NEXAFS Microscopy and Soft X-Ray Resonant Scattering · NC STATE University Ade Research Group (Polymer Physics/X-ray Characterization Techniques) Data: Araki, BL6.3.2. ALS, Berkeley

H. Ade et H. Ade et al.,ERLal.,ERL 20062006

NC STATE University Ade Research Group (Polymer Physics/XAde Research Group (Polymer Physics/X--ray Characterization Techniques)ray Characterization Techniques)

NEXAFS Microscopy and Soft X-Ray Resonant Scattering

Workshop: "Unique Opportunities in Soft Materials and Nanoscience with an ERL"

Cornell, June 19, 2006

H. AdeDepartment of Physics

North Carolina State Universityhttp://www.physics.ncsu.edu/stxm/

Thanks to J. Stoehr, G. Mitchell for their viewgraphs/dataand many other colleagues, collaborators, and users

Thanks to organizers for the invitation and supportand NSF, DOE, Dow Chemical for financial support over the years



Motivation and needs10-11 m 10-7 m10-9 m 10-5 m 10-3 m

nano-structurecrystallography structure

atomic structure

protein

polymers

micelle

bacteria

porous media

precipitates

DIFFRACTIONX-rays, n, e- SAXS, SANS

USAXS, USANS

Light scattering

TEM

Optical microscopySoft matter = lots of carbon!



Microscopy: Structure characterization in real space



Small Angle ScatteringCoherence length larger than domains,but smaller than illuminated area

log (intensity)40

-20

-40

0

20

-40 -20 0 20 40-40

-20

0

20

40

scattering vector q (μm-1)

scat

terin

g ve

ctor

q(μ

m-1

)Structure function:

informationabout domain

statistics.

Scattering/diffraction: Structure

characterization in reciprocal space

Spallation Neutron Source: total cost of $1.4 billion, http://www.sns.gov/



X-ray Microscopy has come a long way!(Examples “biased” towards spectromicroscopy/polymers/magnetic materials. – No bio or cryo)

1895

1993

2003

1 μm

1992

1993 1996

NEXAFS microscopy (STXM) XMCD microscopy (PEEM)

XLD microscopy (STXM) XMCD microscopy (TXM) Dynamics (PEEM)

Still a long way to go before we hit fundamental limit!!!



First NEXAFS imaging in transmission (1992)polypropylene/styrene-acrylonitrile

π* of styrene (C=C, 285.5 eV)

π* of acrylonitrile (286.8 eV)

(286.2 eV)

σ* C-H (287.9 eV)

H. Ade, X. Zhang, S. Cameron et al., Science 258, 972 (1992)

Based on the Stony Brook X1A-STXM (1989-present)

J. Kirz, H. Ade, et al., Rev. Sci. Instrum. 63 (1) (1992). C. Jacobsen, S. Williams, et al., Opt. Commun. 86, 351 (1991).

Stony Brook-STXM not explicitly built to perform spectroscopy, but proved flexible enough to perform carbon NEXAFS.

Polymer Science was “enabled”

Data: Stony Brook STXM at NSLS



X-Linear Dichroism Microscopy

Pattern rotates with rotation of polarization due to radial orientation of phenyl groupsPhenyl and carbonyl groups point (on average) radiallyoutwardπ* and σ* resonances show complementary dichroismDegree of orientation of various fiber grades can be quantitified

Kevlar 149 fibers, 200 nm thick, imaged at 285.5 eV

H. Ade and B. Hsiao, Science 262, 1427 (1993)

Photon Energy (eV)

280 285 290 295 300 305 310 315 320O

ptic

al D

ensi

ty0

1

2

3

4

5

6

7

Kevlar® 149

Kevlar® 49

N N C C

OO

HH][

n

EE

Spectra with horizontal polarization in locations indicated

A. P. Smith and H. Ade. Appl. Phys. Lett. 69, 3833 (1996)

-30°

38°

8°

Data: Stony Brook STXM

Determine degree of orientational order



e-

hν

C 1s edge ~ 290 eVN 1s edge ~ 405 eVO 1s edge ~ 540 eV

CH2 CH2

n[ ]

[ ]C C

n

CH2C

CH3

C

O

O

CH3

n[ ]

N

H

CC

C

C

CC N C

H

C

O

C

C

C

CC C

O

][n

Unoccupied Molecular Orbital

Near Edge X-ray Absorption Fine Structure (NEXAFS) SpectroscopyExample: “”Richness” of Polymer NEXAFS Spectra

Unsaturation C=C

Unsaturation C=O



Photon Energy / eV285 290

C OO

73O

C

O

27

vectra™

CH3

CH3

O CO

O n

PC

PAR

CH3

CH3

O CO C O

O

n

CCO O

NH

NH

n

Kevlar™

CCO

OO

O (CH2)4 n

PBT

PNIC

O

O

CH2 CH3

CH3

CH3

CH2 *CO

O*

n

CCO

OO

O (CH2)2 n

PET


PS

PBrS

SAN

n

78

Br

22

n

*

C N

m

CH2 NH CO

NHNH ** n

MDI polyurea

CH2 NH CO

ONHCO

OCH2 4* *n

MDI polyurethane

CH3

N

HN

H

C

O

n

TDI polyurea

CH3

NH

NH

CO

OC

O*

O

(CH2)4 *n

TDI polyurethane


CH2 CC

CH3

OO

CH3(CH2)3

n PBMA

CH2 CC

CH3

OO

CH3

n PMMA

* NH

C *

O

n

On PEO

Nylon-6

EPRCH3

* n

PP *CH3

n

PE * n

Fingerpt.ppt 8 May 1998

Some Polymer NEXAFS Spectra Dhez, Ade, and UrquhartJ. Electron Spectrosc. 128, 85 (2003)

Data: Stony Brook STXM at NSLS



Basis for Resonant Scattering Contrast:

Carbon edge examples

280 290 3000.000

0.001

0.002

0.003

0.004

0.005

β

E (eV)

PS PMMA P2VP

270 280 290 300 310

-0.003

-0.002

-0.001

0.000

0.001

0.002

δ

E (eV)

PS PMMA P2VP

Scattering factors f’ and f” (optical const. δ and β, respectively) show strong energy dependence

“Bond specific” scattering!Substantial potential as complementary tool!

NP2VPPS PMMA

I ∝Δδ2+Δβ2

Absorption (NEXAFS)

Dispersion


NC STATE University Ade Research Group (Polymer Physics/XAde Research Group (Polymer Physics/X--ray Characterization Techniques)ray Characterization Techniques)Sc

atte

ring

Inte

nsity

[a.u

]

q (1/nm)

PMMA/P(BA-co-S)

Idealized, “consensus” particle

“PS”“PS”

“PS”-”shell”

J.M. Stubbs, D.C. Sundberg Polymer 46 (2005) 1125

P(MA-b-MMA) / PS

Fuzzy TEMmodified core/shell structure?Phase less separated?Where really is the PS?

“PS” effective radius larger than PMMA “radius”

“PS” about same size than PMMA/acrylate!Scattering indicates PS is slightly more in center of nanoparticle relative to PMMA/P(BA-co-S)

Different process and composition

First results from structured polymer samples

(Figure courtesy J. Stubbs UNH)


10-1

100

101

102

103

0.120.080.040.00

200 nm

150 nm

100 nm

280.0 eV285.2 eV288.4 eV320.0 eV

10-1

100

101

102

103

0.120.080.040.00

qR=4.5q=0.39

qR=7.73q=0.68

qR=10.9q=0.95

R=115 nm

q (1/nm)

Scat

terin

g In

tens

ity [a

.u] 280.0 eV

285.2 eV288.4 eV320.0 eV




C OO

73O

CO

27

vectra™

CH3

CH3

O CO

O n

PC

PAR

CH3

CH3

O CO C O

O

n

CCO O

NH

NH

n

Kevlar™

CCO

OO

O (CH2)4 n

PBT

PNIC

O

O

CH2 CH3

CH3

CH3

CH2 *CO

O*

n

CCO

OO

O (CH2)2 n

PET


PS

PBrS

SAN

n

78

Br

22

n

*

C N

m

CH2 NH CO

NHNH ** n

MDI polyurea

CH2 NH CO

ONHCO

OCH2 4* *n

MDI polyurethane

CH3

NH

NH

CO

n

TDI polyurea

CH3

NH

NH

CO

OC

O*

O

(CH2)4 *n

TDI polyurethane


CH2 CC

CH3

OO

CH3(CH2)3

n PBMA

CH2 CC

CH3

OO

CH3

n PMMA

* NH

C *

O

n

On PEO

Nylon-6

EPRCH3

* n

PP *CH3

n

PE * n

Fingerpt.ppt 8 May 1998

Some Polymer NEXAFS Spectra Dhez, Ade, and UrquhartJ. Electron Spectrosc. 128, 85 (2003)

Data: Stony Brook STXM at NSLSLots of contrast from β and hence δ (Kramers-Kronig), for I ∝ Δδ2 + Δβ2

marry spectroscopy with structure from scattering

Reduced need to deuterate!



Low Contrast Example:Developmental Porous SiLK*

ResinsG. Mitchell, Dow Chemical

Decreasing feature size in integrated circuits require lower dielectric constants for insulating layers to:

increase circuit speed, decrease power consumption, decrease crosstalk

Adding pores to SiLK dielectric resin is used to decrease dielectric constant from 2.65 to 2.1 and lessFuture materials require more and smaller pores

* SiLK is a trademark of The Dow Chemical Company

G. E. Mitchell, I. Koprinarov, B. G. Landes, J. Lyons, B. J. Kern, M. J. Devon, E. M. Gullikson, and J. B. Kortright, Appl. Phys. Lett. (in press)



Systematic variation in porogen size

Smallest pores collapse

Wouldn’t be able to do the experiment with STXM, TEM, or SAXS. - Deuteration and SANS would work.

Data: Beamline 6.3.2 ALS



Data: Araki, BL6.3.2. ALS, Berkeley


Complementary Tool to Neutrons and hard X-rays

Soft X-ray Resonant ReflectivityPS/PMMA bilayer

Potential: Diffuse scattering from interfaces

Observed strong photon energy dependenceNeed to reduce uncertainty for δ and β

0.0 0.5 1.0 1.5 2.0

Fit Data

14.2 KeV

PS(15nm)/PMMA(42nm)

320 eV

288.5 eV

285.2 eV

283.4 eV280 eV270 eV

10-12

10-9

10-6

10-3

0

Ref

lect

ance

q (nm-1)

PS

Si

PMMA

air

270 280 290 300 310 320

1E-6

2E-6

3E-6

4E-6

5E-6 PS-PMMA Interface PS-Vacuum Interface

(c)

Ref

lect

ivity

(dδ2+dβ2)/4

C. Wang, T. Araki, H. AdeAppl. Phys. Lett. 87, 214109 (2005)



NEXAFS microscopy and Resonant Scattering

Unique tools for soft condensed (carbonaceous) matter



Brighter Sources!How do we take advantage of them? Space, time, energy?

How about my favorite absorption edge??Carbon, 300 eV?



Types of X-ray

Microscopes

Sample

MagnifiedImage

X-ray Beam

Objective LensProjective Lens

Screen

MCP

Aperture

X-Photoemission Electron Microscope

PEEM

ALS 5.3.2 and 11.0.2 STXM

TXM

Also: Scanning Photoemission X-ray Microscope (SPEM) Ade et al. Appl. Phys. Lett. 56, 1841 (1990)See talk by Maya Kiskinova for details

Best for NEXAFSRelatively slow

FastLimited NEXAFS

Coherent illumination



Zone Plates: Diffractive Optics for X-raysDeveloped by groups in Goettingen/Bessy, LBNL, Stony Brook/Lucent, a few others

Circular Au, Ni, or Ge structuresNanofabrication with E-beam lithographySpatial resolution equal approx. outermost zone widthFocal length is proportional to energy

Diameter: 150-250 μmZone widths presently as small as 15 nmWorking distance: ~150 μm

Image from http://www-cxro.lbl.gov/microscopy/diffractive.html

1975: Modern X-ray Microscopy starts: First TXM at DESSY. B. Niemann, D. Rudolph and G. Schmahl

1978: Move to ACO1982: Move to BESSY-I

1982: First STXM at NSLS: Rarback and Kirz120 nm resolution

stop

zones

first order radiation

zero order radiation

first order focus

OSA

f

Arrangement in a Scanning Transmission X-ray Microscope (STXM)



Position (nm)0 200 400 600 800 1000

Line

pro

file

inte

nsity

(nor

mal

ized

uni

ts)

0

1

2

360 nm1:2

40 nm1:1

30 nm1:1

Position (nm)0 200 400 600 800 1000

Line

pro

file

inte

nsity

(nor

mal

ized

uni

ts)

0

1

2

360 nm1:2

40 nm1:1

30 nm1:1

30 nm 1:1

11% contrast

24% contrast

5.3.2 Polymer STXM , Fall 2001, “35 nm zone plate”Soon to install “25 nm zone plate”

A.L.D. Kilcoyne et al. J Synchrotron Rad. 102, 125-136 (2003)

Recent STXM Technology Innovations:1) Interferometer, 2) polarization control

New 25 nm CXRO zone plate test:25 nm lines of a test pattern obtained at 395 eV. * Polarization control

BL11.0.2.2 STXMSpring 2004, “25 nm zone plate”

Impact: 1) Zone plate are now the spatial resolution limiting factor, 2) enables magnetism, 3) quantification, wider range of samples

Top view

Detector

OSA mount Sample mount

X-interferometer mirrors



TXM: ALS Beamline 6.1.2 XM-1 Spatial resolution record

Weilun Chao, B. D. Harteneck, J.A. Liddle, E. H. Anderson, and D. T. Attwood, Nature 435, 1210 (2005)

15 nm zone width, 1.7 nm placement accuracy

40 nm period test pattern

Imaged with 15 nm ZP

Imaged with 25 nm ZP

30 nm period test pattern

15 20 25 30 35 40 45 50 550.0

0.2

0.4

0.6

0.8

1.0

15

1/period (um-1)half-period (nm) 25 20 12.5 10

Nor

mal

ized

Imag

e M

odul

atio

n ΔrMZP=25nmσ =0.21 to 0.42

Calculated Measured

ΔrMZP=15nmσ =0.19 to 0.38

Calculated Measured

http://www-als.lbl.gov/als/aboutals/virtualtour6.1.2.html



Volume zone plates with tilted structuresVolume zone plates with tilted structures

diffraction efficiency η1(drn)Gerd Schneider, BESSY, Berlin

Volume zone plate very difficult to makeAspect ratio for low efficiency zone plates is easier to achieveERL eliminates need for high efficiency zone plates

Low efficiency is compensated for by increased brightness

Alternative: Use third order: 1/9th the efficiency, 3x better spatial resolution

How do we get to <10 nm resolution?

200 nm Ni520 eV

2%

Thin ZP plates

http://www-cxro.lbl.gov/optical_constants/tgrat2.html

1%



Relative ZP/sample vibrations (5.3.2 STXM)

10 nm

10 nm

X-axis

Y-axis

Photon Energy (eV)

360 365 370 375 380 385 390

Alig

nmen

t shi

ft (n

m)

-50

-40

-30

-20

-10

0

10

20

30

X axis shiftX Axis shift linear fitY axis shiftY axis shift linear fit

Auto-correlate images to detect shiftsRandom shits less than 30 nmsome residual misalignment in y

Image-sequence alignment

A.L.D. Kilcoyne et al. J Synchrotron Rad. 102, 125-136 (2003)

Instrument challenges for STXM



Time resolution real space: STXM

Recall: Mechanical scan of sample or optics“Scan speeds” presently as low as 100 μs/20 nm pixel1000x1000 pixel image in 2 minutes

Difficult to imagine technology that can really take advantage of ERL for imaging speed alone

Pump probe on radiation hard materials (e.g. magnetic materials) in STXM and PEEM



Sine excitation @ STXM H. Stoll et al. MPI, Stuttgart

(sample prep.: I. Neudecker, D. Weiss, C.H. Back, Regensburg Univ.)

B0 ≈ 10 mT

f = 250 MHz

B0 ≈ 16 mT

f = 250 MHz

30°

H

Permalloy sample A(1.5 x1.5) µm², 50 nm

thick

I

H

B = B0 sin (2π f t)

Photon

Change of vortex core chirality (handedness)



Use ERL for improved energy resolution?

NEXAFS already limited by core-hole lifetimeResonant Inelastic X-ray Scattering for improved spectroscopy

But, damage!Need to back off on spatial resolution

Photon Energy (eV)

280 285 290 295 300 305 310 315 320

Opt

ical

Den

sity

0

1

2

3

4

5

6

7

Kevlar® 149

Kevlar® 49

N N C C

OO

HH][

n

Spectra with horizontal polarization in locations indicated

-30°

38°

8°

Determine degree of orientational order



Radiation DamageTypical critical dose ~ 1000 eV/nm3

Rose criterion for detection: S/N=5Transmission, S=-ln(I/I0)(10 nm)3 PS feature has S/N of 5 for 650 incident photons, and 115 absorbed photons in pi* peak at 285 eV

Dose= ~32 eV/nm3

Quantitation/spectra/tomography typically require 100x dose

Near critical at 10 nm for PSDose ∝ 1/(Δα)

Imaging using phase contrastScattering using phase and absorption contrast

T. Coffey et al. J. Electron Spectroscopy 122 (2002) 65 –78



Data: Beamline 6.3.2 ALS

What about scattering?

I ∝ Δδ2 + Δβ2Works with low contrast

Q-range ~ 3 nm-1 (@300 eV)Dose is distributed!!Time resolution could be quite good“SAXS” not intrinsically a brightness driven experiment

Combine low resolution STXM and SAXS: Microbeam scattering

Damage?



“Incoherent“ vs. Coherent X-Ray ScatteringSmall Angle ScatteringCoherence length larger than domains,but smaller than illuminated area

SpeckleCoherence lengthlarger than illuminated area

log (intensity)40

-20

-40

0

20

-40 -20 0 20 40-40

-20

0

20

40


scat

terin

g ve

ctor

q(μ

m-1

)

log (intensity)40

-20

-40

0

20

-40 -20 0 20 40-40

-20

0

20

40


scat

terin

g ve

ctor

q(μ

m-1

)

informationabout

domainstatistics

trueinformation

about domain

structure

Figure courtesy J. Stohr

Resonant scattering and resonant coherent speckle/imaging should be excellent tools for many materials !



New tools(Jan Luning’s talk)

TransmissionX-ray

Microscope

“Reconstruction”from

Speckle Intensities

∅ 5 μm(different areas)

Coherent X-Ray Imaging

Phase problem can be solved by “oversampling” speckle image

S. Eisebitt, M. Lörgen, J. Lüning, J. Stöhr, W. Eberhardt, E. Fullerton (unpublished)

Technique of choice for use at ERLs

damage limit?

Figure courtesy J. Stohr

For more recent data see. Eisebitt et al. Nature 432, 885 (2004)



Conclusions

NEXAFS imaging, Resonant ScatteringHigh, tunable contrast: Compositional, i.e. “bond specific” sensitivity

Technology: STXM, PEEM, SPEM, scattering, speckleGood progress over the last 20 yearsThese techniques will benefit greatly from new sources, more capacity, more competition, better zone plates

New zone plate optics strategySTXM with spatial resolution < 10 nm possible, but challenging

Higher energy offer relatively larger payoff (coherent volume ∝ λ2)Radiation damage a limitation in absorption modeMicrobeam scatteringResonant coherent imaging in phase mode is good idea (Jan Luning’stalk?)

Resonant photon correlation spectroscopyPump probe experiments will have better time resolution



Thank you for your attention!!!!

Post Doc hire by this fall



Resonant scattering/reflectivity: ConclusionsHigh, tunable contrastCompositional, i.e. bond specific, sensitivityMultiple modes: Scattering and reflectivity

Soft X-ray should be good complement to neutrons and hard x-raysGood complement to TEM, SPM, STXM, SANS/SAXS

Reality check!???

Future:Contrast from orientation! Control of sample environmentResonant coherent imaging?

NEXAFS microscopy sensitive to orientation


NC STATE University Ade Research Group (Polymer Physics/XAde Research Group (Polymer Physics/X--ray Characterization Techniques)ray Characterization Techniques)NC STATE University Ade Research Group (Polymer Physics/XAde Research Group (Polymer Physics/X--ray Characterization Techniques)ray Characterization Techniques)

Can use large angles, hence, get good q-range

•At 280, 285.2eV 288.8eV ----- bilayer signature•283.2eV --- only single PMMA layer•320 eV and 8 keV --- essentially only full layer signal

Rapid changes as the function of photon energy.

4)1()1()()( 222

2211

121221212

βδβδβδ

ββδδ Δ+Δ≅

−−+−−−+−

≅=ii

irR

Reflectivity at the PS/PMMA interface is related to the contrast between PS and PMMA

Soft X-ray Resonant Reflectivity

0.0 0.5 1.0 1.5 2.0 2.5 3.0

8 keV

283.4 eV

320 eV

288.8 eV

286.4 eV

285.2 eV

280 eV

PS/PMMA 10/30 nm

Ref

lect

ivity

q (nm-1)

C. Wang, T. Araki, H. Ade

Simulations

270 280 290 300 310 320

1E-6

2E-6

3E-6

4E-6

5E-6 PS-PMMA Interface PS-Vacuum Interface

(c)

Ref

lect

ivity

(dδ2+dβ2)/4

-3E-3

-2E-3

-1E-3

0E-3

1E-3

2E-3

(b)

δ PS PMMA P2VP

1E-3

2E-3

3E-3

4E-3

(a)

β

PS PMMA P2VP



dPS/PMMA has larger interface than hPS/PMMA• Implications for neutron work, comparison to theory, capillary waves, etc.

PS/PMMA interface: hPS and dPS



Direct comparison to Neutron Reflectivity

Method Bilayers D (nm)dPS

D (nm)PMMA

σrms (nm) dPS/Air

σrms (nm)dPS/PMMA

dPS(50nm)/PMMA(200nm) 43.3 187.3 0.55 1.48

dPS(100nm)/PMMA(200nm) 98.5 189.3 0.77 1.48

dPS(50nm)/PMMA(200nm) 41.7 188.8 0.43 1.34

dPS(100nm)/PMMA(200nm) 96.5 190.3 0.49 1.46

Neutron Reflectivity

RXR

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.61E-9

1E-7

1E-5

1E-3

0.1

10

Neutron Reflectivity fit data

dPS(100nm)/PMMA(200nm)

dPS(50nm)/PMMA(200nm)

Ref

lect

ance

q (nm-1)

σrms hPS/PMMA =1.16 nm



ConclusionsOpportunities

Technology: STXM, PEEM, SPEM, scattering, speckleTremendous progress over the last 20 yearsClearly all these techniques will benefit greatly from new sources, more capacity, more competition, better zone plates

SciencePolymers! - Yes, but radiation damage might set a limit♦ Polymer-inorganic bybrids♦ Templating

Magnetic materialsNanoscience in generalBiology

and most importantly new and exited users

Documents

NEXAFS Microscopy and Soft X-Ray Resonant Scattering · NC STATE University Ade Research Group (Polymer Physics/X-ray Characterization Techniques) Data: Araki, BL6.3.2. ALS, Berkeley