Coherence activities at the ESRF: A Practical Approach ... · Scanning (Transmission) X-ray...

1

European Synchrotron Radiation Facility

Jean SUSINI

COHERENT X-RAYS AND THEIR APPLICATIONSA series of tutorial–level lectures edited by Malcolm Howells

Lecture 6 - 9th June 2008

Coherence activities at the ESRF:

Scanning (Transmission) X-ray Microscopy

STXM �� SXM

A Practical Approach

ESRF Lecture Series on Coherent X-rays and their Applications, Lecture 6 - Jean Susini

Scanning X-ray microscope arrangements

XPSanalyser

� Soft X-rays (δ ~ 40nm)

Scanning Photo-Emission Microprobe

� Hard X-rays (δ ~ 80nm)

X-ray Fluorescence Microprobe

Energy Dispersivedetector

Crystal spectrometer

� Soft- X-rays (δ ~ 20nm)

� Hard X-rays (δ ~ 80nm)

Scanning Transmission X-ray Microscope

X-raydetector

λλλλ

λλλλ

λλλλ

• Detection of signals arising from interaction of small probe with sample

• Sample scanned through beam to build ‘image’

pixel by pixel

2


Synchrotron based micro-probe techniques

X-Ray Fluorescence

X-ray spectroscopy

X-rayDiffraction & scattering

InfraredFTIR-spectroscopy

• Composition• Quantification

• Trace element mapping

• Short range structure

• Electronic structure

• Oxidation/speciation mapping• Molecular groups & structure

• High S/N for spectroscopy• Functional group mapping

• Long range structure• Crystal orientation mapping

• Stress/strain/texture mapping

Phase contrast X-ray imaging

• 2D/3D Morphology• High resolution• Density mapping

ID11, ID13, ID22

ID21, ID22, ID24

ID21

ID21, ID22

ID13, ID21, ID22

ESRF instruments


Synchrotron based hard X-ray microprobe

Photodiode

Undulator

Crystal monochromator

X-ray lens

Aperture

Sampleraster scanned

Fluorescence detector

CCDDiffraction

CCDAlignment & imaging

• Spatial resolution : 0.05-1µm

• Spectral resolution : 10 -2 > ∆E/E > 10-4

• Averaged flux : 10 10 – 1013photons/s/µm 2

3


Synchrotron micro-probes

focus

Low ββββz25x930µm2 - 17x29µrad2

High ββββz25x135µm2 - 17x208µrad2

(Secondary) source 50-250m

< 50x50nm2

Diffractive lenses Refractive lenses

X-ray waveguides X-ray reflectors

• Resolution determined by probe size and overall stability

• Size of the probe is a convolution of the geometric image of the source and the point spread function of the lens

• Diffraction limited vs aberration limited?

• Coherent illumination required for diffraction-limited resolution but images are not coherent.

• SXMs are coherent (brightness) experiments


Diffraction limited focusing

qpf

111 +=p

qsG ×Σ=

αsinnNA=

Geometrical demagnification:

Diffraction limited focusing ifπλθ4

=×Σ

Thin lens equation

Numerical aperture

NACsDL

λ= where C=0.61 for a circular aperture

p q

Σ sααααθθθθ

Central Airy disk

NAd

λ22.1=

for circular lens

4


A trade-off

p q

Σ sααααθθθθ

Central Airy disk

NAd

λ22.1=

for circular lens

Optical aberrations often limit resolution before diffraction limit is reached

p

qsG ×Σ=

Geometrical demagnification Coherent illumination

αλ

sin22.1=DLs

• Source size (horiz. vs. vert.)

• Distance source-optics

• Working distance

• Working distance

• Lens aperture

• Beam divergence


Ideal focusing of a point source

� Focusing: equal optical paths source-to-focus

� Ellipse definition: p + q = constant

Courtesy C. Morawe (Optics Group)

pq

5


A fundamental limit: “Liouville’s theorem”

1 - The phase-space density of photons cannot be increased.

2 - In fact, because of absorption and finite efficiency of the optical elements, the phase-space density decreases.

3 - Implications for diffraction experiments.

The emittance εεεεsize x divergence

is constant

ε ε ε ε = ∆= ∆= ∆= ∆i’ x ∆∆∆∆i = const

• Focussing : Size and divergence

• Collimating : Size and divergence

Beam divergence∆x’, ∆y’, ∆z’

Beam size∆x, ∆y, ∆z


Undulator beam and spatial modes

From lecture 4, M. Howells

6


Some source parameters

ESRF highlights 2007 p.121

High beta section (even IDs)

• Size H : 415.0µm

• Div. V : 2.9µrad

• Size V : 8.6µm

• Div. H : 10.3µrad

Electron beam parameters (RMS)

Low beta section (odd IDs)

• Size H : 51.0µm

• Div. V : 2.9µrad

• Size V : 8.6µm


Photon source parameters (RMS)

High beta section

• Size H : 415.0µm

• Div. V ~ 13-5µrad (5-40keV)

• Size V : 8.6µm

• Div. H ~ 16-11µrad (5-40keV)

Low beta section )

• Size H : 51.0µm

• Div. V ~ 13-5µrad (5-40keV)

• Size V : 8.6µm



Some beam parameters

0

500

1000

1500

2000

2500

5 10 15 20 25 30 35 40

0

2

4

6

8

10

12

14

16

18

20

High ββββHorizontal

Low ββββHorizontal

Low-High ββββVertical

Energy (keV)

Num

ber of modes

Vertical

Num

ber

of m

odes

Hor

izon

tal

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

5 10 15 20 25 30 35 40

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

Energy (keV)

Bea

m s

ize

(mm

)H

oriz

onta

l

Beam

size (mm

)V

erticval

High ββββHorizontal

Low-High ββββVertical

Low ββββHorizontal

FWHM Beam size @ 50m from the source

High beta sources provide a slightly more

“coherent” beam

A coherent experiment uses only one mode!!

What about the wasted modes?

7


Source demagnification : Scheme 1

• Requires a small pinhole (difficult to produce for hard X-ray?, heat load?, ….)

• Compromises the energy tunability ?

• Compact and stable instrument

• Overfilling of the slits makes the beamline less sensitive to drifts and vibrations

• Optical optimization is possible

• The secondary slit can be used to clean-up the beam (speckles from upstream components)

• Trade-off flux vs. resolution is tunable

A “ short” beamline with a secondary source

20x100nm225 x 900mm2

30m 1m 30m 0.1m

M1=1/30

30x30mm2

M2=1/300

See also lecture 3 (A. Snigirev)


Source demagnification : Scheme 2

• Exploits the source size (low β or apertured high β )

• Allows long working distance

• Preserves the coherence

• Preserves the energy tunability (spectroscopy)

• Preserves the vertical divergence tunability (diffraction)

• Overall stability

• Flux losses or large aperture optics

• Cost

Long beamline with a direct source demagnification

150m

25x150mm2

M=1/1500

20x100nm2

0.1m

See also lecture 3 (A. Snigirev)

8


High energy = very grazing incidence

+=

+=

qp

qpR

qp

qpR

is

im

θ

θ

sin2

sin

2

Radii of curvature

Source

Focus

• Small θ −> Long mirror (> 1m) or small NA• Off-axis geometry

• Aberrations (spherical aberration, astigmatism, figure errors)

θ < θc ~ 10mrad2θms RR ≈≈≈

kmR

mmR

m

s

]/[ 3][][

83.19cmgcc keVmrad

E ρθ =


Mirror shapes : z = ααααy2(1+ββββy+γγγγy2), Rc ~ 1/2α/2α/2α/2α

Source Focus

pqθθθθ

• Ellipsoidal : point-to-point focussing

−+=

−=

+=

q

p

ppq

q

p

p

q

p

p

e

e

e

116

cos5

4

1

12

cos

14

sin

2

2 θγ

θβ

θα

J. Susini, Optical Engineering, 34 (2) (1995)

“ideal” ellipse are extremely difficult to manufactured

9


Kirkpatrick-Baez geometry

Principle : perpendicular meridian planesP. Kirkpatrick and A.V. Baez,

“Formation of optical images with X-rays”J. Opt. Soc. Amer., 38, (1948)

Kirkpatrick-Baez mirror pair

Source

FocusSpherical aberrations ~ 1/L 2

Multilayer coating : Bragg angle �� shorter mirror or/and larger aperture

L


Kirkpatrick-Baez system at the ESRF

ApertureV X H (µm)

Focus SizeFWHM (nm)

FluxPh/s @ 90mA

200 X 50 92 X 87 5 10 10

400 X 100 70 X 74 2 10 11

600 X 160 90 X 70 4.5 10 11

microfocus on ID19 ( 20.5 keV)

ESRF: 40nm

O. Hignette et al., AIP Conf. Proc. 879 (2007)

Spring8: 25nm

H. Minura et al., APL 90 (2007)

10


Fresnel Lenses – Zone Plates

Augustin-Jean Fresnel1788-1827

Diffractive X-ray Lenses

Circular transmissive diffraction gratings with radially decreasing line width giving focusing effect

Alternate ‘zones’ modify phase/amplitude of incident wavefront:

for material of thickness, t, wavelength, λ, refractive index 1-δ-iβ,

phase shift, ∆φ, is:

λδπφ t2=∆

For X-rays:

Baez (1952), Schmahl (1969), Kirz (1971), Niemann (1974)


Fresnel zone plates: a Gabor hologram of a point object

Planar wavefrontHologram (Fresnel Zones)

Reconstructionby

coherent illumination

focus

Diffraction limit ( δ=1.22∆rN)

• requires coherent illumination

• λλλλ /∆λλλλ > N

11


Fresnel Zone Plates

D

∆∆∆∆rNt

� Resolution:m

rNm

∆= 22.1δ

� Focal length:λm

rDf N

m

∆=

� Depth of focus:λm

rDOF N

24∆±=

• m=1

• ∆rN=50nm

• D=100µm

8µm @500eV

160µm@10keV

2mm @500eV

40mm@10keV


Soft X-ray zone- plates

E. Anderson, Center for X-Ray Optics, LBNL, USA

• ∆∆∆∆rN = 25 nm

• D = 63 µm

• N = 618 zones

• f = 650 µm

• NA = 0.05 @ λ = 2.4 nm

• ∆rN = 15 nmW. Chao et al., Nature 435 (2005)

12


Zone plate aspect ratiostructure height, t, critical for efficiency, ∆∆∆∆rN for resolution

Aspect ratio for ∆∆∆∆rN=50nm

Nickel25%

12 : 1

2.0 keV

Tantalum32%

28 : 1

8.0 keV

Nickel24%6 : 1

0.5 keV

Practical limit for small ∆∆∆∆rN is ~ 10-15:1

∆∆∆∆rN

t

Material t (µm) ε (%)E=0.5keVGe 0.28 16Ni 0.25 24

E=2.0keVNi 0.60 25Au 0.45 24

E=8.0keVTa 1.70 32W 1.50 33

t


( )

2

2

sin1

≈p

p

mpm π

πε

ε1(2) = 40.5%

ε1(3) = 68.5%

ε1(4) = 81.2%

Fresnel zone-plates: several strategies

Amplitude

Phase

Blazed

Ideal 3-levelapproximation

X-ray opaque material

X-ray phase shifting material

• Amplitude : Opaque zones (K→ ∞)

• Phase : weak absorbing and phase reversal zone (Κ→0 and Φ = π)

ε1 ~ 10% and ε3 ~ 1%

ε1 ~ 40 %

( )Φ−+= ΚΦ−ΚΦ− cos211 2

22ee

mm πε

• Grating with equal lines and spaces

m = ±1, ±3, ±5,….λ

δπδβ t

K2

and =Φ=

13


Blazed zone plate: 3 level lens

E. di Fabrizio et al., Nature 401 (1999)

• material : Nickel

• diameter : 150µm

• ∆RN : 150nm

7 keV : ε ~ 57%

0

20

40

60

80

100

5 5.5 6 6.5 7 7.5 8 8.5 9

Effi

cien

cy (

%)

Energy (keV)

calculated (ideal)

calculated (with fabrication errors)

measured


Rejection of the unwanted diffraction orders

m = 1

ORDER SELECTING APERTURE

FIRST ORDER FOCUS

CENTRAL STOP

Zone-plate based SXM = Central stop + ZP + OSA

14


SXM – ID21 – Old version

DET

SAM

OSA

ZP

EW

PZTX-rays


Multilayer Laue Lenses

substrate

Varied line-spacing grating

Depth-graded multilayerson flat Si substrate

• Deposit varied line-spacing grating on flat substrate(thinnest structures first)

• Section to 5-20 µm thickness (high aspect ratio structure)

• Assemble two into a single device (MLL)

H. C. Kang et al., Phys. Rev. Lett. 96 (2006)

15


Tentative of comparison

• Resolution ++++ +++ +++

• Achromaticity -(-) +++(+) --(-)

• Efficiency + +++(+) ++

• Imaging (MTF) ++++ ++ +++

Zones plates Refractive lenses

Kirkpatrick-Baez mirror pair

Mirrors

Energy


X-ray Fluorescence: a brief reminder

� Element specific

� Co-localization

� Quantification

X-rayKen

ergy

continuum

ML

Photo-electroncontinuum

ML

K

KααααKββββ

Eexc

Energy dispersivedetector

energy

Inte

nsity

16


XRF on synchrotron beamline = low detection limit

monochromaticbeam

( 10-5< ∆E/E < 10-2)

High flux

45°

45°

EDS

Horizontal Linearpolarization

Broad spectrumBright

coherent

0.01

0.1

1

10

100

1000

104

105

10 15 20 25 30 35 40

Atomic number Z

MDL (ppm) on NIST 1577 (10sec)

PS

Cl

KCa

Mn

Fe

CuZn Se Br

10-18 g in a cell

• Low background

• Tunable energy (XANES)

Only a few degrees collection angle!

Detector developmentsDwell time < 100ms


µ-XRF in Trichomes of Arabidopsis Thaliana

P

70 µm

S Cd K Ca

Eex: 5.8 keV, probe size: 0.3x0.2 µm2, dwell time: 800 ms/pixel.

M.P. Isaure et al ., Biochimie , 88, 2006

17


Fluorescence tomography (3D-µXRF)

Pixel- by-pixel acquisition Sinogram(s) 2D-Slice or 3D-Volume

2D X-rayFocusing device

EDS

Sample



Pixel- by-pixel acquisition Sinogram(s) 2D-Slice or 3D-Volume

� The reconstruction problem is far more difficult compared to

transmission tomography

� self absorption corrections

� µ(Ea, x) is a priori unknown

� weak fluorescence signal for light elements

18



Pixel- by-pixel acquisition Sinogram(s)

Algorithmic solution:Optimal estimation of attenuation maps by combination of transmission, fluorescence and Compton tomographies

• B. Golosio et al., J. Appl. Phys. 94(1) (2003)

2D-Slice or 3D-Volume

Geometrical solution:Collimation of the detection angle to define a voxel: confocal geometry

• B. L. Vincze et al., Anal.Chem., 76(22) (2004)


Combining several signals

Absorption tomography (σσσσphoto ~Z4)�Absorption coefficient maps

Compton tomography (σσσσCompton ~Z)�Electronic density maps

Fluorescence tomography�Elemental composition maps

B. Golosio et al., J. Appl. Phys. 94 (1) (2003)

EDS

EDS

Integration of the informationAbsorption + Compton + Fluorescence

QUANTIFICATION

B. Golosio et al , APL 84 (2004)

19


Single Fly Ash Particle

Combined Tomographies

Transmission Tomography

Rubidium

Iron

Manganese

B. Golosio et al , Applied Phys. Letters , 84 (2004)


XRF Tomography – “confocal geometry”

Polycapillary

Coincidingfocii

FocussedX-ray beam

XRFdetector

� Spatial resolution: 5-10µm

L. Vincze et al., Anal.Chem. 76 (22) (2004)

Si(Li) detector

I0

Optical microscope

90°

Focusing lensSample

CCDAlignment

& imaging

ESRF - ID18F

20


From 3D-confocal XRF to quantification

unusual high Ca concentration

“Ca-rich lithology in the Earth's deep (> 300 km)

convecting mantle ”

Quantification using NIST SRM 613

CaMn

Fe

PbHf Th

SrZr

Y

Phase 1 (Sr)

Phase 2 (Y,Zr)

Zr

YScatter peaks

Ca: 260 ppm, Sr: 1.2 ppm, Zr: 1.1 ppm

Detection limits (LT=100 s):

B. Vekemans et al., JAAS. 19(10) (2004) F.E. Brenker et al., EPSL 236 (2005)

Y, Zr Sr

Th


X-ray Absorption Near Edge Spectroscopy (XANES)

• Local site symmetry

• Oxidation state

• Orbital occupancy

Energy

Nor

m. f

luo

yiel

d or

-al

ogI/I

0Continuumstates

Rydbergstates

σσσσ

σσσσ∗∗∗∗

ππππππππ∗∗∗∗

1sX-ray

Electronic transitions to bound states, nearly bond states or continuum

21


Chromium chemical mapping in cells

0

1(a.u.)

Potassium Cr (total)

Cr(VI)

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5.96 5.98 6 6.02 6.04 6.06 6.08 6.1

CrCl3PbCrO4

Tra

nsm

itte

d in

ten

sity

(a

.u.)

Energy (keV)

R. Ortega et al. , Chem. Res. Toxicol. , 5, (2005)

Micrograph

10 µm


Absorption and Phase contrasts

Absorption contrast

~ 1 / E3

Phase contrast

~ 1 / E

Refractive index n for X-ray wavelengths: n ~ 0.999999

n = 1 - δδδδ + i ββββ

phaseamplitude

Pha

se v

sam

plitu

de

0.1

1

10

100

1000

0.1 1 10 100

Hard X-rays

Soft X-rays

Energy (keV)

(δδδδ/ββββ) = f(E) for Al

See lecture 5 - P. Cloetens

22


X-ray microscopy techniques using phase shifting

Schmahl et.al(1993)

Zernike phase contrastwith phase shifting annular plate in back-Fourier plane of imaging objective

soft X-rays & keV photons)

Bonse / Hart(1968)

X-ray interferometerbased on double crystal in Laue geometry

keV photons

Morrison et.al(1996)

Differential phase contrastusing a quadrant detector similar to SEM

soft X-rays & keV photons

Polack / Joyeux(1995)

Kaulich / Wilhein(2002)

Differential phase contrastusing Young type slits or a Fresnel bi-mirror

Using 2 zone-plates

soft X-rays (keV photons)

SXM TXM

√√√√

√√√√

√√√√

√√√√

x

x

x x


Phase contrast in SXM ?

Phase gradientsdue to material inhomogeneities or thickness variations

Intensity variation ?

Differential Interference Contrast(DIC)

Phase effects+

interference fringes =

intensity variation

Differential Phase Contrast(DPC)

Beam deflection+

Configured detector =

intensity variation

2 options

23


DPC: SXM and configured detectorsPhase gradients due to material inhomogeneities or thickness variations

cause a deflection of the microprobe beam

� A single element detector would record the intensity but no information on the degree of deflection � Absorption only

� If the detector is split into several pixels, the beam deflection leads to intensity variation can detected� Phase information can be extracted.

See also Lecture 4 (M. Howells)

• STXM and TXM imaging modes can produce equivalent image contrast

• The roles of source and detector can be interchanged• STXM mode allows some imaging conditions to be realized more easily than in TXM mode,

• … and vice versa !

Reciprocity Principle


DPC and fast-readout CCD array detector

CVD DiamondID21

3.3keV100nm pixel size

5 µm

R(k)=1 R(k)=k x R(k)=k y

• 80x80pixels

• Spreads signal across many detector elements

• Extends the dynamic range for which photon counting is possible.

• Allows flexible choice of imaging modes (with simultaneous absorption, phase and dark field contrast)

G.R. Morrison et al., Journal de Physique IV 104 (2002)

W. Eaton et al., AIP Proceedings 507 ( 2000)

24


Configured segmented detector

• M. Feser et al., NIM A. 565 (2006)

• B. Hornberger et al. Ultramicroscopy 107(8) (2007)

• M.D. de Jonge et al., Microsc Microanal 13(2) (2007)

(4+5)-(6+7)Σ Σ Σ Σ segments

Ge siemens star525eV

X1A microscope NSLS

Diatom10keV

Beamline 2-ID-E APS


Differential Interference Contrast (DIC)

Visible light microscopy

G. Nomarski, 1960

The image of DIC microscopes is formed from the interference of two mutually coherent waves with lateral displacements (shear) of the order of the minimum size of the imaged structure and are phase-shifted relative to each other

• Two plane polarizers are involved, one before the condenser and one after the objective.

• Birefringent prisms act on the plane polarized light first to split the beam in two, and later to recombine the two beams.

The intensity distribution in measured DIC images is given by

a non linear function of the spatial gradient of a specimen’s

optical path length distribution along the direction of shear

25


Differential Interference Contrast (DIC)

Visible light microscopy X-ray microscopy

fZP1 - fZP2

G. Nomarski, 1960

• T. Wilhein et al., Appl. Phys. Lett . 78(14), (2001)

• T. Wilhein et al., Optics Communication 193, (2001)

• B. Kaulich et al., J Opt Soc Am A 19(4), (2002)

• B. Kaulich et al., Optics Express , 10(20) (2002)

• E. Di Fabrizio et al., Optics Express , 11(19) (2003)


X-ray Differential Interference Contrast (X-DIC)

∆s < δ

∆f < D.O.F.

� Shear of wave front division or distance of Airy disks in focal plane is smaller than optical resolution (“differential”)

λ=0.3nm

ffffZP1 - ffffZP2

• Distance ∆∆∆∆ffff of both ZPs smaller than depth of field

• Displacement ∆∆∆∆s smaller than resolution δδδδ

Bright field

X-DIC

26


X-ray Differential Interference Contrast (DIC)

10µm

Absorption

-0.01

-0.005

0

0.005

0 5 10 15

microns

C~ 1%

-0.3

-0.2

-0.1

0

0.1

0.2

0 5 10 15 20

microns

10µm

DIC

C > 28%

PMMA test objectID21

E=5keV

B. Kaulich et al., J Opt Soc Am A 19(4), (2002)



Development of multi-spot zone plates by reconstruction of the hologram of a plane wave and several spherical waves

(TASC-INFM, Trieste, Italy)

E. di Fabrizio et al., Optics Express , 11(19) (2003)

27



• E = 5.8keV (λλλλ=2.1Å)

• Dwell time: 50 ms/px

• Probe size: 0.6x0.7µm 2

20 µm

Absorption X-DIC (phase)

Cell membranes of maize plant cells

unpublished


X-DIC+XRF microscopies = Quantification?

Arabidopsisthaliana

X-DIC

Density+thickness map

µ-XRF

Intensity maps

5µm

quantification

high

low

28


Multi-keV X-ray microscopy

� Contrast mechanisms

• Phase contrast

• X-ray fluorescence

• Micro-spectroscopy (XANES)

Lower exposure dose

Trace element detectionQuantitative XRF analysis

Chemical state specificity

� Geometry

• Spatial res.: 1000nm → 30nm

• Field of view: variable

• Depth of field > 10µm

• Working distance > 20mmSample environment (cryo)In-situ

Multi-scale (meso-scale)

Tomography

� Sample

• Hydrated-frozen

• Thick

• Minimum preparation

Cryo is mandatory

No labeling

No sectioning


Needs for a multi- modal approach

Abundance Quality%

ppmppb structural

chemicalelemental

Scale

mm

µm

nm

Multi-modal analysis

Multi-variate data processing

In-situ

Quantification

3D

Detectors

Detection geometry

Source and optics

Coherence activities at the ESRF: A Practical Approach ... · Scanning (Transmission) X-ray...

Documents

Transcript of Coherence activities at the ESRF: A Practical Approach ... · Scanning (Transmission) X-ray...