PubMed]
11. M. Pfiffer, P. Cormont, E. Fargin, B. Bousquet, M. Dussauze, S. Lambert, and J. Néauport, “Effects of deep wet etching in HF/HNO3 and KOH solutions on the laser damage resistance and surface quality of fused silica optics at 351 nm,” Opt. Express 25(5), 4607–4620 (2017). [CrossRef] [PubMed]
12. T. Suratwala, L. Wong, P. Miller, M. Feit, J. Menapace, R. Steele, P. Davis, and D. Walmer, “Sub-surface mechanical damage distributions during grinding of fused silica optics,” J. Non-Cryst. Solids 352(52-54), 5601–5617 (2006). [CrossRef]
13. K. R. Manes, M. L. Spaeth, J. J. Adams, M. W. Bowers, J. D. Bude, C. W. Carr, A. D. Conder, D. A. Cross, S. G. Semos, and J. M. Di Nicola, “Damage mechanisms avoided or managed for NIF large optics,” Fus. Sci. Technol. 69(1), 146–249 (2016). [CrossRef]
14. C. A. Haynam, P. J. Wegner, J. M. Auerbach, M. W. Bowers, S. N. Dixit, G. V. Erbert, G. M. Heestand, M. A. Henesian, M. R. Hermann, K. S. Jancaitis, K. R. Manes, C. D. Marshall, N. C. Mehta, J. Menapace, E. Moses, J. R. Murray, M. C. Nostrand, C. D. Orth, R. Patterson, R. A. Sacks, M. J. Shaw, M. Spaeth, S. B. Sutton, W. H. Williams, C. C. Widmayer, R. K. White, S. T. Yang, and B. M. Van Wonterghem, “National Ignition Facility laser performance status,” Appl. Opt. 46(16), 3276–3303 (2007). [CrossRef] [PubMed]
15. M. L. Spaeth, K. R. Manes, D. H. Kalantar, P. E. Miller, J. E. Heebner, E. S. Bliss, D. R. Speck, T. G. Parham, P. K. Whitman, and P. J. Wegner, “Description of the NIF laser,” Fus. Sci. Technol. 69(1), 25–145 (2016). [CrossRef]
16. C. W. Carr, J. B. Trenholme, and M. L. Spaeth, “Effect of temporal pulse shape on optical damage,” Appl. Phys. Lett. 90(4), 041110 (2007). [CrossRef]
17. M. L. Spaeth, P. J. Wegner, T. I. Suratwala, M. C. Nostrand, J. D. Bude, A. D. Conder, J. A. Folta, J. E. Heebner, L. M. Kegelmeyer, B. J. MacGowan, D. C. Mason, M. J. Matthews, and P. K. Whitman, “Optics Recycle Loop Strategy for NIF Operations Above UV Laser-induced Damage Threshold,” Fus. Sci. Technol. 69(1), 265–294 (2016). [CrossRef]
18. J. Folta, M. Nostrand, J. Honig, N. Wong, F. Ravizza, P. Geraghty, M. Taranowski, G. Johnson, G. Larkin, D. Ravizza, J. Peterson, B. Welday, and P. Wegner, “Mitigation of laser damage on National Ignition Facility Optics in Volume Production,” Proc. SPIE 8885, 88850 (2013). [CrossRef]
19. J. Heebner, M. Borden, P. Miller, C. Stolz, T. Suratwala, P. Wegner, M. Hermann, M. Henesian, C. Haynam, S. Huner, K. Christensen, N. Wong, L. Seppala, G. Brunton, E. Tse, A. Awwal, M. Franks, E. Marley, K. Williams, M. Scanlan, T. Budge, M. Monticelli, D. Walmer, S. Dixit, C. Widmayer, J. Wolfe, J. Bude, K. McCarty, and J. M. DiNicola, “A programmable beam shaping system for tailoring the profile of high fluence laser beams,” Proc. SPIE 7842, 78421C (2010). [CrossRef]
20. N. Fleurot, C. Cavailler, and J. Bourgade, “The Laser Megajoule (LMJ) Project dedicated to inertial confinement fusion: Development and construction states,” Fusion Eng. Des. 74(1-4), 147–154 (2005). [CrossRef]
21. P. Vinini and M. Nicolaizeua, “LMJ: overview of recent advancements and very first experiments,” Photonics West 2015 – LASE, Feb 12 (2015).
22. W. Zheng, “Status of the SG-III solid-state laser facility,” J. Phys. Conf. Ser. 112(3), 032009 (2008). [CrossRef]
23. M. C. Nostrand, T. L. Weiland, R. L. Luthi, J. L. Vickers, W. D. Sell, J. A. Standly, J. Honig, J. Auerback, R. P. Hackel, and P. J. Wegner, “A larger aperture, high energy laser system for optics and optical component testing,” Proc. SPIE 5273, 325–333 (2004). [CrossRef]
24. 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]
25. P. J. Wegner, J. M. Auerbach, T. A. Biesiada Jr, S. N. Dixit, J. K. Lawson, J. A. Menapace, T. G. Parham, D. W. Swift, P. K. Whitman, and W. H. Williams, “NIF Final Optics System: Frequency Conversion and Beam Conditioning,” Proc. SPIE 5341, 180 (2004). [CrossRef]
26. Borofloat glass: Schott North America, Duryea Pa. 18642 http://www.us.schott.com/d/borofloat/23052e97-2e03-43e9-996b-267978964942/1.2/borofloat33_broschure_us_web_ds.pdf
27. Supergrey absorbing glass: supergrey is a highly absorptive soda lime glass doped with several transition metals (Fe,Ti) http://www.pilkington.com/north-america/usa/english/products/bp/byben.
28. P. A. Baisden, L. J. Atherton, R. A. Hawley, T. A. Land, J. A. Menapace, P. E. Miller, M. J. Runkel, M. L. Spaeth, C. J. Stolz, T. I. Suratwala, P. J. Wegner, and L. L. Wong, “Large Optics for the National Ignition Facility,” Fus. Sci. Technol. 69(1), 295–351 (2016). [CrossRef]
30. T. Suratwala, M. L. Hanna, E. L. Miller, P. Whitman, I. Thomas, P. Ehrmann, R. Maxwell, and A. Burnham, “Surface chemistry and trimethysilyl functionalization of Stober Silica Sols,” J. Non-Cryst. Solids 316(2-3), 349–361 (2003). [CrossRef]
31. 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 in SiO2 under multiple wavelength irradiation,” Proc. SPIE 5991, 599108 (2005). [CrossRef]
32. M. Chambonneau, M. Chanal, S. Reyné, G. Duchateau, J. Y. Natoli, and L. Lamaignère, “Investigations on laser damage growth in fused silica with simultaneous wavelength irradiation,” Appl. Opt. 54(6), 1463–1470 (2015). [CrossRef] [PubMed]
33. F. Y. Genin, M. R. Kozlowski, and R. Brusasco, “Catastrophic failure of contaminated fused silica optics at 255 nm,” Proc. SPIE 3047, 987–995 (1997).
34. J. Honig, M. Norton, W. Hollingsworth, E. Donohue, and M. 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]
35. R. N. Raman, S. G. Demos, N. Shen, E. Feigenbaum, R. A. Negres, S. Elhadj, A. M. Rubenchik, and M. J. Matthews, “Damage on fused silica optics caused by laser ablation of surface-bound microparticles,” Opt. Express 24(3), 2634–2647 (2016). [CrossRef] [PubMed]
36. 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]
37. 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]
38. 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]
39. M. Matthews, N. Shen, J. Honig, J. J. Bude, and A. Rubenchik, “Phase modulation and morphological evolution associated with surface-bound particle ablation,” J. Opt. Soc. Am. 30(12), 3233–3242 (2013). [CrossRef]
41. S. G. Demos and R. A. Negres, “Morphology of ejected particles and impact sites on intercepting substrates following exit-surface laser damage with nanosecond pulses in silica,” Opt. Eng. 56(1), 011016 (2016). [CrossRef]
43. S. R. Qiu, M. A. Norton, R. N. Raman, A. M. Rubenchik, C. D. Boley, A. Rigatti, P. B. Mirkarimi, C. J. Stolz, and M. J. Matthews, “Impact of laser-contaminant interaction on the performance of the protective capping layer of 1 ω high-reflection mirror coatings,” Appl. Opt. 54(29), 8607–8616 (2015). [CrossRef] [PubMed]
44. W. Zapka, W. Ziemlich, and A. C. Tam, “Efficient pulsed laser removal of 0.2um sized particles for a solid surface,” Appl. Phys. Lett. 58(20), 2217–2219 (1991). [CrossRef]