COMMENTS ON PAPER BY PEDINOFF, BRAUNSTEIN, AND STAFSUDD The speaker pointed out that anisotropies up to 15% have been measured by this technique. DAMAGE RESISTANCE OF AR-COATED GERMANIUM SURFACES FOR NANOSECOND CO2 LASER PULSES* Brian E. Newnam and Dennis H. Gill An evaluation of the state-of-the-art of AR coatings on gallium-doped germanium, used as a saturable absorber at 10.6 um, has been conducted. Both 1-on-1 and N-on-1 laser damage thresholds were measured with 1.2 ns pulses on bare and coated surfaces. Only front surface damage was observed. With few exceptions, the thresholds for coated surfaces were centered at 0.49 ± 0.3 J/cm2. Bare Ge had a threshold ranging from 0.65 to 0.70 J/cm2. No significant differences due to substrate polish, crystallinity or doping level were evident, and multiple-shot conditioning resulted in the same threshold as for single shot tests. From an analysis of standing-wave electric fields, damage of AR-coated Ge appeared to be limited by the surface properties of Ge. Measurements at both 1.2 and 70 ns indicated that the threshold (J/cm2) of both coated and uncoated Ge increases as the square root of the pulse width. Key words: Antireflection coatings, germanium, laser damage, saturable absorber, standingwave electric field. 1. Introduction Gallium-doped germanium has been developed for use as a saturable absorber to prevent pre-pulse gain depletion in the large CO2 amplifiers of LASL's eight-beam fusion laser [1,2]. For use at saturating intensities for pulses 1-nanosecond in duration, the damage resistance of the AR-coated surfaces must be maximized. Accordingly, a careful evaluation of the state-of-the-art of AR coatings at 10.6 um was performed. Antireflection coatings comprising fourteen coating designs using eight film materials were obtained from nine coating manufacturers. Polycrystalline, p-doped Ge substrates polished by one vendor were supplied to each. Substrates polished by a second vendor were also supplied for comparison. Additionally, single-crystal Ge, p-doped and undoped, and undoped polycrystalline Ge were coated by one vendor to evaluate the effect of crystal structure and doping. The dimensions of the test substrates were 25 mm in diameter and 6 mm thick. Coating depositions, however, were performed in chambers large enough to eventually coat amplifier-size Ge discs (41-cm diameter and 4-cm thickness) with sufficient uniformity to obtain a reflectance per surface of less than 1% at 10.6 um and less than 3% from 9 to 11 μπ. 2. Experimental Procedure Laser damage tests were conducted with 1.15 ± 0.05 ns pulses (FWHM) at the P(20) 10.6 μm wavelength. These short pulses were reliably carved out of a smoothed gain-switched pulse by use of a Pockel cell arrangement. The schematic of the laser diagnostics is shown in figure 1. Pulsewidth measurements were made with a Molectron pyroelectric detector coupled to a 5-GHz bandwidth oscilloscope of LASL design [3]. For supplementary tests with a 70-ns pulsewidth, a photon-drag detector was used. Oscillograms of the temporal pulses are shown in figure 2. The test samples were located prior to the focus of a 1 m F.L. ZnSe lens where the beam spot-size radius was 1.1 mm. The peak value of the irradiance (J/cm2) at the sample plane was measured on each shot by use of a 197-μm diameter pinhole (Optimation, Inc.). The pinhole was located in a split-off beam and placed at the same distance from an identical ZnSe lens as was the sample. The energy transmitted by the pinhole at the center of the reference laser beam was measured by a Laser Precision Energy Meter (isolated from rf noise). Prior to each test series a calibration was performed with an identical 197-μm pinhole centered at the sample plane. During the damage tests the reproducibility of the spatial profile was monitored by comparing the energy focussed through the pinhole reference with the total energy measured by a Scientech calorimeter. The 197-um diameter of the pinhole was chosen to minimize the spatial averaging over the beam profile (at the pinhole perimeter the intensity of the Gaussian profile dropped to 98% of the peak value) while transmitting an adequate amount of energy for easy detection. Also, the ratio of the pinhole diameter to the wavelength was large enough to avoid significant diffraction effects. By use of the equation, 1 Figures in brackets indicate the literature references at the end of this paper. (1) where a = 99.98%. aperture radius [4], the transmittance of the 197 um-diameter aperture was calculated to be Damage was detected visually by the onset of increased surface scattering of a He-Ne laser beam directed on the same site (back and front) as the pulsed laser and by examination under bright whitelight illumination using the He-Ne beam as the locator. After the tests, examination of irradiated areas was also performed with a microscope (300X). 3. Damage Morphology The characteristics of the laser-damaged sites viewed under 200 and 300 X magnification were very interesting. Only front surface damage was observed in these tests. In figure 3, disruption of an AR-coated (Ge/PbF2/ZnSe/Air) surface of p-doped, polycrystalline Ge, caused by a single 1-ns pulse above threshold, is examined. The AR coating has been removed rather uniformly leaving a well-defined perimeter. Linear interference ripples oriented normal to the laser polarization are grouped around circular damage pits. Temple and Soileau have identified these ripples as perturbations in the surface topography due to interference of the incident laser electric field with the time-varying (laser frequency), induced surface charges on surface scratches, voids and inclusions [5]. The diameter of the pits are mostly 8 to 12 um and the ripple spacing is approximately 8.5 um which are close to the laser wavelength. Damage sites in uncoated Ge (not shown) caused by 1-ns pulses did have faint ripples with spacing exactly equal to the laser wavelength (± 0.2 nm). The morphology of damage caused by 70-ns pulses was very different from the above as shown in figure 4. On the coated surface, a random distribution of irregular sites was related to damage at defect sites, and extensive cracking of the AR coating is probably thermally-caused delamination. For bare Ge (fig. 5) the damage sites were all centered on circular pits accompanied by very tightly-spaced (3.5 μm) interference fringes parallel to the laser polarization. The cause of these fringes has not been identified. 4. Results The experimental results for coated and bare Ge substrates are presented in tables 1-3. These thresholds are for pit formation or film disruption which occurred at much lower intensities than a breakdown plasma. Only the mean value of each threshold is listed for the coated surfaces since the range, typically ± 0.02 or less, was unusually small. The absolute accuracy is considered to be ± 10%. All thresholds listed are for front surface damage only as we were unable to damage any rear surface, coated or uncoated. Further, we observed no difference between thresholds for 1-on-1 and N-on-1 tests, where N-1 shots (10 to 15) were fired below the single-shot threshold, before irradiating with a damaging intensity. To compare the effects of two different conventional polishing methods, single- and polycrystalline substrate material and Ga-doping level (undoped, R = 30 N.cm; doped, 3 N.cm), one coating vendor deposited a two-layer ThF4/ZnS Vee-coat on each different substrate during one run. As seen in table 1, no significant differences in thresholds caused by the two polishing methods were manifest. This was surprising since Polish A qualified as better than "40-20" (scratch and dig code) and Polish B was slightly worse than "40-20". Likewise, no real differences were measured between coated single-crystal and polycrystal Ge surfaces. Gallium-doping had no effect on coated single-crystal thresholds, and only a minor 10-15% threshold reduction was measured for Ga-doped polycrystalline Ge. The thresholds for fourteen coating designs on p-doped, polycrystalline Ge are listed in table 2. Multiple entries represent different samples of the same coating. The values ranged from 0.41 to 0.57 J/cm2 and the mean value was 0.49 ± 0.03 J/cm2. Even the two-layer coating of CaF2/ZnSe had the same threshold despite the fact that CaF2 has a large absorption coefficient at 10.6 μm. Due to the relative uniformity of thresholds for coated Ge surfaces, particular attention was paid to the thresholds of uncoated Ge presented in table 3. The values for three different bare surfaces were all greater than those of coated Ge by about 40%. In addition, it is noted that Ga-doping lowered the threshold by 10%. N The experimental results indicate that the damage threshold of AR-coated Ge surfaces is 1) independent of the design and materials of the AR coatings and is 2) lower than uncoated Ge. Furthermore, damage occurred only at the front surface. These results may be explained by considering the electric fields in the Ge, coated and uncoated. Figure 6 represents the standing-wave electric fields, normalized to the incident field E in the vicinity of the front surface of an AR-coated and uncoated Ge substrate. Although, the exact field distribution within the AR coating must be calculated for each design, the gradual decrease of E/E2 from 1.0 at the air-film interface to 0.25 at the film-Ge interface is the same for any design. |