An experimental technique has been utilized to measure the variation of bulk damage scatter with damaging fluence in plates of KH2PO4 (KDP) crystals. Bulk damage in unconditioned and laser-conditioned doubler-cut KDP crystals has been studied using 527 nm (2ω) light at pulselengths of 0.3 - 10 ns. It is found that there is less scatter due to damage at fixed fluence for longer pulselengths. In particular, there is ~4X increase in fluence for equivalent scatter for damage at 2ω, 10 ns as compared to 0.30 ns in unconditioned KDP. The results for the unconditioned and conditioned KDP show that for all the pulselengths the scatter due to the bulk damage is a strong function of the damaging fluence (θ~5). It is determined that the 2ω fluence pulselength-scaling for equivalent bulk damage scatter in unconditioned KDP varies as τ0.30±0.11 and in 3ω, 3ns ramp-conditioned KDP varies as τ0.27±0.14. The effectiveness of 2ω and 3ω laser conditioning at pulselengths in the range of 0.30-23 ns for damage induced 2ω, 3 ns is analyzed in terms of scatter. For the protocols tested (i.e. peak conditioning irradiance, etc.), the 3ω, 300 ps conditioning to a peak fluence of 3 J/cm2 had the best performance under 2ω, 3 ns testing. The general trend in the performance of the conditioning protocols was shorter wavelength and shorter pulselength appear to produce better conditioning for testing at 2ω, 3 ns.
We have demonstrated a simple experimental technique that can be used to measure the nonlinear absorption coefficients in glasses. We determine BK7, UG1, and UG11 glasses to have linear absorption coefficients of 0.0217 ± 10% cm-1, 1.7 ± 10% cm-1, and 0.82 ± 10% cm-1, respectively, two-photon absorption cross-sections of 0.025 ± 20% cm/GW, 0.035 ± 20% cm/GW, and 0.047 ± 20% cm/GW, respectively, excited-state absorption cross-sections of 8.0 x 10-18 ± 20% cm2, 2.8 x 10-16 ± 20% cm2, and 5 x 10-17 ± 20% cm2, respectively, and solarization coefficients of 8.5 x 10-20 ± 20% cm2, 2.5 x 10-18 ± 20% cm2, and 1.3 x 10-19 ± 20% cm2, respectively. For our application, nonlinear effects in 10-cm of BK7 are small (≤ 2%) for 355-nm fluences < 0.2 J/cm2 for flat-top pulses. However, nonlinear effects are noticeable for 355-nm fluences at 0.8 J/cm2. In particular, we determine a 20% increase in the instantaneous absorption from linear, a solarization rate of 4% per 100 shots, and a 10% temporal droop introduced in the pulse, for 355-nm flat-top pulses at a fluence of 0.8 J/cm2. For 0.5-cm of UG1 absorbing glass the non-linear absorption has a similar effect as that from 10-cm of BK7 on the pulse shape; however, the effects in UG11 are much smaller.
We have developed techniques using small-beam raster scanning to laser-condition fused silica optics to increase their damage threshold. Further, we showed that CO2 lasers could be used to mitigate and stabilize damage sites while still on the order of a few tens of microns in size, thereby greatly increasing the lifetime of an optic. We recently activated the Phoenix pre-production facility to condition and mitigate optics as
large as 43 cm x 43 cm. Several full-scale optics have been processed in Phoenix. The optics were first photographed using a damage mapping system to identify scratches, digs, or other potential sites for initiation of laser damage. We then condition the optic, raster scanning with the excimer laser. The first scan is performed at a low fluence. A damage map is then acquired and any new damage sites or any sites that have grown in size are mitigated using the CO2 laser. The process is repeated at successively higher fluences until a factor of 1.7 above the nominal operating fluence is reached. After conditioning, optics were tested in a large beam 3ω laser and showed no damage at fluences of 8 J/cm2 average.
Using the Phoenix pre-production conditioning facility we have shown that raster scanning of 3ω optics using a XeF excimer laser and mitigation of the resultant damage sites with a CO2 laser can enhance their optical damage resistance. Several large-scale (43 cm x 43 cm) optics have been processed in this facility. A production facility capable of processing several large optics a week has been designed based on our experience in the pre-production facility. The facility will be equipped with UV conditioning lasers -- 351-nm XeF excimer lasers operating at 100 Hz and 23 ns. The facility will also include a CO2 laser for damage mitigation, an optics stage for raster scanning large-scale optics, a damage mapping system (DMS) that images large-scale optics and can detect damage sites or precursors as small as ≈15 μm, and two microscopes to image damage sites with ≈5 μm resolution. The optics will be handled in a class 100 clean room, within the facility that will be maintained at class 1000.
In this paper we present the results of bulk damage experiments done on Type-II DKDP triple harmonic generator crystals that were raster conditioned with 351 - 355 nm wavelengths and pulse durations of 4 and 23.2 ns. In the first phase of experiments 20 different scan protocols were rastered into a sample of rapid growth DKDP. The sample was then rastered at damage-causing fluences to determine the three most effective protocols. These three protocols were scanned into a 15-cm sample of conventional-growth DKDP and then exposed to single shots of a 1-cm beam from LLNL's Optical Sciences Laser at fluences ranging from 0.5 - 1.5X of the 10% damage probability fluence and nominal pulse durations of 0.1, 0.3, 0.8, 3.2, 7.0 and 20 ns. The experiment showed that pulse durations in the 1 - 3 ns range were much more effective at conditioning than pulses in the 16.3 ns range and that the multiple pass "peak fluence" scan was more effective than the single pass "leading edge" scan for 23.2 ns XeF scans.