National Synchrotron Light Source II

1 BROOKHAVEN SCIENCE ASSOCIATES

National Synchrotron Light Source II

Status and Progress of 0.1meV R&D at NSLS-II

Xianrong Huang, Zhong Zhong, Yong Cai

Thanks to Yu. Shvyd’ko (designer, APS), S. Coburn (BNL), D. P. Siddons (NSLS), A. Baron (RIKEN), C. Kewish (APS), A. Macrander (APS), J. Hill (BNL)

and all contributors

Feburary 7-8, 2008


Outline

I. Angular Dispersion Optics: Principles and Designs

II. R&D Progress:

Modeling, developing dynamical theory programs

Revisit Shvyd’ko’s preliminary experiment of CDW prototype

Set up semi-permanent R&D facilities at NSLS

0.7 meV prototype development...

III. Technical challenges and possible solutions

IV. Exploring alternative approaches

V. R&D schedule and resources


0.1-meV-resolution Spectroscopy

Inelastic X-ray Scattering (IXS) spectroscopy for low-energy photons (5-10keV):

More photons from undulators while IXS is extremely flux-hungry

Better momentum resolution for the same acceptance angle

Stronger scattering cross section, better match to sample size

Extend high-resolution IXS to low- and medium-energy SR facilities

(e.g. NSLS-II) ...

Unfortunately, lower-energy IXS conflicts with the principles ofbackscattering analyzers and monochromators — spectral Darwin widths of low-index Bragg reflections are too large, E > 20 meV.

Solutions: Select a small portion of the spectrum New Angular Dispersion Optics

Si (008) backscattering E = 9.1 keV


I. Principles of Angular Dispersion Optics

sincos cose d

sin

sine

ed

= const

Rigorous !

Dispersion in asymmetric diffraction: Different wavelengths diffracted

along different directions

sin tan

sin(90 )e

d d

2 tane

H

E

E

2H

hcE

dwhere

90e

Si (008), EH 9.1 keV e = 3 rad, = 89.6

Dispersion maximized in Grazing Backscattering:

E = 0.1 meV

Resolution determined by and e

Independent of spectral Darwin width

Asymmetric angle

Polychromatic


Implementation for 9.1 keV: Basic components(Yuri Shvyd’ko’s designs)

Si (220)

In-line CDDW

h = 500 m1 meV L = 12 cm0.3 meV L = 38 cm0.1 meV L = 140 cm

h = 50 m1 meV L = 1.2 cm0.3 meV L = 3.8 cm0.1 meV L = 14 cm

tane

H

E

E

4

Backward CDW

tane

H

E

E

2

Transmission through thin crystalsdue to Borrmann effect


Dynamical Theory Calculations

1.7

Completely verified the principles

R

same crystallengths and parameters

0.6, 0.16 meV2 tan

e

H

E

E

0.3, 0.08 meV

4 tane

H

E

E


Unique property: Acceptance ~100 rad

Acceptance = 140 rad C/W

Wide acceptance due to collimating by C crystal.

Does not place stringent requirements on stability of incident beam. Works for divergent beam (so can be used as analyzers).

= 89.5o

1 = 1.0o

CDW and CDDW both have large acceptanceup to 200 rad


Incidence 1

E

Acceptance ~ 100 rad

E ~ 0.1 meV

DuMond Diagram


Focused–beam Mono and Analyzer

Monochromator

Analyzer

Major difficulty: Long D crystal length L For beam height h = 0.5 mm L ~ 1-2 m !

Possible solution to shorten L: Using slightly focused beam h < 50 m L ~ 10 cm utilizing wide acceptance

Analyzer side: it is important to develop precise Multilayer Mirror to improve efficiency


Parallel–beam Mono with Channel-cut

< 5 rad

Better resolution function


Energy Tuning

Energy scan within the spectral width E = 34 meV of the backscattering through Rotation of the D crystal in CDW Rotation of both D1 and D2 in CDDW

Out of E range, energy scan is through temperature scan of D crystals

rotating fan

W


Or

0.1 meV spectrometer: Layout

Focusing mirror 1 forreducing the beam size

Choice of Analyzer depends on focusing condition: 1. Segmented CDW (L ~ 20 cm for each segment) if beam size > 0.1 mm 2. A single CDDW with L < 20 cm if beam size < 0.1mm Developing Multilayer Mirror is important

Large-acceptance


Main Features of CDW and CDDW

E/EH purely determined by and e (geometrically), independent of EH or of Bragg reflection

For same and e , E decreases with decreasing EH , working better for lower energy

High efficiency (~ 50%)

Wide angular acceptance ~0.1 mrad, working for divergent or focused beam!

Steeper wings (i.e. cleaner resolution function)

For both Mono and Analyzer (analyzer must be combined with collimating mirror)

New but feasible optics with 1-0.1 meV resolution for low- to medium-energy IXS

tane

H

E

E

e determined by asymmetric factor b

of C/W crystals


II. R&D Progress

1. Modeling optics & developing dynamical theory programs

Simulation programs and tools based on rigorous dynamical theory have beendeveloped from scratch, tested to be error-proof

• Applicable to backscattering, extremely asymmetric diffraction, transmission, Borrmann effect, ...

• Component-orientated, easy plug in (using C++ classes)

• For design, spectrum calculation, simulating experiments (alignment, energy tuning), etc

These codes have been used to verify quantitatively that the CDW and CDDW schemes work and are efficient at achieving 0.1 meV energy resolution.


2. Revisit Shvyd’ko’s preliminary experiment of CDW prototype

• Successfully observed angular dispersion effect• Darwin width 34 meV narrowed to 2.2 meV Very promising

• Why the discrepancy???

Yu. Shvyd’ko et al, PRL 97, 235502 (2006)

SRI (2007)

E=9.1 keV

E=2.2 meV


Problems: Damages caused by

cutting and polishing Lattice bending caused

by machining or weak-linking, broadens E dramatically

Surface roughness (next slide)

X-ray topography of thin W crystals

on peak

off peak

Lattice is bent (bent part out of the Borrmann effect angular range). Crystal is not acting as a flat wavelength selector

Transmission Image taken at 33 BM, APS

Defects destroy Borrmann effect


X-ray topography (cont.)

Reflection topograph taken from the upper surface of W crystal showing surface roughness

Topograph of the long D crystal. Almost perfect except scratches.

Conclusion: Failure to achieve theoretical performance is mainly due to the bad quality of W crystals, particularly the bent areas that select different wavelengths.

Solution: Completely redesign it


Experiments underway to measure Borrmann effect and to determine the thickness t and the feasibility of free-standing crystal (next by Zhong Zhong).

Structure factors calculated using Sean Brennan’s program

3. Redesign W crystal, measure Borrmann effect

Our simplified W crystal that works better. Final design will be a free-standing thin crystal

Redesign it to an independent component, used in both CDW and CDDW Transmission through W crystal is due to the Borrmann effect

for measuring Borrmann


II. R&D Progress (cont.)

4. Setting up semi-permanent R&D facilities at NSLS

5. Recent progress of 0.7 meV CDW experiments

(see next by Zhong Zhong)


III. Technical challenges and solutions

1. Lattice constant homogeneity of Long D crystal: ~20 cm per segment, up to 10 segments, but using a strip detector, perfect alignment or precisely the same temperature is not required.

Lattice constant homogeneity is required d/d ~ 10-8 throughout each segment

Solution: Acquire highest-quality Si crystals, need accurate measurements before use by the Laue-case lattice comparator setup designed by D. P. Siddons (underway).

Use focused beam (h <50m) to shorten crystals to < 20 cm, down to a few cm Less challenging on Mono side. On analyzer side, depends on development of Multilayer mirrors

D crystal


Challenges and solutions (cont.)

2. Temperature homogeneity and stability of Long D crystal: < 1 mK, achievable from previous experiments

3. Surface roughness (slope error) for grazing diffraction:• Calculations shows the C/S crystal can tolerate local surface

slope error < 0.05º, achievable.

• Roughness of D crystal affects the bandwith for grazing angle ~0.5º ( = 89.5º). E depends on .

• Need high-precision fabrication of crystal surfaces, slope error <0.05º.

4. D crystal mounting: Strain-free mounting, no bending of D crystal (< 0.5 rad), achievable.

5. Stability: 0.5 rad stability of D crystals.

Thermal coefficient of Si: 2.5610-6 K-1

4 tane

H

E

E


Challenges and solutions (cont.)

6. Developing high-efficiency multilayer mirrors Large acceptance (~5 mrad), collimating/focusing to <0.1 mrad and 50 m

Focusing determining the choice of analyzer and D crystal length.

Solution:

Similar mirrors are already achievable at commercial companies.

or R&D effort will be made in collaboration with APS, Spring-8, ESRF etc to develop specifically designed high-efficiency multilayer mirrors.

Mirror efficiency reduce performance but does not affect resolution.


IV. Alternative approaches of 0.1 meV

Cons:Still need long crystals Efficiency less than CDDW

Acceptance > 100 rad

Pros: More flexible, many variants Arbitrary energy

sin

sine

ed

E = 9.3 keV

1. Four-bounce, for both mono and analyzer


2. Multi-cavity Fabry-Perot Interferometer (funded by BNL’s LDRD)

Si (12 4 0) E = 14.4 keV

5meV 0.5meV 0.1meV Multi-cavity FB can increase both Free Spectral Range and Finesse.

Reduce bandwith from ~10 meV to sub-meV

Pre-monochromatization not difficult.

Ultra-compact, single-component, powerful

In principle with no limit neV

Pre-mono

Courtesy of D. P. Siddions, K. Evans-Lutterodt,

A.Isakovic


3. Massive Parallel Analyzer

500 pieces, covers 10 rad each

5 mrad total

0.2 mm

0.2 mm

1 m

0.3 mm

3.5 mm

Concept is similar to back-scattering analyzer. Each crystal is a small D crystal. Collimation provided by small beam size on

sample, small crystal surface and distance. Angular dispersion by D crystal allows

acquiring all energies by a PSD detector. Sample

Shielding

PSD Detector


V. R&D Schedule

FY 08 — to achieve 0.3 meV

Improve facilities (monos, slits, detectors/CCD camera, motors etc) at X12A

Fabricate crystals and repeat 0.7 meV CDW

Set up temperature-controlled chambers for long D crystals

Measure Borrmman effect to optimize thickness of thin W crystals, repeat 0.7 meV CDW (optics can be used for 1meV endstation)

Build inline CDDW to achieve 0.3 meV resolution

Test monochromatization with divergent beam

Design multilayer mirrors


FY 09 — to achieve 0.1 meV prototype

Full work on CDW and CDDW prototype to achieve 0.1 meV resolution (with limited flux)

Detailed exploration of crystal quality (defects, impurities, inhomogeneities)

Detailed study of fabrication issues (figure error, roughness)

Detailed exploration of temperature control

Develop collimating/focusing multilayer mirrors (collaboration with OFM group at APS, other labs, commercial vendors etc)

Dynamical theory calculations of alternatives (interferometer, four-bounce, multi-crystal analyzer), concentrating on feasibility as analyzer

Test alternative four-bounce design (if necessary)


FY10-FY11

FY10

Design and develop engineering solutions for adequate temperature homogeneity and stability control

Achieve full-scale 0.1 meV CDW/CDDW components with limited flux Test and improve collimating/focusing multilayer mirrors Build and test full-scale CDW/CDDW components (particular mono optics)

FY11

Test and improve analyzer optics with multilayer mirrors (10 x 5mrad2 H&V angular acceptance)

Test alternative approaches Finalize design


0.1 meV R&D Resources

Staff (we are hiring)

FY08 FY09 FY10 FY11Scientist 1.0 1.0 1.0 1.0Assist Sci 1.0 2.0 2.0 2.0Post-doc 0.0 1.0 1.0 1.0Designer 0.5 0.5 0.5 0.5

Crystal FabTechnician 1.0 1.0 1.0 1.0

Crystal Fabrication Equipment• Diamond rough saw• Diamond mill• Lapping machine• Chemical polisher• Table top x-ray source• Chemical hood, cabinet etc

Access to NSLS beamlines:• X12A: dedicated R&D• X19C: crystal characterization (regular beamtime)• Many other beamlines at NSLS• PUP at APS


Summary

New optics achieving 0.1 meV resolution at ~9 keV photon energy have been designed for IXS at NSLS-II based on simple and well-established angular dispersion effect in asymmetric Bragg diffraction.

The main optical components are CDW and CDDW mono and analyzer; their detailed operation principles have been proved by rigorous dynamical theory calculations; the CDW mono has been experimentally tested to 2.2meV.

All mechanisms involved are fully understood.

There are technical challenges (e.g. the long crystal issue). These challenges are also well understood and we have feasible solutions (including alternative approaches).

The R&D have already been underway, including detailed theoretical modeling and preliminary experiments. Further R&D are well planned with adequate resources.

National Synchrotron Light Source II

Documents