14–17 October 2008
Edited by Josiane Costeraste, Errico Armandillo, and Nikos Karafolas
The European Space Agency (ESA) plans to launch several LIDAR systems in space within the next five years (ADM-Aeolus, EarthCare, Bepi-Colombo). The stated aims of these missions are detailed earth atmosphere observations for improved climatology and meteorology predictions as well as planetary surveying. The suited LIDAR systems comprise high energy pulsed lasers in the IR and UV spectral range. Under the vacuum conditions of space especially the outgassing of organic materials is inevitable, in spite of using only materials with low outgassing rates (total mass loss (TML) < 1%, collected volatile condensable material (CVCM) < 0.1%). The interaction of intense laser radiation with volatized material’s constituents gives rise to a high risk of deposit formation on the optics and consequently to a drastically reduced lifetime.
Space conditions were simulated using an oilfree pumped vacuum chamber with a base pressure better than 10-6 mbar. Coated and uncoated fused silica windows (diameter 1”, thickness: 3mm) served as optical test specimens. The total amount of the contamination material was in the range of 0.3 – 0.5 g, coated on 5 x 5 cm2 aluminium sheets. Heating ribbons are mounted on the outside of the chamber for indirect heating of the contamination samples. Before each new run the chamber was baked out at 165°C for 24 hours to prevent cross-contamination between consecutive tests. Prior to each new material test blank tests without any contamination sample were performed, using the same temperature, pressure and laser parameters. In case of detecting any fluorescence or depositions on the windows, the bake-out was repeated until the blank test was successful. In case of deposit formation under UV irradiation, the organic compounds in the deposit are excited to fluorescence. The fluorescent deposits were recorded in-situ and online by a high sensitive electron multiplying CCD (EM-CCD) with a Peltier-cooled sensor (-20°C) and two pre-amplifiers (Andor, Luca DL-6588M-TIL). The signal to noise ratio is very high. The quantum efficiency reaches 50% in the 400 nm - 700 nm wavelength range. After each test the surface structure and thickness of the deposits was analyzed by white light interference microscopy. By comparison of the surface contour maps with the EM-CCD fluorescence images a very good correlation between deposit thickness and fluorescence intensity was found . The total amount of deposit in dependence of laser exposure time was estimated in the following way from the fluorescence pictures: the gray scale values in a circular area at the beam position were integrated and recorded online. The diameter of the integration circle was chosen five times larger than the beam diameter.
DYNAMIC OF DEPOSIT FORMATION
The evaluation of in-situ recorded fluorescence images offers the possibility to gain insight into the dynamic of deposit growth. Fig. 2a shows a Nomarski micrograph of a typical deposit grown on an uncoated fused silica optic after a test with a polyurethane based contaminant. The irradiation parameters were the following: laser fluence: 0.4 J/cm2, wavelength: 355 nm, pulse width: 3ns, repetition rate: 100 Hz, shots: 10 Mio. The beam profile of the laser is Gaussian. The contaminant temperature was 40°C. Fig. 2b shows the linescan of the surface profile of the deposit. The crater or doughnut like structure is very typical for laser induced deposits at high fluences. Fig 3. shows a series of fluorescence pictures, taken at the onset of a deposition process. By comparison of the surface profile of the deposit after the test with the corresponding fluorescence picture a calibration of the fluorescence intensity to deposit thickness was performed. At the beginning a steady growth of the deposit is observed with a pancake-like structure. Then a competing process of deposit removal starts at the center of the deposit, i.e. the region of highest laser intensity. After this the removal of material dominates and the “doughnut” structure develops with an increasing central hole. The removal process is probably caused by increasing absorption and thus increasing temperature with rising deposit thickness. Potentially the absorption is further enhanced by material decomposition and formation of carbon rich residues. We found in our tests that the transition from “pancake” to “doughnut” growth occurred earlier when applying higher laser fluences. In case of very low fluences (< 0.1 J/cm2) “pancake” growth dominates for most contaminant materials. The size of the central hole in the doughnut like deposits also increases with increasing laser fluence (Fig.4). From a comparison of deposit surface profiles with the corresponding beam profiles (1/e2 beam diameter 360 µm) we estimated the threshold value above which no deposit occurs in the center of the beam to approximately 0.12 J/cm2 for a test series with a silicone based glue.
For an unambiguous comparison of different contaminants in terms of deposit formation activity it is favourable to investigate pure pancake growth without the competing removal process. As already mentioned, this can be realized in case of a pulsed laser by irradiating with low fluence. An alternative is to use a cw laser diode. Fig. 5 shows the appropriate deposit growth dynamic calculated from the corresponding fluorescence pictures. The wavelength of the laser diode was 375 nm, the mean power 10 mW. The mean laser intensity was similar to the test with pulsed irradiation, shown in Fig. 3. Fig. 6 shows the integrated fluorescence gray scale values of the deposit as a measure for the total amount of deposited material. A nearly linear growth as a function of laser exposure time is observed.
INFLUENCE OF OPTICAL SURFACE PROPERTIES ON DEPOSIT FORMATION
The interaction of volatized molecules with the surface of the optical sample is one of the initial processes for deposit formation. We therefore investigated the influence of different surface properties (temperature, roughness and chemical composition) on this interaction. MgF2 and SiO2 are typical materials used in optical coatings. Fig. 7 shows the results of contamination tests with these compounds as top layers for AR coated samples. The tests were performed with peak fluences between 0.2 and 1.3 J/cm2. As contaminant a two part polyurethane based glue was used. The contaminant temperature was 40°C. The surface profiles of the deposits show, that the growth rate on MgF2 is clearly higher than on SiO2. Similar results were obtained with a silicone based contaminant.
To investigate the influence of surface roughness on deposit growth, tests with uncoated fused silica samples with mean roughness Ra = 0.1, 0.3, 0.5 and 1.5 nm were performed. There was no significant difference in deposit growth rate observable.
Fig. 8. shows the influence of surface temperature on the deposition process. The optical samples were uncoated fused silica windows, which were cooled resp. heated to 17, 30 and 50°C. The contaminant, an epoxy based glue was heated to 100°C. The deposit growth rate clearly increases with decreasing surface temperature. The adsorption of volatized atoms and molecules on surfaces is a temperature dependent process and seems to play an important role in the laser induced deposition process.
The chemical composition of the deposits was investigated by ToF-SIMS. Fig. 9 shows a distribution mapping of a deposit grown on a MgF2 coated fused silica optic. The contamination sample was a silicone based material heated to 40°C. The laser parameters were the following: wavelength 355nm, pulse length 3 ns, repetition rate 100 Hz, peak fluence 1.5 J/cm2, 107 shots. The distribution of the chemical components reproduce the crater-like structure found in the surface profile analysis. In the ring hydrocarbons are dominant (dark-blue). In the non-irradiated zone unaffected MgF2 coating was detected (green). At the outer edge of the central hole degraded and non-degraded silicone from the contamination sample is dominant (red). Obviously silicone can exist at higher fluences than hydrocarbons. A similar behaviour was found for a deposit produced in a test with IR-irradiation at 1064 nm.
INFLUENCE OF DEPOSITS ON OPTICAL PROPERTIES
The quality reduction of optical components through deposit formation was estimated by transmission measurements. Fig. 10 shows a comparison between an unaffected area and a 50 nm thick deposit on an uncoated fused silica sample. The transmission in the UV wavelength range is reduced by 8%. Even for a 15 nm thick deposit a transmission reduction of 5% was found.
Laser induced deposit formation was investigated under space conditions. Epoxy, silicone and polyurethane based contaminants were tested. Different growth modes were observed. At high fluences, a deposit removal in the center of the beam is evident, leading to characteristic crater-like structures. At low fluences a homogenous growth was found. The formation of depositions depends on the surface composition of the optical samples. For MgF2 coated optics the deposition growth rate is 2 - 3 times larger than on SiO2 coated ones. The deposition rate increases with decreasing surface temperature. The surface roughness has no significant influence on the deposition process. Surface analysis of the deposits showed that silicone constituents can exist at higher laser fluences than hydrocarbons.
The authors wish to thank F. Hadinger (DLR) for his support in performing the contamination tests and H. Kheyrandish (CSMA-MATS, UK) for surface analysing of the deposits.