COMMENTS

Question:

Answer:

Question:

Answer:

Question:

Answer:

Question:

Answer:

Question:

Answer:

Can you say anything about the absorption coefficients of these hot sites that
you saw?

Well, I can tell you that the thermal signatures vary by factors of 10, 20, 30
above the surrounding area. We have no way of directly measuring the response
of those sites using, say calorimetry. So, when we do calorimetry, we get an
average response over the whole surface of the sample and there's no way for me
to give you an absolute number. But the thermal signature does vary by factors
of 10, 20, 30 and that seems to be a good indication that there are substantial
differences between areas on those surfaces.

Did you look at the large anomalies you said were related to pits, did you look
very carefully at the pits, were they contaminants or are you pretty sure of what
the nature was?

I can tell you that the larger sites all had an identical appearance, they all
looked like, to me, larger versions of the micropits that we see in single shot
damage. In terms of going back inside and looking for specific contaminants or
something like that, no we didn't do any analysis. The similarity to the
microscopic pit formation was really quite remarkable and it didn't hit me until
I had 60 or 70 of these things sitting on the light table that this seemed to
be a very consistent morphology.

How did you determine the emissivity values that you gave. Did You look those
up or did you figure them out?

When you're doing thermal measurements, you have to be careful of the bandwidth
of your instrument that you are using for measurements and we tried to be as
self-consistent as possible. So we measured the emissivity values for these
samples comparing them directly at several different temperatures to reference
surface with an emissivity that we assumed to be one or very close to it.
that was how we determined those.

And

Have you thought about the possibility that some of these small anomalies may
not be absorption anomalies, but rather anomalies in the spectral emissivity of
the region.
Because the thermal camera can't tell the difference if there's a
change in the spectral emissivity, it will represent that as a temperature
change.

That is quite correct. If the coating had an emissivity of 0.8 then a higher
temperature would appear if the emissivity on a localized spot were 0.1. That
is not a significant change between 0.8 and 0.1 in terms of the temperatures that
we would measure. It does not correlate in magnitude to the changes that we are
seeing on these sites. Now for the enhanced metal reflectors, it is possible
that could play a role, but again, I would expect to be able to see a difference
in the emissivity just visually using the microscope and maybe that is being
naive, but there is a big difference between a surface emissivity of 1 versus
0.8, at least in visual light most of the time.

It would also seem that local debonding or local variation in the thermal contact
resistance between the coating and substrate could give a hot spot. Do you have
any way of checking for that, for example, visually to look for elimination?
Yes, I would agree entirely with that analysis, I would say that these coatings
that we used for the mapping experiments were commercial reflectors and they have
their fair share of cosmetic defects. Very frequently the surface of the sample
was masked completely with observable cosmetic defects that look like stains in
the coating or by normal pits right where we wanted to look. Within the
resolution of our transport between the various systems, that is the mapping
stage and the microscope and the scatter stage, there was nothing to be seen.
Under a Nomarski microscope I would normally expect to be able to see a debonded
area or something like that. I can't say for sure, and it is worth checking out.

Laser damage threshold improvement was observed in selected thin film coatings that were repeatedly irradiated by 1.06μm, 16 nsec pulse emitted by a laser with 30 Hz repetition rate. Each sample was first irradiated for one minute at separate sites (1 on 1 testing) to determine the laser damage threshold of the thin film. A separate site was multiply irradiated (n on 1 testing) using a pulse fluence that initially was arbitrarily low, then subsequently raised in a step-wise fashion until damage occurred.

Results indicate a significant increase in laser damage threshold of some thin film coatings due to this treatment.

Introduction

High damage threshold high reflective (HR) optical coatings for Lawrence Livermore's next generation laser is one of the research efforts being pursued by the Advanced Laser Development Group at Lawrence Livermore National Laboratory. These goals include: 99.7% minimum reflection, 40 J/cm2 damage threshold at 1.06μ and 10 nsec, and at 20% the cost of the present HR coatings presently used in Livermore's Nova laser.

Previous studies have indicated that optical films can be altered by subthreshold laser irradiation. M.E. Frink, et al. performed post deposition laser treatment of antireflection (AR) coatings of unknown composition. They concluded that the increased damage levels measured as a result of this treatment were temporary. J.E. Swain, et al.2 studied post coating laser irradiation of neutral solution AR coatings and concluded that a damage threshold level improvement of 2.5 times that of the pretreated coating could be realized. James Rowe reported that laser irradiation during evaporation deposition of selected coatings produced films with absorption one half that of the same coatings deposited without co-irradiation.

Alteration of bulk properties of optical materials through multiple shot irradiation has also been demonstrated. Multiple irradiation of new crystals of potassium dihydrogen phosphate (KDP) increased its bulk damage threshold by about a factor of two4. 5

Shimizu, et al. reported an innovative experiment where a disseminated absorbing species (silicon) was removed from a silica matrix via irradiation with a 308 nm and 575 nm laser. This effort reports the results of some initial experiments of a post deposition laser treatment process on selected reflective (R) and HR coatings.

Samples

Six R or HR coatings deposited on 5 cm diameter high purity silica substrates were tested. Two were electron beam (e-beam) dielectric HR coatings consisting of a 15 layer quarter wave design with a half wave silica overcoat. The compositions were hafnia/silica and zirconia/silica, and the primary reflection was tuned to 1.06μm. Four coatings, three sol-gel, and one made by evaporation, were produced by the authors. The sol-gel coatings were a single layer zirconia, a four layer alumina/silica and an eight layer alumina/silica, all having the primary reflection at 1.06 μm. The evaporated coating was a single layer lead fluoride of undetermined thickness.

Experiment

The 1 on 1 damage threshold of each virgin coating was first determined. One on 1 damage testing is defined as a single location being irradiated by 1.06μm, 16 nsec, 30 Hz repetition rate laser for one minute at a single fluence. The n' on 1 threshold for each coating was then measured. N on 1 damage testing consisted of a single location being irradiated as per 1 on 1 tests; however, after a one minute exposure to sub-damage threshold intensity radiation the fluence was increased by 1-3 J/cm2, and the procedure repeated at the same location. The results are graphically illustrated (Figs. 2 thru 7). This step wise procedure was repeated with microscopic inspection between fluence increases until at least 5% of the irradiated area was damaged. Each irradiated area was visually inspected with a 100X Nomarski microscope. The area damaged was determined by measuring the size of the largest damage sites, totaling the number of sites, and reading total area percent of damage versus fluence of these samples (Fig. 1). Arrows extending above the data points in Fig. 2 thru 7 indicate catastrophic damage covering at least 50% of the irradiated area.

Results and Discussion

On all e-beam samples, n on 1 treatment resulted in a sizable improvement in laser damage threshold. The improvement was most dramatic with the ZrO2 HR sample. The HfO2 HR sample had as high a damage threshold as the ZrO2 but the increase from 1 on 1 to n on 1 was less. In both cases, shown in Figs. 2 and 3, damage had not occurred at the laser's maximum fluence, of 50 J/cm2.

The single layer ZrO2 sol-gel was unaffected by the n on 1 treatment, and the damage thresholds were not impressive. Subsequent to these tests, it was discovered that the ZrO2 solgel precursors were contaminated with iron, which probably is the reason for the low damage threshold of this sample compared to the ZrO2 e-beam HR coatings.

The A1203 sol-gel coatings were also improved by the n on 1 treatment. This improvement was not quantified because the n on 1 damage threshold was above the 50 J/cm2 maximum fluence of the testing laser.

The PbF2 damage thresholds were identical and well defined. This result may indicate an absorption in PbF2 that is intrinsic.

Results of these initial experiments do not allow determination of the mechanism that caused n on 1 and 1 on 1 thresholds to differ for some, but not all, of the coatings tested. If repeated irradiation increases the threshold, the irradiation must either reduce absorption of laser light or increase the mechanical toughness of the sample. Since sol-gel samples, which are mechanically weak, have thresholds comparable to those of e-beam deposited films, it is unlikely that thresholds are strongly affected by mechanical strength. Therefore, we believe absorption was reduced.

If we assume that damage results from laser heating and rapid volatilization of isolated absorbing volumes, removal of these discrete impurities would increase the threshold. Slowly evaporating the absorbers by repeated, low level irradiation might allow their removal without the explosive evaporation that occurs when the absorbers are exposed to a single intense pulse. Coatings that were porous and contained only a few, widely separated absorbers would be most likely to exhibit threshold improvement. Dense coatings, or those with uniformly distributed absorption, would probably be severely damaged even during slow removal of absorbing material. This model would allow the improvement by laser treatment of some coatings and lack of improvement in others, in keeping with our observations.

*Work performed under the auspices of the U. S. Department of Energy by the Lawrence Livermore National Laboratory under contract number W-7405-ENG-48.