AXAF is an x-ray observatory designed to study x-ray emission from al categories of astronomical objects, from normal stars to quasars. AXAF has broad scientific objectives and outstanding capability to provide high resolution images, spectrometric imaging and high resolution dispersive spectroscopy over the energy bandwidth from 0.1 to 10-keV. This is a significant year in the development of AXAF, to be launched in late 1998. Major elements of the observatory, the optics and the scientific instruments, are now nearing completion in preparation for calibration later this year.

The Mission Support Team SAO Reflectivities studies laboratory was responsible for the verification of the coating performance specification during the coating of the eight AXAF flight optics. Prior to the start of the coating of the flight optics, it was necessary to verify the scaleup of the coating chamber parameters from the test chamber to the flight optic coating chamber as well as to verify repeatability of coating quality. Immediately prior to the coating of each flight optic, witness samples were coated to verify the coating specification for each particular geometry. Similar witness samples were coated to verify the coating specification for each particular geometry. Similar witness samples were also coated with each flight optic. An overview of this monitoring program is presented along with a description of how the measurements are made, what tests are used to verify performance and a description of the witness sample deployment. Preliminary data on coating uniformity will also be presented.

The iridium coating of the AXAF flight optics by Optical Coating Laboratories, Inc. (OCLI) has taken place over the past eight months. During the coating effort, it was necessary to verify the performance of the coatings continually as deposition parameters were changed to accommodate the different geometry of each optic. In addition, quick look reflectivity measurements have been completed for witness samples coated during each run. The minimum reflectivity to be met was 82 percent at 20 arcminutes grazing angle and 50 percent at 34 arcminutes grazing angle. THis can be compared with a theoretical reflectivity, based on the Henke coefficients, of 88 percent and 62 percent at 20 and 34 arcmin respectively. Previous measurements by P. Slane, et al., gave mean values of 84 percent at 20 arcmin and 51 percent at 34 arcmin grazing angle. Measurements reported here were carried out by the SAO reflectivity laboratory, which is part of the AXAF mission support team. The quick look results of both the qualification samples and the production samples are presented along with the test results from the scaleup and verification runs that were used to set the coating parameters.

In-house optical performance modeling is being done at Marshall Space FLight Center (MSFC) in support of the development, fabrication, and testing of the Advanced X-ray Astrophysics Facility-Imaging (AXAF-I). Specialized software for modeling grazing-incidence x-ray telescopes is being developed by MSFC with assistance from the University of Alabama in Huntsville. Interfaces between the optical model and the mirror surface metrology and the predictions of structural/thermal deformation models have been developed. The mirror surface measurements were provided by the mirror manufacturer. A structural model of the AXAF-I High Resolution mirror Assembly has been developed at MSFC. The AXAF-I system, optical modeling software, and the image effects of specific mirror figure errors are discussed. In particular, the optimization of the alignment of the as- built mirrors and prediction of the x-ray image effects of the x-ray ground test configuration are described.

Measurements of the transmission properties of the AXAF CCD imaging spectrometer (ACIS) UV/optical blocking filters were performed at the National Synchrotron Light Source at Brookhaven Laboratories. The X-ray transmissions of two Al:Si/LEXAN/Al:Si three layer filters were measured between 260 and 3000 eV. The main purpose of the calibration was to determine a model transmission function with an accuracy of better than 1 percent. We present results from fits of model transmission functions to the measured x-ray transmission data. Detailed fine energy scans above the Al-K and C-K absorption edges revealed the presence of fine oscillations of the x-ray transmission. These features are most likely extended x-ray absorption fine structures (EXAFS). The amplitude of the EXAFS oscillations above the Al absorption edge is about 5 percent of the mean value of the x-ray transmission. EXAFS theory predicts a temperature dependence on the amplitude of the EXAFS oscillations. This dependence arises from the fact that thermal vibrations of the atoms in a solid produce a phase mismatch of the backscattered electron wave function. Since the ACIS filters will be at a much lower temperature on orbit we provide a prediction of the EXAFS component for the expected on orbit temperature of the ACIS filters.

The joint European x-ray telescope (JET-X) is one of the core scientific instruments of the RUssian SPECTRUM X-(gamma) astrophysics mission. JET-X is designed to study the emission from x-ray sources in the band of 0.3-10 keV; in particular to meet primary scientific goals in cosmology and extragalactic astronomy. JET-X consists of two identical, coaligned x-ray telescopes, each with a spatial resolution of better than 30 arcsec half energy width. Focal plane imaging is provided by cooled x-ray sensitive CCD detectors which combine high spatial resolution with good spectral resolution, including coverage of the iron line complex around 7 keV at a resolution of (Delta) E/E approximately 1.5 percent. Each telescope is composed of a nested array of 12 mirror shells with an aperture of 300 mm and focal length of 3500 mm; the total effective area is 330 cm² at 1.5 keV and 145 cm² at 8.1 keV. The mirror shells have a Wolter I geometry and are manufactured by an electroforming replica process. The paper presents the characteristic of the flight model x-ray optics.

W/Si and Co/C multilayers have been deposited on epoxy- replicated Au mirrors from the ASTRO-E telescope project, SPectrum Roentgen Gamma (SRG) flight mirrors, DURAN glass substrates and Si witness wafers. A characterization of the multilayers with both hard x-rays and soft x-rays is presented. The roughness value obtained from the Si wafers and the DURAN glass are in the range 3.0-4.2 angstrom and 4.4-4.6 angstrom, respectively. For the epoxy-replicated Au mirrors roughnesses of 5.0-5.8 angstrom are achieved, while the roughness of the SRG flight mirrors are in the range of 8.5-11.0 angstrom. This clearly indicates the effectiveness of the epoxy-replication process for the production of smooth substrates for multilayer deposition to be used in future x-ray telescopes.

The objective crystal spectrometer (OXS) on the forthcoming Spectrum-Roentgen-Gamma satellite is designed to carry three kinds of crystals: LiF(220), Si(111) and RAP(001), placed in front of the SODART telescope. Thirty six super polished (RMS roughness < 0.1nm) Si(111) substrates were coated with 65-80 periods of Co/C multilayers using electron beam evaporation deposition combined with ion polishing for the metal layers. These crystals are to be used in the energy band immediately below the C-K absorption edge of 0.284 keV. Because the crystals are to be assembled as one crystal on the OXS, the reflectivity performance as a function of energy and angle of incidence of all crystals has been measured using line radiation from an x-ray tube which provides 1.487 keV and 0.277 keV and using synchrotron radiation from 0.16 keV to 0.28 keV at the Synchrotron Ultraviolet Radiation electron storage ring a t the National Institute of Standards and Technology. The results from these measurements are discussed.

Wide field x-ray telescopes based upon the 'lobster eye' geometry can play an important role in future astronomical investigations. Our approach is based upon the geometry of Schmidt, 1975 in which a telescope is composed of two similar orthogonal stacks, each made from equally spaced identical flats that reflect on both faces. Its advantages are compatibility with large aperture and large field of view. Construction of a mirror reduces to two tasks, (1) mass producing thin, smooth, two sided flats and (2) securing them along two edges by a method that preserves their flatness. We are investigating both tasks in parallel. To address the first problem we are attempting to improve the flatness and reflectivity of an inexpensive, thin, commercial glass by replicating a smoother, flatter, gold coated surface on both faces. To address the second task, a small, 1D mirror assembly covering a 5 degree field of view was constructed from inexpensive, thin commercial glass that was coated on two sides with nickel. X-ray testing indicated that its resolution was consistent with theory and the approximately two arcminute deviation from flatness of the glass.

The effects of topographic finish errors on the quality of the image formed by a simple focusing mirror and the performance of different laboratory roughness-measuring instruments can be described in terms of a common formalism involving statistical operators or 'estimators'. In effect, conventional finish specifications and the test measurements used to ensure compliance with them, are stated in terms of the average values of these estimators. Int his paper we calculate correction factors to be applied to such average- based specifications and test measurements to account for the fluctuation effects inherent in the statistical nature of the processes considered.

In a previous paper, we described the experimental set-up and requirements used to study an XMM mandrel by x-ray angle-resolved scattering (ARS). We presented first results and compared them to micro-profilometry data. Here we complete the description of the experimental method and the data analysis of the x-ray ARS studies. We point out several pitfalls and propose solutions to avoid them. We emphasize the need to span a wide intensity dynamical range and the importance to separate the intensities form the specular and the scattered beams. This separation is of particular interest for estimating the rms-roughness from the power spectral density, modeled by a power-law of the spatial frequency. We then compare the results for the roughness with those obtained from profiler measurements. In a second part, the figure measurements of the XMM mandrel are described and analyzed in detail. They have been carried out with both an x-ray pencil beam and an optical long trace profiler. In particular, much attention has been given to the determination of the angle between the two sections of the Wolter I optical configuration and to the effect of the mandrel mounting supports. The PSD was completed with the low-frequency results. Finally, the surface data from the earlier experiment on the Ni-coated normally-polished paraboloid and the superpolished hyperboloid were used to predict the image quality of a Wolter I type optics having the same surface characteristics. The influence of different surface finishes on the image point spread function of a grazing incidence mirror is discussed.

The long trace profiler (LTP) is in use at a number of locations throughout the world for the measurement of the figure and mid-frequency roughness of x-ray mirrors. The standard configuration requires that the surface tested lie in a horizontal plane as the optical head is scanned along a horizontal line. For applications where gravity-induced sag of the surface cannot be tolerated, such as in x-ray telescope mirror metrology, it is desirable to measure the mirror as it is mounted in a vertical configuration. By making simple modifications to the standard LTP system, we have demonstrated that it is possible to use the LTP principle to measure the surface of x-ray mirrors and mandrels mounted in the vertical orientation. The major change in the LTP system is the use of a penta prism on a vertical translation stage to direct the probe beam onto the surface and the addition of a precision rotation stage to hold the test object. A 3D map of the surface topography of the complete cylindrical asphere can be generated quite easily with this technique. Measurements with a prototype system indicate a slope error accuracy of better than 1 microradian is possible, with a figure error repeatability of better than 50 nm.

The solar x-ray imager (SXI) telescope has a Wolter-I design with both grazing-incidence mirrors made from a single Zerodur substrate. The mirror fabrication requirements were for 40 percent encircled energy within 5-arc-seconds image diameter at an x-ray wavelength of 8.34 angstrom and for 56 percent, at 44.8 angstrom. The low-spatial-frequency mirror- surface error allocations (including on-orbit effects) were predicted to contribute (Using the ray-trace program GRAZTRACE) a geometric-image core of 3.34 arc seconds RMS diameter. The allocated mid- and high-frequency mirror surface errors had RMS values of 94 angstrom and 5 angstrom, respectively. The x-ray scattering program EEGRAZ was used to predict the resulting 5-arc-second diameter image encircled energies of 48 percent and 67 percent at 8.34 angstrom and 44.8 angstrom, respectively. These predictions exceeded the requirements. The characterization of the mirror as actually fabricated was accomplished using a ZYGO interferometer, a circularity test stand and a WYKO interferometer. Based on this data, the predicted geometric- image core RMS diameter was 4.03 arc seconds. The measured mid- and high-frequency surface errors had RMS values of 104 angstrom and 7.2 angstrom, respectively. Thus the allocated tolerances were not met. The resulting predicted encircled energies at 5-arc-second diameter were 32 and 52 percent at 8.34 angstrom and 44.8 angstrom, respectively. THese values fall short of the requirements for mirror fabrication by a small margin. The actual performance of the SXI mirror will be experimentally tested. The results will then be compared with the predictions.

We have developed an x-ray telescope that uses a new technique for focusing x-rays with grazing incidence optics. The telescope was built with spherical optics for all of its components, utilizing the high quality surfaces obtainable when polishing spherical (as opposed to aspherical) optics. We tested the prototype x-ray telescope in the 300 meter vacuum pipe at White Sands Missile Range, NM. The telescope features 2 degree graze angles with tungsten coatings, yielding a bandpass of 0.25-1.5 keV with a peak effective area of 0.8 cm² at 0.83 keV. Results from x-ray testing at energies of 0.25 keV and 0.93 keV (C-K and Cu-L) verify 0.5 arcsecond performance at 0.93 keV. Results from modeling the x-ray telescope's response to the SUn show that the current design would be capable of recording 10 half arcsecond images of a solar active region during a 300 second NASA sounding rocket flight.

The AXAF CCD imaging spectrometer (ACIS) consists of ten CCDs arranged in two arrays, one for imaging astronomical fields and one to be used in conjunction with transmission gratings for spectroscopy of astrophysical sources. ACIS uses Lexan/aluminum meshless films placed above these two CCD arrays to filter by mapping their soft x-ray transmission on fine spatial scales, so that the filter response can be removed from the CCD data and a more accurate estimate of the true sky recovered. We measured engineering and flight versions of these filters at the University of Wisconsin Synchrotron Radiation Center between June 1995 and April 1996. For all data, better than one percent accuracy in transmission as a function of energy was maintained over the entire filter area. The resulting transmission maps reveal spatial non-uniformities in the filters of about 0.5 percent to 2 percent. These transmission maps provide the finest spatial calibration ever achieved on such filters.

A novel soft x-ray grazing incidence spectrometer for studying laser produced plasma has been developed. The spectrometer is a complex system which is composed of three units: a grazing incidence grating spectrography, 4 dimensions micro adjustment equipment, and an aligning system with optical fiber bundle, electronic camera, TV monitor and laser transit. The spectrograph is used to resolve the lines of a soft x-ray spectra. The extreme ultraviolet spectra (time-integrated) region in the wavelength of 2-32nm. The resolving resolution of the spectrograph is 0.005nm.

Multilayer coatings have been applied to holographic diffraction gratings that have ion-etched blazed groove profiles. For each grating, the multilayer coating was designed to match the blaze angle so that the grating had high efficiency near normal incidence and in a wavelength region of choice in the 125-300 angstrom range. The grating efficiencies as functions of wavelength and incidence angle were measured using synchrotron radiation and were compared to the expected results. Normal-incidence efficiencies up to 13 percent were achieved at a wavelength of 128 angstrom. The performance characteristics of multilayer holographic and ruled replica gratings were compared and were related to the shape and roughness of the groove profiles as determined by atomic force microscopy.

A W/Si double multilayer monochromator has been installed on the ESRF BM29 (multipurpose XAFS) beamline with the specific task of suppressing higher order harmonics from the synchrotron x-rays of a bending magnet. This novel technique uses an identical pari of multilayers which were deposited by a method based on distributed electron cyclotron resonance plasma sputtering. In situ growth monitoring enabled the clear identification of a WSi chemical interface with an approximate width of 1nm: the relative size of the interface was found to severely limit the ability of the multilayer to completely reject specific harmonics. In total, the monochromator enabled suppression of all higher order harmonics to less than six orders of magnitude with an approximately 30 percent throughput in the first Bragg peak (at 8 keV). The use of such a multilayer monochromator for harmonic rejection in tandem with a crystal monochromator replaces the traditional method of deploying large grazing incidence mirrors; particular advantages for the multilayers are the significant reduction in size and cost.

The best results on compensation of stigmatism in optical schemes of grazing incidence monochromators-spectrographs, based on Rowland Circle, have been obtained by simultaneous optimization of correlation between curvature radii in meridional and sagittal planes of illumination mirror and diffraction grating. In this scheme the grating should be installed in divergent pencil of beams for meridional plane and in convergent pencil of beams for sagittal plane. This pencil of beams is formed by additional concave mirror, installed in front of the entrance slit of spectrometer. This grating has a concave shape in meridional plane and a convex shape in sagittal plane. It has straight grooves. This kind of grating may be grooved mechanically.

The two types of multilayer gratings used n x-ray domain, i.e. multilayer coated gratings and Bragg-Fresnel multilayer gratings are analyzed with the electromagnetic theory. The analysis starts from the differential formalism previously developed for a bare (single profile) grating. The method is generalized to an arbitrary thick modulated stratified medium thanks to the use of a suitable S-matrix propagation algorithm and gives stable and accurate numerical results. Thus, new diffracting devices which were out of reach of all existing theories can be analyzed, such as x-ray multilayer echelles. The method is then extended to study unperiodic focusing devices such as Bragg-Fresnel linear zone plates.

We have constructed an x-ray optical test bench which has been used to characterize transmission and manufacturing artifacts associated with monolithic polycapillary optics of the Kumakhov type. The test bench incorporates a conventional 8 keV x-ray source and a sample holder which is capable of high precision translation and rotation. Both non-imaging and imaging detectors are used, including a magnifying x-ray camera capable of resolving the transmission of individual capillary channels in a polycapillary bundle. We demonstrate the use of the bench to determine imperfections in the fabrication of early prototype capillary optics. The images illustrate the anomalies associated with the forming of the capillary bundles, the effects of extended exposure to synchrotron radiation, and the x-ray transmission lateral uniformity. We also show the intense focused beams obtained with complete optics.

A 1D Bragg-Fresnel lens for monochromatizing and focusing hard x-rays has been developed using a multilayer zone plate and a crystal. In the present study, Ag and Al have been chosen as material for opaque and transparent layers, respectively. About 300 layers of Ag and Al were alternately deposited on a plane substrate according to the Fresnel's formula using a helicon plasma sputtering technique, in which Ar gas pressure was less than 1 mTorr. The total thickness of the layers is about 100 micrometers. The multilayered plane was sliced vertically and glued onto Ge(211) crystal, and thinned to about 15 micrometers. By combining the dispersive characteristics of the crystal with the focusing capability of the multilayer zone plate, a new type of Bragg-Fresnel lens has been developed. Experimental tests have been performed at BL-17A of Photon Factory. The projected image of the synchrotron light source has been recorded on nuclear plates.

Capillary optics have been suggested some time ago to confine x-ray beams to the size of a few microns. First experiments using glass capillaries of various sizes and shapes have been reported recently. We discuss the design and fabrication of capillary optics at CHESS, the Cornell High Energy Synchrotron Source, for applications such as Laue diffraction, x-ray fluorescence, and small angle scattering. Early results in the production of the glass optics led to the construction of a glass pulling device to ensure a controlled fabrication process. Large efforts have been made to carefully evaluate the quality of these x-ray optics. MEtrology tests show good agreement between the design figure (e.g. linear or parabolic taper) and the actual profile, and indicate a high reproducibility in the production. X-ray tests have been carried out at CHESS employing white synchrotron radiation and monochromatic beams at different energies. Using a variety of glass materials it was discovered that surface roughness is often limiting the performance of these glass optics. Details on a microscopy analysis of the reflecting inner surface from capillaries are presented. In a survey of several glasses, soda-lime material was found to have the smoothest surface structure.

Soft x-ray optics based on slumped, profiled microchannel plates (MCPs) are under development. Their purpose is to collimate the expanding beam from a laser-plasma x-ray source, resulting in a parallel beam suitable for lithography of semiconductor devices. We present a prototype design for such an optic with a radius of curvature optimized for maximum beam intensity. The plate thickness (and hence the channel length) is varied as a function of distance from the plate center. The resulting plate profile is determined by the radius of curvature but not by the x- ray energy. This design is shown by Monte Carlo ray-trace modeling to give a circular illuminated field with a diameter of 36mm for 1 keV x-rays. The calculated intensity of the beam is equivalent to that of the divergence beam at a distance of 0.35m from the source. We also present preliminary x-ray measurements from the optic. These optics, while designed with the semiconductor industry in mind, may find application in any field where a uniform or parallel soft x-ray or VUV beam is required.

The design for a light-weight x-ray astronomical telescope that uses a diffractive optical element is described and simulated. The design provides tunable narrow-band imaging with a spatial resolution of several arc seconds over wide fields of view, greater than 10 arc minutes. The x-ray energy range from 0.56 keV to 1.86 keV, containing astrophysically interesting lines of O, Ne, Fe, Mg and Si, would be covered with high throughput through the use of a blazed diffractive optic. State-of-the-art grating technology is adequate for the optic design and a variety of x-ray imaging detectors could be used for the readout.

We present an approach to obtaining a high throughput hard x-ray telescope based on combining present replication technologies with multilayer supermirror coating. Fabrication of Pt/C supermirror using specially designed sputtering system has been tried and properties of x-ray reflection such as 1-st and 2-nd order bragg reflection peak profile, energy dependence of x-ray reflection, interfacial roughness were measured. Optimization of coating para meters such as 2D, total pair number, thickness ratio of Pt and C were investigated from points of view of theoretical simulation and fabrication technique in order to get maximum effective area and field of view of telescope system. Based on currently optimized parameters, we designed telescope system with 45 cm outer diameter, 2-stage 20 cm mirror length and focal length of 8 meters. It was found that expected effective area of four such telescopes is 320 cm² for the x-ray energy < 40 keV. Application of this type of telescope of balloon experiment to reveal hard x-ray image of clusters of galaxies or supernova remnant is also planned.

The peculiarities of the space absolute radiometric measurements of the solar radiation flux from full disk in the range from 0.14 to 135 nm are discussed in details. The experience of similar radiometer functioning at the Soviet satellite 'Cosmos-381' in 1970/71 is used. This radiometer together with two grating spectrometers of grazing and oblique incidence will form the complex of apparatus for the permanent space patrol of the solar EUV and x-ray radiation.

Multilayer carbon interference structures (MCIS) obtained by interleaving hydrogenated carbon layers were studied. MCIS with period from 115 A down to 20 A were obtained by ion- plasma methods. MCIS parameters in the wavelength range 1.54 Q-44.7 A were studied using x-ray diffraction methods. With the help of a two-crystal spectrometer the MCIS rocking curves for the first Bragg peak were measured. The relations R equals f((Theta) ) were calculated for various wavelengths. The differences between experimental and theoretical results are discussed. It is shown that the use of MCIS leads to the same high resolution as that of KAP crystals. The parameters were calculated for the MCIS that are expected to have a reflection coefficient of up to 70 percent at (lambda) equals 1.54 A and a resolution several times greater than that of the traditional Me/C-structures.

One of the specifications used to polish the AXAF witness samples was that the rms surface roughness be <EQ 5 angstrom as measured by optical profilometry. This specification was chosen based on the cost of polishing and the necessity to keep scatter to a minimum. However, it is not necessarily the best indication of the expected performance of the soft x-ray reflectivity of the surfaces. In particular, the reflectivity data from the AXAF flight optic witness samples indicate sample to sample differences of a few percent which do not correlate with the optical profilometry results for these samples. Further investigations were carried out to measure rms surface roughness using atomic force microscopy (AFM). The differences shown by AFM surface roughness measurements correlates to differences found in reflectivity for these same samples. One-dimensional power spectral density data is presented from both AFM and WYKO measurements along with the reflectivity results at 8 keV for the AXAF witness samples. The results indicate that to obtain accurate prediction of x-ray performance it is necessary to look at the scanning probe metrology data provided by the AFM, in addition to the optical profilometry data.

A spectrometric facility for measurements of reflection, absorption and excitation spectra of films and massive samples in 50 to 300 nm wavelength region with the spectral resolution up to 10³ is designed and tested. Plasma produced EUV radiation from the laser beam focused onto a gaseous or a solid target makes it possible to achieve the spectral intensity on a sample up to 3(DOT)10⁶ photons/s(DOT)cm^-1. Application to measurement of reflection spectra of some phthalocynines is reported here.

X-ray beams delivered by third generation synchrotron radiation machines have a high degree of coherence that allows to use different coherent imaging techniques such as phase contrast microscopy, holography, and phase-contrast tomography in the high-energy domain. The question arises to what extent the existing optical elements such as mirrors and multilayers are capable of preserving this high spatial coherence. A theory of partially coherent x-ray scattering by a slightly modulated mirror surface under total reflection is developed. This gives estimations for intensity contrast as a function of the surface modulation. X-ray topography with coherent light is proposed and applied for mirror diagnostics. X-ray interferometry in an in-line holography setup is used to describe the coherence properties of the x-ray beam including optical components like mirror and multilayers. It is shown that significant improvements of the polishing process are still needed for the long mirrors while small state-of-the-art mirrors (less than 100 nm) are nearly adequate.

High performance x-ray Wolter I telescopes require especially at high energies excellent specular mirror shell reflectivity features to collect grazing incidence radiation. Applying galvanoplastic techniques the surface quality of complementary mandrels can be replicated on the mirror shells with negligible degradation only. Therefore, adequate surface finish by advanced superpolishing of the mandrels is a strong prerequisite in order to improve the specularly reflected portion of the incoming x-ray light. For that reason, a new surface technology was developed within the frame of the JET-X mirror assembly project providing reduced high frequency components of the surface roughness correlated with the reduction of mandrel surface scattering (diffuse reflection). At high x-ray energies (8.1 keV) the measured encircled energy image of mirror shells replicated from these mandrels was up to 3 times better than from untreated mandrels.

For the past six years, a high-accuracy reflectance calibration system has been under development at the National Synchrotron Light Source at Brookhaven National Laboratory. The system utilizes Los Alamos National Laboratory's Beamlines X8A and X8C. Its purpose is to calibrate the reflection efficiencies of witness coupons associated with the coating of the eight mirror elements composing the High Resolution Mirror Assembly for NASA's Advanced X-ray Astrophysics Facility (AXAF). During the past year, measurements of reflectances of numerous iridium- coated witness flat mirrors have been obtained to a relative statistical precision of 0.4 percent, and an overall repeatability within 0.8 percent in the overlapping energy regions. The coating processes are strikingly repeatable, with reflectances in the 5-10 keV range for off-end witness flats nearly always being within 1 percent of one another, excluding interference fringes. The comparison reflectances between flats obtained from qualification coating runs and production runs of the Wolter Type I mirror elements are in turn nearly equal, indicating that the qualification run witness flats provide a good representation of the flight optics. Results will produce a calibration of AXAF with extremely good energy detail over the 2-12 keV range, which includes details of the M-absorption edge region for Ir. Development of the program to cover 0.05-2 keV continues.

For the best flexibility in ground and on-orbit calibration modeling of the AXAF telescope over its entire field of view, including off-axis calibration evaluations, AXAF synchrotron reflectance calibrations require that the measured reflectance data be reduced to optical parameters analogous to n and k. We have developed a method for AXAF witness mirror analysis which is a modification of the NKFIT optical constants algorithm published by D.L. Windt. The algorithm assumes uniform layer thicknesses using a recursive, exact formation of Fresnel's equations, with a modified Debye-Waller roughness correction factor. The recursion formula has been modified to include an explicit double-precision formulation. The results of most of the fits of AXAF calibration measurements yield residuals less than 1 percent of the reflectance value levels down to R approximately .03. The precision of the measurements is smaller still, which compromises the (chi) ² fitting algorithm; however, the results will most likely prove adequate for AXAF witness mirrors calibrated in the 5-12 keV range. Coating density determined from the refractive index n is approximately 98.5 percent of the bulk for iridium. Derived coating thicknesses are extremely consistent with the photon energy, giving still more significant calibration information to the program.

Precise transmission measurements of free standing iridium foils and of iridium coatings supported on thin polyimide film have been made at the X8 beamline of the National Synchrotron Light Source, at energies from 2 to 12 keV. These measurements were conducted to provide iridium optical constants in support of NASA's Advanced X-ray Astrophysics Facility (AXAF). Transmission data were collected at small energy increments across iridium M and L absorption edges to study detailed edge structures. From these data, the imaginary components of the index of refraction were computed. The data also allows computation of the real parts, using the Kramers-Kronig dispersion integral. Preliminary results indicate a measurement accuracy of better than one percent for transmission. Absorption coefficients deviate by varying amounts from values predicted from Henke data table, which is consistent with the accuracy claimed for those table.s Nonuniform thickness in our iridium foils may be a source of errors in our analysis and improved foils are being sought. Additional measurements will also be needed with foils of different thickness to account and correct for possible effects of sample thickness on the determination of absorption edge fine structure. To our knowledge, transmission measurements of this degree of accuracy and precision have not been previously reported in the literature for iridium in the 2- 12 keV energy range.

In soft x-ray spectroscopy the region from 2 to 6 keV is the most critical, because crystal properties are not so good as at higher energies and optical gratings become inefficient. We have investigated the performances of ion etched blazes gratings as components for soft x-ray spectroscopy. The gratings were mounted in a Rowland geometry at angels of incidence varying from 89 degrees to 89.45 degrees. As source, we used a microfocus soft x-ray source and as detector an uncoated channel electron multiplier (CEM) sliding along the Rowland circle to scan the spectrum. We recorded several spectra between 50 eV and 6 keV and among them the L lines of Zr, Mo, Pd and the K lines of Ti, Cr, Fe. The spectral resolution obtained shows essentially almost slit limited line widths. The diffraction efficiency ranges from 1 to 3 percent at 89.35 degrees and up to 4.2 percent at 89 degrees. This experiment proves that it is possible to extend the classical spectroscopic techniques to the soft x-ray region.