NSLS II Metrology R D Activities Peter Z Takacs

NSLS II Metrology R&D Activities Peter Z. Takacs Experimental Facilities Advisory Committee Review 20 Oct 2006 1 BROOKHAVEN SCIENCE

Rationale for Metrology R&D for NSLS II Every new advance in SR source design has driven improvements in optical components: Pre-NSLS (<1980) SR mirror slope error quality: >2 arc sec (10µrad) Original NSLS mirror specs: 1 arc sec (5µrad) for <10µm spot size. – This was difficult for manufacturers to achieve NSLS upgrade: 1 µrad for <1µm spot size. – This is now routine NSLS II requirements are now 100 nrad (!) “If you can’t measure it, I can make it. ” - Norm Brown, LLNL, 1980 s BNL developed the metrology to force manufacturers to improve their fabrication processes: surface roughness, then slope error We need to do it again. 2 BROOKHAVEN SCIENCE

Where are mirrors in NSLS II beamlines? X-ray scattering/crystallography - KB mirrors, bendable Small-angle x-ray scattering (SAXS) - KB primary pair and KB secondary pair, both bendable Scanning transmission x-ray microscope (STXM) - spherical grating monochromator (SGM) and steering mirrors for beam line branches. High resolution inelastic x-ray scattering (IXS) beam line - spherical collimating mirror after pre-mono and KB pair after the high-res mono. Superconducting wiggler - vertical focusing mirror for 50 -100 ke. V photons. Soft x-ray beam lines - collimating and focusing optics Gratings, spheres, cylinders, paraboloids, ellipsoids, etc. Add “real world” slope error to soft xray beamline KB mirror surface: 3 BROOKHAVEN SCIENCE

Challenges in NSLS II mirror development Need to develop reliable source(s) of nm-quality mirror components. • • Work with vendors to insure required mirror parameters are met. Provide metrology feedback Need to develop in-house metrology instrumentation and techniques adequate for nm figure and 100 nrad slope errors. Plan for mirror metrology R&D -- near-term and longer-term tasks: 1. Develop Next-Generation Long Trace Profiler for reliable 100 nrad measurements. 2. Develop stitching interferometry system for high-resolution figure over complete 2 D surface area of mirror. 3. Evaluate new polishing techniques Ø QED Magneto. Rheological Finishing (MRF) 4. Develop in-situ LTP for beam line diagnostics 5. At-wavelength testing capability 4 BROOKHAVEN SCIENCE

1. Next Generation LTP Present LTP III limited by systematic errors at the 1 -2 µrad level. Need to improve internal optical components and air bearing stage. BNL LTP III measuring Si cylinder • Glass quality affects measurement accuracy – Replace commercial PBS with custom PBS - $5 K to $20 K estimates • Replace Al beam (100µrad err) with ceramic beam (<5µrad err) • New linear motor drive system • 2 D camera Explore high resolution LTP options for spatial periods <1 mm 5 BROOKHAVEN SCIENCE

NG-LTP: Resources required Zygo Wavelength-shifting PMI required for internal glass quality measurement. • New technique allows separation of front and back surface from interior New. View Micro-PMI (or equivalent) required for surface roughness control of internal LTP components. • Essential for replacement of defunct Micro. Map profiler (vintage 1985) • Also use for profilometry of mirrors and quantitative topography of nanostructures, e. g. refractive kinoform optics Also requires software development to add 2 D camera and speed data acquisition. Collaboration with LBL • Software development • Quantity discount in custom optics procurement 6 BROOKHAVEN SCIENCE

2. Stitching metrology development Subaperture stitching interferometry (SSI) necessary for 2 D surface map • Required for deterministic surface info at ~50µm spatial periods QED has the SSI metrology - companion to MRF machine. QED-developed algorithm solves for test surface error AND reference optics errors => self-calibrating, < 2 nm residual errors. Combine high resolution Fizeau PMI with LTP optical head. Estimate 3 years to develop operational stitching system. QED SSI uses conventional Zygo Fizeau interferometer head combined with 6 -axis positioning manipulator. View from interferometer in stitching of Osaka elliptical cylinder Measured and predicted 40 nm image shape from SSI on 100 mm long elliptical cylinder. Yumoto, et al. , RSI 76, 063708 (2005) 7 BROOKHAVEN SCIENCE

3. Polishing R&D Lessons from Osaka: • • EEM technology 20 years in development EEM can produce nm-level figure accuracy Requires novel metrology techniques Not (yet) available commercially. Elastic Emission Machining Y. Mori, Osaka BNL will NOT go into fabrication business. • We must rely on commercial optical fabricators Need to explore new polishing technologies to achieve 100 nrad optics Need for in-house metrology instrumentation • Replacement needed ASAP for defunct Micro. Map (former NCP-1000) for surface roughness measurement High speed rotating tool Magnetorheological finishing (MRF) is most promising new technique for SR optics • Developed by QED Technologies, Rochester 8 BROOKHAVEN SCIENCE

3. MRF polishing evaluation Need to demonstrate Angstrom-level surface finish capability of MRF process. • Basic process limitations? uses diamond grit • Need to fine-tune machine parameters: slurry chemistry, dwell time Establish collaboration with QED 1. Produce super-polished Si flat 2. Produce KB elliptical cylinders 3. Develop SSI metrology for non-rot symmetry parts We will evaluate surface quality in lab and performance in NSLS and/or APS beam line. • Use New. View surface profiler and NG-LTP. • Requires stitching metrology software development for rectangular substrate shape. Successful results => transfer technolgy to MRFcapable vendor. • Zeiss has expressed interest in meeting our needs – Heavily invested in MRF machines 9 BROOKHAVEN SCIENCE

Long-term metrology tasks 4. In-situ LTP • In-situ LTP needed for beam line diagnostics. – Measure thermal and mechanical distortion on high heat load optics. – Look for transient heating effects on rigid body alignment. – Locate beam footprint on optical surface for alignment check. • Need to design essential interface ports into mirror chambers. – View through window normal to surface – Use scanning penta prism inside chamber 5. At-wavelength metrology • Develop phase retrieval image evaluation system for x-ray wavelengths. – Similar to Souvorov technique at SPring 8 • Useful for evaluating wavefront quality of various microfocusing optics – Zone plates, refractive optics, Bragg-Fresnel • Potential collaboration with J. Fienup at Rochester – Postdoc will be available in ~2 yrs • Requires wavelength converter and camera hardware, software development • Test beam line would be useful for this and other at-wavelength methods. 10 BROOKHAVEN SCIENCE