XX--rayray optics Crystal optics
Jürgen Härtwig
ESRF X-ray Optics Group, Crystal Laboratory
X OOG ESRF X-ray Optics Group What was already presented (among others)?
Physics of the electron beam source (Boaz Nash)
Physi cs of X-ray radia tion prodtiduction and transpor t (Ml(Manuel Sánchez del Río) Multilayers in synchrotron optics (Christian Morawe)
Energy resolving detectors for X-ray spectroscopy (John Morse)
So we continue today with the X-ray optics Outline
1. Introduction 2. Monochromators 3. Some properties of asymmetr ica l refltiflections 4. Shortly about high energy resolution 5. Crystal quality and how to measure it (6. Plane or divergent, monochromatic or polychromatic waves in our experiments?) 1. Introduction Some questions I plan to discuss and maybe to answer: Which kind of monochromators are used? How may I change the energy resoltilution, beam divergence, beam dii?dimension? May the Bragg diffraction geometry have an influence on the coherence? RlRole of source size an d angu lar source size. IflInfluence on transversa l coherence, resolution etc. ? Are “Imaging quality”, “focusing properties”, “coherence preservation” related? What is a “highly perfect” crystal? What are the lowest strains that we can measure? Is there a “highly parallel (monochromatic)” beam? Is there a “nanometric parallel” beam? We need to define what a “plane” or a “monochromatic” wave could be in the real exppypperimental life. How may we approximate them? etc. Optical system / experimental set-up: Source optical elements sample optical elements detector
The task of the optics: To transform the beam to obtain the best matching with the experiment; not loosing the good properties of the beam after its creation.
It acts on: - shape - wavelength/energy - divergence - polarisation - coherence “No optics is the best optics”!? Yes, btbut … In principle possible - all in vacuum, working with non-modified “white” or “pink” beam. However, not very useful. We may need e.g. monochromatisation, focussing, ... . A whole zoo of optical elements • slits, pinholes • filters, windows • mirrors (reflectivity based) • beam splitter monochromators (crystals) • monoch/ll/lhromators/collimators/analysers (crystals, multi-layers – often also called “mirrors”) •phase plates ((hphase retard er, pol li)arizer) (crystals) • lenses, zone plates •combine d e lemen ts (ML gra tings, Bragg-FlFresnel-l)lenses) • etc.
Mirrors and monochromators, collimators, analysers ffatlat, but asoalso bent for collimat ion, focussing, imag e magnification, ... Main physical effects used in X-ray optics were discovered in the first years starting from the discovery of X-rays by C. W. Röntgen: absorption (Röntgen 1895/96 filters) Bragg-diffraction (Laue 1912 monochromators, etc.) specular reflection (Compton 1922 mirrors) refraction (Larsson, Siegbahn, Waller 1924 later lenses)
PtiProperties of dbldouble and many-crystltal set-ups: Jesse W. M. DuMond, Physical Review, 52, 872-885 (1937) DuMond graphs, dispersive and non-dispersive set-ups, channel cut crystals (later invented as “Bonse-Hart-camera”), four crystal spectrometer (later invented as “ Barthels monochromator”), etc.
But -many newer ddlevelopments Quite a lot of literature
Overviews:
Tadashi Matsushita, X-ray Monochromators, in Handbook on Synchrotron Radiation, Vol. 1, ed. E. E. Koch, North-Holland Publi s hing Company, 1983
Dennis M. Mills, X-Ray Optics for Third-Generation Synchrotron Radiation Sources, in Third-Generation Hard X-ray Sources, ed. Dennis M. Mills, John Wiley & Sons Inc., New York, 2002 Short remark concerning beam dimensions
Few years ago – micro-beams were modern, now – nano-beams are in vogue.
But - we need all kind of beams: large beams (()decimetre sized) and small ones (nanometre sized), “parallel”, divergent and focussed beams. Beam dimensions – examppple: paleontology
Examples from Paul Tafforeau
Nearly 4 orders of magnitude in dimension Without scanning Without magnification
Multi scale experiments! Further large scale objects: the monochromator crystals
For their tests we like to use: wide beams, if possible at least 10 x 45 mm2 (V x H), with a “good” spatial resolution ~ 1 μm
The above field of view and resolution needs sensors with: 10,000 x 45,000 pixels, this is 450 Mega pixels
Not yet on the market (?) 2. Monochromators but also by: filters + 100 source spectrum + scintillator screen E = 8 keV response spectrum 10-1 (Pau l Ta fforeau/ID19)
forbidden area 10-2
R = 1
-3 rs) rs) reflectivity reflectivity oooo dddd 10 (Mirr(Mirr High-Z
Integrate Integrate Ge -4 111 ML's ML's 10 ML's ML's Low-Z dddd Si nnnn 111 ML's * C 111
-5 Crystals 10 -resolutio-resolutio pth-grade pth-grade eeee hhhh Be dd 110 highig
10-6 10-6 10-5 10-4 10-3 10-2 10-1 100 High resolution E/E 10-8 and below Ch. Morawe Bragg diffracting X-rayyp optical elements like monochromators, analysers, etc.
MftdManufactured from diltidislocation free crys tltals. Mostly used: Si, Ge, C* (locally dislocation free diamond).
We mainly use silicon.
They must be tailored into monochromators etc. Orientation, cutting, lapping, polishing and etching. Strain free crystal preparation. Accurate and stable mounting. Adequate cooling scheme. Crystal laboratory:
Manufacturing of nearly all perfect Si (and few Ge) crystal monochromators & analysers, etc for all ESRF beam lines, CRG blibeamlines and externa l laboratories. Silicon pieces are made from float zone silicon ingots with 100 mm diameter (Wacker). More than 1.5 tons of silicon single crystal material has been processed in more than twenty years.
Example manufacturing crystal monochromators and analysers Which types of monochromators are used? Single crystal monochromators - beam splitter monochromators
white beam white beam white beam
monochr. beam 2 monochr. beam 1
Double crystal monochromators
monochr. beam Reflection (Bragg) and white beam tra nsmissio n (Laue ) geometry used Single crystal monochromators - beam splitter monochromators
white beam white beam white beam
monochr. beam 2 monochr. beam 1
Double crystal monochromators
Reflection (Bragg) and white beam monochr. beam tra nsmissio n (Laue ) geometry used Few theory and definitions
Reflectivity (and transmissivity) curve of a crystal
plane R Theory monochromatic Plane & monochromatic wave Darwin incoming wave, width varying the angle of incidence counting the diffracted photons RRTA1+T+A=1 Real situation - experiment Rocking curve Convolution (autocorrelation) of reflectivity (or transmissivity) curve any wave with other reflectivity curves, or/and wavelengg(th (energy) distribution, or/and divergence distribution (instruments/apparatus f unction) Numerical example (using XOP): dbldouble monochromator non-dispers ive (+,-) double crystal set-up in Bragg case two 111-Si reflections
Darwin curve two Si 111 crystals Si 111,Si 111,60keV, 60keV, 5cm 10mm thick thick plateplate 1.0 1,0
g
080.8 gg wh 080,8 0.6 wh 0,6 position reflectivity reflectivity 0.4 0,4 matical Bra ee
0.2 kin 0,2 0 0.0 0,0 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 angle - (arc seconds) angle - (arc seconds) B B
known since: C. G. Darwin, However, in operation The theory of X-ray reflection, – mostly fixed angle Phil. Mag. 27, 315, 675 (1914) Reflectivity curves, with stronggper absorption
Prins-Darwin or reflectivity curve Si 111, 8keV, 5cm thick plate 1,0
080,8 wh
y 0,6 ivit tt
reflec 0,4
0,2
0,0 -5 0 5 10 15 20 angle - (arc seconds) B
How to measure it? Not so easy! See later. Double crystal monochromators
Incident and exit beams have the same direction
R() In working position: R
fixed angle (mostly); 2 R multiplication of two reflectivity curves Two vari ants
Fixed exit monochromator - more than one movement necessary Two vari ants
Fixed exit monochromator - more than one movement necessary Two vari ants
Fixed exit monochromator - more than one movement necessary
Channel-cut monochromator - NOT fixed exit - naturally aligned - weak link plus piezo movement for detuning etc. Two vari ants
Fixed exit monochromator - more than one movement necessary
Channel-cut monochromator - NOT fixed exit - naturally aligned - weak link plus piezo movement for detuning etc. Two vari ants
Fixed exit monochromator - more than one movement necessary
Channel-cut monochromator - NOT fixed exit - naturally aligned - weak link plus piezo movement for detuning etc. “Generic” cryogenically cooled channel-cut double crystal monochromator no fixed exit high heat-load applications Pusher mechanism
Steel flexure
Channel-cut crystal
Crystal assembly (simplified for visibility) Possible problem – “higher harmonics”
if (h,k,l)=(nh’,nk’,nl’) 2dnh’,nk’,nl’ sinB = n
Si (2n,2n,0), = 1.6Å, = 0.8Å, etc.
Refraction correction (middle of the reflection domain): r 2F 0 0 h 0 1 0 sin B , h sin B 2Vuc sin 2B 0 Re duc tion of hig her harm oni cs (if no mirrors may be us ed) monochromator detuning
At dddetuned position (slightly misaligned)
SllSmaller bdband in angle and energy Synchrotron optics: Multilayer high flux monochromators • Two bounce optics • 100x larger bandwidth compared with Si(111) • Harmonics suppression due to refraction and filling factor
Ch. Morawe – SPring-8 11.09.07 How mayyg I change the beam dimension, the beam divergence, the energy resolltiution? 3. Some ppproperties of asy mmetrical reflections
Up to now we looked at symmetrical cases of Bragg diffraction
symmetrical Bragg case symmetrical Laue case (reflection case) (transmission case)
– angle between lattice planes and surface With asymmetric reflections - ppgossibilities to change the beam width and the divergence for a single crystal reflection
w in h K L 0 in K h out wh Lout
Of course, the same works also other way around.
So we have possibilities to act on the beam dimension (expansion, compression), as well as on the beam divergence (ll(smaller, larger) w out h L Kh out K0 in wh Lin
But all is related and things have their price w in h K L 0 in Asymmetry factor: K h out wh Lout sin( ) b B 0 si(in( B ) h
This example: >0, |b|<1: w out =|b|w= |b| w in in out h h Lin > Lout and wh < wh Relation to symmetrical in out wh Lin = wh Lout = constant reflections in -1/2 sym right asymmetry (<0) – less wh = |b| wh out 1/2 sym diverggpence possible to obtain wh = |||b| wh Some “pppyhilosophy” With flat crystals we may change the divergence. We mayy( decrease (or increase) it. Not “more p p,arallel”, but less divergent! This is collimation*). Focussing needs convergent beams. We can not focus wihith flat crystal s in a cl assi cal way. It work s wi ihth bent crystals. Often other X-ray optical elements are more efficient for this (future lectures?!).
*) The word " collimate" comes from the Latin verb collimare, which originated in a misreading of collineare, "to direct in a straight line". (Wikipedia) BUT …
in out wh Lin = wh Lout = constant
This is too simple, hand-waving “derivation”. We need an additional dimension for the phase space. Besides size and angle also wavelength (or energy) is needdded
LtLet us st art from the basi cs 0 = B + - in K 0 h h = B - + out t n
B From electrodynamics (and B dynamical diffraction theory) K K we know that: 0 h h 0
For the wave vectors outside the crystal: K h K 0 h
But for the tangential components - continuity: Kht K0t ht
And remember,,pg wave vectors depend on wavelength: K f() For small in, out and K we obtain (for Bragg and Laue case!):
sinB cos(B ) cos(B ) K | out | in sinB sinB K
The divergence in and polychromaticity K/K of the incoming beam contribute to the divergence out of the outgoing beam
An increased divergence out of the outgoing beam (with respect to that if the iiincoming b)beam) means: the source is virtually closer, or the source size is virtually large r, or the angular source siz e is virtuall y larger.
Special case 1 Monochromatic, divergent incoming radiation
K = 0, in 0 out = |b| in out in or wh = |b| wh Special case 2 PlPolych romati c, paralllllel iiincoming radiditiation
K 0, in = 0 cos(B ) cos(B ) K out sin(B ) K
Remember, “our” beams often are rather close to plane waves, but rather polychromatic
2.1. out = 0 if cos(B - ) = cos(B + ) if (sym. Bragg case!) Only in the symmetric Bragg case the beam divergence is conserved for a polychromatic beam!!! Only in that case coherence is conserved!!! Only in that case focussing is not perturbed! Only in that case highest geometrical resolution possible!
2.2 out 0 for all other cases A divergent, polychromatic beam is transformed in a even more divergent, po lhlychromatic b!beam! Source size and angular source size are crucial parameters with respect to the character of the wave “seen” by the sample. angular source size ( = s/p) s p δ not source divergence!!
The angular source size is important for further physical properties: the geometrical resolution for imaging, the transversal coherence length, the demagnification limit of a “lens”. Image blurring due to non-zero source size angular source size: δ = s/p
s δ p q
geometrical resolution: ρ = q s/p = q δ
s δ p q
Spatial coherence
Transversal coherence length: lT = ½ λ p/s = ½ λ/δ Magnification, demagnification, focussing properties/quality
Geomet ri cal demagn ifica tion, Diffrac tion li m ite d focus ing: source size limit: ρDL = 1.22 λ / sinα
ρG = q s/p = q δ
s ρ
p q (graph: J. Susini) 4. Shortly about high energy resolution
One way: Sophisticated many crystal set-ups.
Example ID18: A two-step collimation with low-index asymmetric reflections followed by a two-step angular analysis with high-index asymmetric reflections.
Idea: M.Yabashi, K.Tamasaku, S.Kikuta, and T.Ishikawa, Review of Scientific Instruments, 72, 4080 (2001). Alexander Chumakov ID18 ESRF ID18/ESRF High-resolution optics for Nuclear Resonance Scattering “0.5 meV” monochromator (∆E/E ≈ 3.5·10-8) (theoretically expected performance)
acceptance: 130 nano-rad divergence: 65 nano-rad
31 mm 3 2
vertical size: 0.6 mm vertical size: 0.6 mm divergence: 2 rad divergence: 2 rad
4 1
1st crystal: 2nd crystal: 3rd crystal: 4th crystal: Si(400) Si(400) Si(12 2 2) Si(12 2 2)
B = 18.469º B = 18.469º B = 77.533º B = 77.533º asymmet|b|018try: |b| = 0.18 asymmet|b|018try: |b| = 0.18 asymmet|b|98try: |b| = 9.8 asymmet|b|32try: |b| = 3.2
in = 5.4º in = 5.4º in = 27.6º in = 35.4º out = 31.6º out = 31.6º out = 2.7º out = 10.4º angular acceptances: angular acceptances: angular acceptances: angular acceptances:
in = 20 rad in = 20 rad in = 0.72 rad in = 1.3 rad out = 3.6 rad out = 3.6 rad out =7 rad out = 4 rad footprint: 6.4 mm footprint: 33 mm footprint: 41 mm footprint: 3.3 mm Other possibility – back scattering geometry (Bragg angle close to 90 deg) with high order reflections (large h k l)
= 2dhkl sinB ∆ = ∆ · 2dhkl cosB
∆= ∆E/E = ∆/tanB
a) → a f10few 10-8 radb)d b) → d[t(deg [cotg() → 0]
Silicon @@, 20KeV, h=k=l=13, ,( (E/E) ~10-8E ≈ 0.5 meV, ID16, ID28 analyser Sample
Detector source monochromator 5. How higgyh crystal q qyuality and how to measure it?
Crystal quality - limit in high energy resolution, …(?)
IflInfluence on coherence preservati on, image qua lity, focussing efficiency, ...
∆d/d= 5x10-9 Crystal interferometers Diamonds at the ESRF From the very beginning of the ESRF used as phase plltates and monochthromators.
Now locally dislocation- and stacking fault free material available.
Which is the level of local residual strains? Future MX BL ID30A (MASSIVE)
~100 μm2 ~15 μm2
56 X-ray diffraction topography White beam topograph in transmission (work with Fabio Masiello) 110-oriented plate supplier: Paul Balog/Element Six
Dislocation free areas of 11. 3 mm 6x4mm2 and more!!!
Locally crystal quality close to that of silicon, quantitatively confirmed by double crystal topography
But are we able to measure weak strains quantitatively? QQyyuantitative analysis strain analysis 110-oriented crystal plate effective misorientation map basing on one topograph
20keV, Si [880] C* [660], detection limit: δθ > 8·10-9
The effective misorientation is of the order of 4 × 10−8 for a region of interest of -8 050.5 × 050.5 mm2 and × 10 1 × 10−7 in a region of 1 × 1 mm2
Sample is slightly bent due to the non-homogeneous dislocation distribution! work with Fabio Masiello 6. Plane or divergent, monochromatic or polhlychromati c waves Twwquo basic questions: What are “plane” or “divergent” waves? What are “hi”“monochromatic”, “lh“polychromati i”c”, “hi”“white” beams/waves? For our monochromator and/or single crystalline sample! Reminder of basic physics: Plane wave – infinite extend, wave front is plane, one wave vector perpendicular to it, delta-function in k-space. Monochromatic wave – wave train of infinite length, infinitely sharp spectral line, delta-function in ω-space. They do not exist in nature!!! The full width at half maximum of the reflectivityy, curve, good reference for our sample, monochromator, ... , to define the character of a wave
FWHM in the angular space: R()
wh 2 P h h h wh sin2B 0
FWHM in the wavelength space: R()
wh ww cot hh B Example of a typical crystal/monochromator reflection: silicon, 111 reflection, Bragg case, thick crystal
energy wh wh / 8 keV 7.6 arcsec 1.5·10-4 20 keV 292.9 arcsec 151.5· 10-4
Those are to be compared with source properties: wave length spread of the source ()
angular source size ( = s/p) s p δ not source divergence!! relative wave length spread of the source (E/E)
<< FWHM in the wavelength space wh
“h“monochromatic” wave
angular source size = s/p
<< FWHM in the angular space wh
“plane” wave Practical examples: relative wave length or energy spread of the source
source energy w / h
-4 -4 Si 111 double mono 2·10 8 keV 1.5·10
-4 -4 laboratory (e.g. CuK1) 3·10 20 keV 1.5·10
white beam 1···10
SR–white beams: >>> wh really polychromatic
”monochromatic” beam (SR or laboratory): wh often not monochromatic (for all reflections narrower than the 1.5·10-4 for silicon 111)
sppyppyecial effort is necessary to approximate “monochromaticity” Practical examples: angular source size ( = s/p)
source source size ssource dist. D δ
-3 class. lab tube 400 µm 0.4 m 1·10 3.5 arcmin energy wh
µ-focus tbtube 5 µm 1 m 1·10-5 1 arcsec 8 keV 767.6 arcsec
SR 100 µm 150 m 6.7·10-7 0.14 arcsec 20 keV 2.9 arcsec
SR 50 µm 75 m 6.7·10-7 0.14 arcsec
laboratory: >< wh divergent & quasi plane waves possible
SR - EEFSRF: < wh qqpuasi plane wave (mostl y)
Source divergggences > angular source sices For yyyour curiosity: First published (direct) measurements of reflect iv ity curves, (Not rocking curves! Those were measured earlier), the ones that we are able to calculate since Darwin’s first dynamical theory published in 1914, was in??? 1962
nearly 50 years after Darwin’s results!!!
C. G. Darwin, The theory of X-ray reflection, Phil. Mag. 27, 315, 675 (1914) Charles Galton Darwin (1887-1962) grandson of Charles Darwin How can we measure a reflectivity curve???
Not trivial. One needs two ingredients.
1. Narrow instruments function in anggpular space – asymmetrical ref lections in wavelength space – high-resolution mono.
2. Crystals of good quality. They became available with development of electronics, later micro-electronics and opto-electronics Double (or triple) crystal monochromator-collimator with one asymmetrical Bragg reflection
1st experimental reflection curves
=-4.48
=0
=4.48
111-reflections of Ge, different asymmetries
CuK1-radiation, triple crystal set-up, silicon double monochromator R. Bubáková,,Jy Czech. J. Phys. B12,,( 776 (1962) By the way – the prob abl y fir st publish e d measurements in a double crystal set-up with an asymmetric monochromator/collimator crystal were about X-ray diffraction topography in 1952.
W. L. Bond J. Andrews, Structural Imppyerfections in Quartz Crystals, Am. Mineral. 37, 622-632 (1952)
This technique was later developed further to the “Plane” Wave Topography, to dtdetec t extreme ly small stitrain fildfields in nearly perfect crystals. Thank you for your atten tion !