Coherent Hard Xray CHX Beamline Update Andrei Fluerasu

Coherent Hard X-ray (CHX) Beamline Update Andrei Fluerasu Coherent Hard X-ray Scattering Group Experimental Facilities Division, NSLS-II Experimental Facilities Advisory Committee Meeting April 23 -24, 2009 1 BROOKHAVEN SCIENCE

CHX Team • BL scientists: AF, Lutz Wiegart (to join Summer 2009), Lonny Berman, Lin Yang • • Management and engineering support: Andrew Broadbent, Qun Shen, Konstantine Kaznatcheev, Mary Carlucci -Dayton Michael Loftus, Lewis Doom, Viswanath Ravindranath, Sushil Sharma Beamline Advisory Team (BAT) Robert Leheny, Associate Prof. , John Hopkins Univ. (spokesperson) Karl Ludwig, Professor, Boston University Laurence Lurio, Associate Professor, Northern Illinois University Simon Mochrie, Professor, Yale University Lois Pollack, Associate Professor, Cornell University Aymeric Robert, Instrument Scientist, LUSI/LCLS, SLAC Alec Sandy, Physicist, 8 -ID, APS, ANL Oleg Shpyrko, Assistant Professor, University of California San Diego Mark Sutton, Professor, Mc. Gill University 2 BROOKHAVEN SCIENCE

Outline • • • Scientific Mission and technical requirements Recommendations from recent reviews Beamline Layout • Overview • Source and front end undulator; requirements for filling modes • Optics Enclosure coherence preservation by mirrors and multilayers; focusing with Be CRLs; Pink beam operation; DCM – heat load • Experimental Station general layout of the experimental station • Summary and Outlook 3 BROOKHAVEN SCIENCE

CHX beamline: Technical Requirements and Scientific Mission Flexible instrument optimized XPCS in SAXS, GI-SAXS and WAXS geometries. Will also provide an excellent instrument for m-beam SAXS & CDI on “large” samples – e. g. cells Scientific opportunities for XPCS @ NSLS-II • Galssy materials; Driven and out-of-equilibrium systems • Nanostructured complex fluids: polymers, colloids • Biological systems: proteins in solution, biomembranes • Fluid surfaces and interfaces • Nanoscale dynamics in inorganic materials: WAXS • Molecular dynamics in metallic and orientational glasses • -beam SAXS • CDI on “large” (e. g. a cell) samples P. Falus, S. G. J. Mochrie et al. 4 BROOKHAVEN SCIENCE O. Shpyrko et al.

CHX: Main Design Objectives Avoid degrading the source brilliance • minimum number of windows; materials chosen carefully • Optics – polished to the best figure, manufactured from defect-free Si, Ge or Si. C to minimize the disturbance of the wavefronts XPCS is a signal-starved technique. Every coherent photon must make it to the sample • Vertical focusing Stability • minimize vibrations, heat load, etc 5 BROOKHAVEN SCIENCE

Recent Recommendations from EFAC and DOE reviews Comment Response investigate whether accommodation of a microbeam SAXS capability compromises the primary XPCS mission of the beamline. . . BL design focused on XPCS. However, with a slightly different tuning and sample environment the BL will also be excellent for m-beam SAXS and CDI investigate the possibility to accommodate limited CDI } take note of the structure of the electron beam, especially Ongoing interaction with for studies of the fastest time scales, and consider the ASD concerning filling need to normalize incident beam fluctuations mode, bunch structure. take note of the importance of “smart” detectors to ensure the success of the XPCS program at the fastest time scale On-going R&D program P. Siddons. A first prototype (100 x 100 pix) is in development devote early attention to developing mirror and ML optics Need for 100 nrad optics. specifications, and how to characterize them. . . R&D - coh. preservation. 6 BROOKHAVEN SCIENCE

st 1 BAT Meeting – Dec. '08 Major Recommendations: • • Main goal of CHX beamline - studies of dynamics by XPCS. Other techniques ( -beam SAXS, CDI) only if possible without compromising performance Detectors are the MOST important (and highest risk) part of the beamline. NSLS-II should advocate for detector development XPCS in a wide angle scattering geometry (WAXS) should be included in the initial scope Limit the power load by a first, high heat load aperture Be CRLs appear as promising for vertical focusing – more studies regarding coherence preservation will be required Role of mulitlayers as a “wider band gap monochromator” was questioned in a layout that includes also a mirror Investigate the option of cryogenically cooling the DCM (with reference to Petra-III which adopted this solution) BL scientist should be encouraged to remain active in the field and engage into collaborations with other researchers / other facilities 7 BROOKHAVEN SCIENCE !

Beamline Layout • • Relatively short source-sample distance and vertical focusing allow to use a full coherent mode e. g. 200 m (V) x 20 m (H) by focusing the beam to a 20 m (V) x 20 m (H) spot Large sample detector distance allows to have (relatively) large speckle sizes and resolve them with fast detectors which will, very likely, have relatively large pixels e. g. < 100 m. The WAXS instrument will use additional focusing (H and V) to increase speckle size and resolution. 8 BROOKHAVEN SCIENCE

Source and Front End Source properties • Low- straight → B=2 x 1021 [ph/(s·mrad ·mm · 0. 1% 2 2 bw)] • U 20 IVU (3 m); E=6 -15 ke. V (Ic= 2 B/4) 9 BROOKHAVEN SCIENCE

Source: requirements for filling mode and uniformity fast XPCS requires a quasi-DC source (uniform filling) Example of noise in g(2) form a 7/8+1 filling mode at ESRF g(2) = g(2)SR x g(2)sample ESRF 7/8+1 filling: train of 868 bunches (7/8 of the SR circ. ) filled with 200 m. A (0. 23 m. A / bunch). Both edges of the train are filled with 1 m. A single bunch. The remaining 1/8 gap is filled in its center with a cleaned 2 m. A single bunch. Refill time ~ 5 min. Need to perform simulations in order to determine how to perform XPCS with the baseline filling mode: ~1000 stored bunches; average current stability 1% over all the stored bunches; intensity variation between bunch that was stored for the longest time and the most recently filled – 20 %; single bunch train injection every ~ 1 minute; the injected pulse train “walks” around the bunch patters shifting the boundary between the oldest and the most recently injected bunches; each injection will result in a disturbance to the beam which will damp in 5 to 30 ms; (more DWs – faster damping time) Slow XPCS @ NSLS-II will benefit of the long lifetime (topup mode) 10 BROOKHAVEN SCIENCE

Optics Enclosure (“FOE”) • Primary slits ~100 m(H) x 500 m(V) P<8 W (160 W/mm 2) note: placed in Front End • White beam mirror OE • • • H deflecting stops bremsstrahlung in provides (some) high harmonic rejection heat sink: ~30 mm long footprint P ~ 0. 6 W/mm 2 Secondary slits Be Compound Refractive Lenses – vertical focusing Monochromator: Vertical Si(111) DCM; small offset; cryo-cooling (P/A > 20 • W/mm 2, P ~ 6 W) Pink Beam: H-deflecting 2 x Multilayer mono, H 2 O cooling, q=1 -2˚, L~ 5. 5 mm, 2 P < 6 W/mm White beam stop, pink beam shutter 11 BROOKHAVEN SCIENCE

Mirrors: slope error requirements h(x)=h Sin(p x/L) => h'(x)= h * p/ L Cos(p x/L) Slope error: PV=2 hp/L=1. 25 rad; RMS=hp/L*1/21/2=0. 88 rad Max. deflection angle: 2*2 hp/L =2. 5 rad Slope errors of 100 nrad will be required for Smallest speckle size to resolve: s= /D~1 Å / 30 m =3. 3 rad • coherence or high resolution applications @ NSLS-II ! Best figure error that manufacturers can guarantee today (3 -2 nm) is not optimal for XPCS • Some R&D effort is required in order to achieve the desired figure errors 12 BROOKHAVEN SCIENCE

Coherence Preservation by Mirrors & Multilayers • On-going R&D project aiming at evaluating / controlling the effect that mirrors and MLs have on the wavefront of the on coherent X-ray beam. E=11 ke. V (ID 06, ESRF), WSi 2/Si ML, 100 m B fiber, 1 st order reflection (reflectivity ~ 0. 7) A. Fluerasu, O. Chubar, R. Conley, L. Berman, A. Snigirev (ESRF), work in progress 13 BROOKHAVEN SCIENCE

At wavelength tests - further steps • Coherence preservation by multilayers • Theoretical work on phase retrieval from in-line holograms O. Chubar, A. Snigirev, A. Fluerasu et al. I. Robinson et al. Phys. Rev. B 52, 9917 (1992) (1 D speckle from ML) • Aim: retrieve the surface profile, power spectral density function Coherence preservation by mirrors, focusing elements • Perform tests on “test” samples • • • purchased from different vendors, polished by different methods Perform test with novel focusing optics: large acceptance Be lenses, linear Be lenses, coherence preservation, etc Perform more “realistic” tests – high heat load in white or pink beam, use the elements under test as “source” as opposed to as “samples”, etc Need for an NSLS-II R&D activity (all A. Madsen et al. , ID 10 ESRF BLs) & operating budget 14 BROOKHAVEN SCIENCE

Focusing with Be CRLs • Be compound refractive lenses offer the best and most reliable way to focus the beam for SAXS-XPCS they are in user operation at ID 10, ESRF and provide an efficient way of using the full vertical coherent beam without any noticeable loss in contrast - no focusing - V. focusing (CRL) ID 10, ESRF 15 BROOKHAVEN SCIENCE

Pink Beam: Double Multilayer Monochromator • Pink beam device Natural line width e. g. for 3 m long U 20 IVU most typically working on the 3 rd or 5 th harmonic DE/E~1/n. N<1/450 smaller than the typical bandwith of ML structures (~ 1%) • Multilayers: provide the only practical way to obtain an efficient pink beam operation with good harmonic rejection (e. g. < 10 -3 -10 -4) at medium energy SRs ! vs. 16 BROOKHAVEN SCIENCE

DCM - Heat Load Conclusions of FEA for the DCM: • Sharp “thermal bump” if water cooling is used due to small beam. Slope errors are ~ 20 rad irrespective of cooling geometry • Cryogenic cooling will be required to maintain the slope error < 0. 2 nrad • Vibrations issue will be addressed by the mechanical design (seek advice and inspiration from NSLS, 26 -ID- APS, Petra-III, …) V. Ravindranath, L. Berman 17 BROOKHAVEN SCIENCE

Beamline layout – Experimental station Local optics on granite block (or optical Detector WAXS -beam SAXS L=0. 5 -2 m, full 90 f Detector SAXS Detector stage Beam stop, etc table) • Vert. and Horiz. Focusing (KB, Z. plates. . . ) • Local mirror(s) for GI-SAXS on liquid surfaces • BPM (quadrant or 32 -ant – P. Siddons) Sample stage • exit window, slits, etc. • goniometer • on-axis microscope • guard slits, etc. 18 BROOKHAVEN SCIENCE

Need for detectors for XPCS • Development of a fast (0. 1 s) and “smart” (integrated correlators) pixel detector Pete Siddons, NSLS – exploratory study related to detectors that are of importance to NSLS-II, particularly an XPCS detector – First prototype in progress! • Other detectors of interest photon-ccounting det. eff~100% @ 8 ke. V Medipix detector: 1 k. Hz, 55 m pixel size Pilatus detector: 100 Hz, 172 m pixel size Pilatus-II detector: ~1 k. Hz, ~80 m pixel size • Array of APDs 'collection' of point detectors - > limitations in stud. non-ergodig systems - > available, possibility to implement with an array of hardware (e. g. FPGA) correlators 19 BROOKHAVEN SCIENCE

Summary and Outlook • • • Beamline conceptual design is on-track and advancing well towards its completion (09/2009) Beamline budget ($10. 04 M) seems adequate even though it was calculated for a beamline (CD-2) which bears little resemblance with the current layout Risk factor: availability of fast (1 MHz) photon-counting pixel detectors with small-enough pixel size. Mitigation: possible use of APD arrays (albeit 0 D and not 2 D detectors) On-going research program (not yet funded through an official R&D) aiming at obtaining mirrors and multilayers with required figure errors Future steps: • • • Finalize the comprehensive cost estimate Bremsstrahlung and SR ray tracing (stopping white beam in FOE) Establish (more) precise requirements for filling modes and bunch BROOKHAVEN SCIENCE to bunch variations 20