1

Chemical and magnetic imaging with x-ray

photoemission electron microscopy (XPEEM)

Andrea Locatelli

Elettra – Sincrotrone Trieste, Italy

9/23/2013 1

9/23/2013 2

Why do we need photoelectron microscopy?

• To combine SPECTROSCOPY and MICROSCOPY to characterise the structural, chemical and magnetic properties of surfaces, interfaces and thin films

• Applications in diverse fields such as surface science, catalysis, material science, magnetism but also geology, soil sciences, biology and medicine.

BiologyMagnetismSurface Science

2

Outline

• Synchrotron radiation and x-ray spectro-microscopy: basics

• Cathode lens microscopy: methods

• Applications

– Chemical imaging: reactivity, characterization of nano-and micro- structured materials

– microXPS, microARPES: examples in Graphene research.

– Biology

– Magnetism

– Time-resolved XPEEM

9/23/2013 3

Why does PEEM need synchrotron radiation?

• High intensity of SR makes measurements faster

• Tuneability – very broad and continuous spectral range from IR to hard X-Rays

• Narrow angular collimation

• Coherence!

• High degree of polarization

• Pulsed time structure of SR – This adds time resolution to photoelectron spectroscopy!

• Quantitative control on SR parameters allows spectroscopy:• Absorption Spectroscopy (XAS and variants)

• Photoemission Spectroscopies (XPS, UPS, ARPES, ARUPS)

9/23/2013 4

),,,;,,,(ee

e

kinEhfJ ϕθσεν ΦΘ=

3

X-ray Absorption Spectroscopy

9/23/2013 5

)(ˆˆ42

2fffi

Er ρψεψωαπµ ⋅= h

The measurement of the absorption warrants:

� Elemental sensitivity.

� Sensitivity to final state

� Sensitivity to valence state, bondorientation, nearest-neighbour

� Magnetic sensitivity

� Buried layer and interfaces accessible

• SR-PEEM can easily implement XAS and its variants (NEXAFS, XMC/LD) by measuring the secondary electron yield as a function of the varying photon energy or photon helicity.

• NO ENERGY FILTER IS NEEDED IN PEEM

Resonances arise from transitions from core levels into unoccupied valence states via excitation processes occurring during the filling of the core hole.

X-ray Photoelectron Spectroscopy

9/23/2013 6

Φ−−= Bk EhE ν

The absorption of a photon ionizes the system, exciting one of the electrons into a free state in the continuum. The transition probability from the initial state to the final is proportional to:

)2

(ˆˆ4222

2

m

kEere

i

jkr

iik

hhh −+⋅∝ ωδψωαπτ

We measure the energy distribution of the photoelectrons emitted from the specimen.

EB is the binding energy of the initial state. EB is a unique fingerprint that allows determining the composition and chemical state of the specimen surface. This is the founding principle of ESCA and XPS (K. Siegbahn’s Nobel prize).

Main features: Elemental and chemical sensitivity (surface core level shifts), sensitivity to the electronic structure, sensitivity to local structure (micro-XPD), highest surface sensitivity

XPS mode: hv consthv in / e- out

Energy filter required!

4

Angle-Resolved Photoelectron Spectroscopy

9/23/2013 7

if

if

kk

hEE

≈+= ν

Another method, also possible using many XPEEM instruments, is that of measuring the momentum of the initial state. The electron analyzer allows us to measure both E and k of the final state, and thus to fully determine the initial state k.

When escaping into vacuum, the k component parallel to the surface is conserved.

2/2 hkinkinf

ikin

mEkk

hEE

==

Φ−+= ν

if kGk ////// =+

Φ−−=−−= ⊥⊥

kinb

kin

EhE

GWEmk

ν

20 /)(2 h

Φ−−=−=

kinb

ekin

EhE

WEmk

νθsin/)(2 2

0// h

3D ordered crystalline system 2D ordered crystalline system

In normal emission k// = 0

Examples

The two main approaches of x-ray microscopy

• Scanning: sequantial indirect imaging

• Lateral resolution determined by X-ray (diffractive) optics: 20-30 nm.

• Combination with TXM

• Excellent spectroscopic ability

• High pressure variants do exisits

• Rough surfaces

9/23/2013 8

e-

sample hv

electron opticswith energy filter

detector

hv

x-ray optics

sample scanning

stage

x-raydetector

electron analyzer

Scanning photoemission electron microscopy (SPEM)

X-ray photoemission electron microscopy (XPEEM)

• Direct imaging, parallel detection

• Lateral resolution determined by electron optics: with AC, few nm possible

• Combination with LEEM/LEED

• Intermediate spectroscopic ability

• Pmax < 5·10-5 mbar

• Flat surfaces

5

The probing depth is tuned by the electron energy

Inelastic mean free path (“universal curve”)

9/23/2013 9

By measuring

secondary Electrons

we probe thin films and buried interfaces to a maximum depth of several nm. This is the case of X-ray absorption spectro-scopy and its variants (NEXAFS, XMCD, XMLD).

By measuring Photoelectrons

emitted from core levels or the

valence band (XPS, ARPES, UPS, ARUPS) we achieve sensitivity to the topmost surface layers, especially when the K. E. Is in the range 50to 150 eV.

SR tuneability & photoionization cross sections

9/23/2013 10

Choosing the best photon energy for the experiment is of crucial importance to maximize surface / elemental sensitivity as well as obtaining favourable acquisition times

Graphene / Ru(0001)

Yeh and Lindau, Atomic Data and Nuclear

Data Tables 32, l-l 55 (1985)

6

Synchrotron Radiation basics

Beamline design

9/23/2013 11

Photon Sources at synchrotrons: undulators

9/23/2013 12

Insertion devices at the Elettra Synchrotron: Courtesy: B. Diviacco( ) nN

BBmc

e

n

uu

n

/1/

polesof numberN with ,N Intensity

1K [T];[cm]934.02

K

E(GeV)1957

)2

K1(

2

2

222

20

=∆=∝

<==

⋅=

++⋅=

λλ

λλπ

γ

θγγλλ

� High flux and wide energy range

� Horizontal, vertical and elliptocal polarization available

a period array of dipole magnets with alternating field direction normal to the plane of the ring;

γα /1<

7

Undulator spectrum

9/23/2013 13

An XPEEM Beamline layout: Nanospectroscopy

9/23/2013 14

Main concept: medium energy resolution, high flux beamline (1012-1013 ph/sec)

Highest flux density on the sample, micro-focus on sample!

Monochromator VLS grating design allows versatility, covering a wide photon enery range with few gratings.

8

Refocusing Optics

• Requirements

– Independent horizontal and vertical focusing

– Homogeneous micro spot

• Solution: Bendable Mirrors

– Adaptive mirrors (or “bendable” mirrors): a plane (or spherical) surface is bent to form an elliptical surface by applying unqeual moments to the mirror ends.

– Kirkpatric-Baez configuration

– Spot size: 2x6 μm2 HxV

9/23/2013 15

horizontal + verticalfocalization

P

Q

Kirkpatrick-Baez

A. Bianco, G. Sostero, and D. Cocco:Proc. SPIE 4782, 74 (2002).

Cathode lens microscopy methods

PEEM, LEEM, SPELEEM, AC-PEEM/LEEM

9/23/2013 16

9

PEEM basics: operating principle

9/23/2013 17

PEEM is a full-field, direct imaging technique. The microscope images a restricted portion of the specimen area illuminated by x-ray beam. Photoemitted electrons are collected at

the same time by the optics setup, which produces a magnified image of the surface. The key element of the microscope is the objective lens, also known as cathode or

immersion lens, of which the sample is part

dDiffdSP

dCH

The cathode lens

9/23/2013 18

1. In emission microscopy θ (emission angle) is large. Electron lenses can accept only small θ because of large chromatic and spherical aberrations

2. Solution of problem: accelerate electrons to high energy before lens → Immersion objective lens or cathode lens

Example for E = 20000 eV:E0 2 eV 200 eV

α for α0 = 45o 0.4o 4.5o

αααα 0 αααα

E

n sinθ = constn ∼ √E

θ → αsin α/sin α0 = √ E0/E

3. The aberrations of the objective lens and the contrast aperture size determine the lateral resolution

d = √ dSP2 + dCH

2 + dD2

dD = 0.6 λ/ rA

10

The different types of PEEM measurements

PEEM Probe Measurement

• threshold microscopy Hg lamp photoelectrons

• Laterally resolved XPS, micro-spectroscopy X-ray core levels or VB ph.el.

• Laterally resolved UPS, microprobe ARUPS and ARPESX-rays, He lamp VB photoelectrons

• Auger Spectroscopy X-ray, or electrons secondary electrons

• XAS-PEEM (XMC/LD-PEEM) X rays secondary electrons

9/23/2013 19

Require

energy filter

Simple PEEM instruments

9/23/2013 20

11

PEEM instrments with energy filter: NanoESCA

9/23/2013 21

9/23/2013 22

Low energy electron microscopy

LEEM is a cathode lens microscopy technique which probes surfaces and interfaces with low energy electrons, using the elastically backscattered beam for imaging. LEEM is particularly well suited to monitor dynamic processes, such as surface reconstructions, epitaxial growth, step dynamics, self-organization

E. Bauer: Low Energy Electron Microscopy, Rep. Prog. Phys. 57 (1994) 895-938.

12

Which properties does LEEM allow to probe?

9/23/2013 23

Different contrast mechanisms are available in LEEM for strucutre characterization

geometricphase contrast

STEP MORPHOLOGY

Mo(110)

quantum sizecontrast

d

FILM THICKNESS

Co/W(110)

diffractioncontrast

sample

contrastaperture

objective

[0,0]

[h,j]

SURFACE STRUCTURE

)

Imaging dynamic processes at surfaces

[001]

very favourable backscattering cross section at low electron energy!

Ni growth on W(110): formation of a striped phase above 1 ps ML Ni

Ni growth on W(110): step flow and completion of ps ML

540 < T < 750 C

13

sample

separator

analyzer

e-gun

SPEL

EEM

inst

rum

en

t o

n N

ano

spe

ctro

scop

y, e

lett

raLo

cate

lli e

t al

, Su

rf. I

nte

rfa

ce A

na

l. 3

8, 1

554

(20

06).

LEEM (low energy electron microscopy) Structure sensitivity

SPELEEM microscope (XPEEM + LEEM)

XPEEM (x-ray photoemission electron microscopy) Chemical and electronic structure sensitivity

Graphene formation on Ir(100)• Growth: large crystals for T > 800 C,

P=2·10-8 mbar ethylene.

• lattice gas � (1x5) lifted � enucleation of graphene islands

• Upon cooling a second distinct phase forms, appearing at about 520 °C.

9/23/2013 26

Fov 10 um, S.V. 6 eV, then 13 eV

Fov 4 um, S.V. 13 eV

14

Spectroscopic Photoemission and low energy Electron Microscope

9/23/2013 27

microprobe-diffraction

ARPES / LEED

microprobe-spectroscopy

XPS

Spectroscopic imaging

XPEEM / LEEM

Microprobe measurements are limited to surface areas of 2 microns in diameter!

Spectroscopic Photoemission and low energy Electron Microscope

9/23/2013 28

Performance: lateral resolution in imaging: 10nm (LEEM)

30 nm (XPEEM)

energy resolution: 0.3 eV (0.2 in microprobe spectroscopy)

Key feature: multi-method instrument to the study of surfaces and

interfaces, based on imaging and diffraction techniques.

Probe: low energy electrons (0-500 eV) structure sensitivi ty

soft X-rays (50-1000 eV) chemical state, magnetic s tate

electronic structure

Applications: characterization of materials at microscopic level

magnetic imaging of microstrucutres

dynamical processes

15

Correction of spherical and chromatic aberrations

9/23/2013 29

focalpoint

focalpoint

Round convex lenses

Chromatic aberration

Spherical aberration

Round concave lenses

Light optics

Correction of spherical and chromatic aberrations

9/23/2013 30

focalpoint

focalpoint

Round convex lenses

Chromatic aberration

Spherical aberration

Round concave lenses

Electron optics

V.K. Zworykin et al, Electron Optics and the Electron Microscope, John Wiley, New York 1945

Electron Mirror

16

9/23/2013 31

Resolution limit without correction

with correction

Spherical α3 + … α5

Chromatic ∆E α + … ∆E α2

+ ∆E2 α

Diffraction 1/α 1/α

ααααd

D. Preikszas, H. Rose, J. Electr. Micr. 1 (1997) 1Th. Schmidt, D. Preikszas, H. Rose et al., Surf.Rev.Lett 9 (2002) 223

Improvement with mirror corrector

Simultaneous improvement in

Transmission and Resolution!!!

9/23/2013 32

XPEEM-LEEM with aberration correction: SMART

17

9/23/2013 33

First results of the SMART microscope @BESSY

Courtesy of Th. Schmidt et al.; 5th Int. Conf. LEEM/PEEM, Himeji, 15.-19. Oct. 2006

50 nm

-2 0 -15 -1 0 -5 0 5 1 0 15 2 0

170

180

190

200

210

220

230

240

250

inte

nsity

d ista nce (nm)

3.1 nm

Atomic steps on Au(111), LEEM 16 eV, FoV = 444 nm x 444 nm

(18.09.06)

Lateral resolution limitations: space charge

9/23/2013 34

XPEEM: photocurrent

estimate for Au

440 bunches

rev. frequency: 1.157 MHz

bunch length: 42 ps (2GeV)

• 1 1013 ph./s on sample =

= 20000 ph./bunch

• Total photoionization yield: about 2% photons result in a photoemission event

• I peak ≈ 400 e-/ 42 ps

≈ 1.5µA vs 20 nA (LEEM)

13 pA/μm2 versus 20 nA/μm2

1. Image blur can be observed with SR, but only under very high photon fluxes. Keep into account in beamline design.

2. Both the lateral and energy resolution are strongly degraded by Boersch and Loeffler effects occurring in the first part of optical path.

Ultramicroscopy 111, 1447 (2011).

18

Space charge in XPEEM and microXPS with FEL radiation

SiO2/Si

patterns

FEL-PEEM

X-PEEM

5 µm16

14

12

10

8

6

4

2

0

1086420-2

kinetic energy (eV)

exit slit 80 µm exit slit 40 µm exit slit 20 µm

liner polarizationgap1 = 23.56gap2 = 41.4phase1 = 0phase2 = 0energy: 9.1 eV

circular

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0

1086420

kinetic energy (eV)

x30

Ag/W(110)

X-PEEM

CHG-PEEM

Inte

nsity

(arb

. uni

ts)

L. Gregoratti, T.O. Mentes, A. Locatelli, M. Kiskinova,

J. Electron Spectrosc. Relat. Phenom. 170, 13-18 (2009) .

XPEEM Applications

Chemical imaging

9/23/2013 36

19

Au/TiO2(110): controlling growth by substr. stoichiometry

9/23/2013 37

Creation of ordered oxygen vacancies

MEMWork Function

(1x1) (1x2)

Stochiometric Irradiated

µ-LEEDstructure

Condens. Matter 19, 082202 (2007); Phys. Rev. B 76, 155413 (2007).

Collaboration with A. Kijena and T. Pabisiack (Wroclaw)

Irradiation at 720 K13 pA/μm2

Structure of the (1x2) TiO2

micro-LEED/IV

ongoing collaboration wih G. Held and Z.V. Zheleva

University of Reading

Au growth on TiO2(110)

1x2

1 ML

1x1 XPEEM@ Au 4f

µ-XPS

9/23/2013

Surface Oxygen on Ag : e-beam “Lithography”

Ful l oxidation of Ag using NO2 does not occur:

Low T: NOad stays , prevents oxidation.

High T: NOad desorbs,

but Ag2O unstable.

LEED reveqas path towards Ag2O under e-beam

S. Günther et al., Chem. Phys. Chem. 2010.

Ins tead: e-beam (60 eV) stimulated desorption of NOad works at RT!

S. Günther et al., App. Phys. Lett. 93, 233117 (2008).

A: metallic Ag

B: Ag2O

NO2 � NOad+Oad

20

9/23/2013

Surface Oxygen on Ag : photon-beam “Lithography”

S. Günther et al., Chem. Phys. Chem. 2010.

hv = 40 eV

(a) Start of NO2 adsorption, t = 0 s,

(b) t = 210 s, p(NO2) =1.8×10-7 mbar, 17 L NO2,

(c) t = 540 s, p(NO2) =2.5×10-7 mbar, 67 L NO2.

-9 eV 4 eV

MEM 28 μm x 350 μm; after 130 L NO2;

39

9/23/2013 40

Reactivity and electron confinement in Mg thin films

Thickness dependent reactivity? Film oxidation

inte

nsity

(a.u

.)

8 6 4 2 0binding energy (eV)

clean

~ 4.5L

~ 9L

~ 11L

~ 13L

valence band9ML Mg/W(110)

hν ν ν ν = 56 eV

Oxy

gen

dose

Inte

nsity

(a.

u.)

-4 -3 -2 -1 0 1E-EF (eV)

-6-4-2

E-E

F

5

6/7

8

9

10

11

12

EF

Γ Α Γ

SS

QWSQWSQWS

atomiclayers:

L. Aballe et al., Phys. Rev. Lett. 93, 196103 (2004)

Inte

nsi

ty (

Arb

. Un

its)

Inte

nsi

ty (

Arb

. Un

its)

Binding Energy (eV)

21

9/23/2013 41

Quantitative characterization of film oxidation

Iox/I tot

inte

nsi

ty (

a.u

.)

-54 -52 -50 -48 E-EF (eV)

-54 -52 -50 -48

Mg2phνννν = 112 eV

clean

~ 6L

~ 9L

~ 11L

~ 13L

7 ML 9 ML

O2e

xpo

sure

13

109

56

77

7

5 6

7

8

9

8

9

9

10

11

87

912

11

1310

12

15

10

9

11

12

11

14

10

7/8

9

12

9 10

LEEM reveals morphology

atomic thickness

1 µm

oxide component imaged by XPEEMreveals chemistry!

109

65

5

7 7

6

7

7 - 8

108-9

9

11

1012

7

89

6-8

7

11

13

13

12

15-14 12

9-10

11

6-812

9-10

8

L. Aballe et al., Phys. Rev. Lett. 93, 196103 (2004)

9/23/2013 42

Oxidation of Mg film and QWR

FACTS

� Strong variations in the oxidation extent are correleted to thickness and to the density of states at EF

� XPEEM is a powerful technique for correlating chemistry and electronic structure information

SIGNIFICANCE OF THE EXPERIMENTS

� Control on film thickness enables modifying the molecule-surface interaction

� Theoretical explanation: Decay length of QWS into vacuum is critical: it reproduces peak of reactivity in experimental data. See Binggeli and M. Altarelli, Phys.Rev.Lett. 96, 036805 (2005)

oxi

dat

ion

ext

en

tD

OS

at E

F

L. Aballe et al., Phys. Rev. Lett. 93, 196103 (2004)

22

9/23/2013 43

Spatio-temporal patterns in surface chemical reactions

Pattern formation

in surface chemical reactions

Jakubith et al., PRL 65, 3013 (1990)

rotating spirals

standing fronts

target waves

Belousov-Zabatinski reaction

(solution of, acidified bromate,

malonic acid, ceric salt)

See also: W. Engel, et al., Ultramicroscopy 36, 148–153 (1991).

9/23/2013 44

Reaction diffusion patterns: NO+H2 /Rh(110)

• First quantitative measurement of concentration profiles

Schaak et al Phys. Rev. Lett. 83, 1882 (1999)

O 1s

N 1s Rh 3d5/2

WF

Th. Schmidt et al.,

Chem. Phys. Lett. 318, 549 (2000)

(LEEM, micro-LEED )

23

9/23/2013 45

Reaction diffusion patterns: NO+H2 /Rh(110)

Structure + composition

(LEEM, micro-LEED )

Reactive phase separation processes

9/23/2013 46

F. Lovis et al., J. Phys. Chem. C 115, 19149 (2011)

Spectroscopic determination of the oxidation state

LEEM XPEEM Au 4f 7/2

2µm85.0 84 .0 83.0 82.0binding energy (eV)

Spectroscopic determination of reaction inducedredistribution

Au/Rh(110) during water formation reaction V/Rh(110) during water formation reaction

24

XPEEM Applications

Graphene and ARPES

9/23/2013 47

Micro-XPS mode - an example: C lattice-gas on Ru

9/23/2013 48

LEEM: changes in reflectivity allow to determine C lattice gas concentration

E. Loginova et al., N.J.Phys. 10, 093026 (2008)

XPEEM: direct measurement of C 1s spectrum

Guenther, Locatelli, Mentes; unpublished

Different BE for C 1s in Gr and C adatoms

Quantitative analysis!

25

Microprobe ARPES: Gr/Pt(111)

9/23/2013 49

Weak coupling! No hybridization of graphene’s π band with substrate d

states

Only Pt

Single layer Graphene on

Pt

Sutter et al., Phys Rev. B 80, 245411 (2009)

Micro-LEED

LEEM ARPES

microprobe-LEED

Graphene/Ir(100): existence of two distict phases

9/23/2013 50

–(b)

BrightfieldLEEM

IrBG FG

max

0

<010>

<001>

1 µm darkfieldLEEM

BG

FG

FG: flat graphene

BG: buckled graphene

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

26

Graphene/Ir(100): existence of two distict phases

9/23/2013 51

BrightfieldLEEM

IrBG FG

max

0

<010>

<001>

1 µm darkfieldLEEM

BG

STM

FG

FG: flat graphene

BG: buckled graphene

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

Ab-initio results: buckled graphene

9/23/2013 52

<100>

<010>

buckled graphene unit cell

Buckled Graphene

Exceptionally large buckling

GGA:

� Min Ir-C distance of 1.9 Å

� Max Ir-C distance of 4.0 Å

DFT-D:

� Min Ir-C distance of 2.1 Å

� Max Ir-C distance of 1.7 Å

� 18 atoms over 160 (i.e. 11%) are chemisorbed, the others are physisorbed

5.5 Ir units = 21.12 Å

5 Ir

uni

ts =

1.9

2 Å

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

27

Laterally resolved XPS

9/23/2013 53

STM (200x200) RT

<001>

C 1s spectro-microscopy

chemisorbed

physisorbed

89%

284.82 eV

283.95 eV

11%

� Excellent agreement with theory, which predicts that 18 atoms over 160 (i.e. 11%) are chemisorbed!

� XPS says that the chemisorbed phase is present in small amount in FG (not resolved by LEEM, but in STM)

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

Electronic structure by ARPES

9/23/2013 54

Issues: what i s the difference in electronic s tructure between FG and BG?

do they both show the same Dirac-like dispersion?

Diffraction Imaging microprobe-ARPES at EF

µ – ARPES

Γ

K

M

Experimental value of 0.42 eV

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

28

Dark-field PEEM

9/23/2013 55

Menteş and Locatelli, J. El. Spec. Relat. Phenomena 2012.

Dark-field PEEM: DOS at the Fermi level point K

9/23/2013 56

� DF-PEEM maps the emission from the π band at K allowing us to quantify the local DOS of flat and buckled graphene phases.

� DF-PEEM images at EF clearly indicate that only FG shows high DOS at K. BG shows a much reduced DOS and inversion of contrast in the image at Γ. This points to strong hybridization, and thus to metallicity.

� Microprobe-ARPES data are thus representative of the FG phase, not the BG one!

µ – ARPES

Γ

M

XPEEM at Γ, EF

max

min

K

df-XPEEM at K, EF

FG

BG

Ir

A. Locatelli et al; ACS Nano, ACS Nano, 7, 6955–6963 (2013)

29

Application to life sciences

Biomineralization

9/23/2013 57

9/23/2013 58

Applications of XAS in biology: biomineralization

• Bio-mineralization resulting from microbal activity

• X-PEEM images of (A) non mineralized fibrils from the cloudy water above the biofilm (scale bar, 5 um)

• (B) mineralized filaments and a sheath from the biofilm (scale bar, 1 um); (bottom)

• X-PEEM Fe L-edge XANES spectra of the FeOOH mineralized looped filament shown in (B), compared with iron oxyhydroxide standards, arranged (bottom to top) in order of decreasing crystallinity.

P.U.P.A Gilbert et al. (ALS group),

Science 303 1656-1658, 2004.

30

Nano-scale architecture of Nacre

9/23/2013 59

Contrast is observed between adjacent individual nacre tablets, arising because different tablets have different crystal orientations with respect to the radiation’s polarization vector.

The 290.3 eV peak corresponds to the C 1s � Pi* transition of the CO bond. Synchrotron radiation is linearly polarized in the orbit plane. Under such illumination, the

intensity of the peak depends on the crystallographic orientation of each nacre tablet with respect to the polarization. This was the first observation of x-ray linear dichroism in a bio-mineral.

R.A. Metzler et al., Phys.Rev.Lett. 98, 268102 (2007)

Oxygen K-edge XAS image

Carbon K-edge XANESCarbon K-edge image

Absorption intensity at resonance

Magnietic imaging

XMCD and XMLD PEEM

9/23/2013 60

31

Magnetic imaging: XMCD

9/23/2013 61

X-ray magnetic circular dichroism (XMCD) is the dependence of x-ray absorption on the relative orientation of the local magnetization and the polarization vector of the circularly polarized light

�Element sensitive technique

� Secondary imaging with PEEM determine large probing depth (10 nm), buried interfaces.

�sum rules allows measuring orbital and spin moments

At resonance, the secondary electron yield is proportional to the dot product between the magnetization direction and the photon helicity vector, which is parallel or anti-parallel to the beam propagation direction according to the handedness of the circular polarizationhv

MnAs/GaAs Magnetic domain imaging SPELEEM microscopehv

FM

PM

16°

Owing to the illumination

geometry, we are sensitive to

the in plane component of M

XMCD principles

9/23/2013 62

– The spin moment is given by the

imbalance of spin-up and spin-down

electrons (holes).

– By using circularly polarized radiation, the

angular momentum of the photon can be

transferred in part to the spin through the spin-orbit coupling. Photoelectrons

with opposite spins are created in the

cases of left and right handed

polarization. Spin polarization is opposite

also for p3/2 (L3) and p1/2 (L2) levels.

– The spin-split valence shell is thus a detector for the spin of the excited

photoelectron. The size of the dichroism

effect scales like cosθ, where θ is the

angle between the photon spin and the

magnetization direction.

– Refs: IBM. J . Res. Develop. 42, 73 (1998) and J. Magn. Magn. Mater. 200, 470

(1999).

• We PROBE 3d elements by exciting 2p into

unfilled 3d states

Dominant channel: 2p � 3d

White line intensity of the L3 and L2 resonances

with the number N of empty d states (holes). So 3d

electrons determine the magnetic properties.

32

Image algebra

9/23/2013 63

The size of the dichroism effect scales like cosθ, where θ is the angle between the photon spin and the magnetization direction. Hence the maximum dichroism effect (typically 20%) is observed if the photon spin and the magnetization directions are parallel and anti-parallel.

9/23/2013 64

Nano-characterization by XMCD-PEEM (imaging)

MAGNETIC STATE using XMCD & XMLD

Co nanodots on Si -Ge

A. Mulders et al,

Phys. Rev. B 71,

214422 (2005).

patterned s tructures

1.6 µmhv

M. Klaeui et al,

PRL , PRB 2003 - 2010

pulse injection

Laufemberg et al,

APL 88, 232507(2006).

domain wall motion induced by spin currents

33

Spin configturations in Heusler alloys

9/23/2013 65

Co2FeAl 0.4Si0.6 rings

C.A.S. Vaz et al

Appl. Phys. Lett. 99, 182510 (2011)

Multi-domain to flux-closure states (favored by the shape anisotropy) with decreasing element size, with a thickness-dependent crossover at the micrometer scale.

La 0.7Sr0.3MnO3

J. Rhensius et al

Appl. Phys. Lett. 99, 062508 (2011)

Magnetic Imaging – present capability

9/23/2013 66

I Kowalik, D. Arvanitis, M.A. Niño et al.,

in preparation

Fe L3 edge (chemical) Fe L3 edge (XMCD)

nano-magnetism of (Ga,Fe)N films Magnetization in NiPd nanostructures

J.-Y. Chauleau, Phys. Rev. B 84, 094416 (2011)

34

Magnetic imaging basics: XMLD

9/23/2013 67

In the presence of spin order the spin-orbit coupling leads to preferential charge order relative to the spin direction, which is exploited to determine spin axis in antiferromagnetic systems.

�Element sensitive technique

� Secondary imaging with PEEM determine large probing depth (10 nm), buried interfaces.

�Applied in AFM systems (oxides such as NiO)

Absorption intensity at resonance SPELEEM microscopehv

16°

Linear vertical and linear

horizontal polarization of the

photon beam

1st term: quadrupole moment, i.e.electronic charge (not magnetic!)

2nd term determines XMLD effect; Ө is the angle between E and

magnetic axis A; XMLD max for E || A; M reflects long range

magnetic order

9/23/2013 68

Applications of XMCD and XMLD

Nature, 405 (2000), 767.

770 775 780 785 790 795 800

Photon Energy (eV)

705 710 715 720 725 730

Nor

mal

ized

Int

ensi

ty (

a.u.

)

Photon Energy (eV)

2 µm

LaFeO3 layerXMLD Fe L3

Co layerXMCD Co L3/L2

ferromagnet/antiferromagnet Co/LaFeO3

bilayer, demonstrating interface exchange

coupling between the two materials

35

9/23/2013 69

Limited probing depth of XMCD: MnAs/GaAs

180 nm MnAsExperiment: Straight walls; Head to head domains

Simulation: Cross sectional cut: diamond s tate R. Engel-Herbert et al, J. Magn. Magn. Mater. 305 (2006) 457

Adding the time domain to PEEM

TR-PEEM methods

9/23/2013 70

36

9/23/2013 71

Time-resolved PEEM: the stroboscopic approach

Stroboscopic experiments combine high lateral resolution of PEEM with high time resolution, taking advantage of pulsed nature of synchrotron radiation

Choe et al., Science 304, 420 (2004)

Detector gating

9/23/2013 72

A. Locatelli et al., in preparation

37

9/23/2013 73

Time resolved XMCD-PEEM: applications

• Switching processes (magnetisation reversal) in magnetic elements ( in spin va lves, tunnel junction)

– Nucleation, DW propagation or both?– Effect of surface topography, morphology crystalline structure

etc.

– Domain dynamics in Landau flux closure s tructures.

• response of vortices, domains, domain walls in Landau closure domains in the precessional regime

• Stroboscopic technique:

– only reversible processes can be s tudied by pump – probe experiments

– Measurements are quantitative

9/23/2013 74

Magnetic excitations in LFC structures

xmcd

ti-t0

38

9/23/2013 75

Summary

• XPEEM is a versatile full-field imaging technique. Combined with SR it allows us to implement laterally resolved versions of the most popular x-ray spectroscopies taking advantage of high flux of 3rd generation SR light sources.

• In particular, XAS-PEEM combines element sensitivity with chemical sensitivity (e.g. valence), and, more importantly, magnetic sensitivity. Magnetic imaging has been the most successful application of PEEM (next tutorial lecture!).

• XPEEM or energy-filtered PEEM adds true chemical sensitivity to PEEM. Modern instruments allow to combine chemistry with electronic structure using ARUPS.

• XPEEM can be complemented by LEEM, which adds structure sensitivity and capability to monitor dynamic processes.

• Lateral resolution will approach the nm range as AC instruments become available. Limitations due to space charge are not yet clear

• Novel application field are being approached, such as biology, geology and earth sciences. HAXPES will increase our capabilities to probe buried structures (bulk).

9/23/2013 76

Review works

Some reviews and topical papers on x-ray spectromicroscopy and XPEEM• S. Guenther, B. Kaulich, L.Gregoratti, M. Kiskinova, Prog. Surf. Sci. 70, 187–260 (2002).

• E. Bauer, Ultramicroscopy 119, 18–23 (2012).

• E. Bauer, J. Electron. Spectrosc. Relat. Phenom. (2012): http://dx.doi.org/10.1016/j.elspec.2012.08.001

• G. Margaritondo, J. Electron. Spectrosc. Relat. Phenom. 178–179, 273–291 (2010) .

• A. Locatelli, E. Bauer, J. Phys.: Condens. Matter 20, 093002 (2008) .

• G. Schönhense et al., in “Adv. Imaging Electron Phys.”, vol. 142, Elsevier, Amsterdam, P. Hawkes (Ed.), 2006, pp. 159–323.

• G. Schönhense, J. Electron. Spectrosc. Relat. Phenom. 137–140, 769 (2004) .

• C.M. Schneider, G. Schönhense, Rep. Prog. Phys. 65, R1785–R1839 (2002) .

• W. Kuch, in “Magnetism: A Synchrotron Radiation Approach”, Springer, Berlin, E. Beaurepaire et al. (Eds.), 2006, pp. 275–320.

• J. Feng, A. Scholl, in P.W. Hawkes, “Science of Microscopy”, Springer, New York, J.C.H. Spence (Eds.), 2007, pp. 657–695.

• E. Bauer and Th. Schmidt, in “High Resolution Imaging and Spectroscopy of Ma-terials”, Springer, Berlin, Heidelberg, F. Ernst and M. Ruehle (Eds.), 2002, pp. 363-390.

• E. Bauer, J. Electron Spectrosc. Relat. Phenom. 114-116, 976-987 (2002).

• E. Bauer, J. Phys.: Condens. Matter 13, 11391-11405 (2001).