Figure 4. Ag/A1203 on Si Linear Polarized Light. E-Vector Perpendicular to plane of incidence. Measured reflectance of a protected silver reflector at various angles. 80 MANUSCRIPT NOT RECEIVED RADIATION EFFECTS IN MIRROR SUBSTRATE MATERIALS E.J. Friebele, J.A. Ruller, and P.L. Higby ABSTRACT Large mirror substrates in surveillance and directed energy applications will be exposed to nuclear radiation from the natural space environment and hostile activity. Ionizing radiation is known to alter the physical properties of ceramics. For example, changes in the density and/or coefficient of thermal expansion result in surface deformation and a resultant loss of optical figure. In contrast to the radiation-induced compaction measured in silica and Zerodur at high dose (>106 rads), there has been a recent report of expansion in silica exposed at lower doses. We report here the effects of radiation on the density of various silicas, Sic, ULE, and Zerodur over a wide range of dose (104 - 109 rads). The changes in density have been correlated with the concentration of electronic defect centers such as the oxygen vacancy E'center and the non-bridging oxygen hole center, leading to the development of structural models for the densification process. In addition, the changes in the thermal expansion behavior of these materials have been measured with a high-resolution Michelson interferometer. Results of the study indicate the possibility of significant surface deformation and loss of optical figure. Manuscript Received 3-29-89 1. Laser Polished Fused Silica Surfaces: Absorption Data Alan F. Stewart and Arthur H. Guenther* Air Force Weapons Laboratory Simultaneous laser polishing of both faces of fused silica windows has produced exceptionally uniform high quality surfaces. Thin wafers of fused silica were laser polished and the absorption measured at 351 nm using a laser calorimeter. The data shows virtually no change in the measured absorption even with the highest power levels used in the polishing process. This data can be related to the conventional polishing process used to prepare these samples before laser polishing. Key words: absorption; contamination; fused silica; laser polishing Introduction The process of laser polishing has been proposed as an effective method for finishing of fused silica surfaces. [1-5] Several authors have reported that laser polished surfaces exhibit higher laser damage thresholds. [1-3] It has also been determined that these surfaces can exhibit extremely low scatter levels comparable to those produced by the most advanced superpolishing techniques. [5] In an earlier study, we have observed high residual stresses in laser polished surfaces and unusual chemical bonding of the silica which have been modelled by others. [4,5] These changes have been related to the extremely high temperatures of the surface during the polishing process. It is widely assumed that the intense heating heals or fuses microcracks in the surface and/or simply burns off residual contamination which control laser damage thresholds and scatter levels. We report here the results of a test series intended to demonstrate the correlation of laser polishing with the removal of adsorbed contamination. Specifically, absorption at 351 nm was measured on a series of laser polished substrates. Our data does not show any correlation between laser polishing parameters and the measured absorption of these parts. We believe this null result is due, in part, to residual contamination from the pre-laser irradiation polishing process which is or becomes chemically bonded to the surface. 2. Experiment Laser polishing of fused silica wafers was performed using a double-sided technique. Earlier data had suggested that thermal cycling of laser polished surfaces might lead to a degradation of the surface and low laser damage thresholds. [5] Simultaneous polishing of both surfaces of a substrate eliminated this effect as a possible complication. The optical bench in the laser laboratory was set up for double sided polishing as shown in figure 1. The high power CW CO2 laser used in this study, the beam diameter (6-7 cm), and beam dwell time (6.5 sec) were the same as described in reference [5]. The arrangement of optics in figure 1 split the laser output into two opposing beams of equal power which were focussed to the same diameter on the sample surfaces. The substrates used in this study were commercially available fused silica wafers 2.54 cm in diameter and 0.025 cm thick. These substrates were the standards used for laser calorimetry where low mass is of extreme importance. Each substrate was held in a split Current addresses are: (A.F.S.) Battelle Pacific Northwest Laboratories, P.0. Box 999, Richland, WA 99352 (A.H.G.) Los Alamos National Laboratory, MS A110, Los Alamos, NM 87545 copper ring which loosely contacted the part on the faces near the rim. Because of the temperatures involved in the laser polishing process, the substrate needed room for expansion and thus was not constrained. Laser polishing of the parts proceeded at power levels ranging from 0 to 40 Watts per square centimeter. As in the earlier studies, an infrared camera system was used to monitor the surface temperature of the parts during the polishing process. temperatures recorded varied in a nearly linear fashion to 2900 degrees Centigrade at 40 These values were considerably higher than the linear dynamic range (1500 degrees) of the camera system. As such, incident power levels were recorded as they were considered to be more reliable and reproducible. Following irradiation, the polished substrates were measured for total absorption at 351 nm using a laser calorimeter. The instrument, which has been described extensively in reference [6], has a baseline resolution of approximately 10 ppm. No additional cleaning process was performed prior to measurement. Laser calorimetry data of absorption for the laser polished fused silica at 351 nm is shown in fig. 2. Given the resolution of the instrument, this data indicated there was essentially no change in the absorption of the parts with laser power employed in the polishing. These results were quite surprising because residual hydrocarbon contamination was suspected to be a major component of surface absorption at 351 nm. The intense surface temperatures generated in the laser polishing process should have desorbed or oxidized all surface contamination. Laser polished surfaces should be among the cleanest of optical surfaces. This surface would definitely remain clean if the ion polishing were performed under UHV conditions as in reference [7]. Based on scattering data, we expected that fractures in the surface left from the conventional polishing process would have "healed" driving out impurities and reducing the number of dangling surface bonds. With no change. in absorption observed, either hydrocarbon or other contamination managed to reestablish themselves as resident on the surface between polishing and measurement, or some other agent extrinsic or intrinsic was active. In fig. 3, contamination data on a zirconia film deposited on a fused silica wafer from the same manufacturer. This data was obtained using a quantitative technique called Secondary Neutral Mass Spectroscopy (SNMS) [8]. An anomaly appears in this data as well that we believe correlates with our data from the laser polishing experiments. As the SNMS instrument sputtered through the film after 2300 seconds, peaks for aluminum and titanium rose to nearly the 1% level of total counts. These contaminants could not be traced to any aspect of the coating process such as the crucible material used or previously deposited species in the chamber or the cleaning process used prior to coating. Tracking backwards to locate the source of this contamination, the manufacturer provided us with a sample of their grinding compound used to fabricate the fused silica substrates. As shown in fig. 4, an energy dispersive x ray scan shows that the compound is 95% alumina with approximately 5% titania. Clearly from fig. 3, conventional cleaning could not remove this contamination layer. As alumina and titania are refractory materials, they would not decompose during laser polishing. It is suspected that this layer is chemically bonded to the surface and could not be removed except by an ion or chemical etching process. The presence of a bonded contamination layer of refractory materials would reduce the utility of laser polishing as an effective surface cleaning process. No measurable changes in absorption would be expected either. Further studies will be required to isolate this variable in the fabrication, cleaning, and polishing process. 4. Conclusion Double sided laser polishing of fused silica windows has been performed. In absorption measurements performed on windows polished at different incident power levels, no significant changes were observed. Contamination data obtained on identical substrates |