Figure 8. Magnitude of threshold voltage shift vs. fluence for various ratios of channel width to length, where the gates are poly-2, Ids = 0.1 μA, and Vds = 1 V. Manuscript Received A Multi-Facet XUV Aluminum Mirror for the FEL We have investigated Marion L. Scott Materials Science & Technology Division Los Alamos National Laboratory a new Los Alamos, NM 87545 12-12-88 concept in retro-reflectors for use in the extreme ultraviolet, namely, UHV aluminum coated, multi-facet, grazing-incidence mirrors. Our results indicate that this type of mirror, which utilizes total-external-reflectance, works very well in the wavelength range from 35 nm to 100 nm (89 ± 3% measured retroreflectance at 58.4 nm for a 9-facet mirror). However, the coated mirror surfaces must not be allowed to oxidize after deposition, which implies that the retro-reflector must be coated and used in situ or the oxide layer must be removed in an ultra-high vacuum (UHV) system. Key Words: XUV; extreme ultraviolet; aluminum; reflector; multi-facet. INTRODUCTION thin The search for good reflectors in the extreme ultraviolet (XUV) spectrum has gone on for many years. Some samples of silicon carbide and diamond have been shown to exceed 40% normal incidence reflectance at wavelengths greater than 60 nm. The silicon carbide results have not proven to be easily reproduced with thin film depositions. Diamond film technology has greatly improved in recent years, however, XUV reflectance measurements on these films have not yet come up to bulk values. Near-normal incidence reflectance measurements on multilayer coatings have reached 60% at 17 nm [1] but reproducibility has been a problem and multilayers at longer wavelengths have not been very successful. Also, multilayer reflectors will only reflect over a narrow spectral band. Our present work on multi-facet reflectors was motivated by the resonator mirror requirements of an XUV free-electron-laser to be constructed at Los Alamos National Laboratory [2]. The resonator mirrors are required to equal or exceed 40% retroreflectance over the wavelength range from 10 nm to 100 nm. It is possible to construct the XUV FEL in segments so that one resonator does not have to span the entire wavelength range, however, the wavelength tunability of the FEL necessitates that the resonator mirrors be as broad-band as possible. The total-external-reflectance of UHV aluminum at 58.4 nm measured in our previous work (98.66% ± 2% for a single bounce at 80 degrees) indicated that a 9-facet reflector with this coating would yield a retro-reflectance between 74% and 100% at this wavelength [3]. MULTI-FACET MIRROR A 9-facet mirror structure was designed on the Los Alamos computer-assisted-design system and fabricated with the aid of a numerically-controlled mill (see Fig. 1). The nine silicon mirror substrates were fabricated by Laser Optics to a figure accuracy of 1/20 wave in the visible and 0.5 nm rms surface roughness. The assembled 9-facet mirror aligned with a visible HeNe laser is shown in Fig. 2. Note that the visible beam accurately strikes the center of each of the nine facets before exiting the array parallel to the input beam. This 9-facet structure is supported in our UHV deposition and analysis system by UHV alignment manipulators attached to each side of the three strut frame. UHV DEPOSITION SYSTEM Our UHV deposition system (illustrated in Fig. 3) has been previously described and only a cursory description will be given here. The UHV chamber has a base pressure of ·10 5x10 Torr and is pumped with hydrocarbon free pumps to avoid carbon contamination of the coated reflector. A water-cooled e-gun source located in the lower part of the deposition chamber is shielded with a cylinder to avoid wide angle coating in the chamber. Some additional shielding inside the 9-facet array requires lateral movement of the array between the position for deposition and the position used in later reflectance measurements. The deposited coating thickness is monitored with a quartz crystal monitor connected to an IC 6000 rate controller. Figure 1. The 9-facet XUV mirror structure fabricated and measured at Los Alamos. Figure 2. The 9-facet XUV mirror shown with a retroreflected HeNe alignment laser. Figure 3. UHV deposition and analysis system utilized in demonstrating the 9-facet XUV reflector. UHV ALUMINUM DEPOSITION onto A thin film (70 nm) of high purity aluminum is deposited through appropriate masking three of the nine silicon mirror substrates simultaneously, before rotating the structure to coat the remaining substrates in groups of three. We have achieved good results with a deposition rate of approximately 0.1 nm/sec although we have not performed measurements to determine whether this rate is optimum. After coating deposition on all nine mirror substrates, the mirror assembly is moved into position and aligned to make the XUV reflectance measurement at 58.4 nm. IN SITU XUV REFLECTANCE MEASUREMENT Many modifications were made to our UHV system to adapt the in situ XUV reflectometer (previously used to measure single-surface reflectance) to measure the reflectance of a 9facet retroreflector. The XUV beam entrance to the UHV chamber had to be lowered and the sample holder had to be raised. The detector was mounted on a linear feedthrough in the top of the chamber and is capable of rotation about the axis of this magnetically coupled feedthrough. The high voltage and signal leads to the imaging microchannel plate detector are also in the top of the chamber for easy removal. The mirror array can be accurately aligned for reflectance measurement by use of the Huntington x,y,z rotatable feedthroughs on either side of the array. The measurement of the reflectance of this 9-facet mirror at 58.4 nm was accomplished by first positioning the imaging detector in front of the XUV beam entrance to the UHV chamber and carefully noting the x, y position of the beam on the detector and recording the count rate of the source. Secondly, the detector is raised and rotated to intercept the XUV beam exiting the properly aligned, 9-facet mirror. The detector is positioned to place the beam image at the same spot on the detector as in the previous measurement and the count rate in this detector position is recorded. on the A correction must be made to the raw count rates recorded in this manner because the detector responds to counts that are outside of the well defined beam image seen oscilloscope. These out-of-beam counts arise from scattering in the system, as well as dark counts from the detector. This correction is determined for each detector position by electronically gating out those counts which are outside the beam image to determine the ratio of beam to scattered counts. The reflectance of the 9-facet array is then computed from the ratio of the corrected counts recorded from the beam exiting the array to those entering the array. The result of averaging several such measurements on the 9-facet mirror array was 89 ± This value is within the uncertainty of our previous single surface reflectance measurement on UHV aluminum at 80 degrees incidence angle raised to the ninth power (74% to 100%). 38. CONCLUSIONS We conclude that an unoxidixed, UHV aluminum, 9-facet reflector is an excellent XUV retroreflector at 58.4 nm. We also conclude that an XUV mirror constructed in this fashion will provide an adequate reflectance over a significant portion of the XUV spectrum (35-100 nm) to operate an XUV free-electron laser resonator. ACKNOWLEDGEMENTS The author would like to acknowledge the support of this work by the DOE Office of Basic Energy Sciences, Advanced Energy Projects Division. REFERENCES 1. 2. 3. T. W. Barbee, Jr., S. Mrowka, and M. Hettrick, "Molybdenum Silicon Multilayer Mirrors for the Extreme Ultraviolet, Appl. Opt., 24, 883-886 (1985). "1 B. E. Newnam, "Multifacet, Metal Mirror Design for Extreme Ultraviolet and Soft X-Ray M. L. Scott, P. N. Arendt, B. J. Cameron, J. M. Saber, and B. E. Newnam, "Extreme |