Comparison of laser and C 0 2 snow cleaning of astronomical mirror samples

The cleaning of mirrors in large, remote telescopes is essential if these telescopes are to meet their performance goals, including low (2%) thermal emissivity in the infrared. Aluminum-coated mirror samples, which were naturally or artificially contaminated with materials representative of those at observatories, were cleaned with a UV laser beam or CO2 snow. Cleaning effectiveness was determined from studies of residual particle densities and size distributions (measured from low-magnification optical imaging). For exposures under two weeks both laser and CO2 cleaning yield comparable results; however, for longer exposures (up to three months) UV laser cleaning is about twice as effective in our tests. At the laser energy densities required for effective cleaning, no surface changes or damage was observed even after cleaning the same spot 200 times. For 8-m class telescopes, the annual consumption cost of sufficiently pure CO2 is comparable to the capital cost of a UV laser. Both methods clean such surfaces in less than one day. Two attractive features of laser cleaning are that the method can be fully automated and run frequently without significant cost of manpower or expendables, and that by focusing the laser beam tighter it can be used to strip old surface coatings prior to recoating.