A users viewpoint: absorption spectroscopy at a synchrotron

Frithjof Nolting

Bottom layer “pinned”Hard layer orExchange coupling

Top layer “free” can be switched

Ferromagnet

Ferromagnet

Antiferromagnet

(1) (0)

MRAMHard disk head

Magnetic data storage and recording

Low resistance high resistance

GMR Element

FM

AFM

Interface

- Different models for AFM/FM coupling exist.

- Different assumptions on AFM structure lead to complete different results.

- Spin arrangement at the interface is important

•High Spatial Resolution

•Elemental Sensitivity

•Ferromagnetic Contrast

•Antiferromagnetic Contrast

•Surface/Interface Sensitivity

How?

X-ray absorption spectroscopy (XAS) with spatial resolution

Photoemission Electron Microscope (PEEM)

Photoemission Electron Microscope

I(nA)

h

Energy

EF

~~

h(x-ray)

core level

valence band

2p 3/2

2p1/2

Magnetic lenses e-

16°

analyzer

20 kVX-rays

Sensitive to:• elemental composition• chemistry• structural parameters• electronic structure• magnetic properties

ELMITEC GmbH

Magnetic microscopy

S

E

XMCD (X-ray Magnetic Circular Dichroism)

XMLD (X-ray Magnetic Linear Dichroism) LaFeO3

775 780 785 790 795 800

L2

L3

Photon energy (eV)

TE

Y (

a.u.

)

Co

5 µm

5 µm

J. Stöhr et al. Science 1993 A. Scholl et al. Science 2000 F. Nolting et al. Nature 2000

840 850 860 870 880 890

0.6

0.9

1.2

1.5

left polarized right polarized

Tot

al Y

ield

(a.

u.)

Photon Energy(eV)

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Asymmetry smoothed

Asy

mm

etry

(%)

Exchange biased Co/NiO multilayer

EPU beamline 4 at ALS

.

.

3 nm Co50 nm NiO

1.5 nm Ru

3 nm Co 30°

Magnetic field

Left and right circularly polarized light

XMCD

Hysteresis of Co and NiO

XMCD spectra – switching polarization

-3000 -2000 -1000 0 1000 2000 3000-20

-15

-10

-5

0

5

10

15

20

XM

CD

(%

)

Applied Field (Oe)

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Co Ni

Pinned Moments ?

H. Ohldag, A. Scholl, F. Nolting, E. Arenholz, S. Maat, A.T. Young, M. Carey, and J. Stöhr, Phys. Rev. Lett. 91(1), 017203/1-4 (2003).

Ratio of A/B

Antoine Barbier et alUser experiment at SLS, July 2004

SIM Beamline Layout

UndulatorT. Schmidt

• 2 Elliptical undulators

• Pure permanent magnet

• 95eV < h < 2000eV• >1019 photons/s/mrad2/mm2/400mA

• 100 % circular polarization[125 - 900 eV]reduced on higher harmonics

• Hor. & vert. linear polar.

OpticsU. Flechsig

• Plane grating monochromator

• E/E > 8000

• <5% 2nd order light

• Switch helicity

• Focus 30x100m2

EndstationC. Quitmann & F. Nolting

• SLS endstation:

– PEEM & LEEMx~25 - 50nm spatialE~150meV energy

– Sample Prep chamber

– User endstation

Front end

Prepchamber

ID1 ID2 Monochromator Chopper

Refocusingoptics

MicroscopeUser

endstation

705 710 715 720 725

0.44

0.46

0.48

0.50

0.52

0.54

0.56

0.58

0.60

Inte

nsi

ty (

a.u

.)

Photon energy (eV)

reference signal sample signal

705 710 715 720 7250.78

0.80

0.82

0.84

0.86

0.88

Inte

nsi

ty (

a.u

.)

Photon energy (eV)

normalized signal

TEY

Reference signal Sample signal

Magnet

Sample = Norm

Reference

Moving gap and monochromator, stop, measure (1s – 1minute), moving …

How do we measure

710 720-0.0020.0000.0020.0040.006

Diff

. [a

.u.]

Photon Energy [eV]

difference0.78

0.80

0.82

0.84

0.86

0.88

1 %

Inte

nsi

ty [a

.u.]

circular plus circuls minus

How do we measure

•Change polarization and repeat

•Take difference of spectra

Absorption spectrum requires frequent moving of gap and shift

must not effect other beamlinestransparent!

switching by moving phase

switching by moving the gap

tuned detuned

move up 56 mm

move 2 mm

move 0 mm

detuned tuned

circ plus circ minus

chopper

120 s

1s - 8 s

100 Hz

Modes for switching the polarization

switching by using a chopper

Alignment of IDs - Horizontal1. Horizontal overlap at Frontend

ID1

ID2

ID2Asymmetric bump

2D Frontend scan

Closed bump using chicane magnets

2. Horizontal overlap at focus of experiment ID2In

tens

ity

Horizontal position

Alignment of IDs - Vertical

3. Vertical (energy) overlap at focus of experiment

ID1

ID2

not yet finished !

Photon energy

Inte

nsity

Inte

nsity

Inte

nsity ID2

Absorption spectrumschematic

778.3 eV

source mono Exit slit

Beam variation - Noise

horizontal movement intensity variationXAS normalization reduces it by a factor of 10-100PEEM no normalization

days no problemhours no problemsec - minutes badmili seconds ok

vertical movement energy variationno normalization!

days no problemhours badseconds badmili seconds ok

10 μm about 2%

10 μm about 10 meV

Energy shift

Energy shift of 10 meVIdentical spectra

1 %

Orbit feedback

Fast orbit feedback slow orbit feedback

beam positionin ID(Bergoz)

X-ray positionin beamline

Shift and gap

Orbit correctors?

705 710 715 720 725

-0.001

0.000

0.001

0.002

0.003

0.004

0.005

0.006

Inte

nsi

ty (

a.u.

)

photon energy (eV)

with fast orbit feedback with slow orbit feedback with no orbit feedback

705 710 715 720 725

0.80

0.82

0.84

0.86

0.88

No

rm s

ign

al (

a.u

.)

Photon energy (eV)

with fast orbit feedback with slow orbit feedback

Orbit feedback – effect on measurement

Normalized signal, circular plus Difference circular plus and minus

Increased noise!!!

Slow Orbit feedback

Circular plus Circular minus

beam positionin ID(Bergoz)

X-ray positionin beamline

Shift and gap

Orbit correctors?

705 710 715 720 725

0.55

0.56

0.57

0.58

0.59

0.60

raw signal

Inte

nsi

ty (

a.u

.)

photon energy (eV)

measurement one measurement two

Slow Orbit feedback

Not transparent !

10 μm

690 700 710 720 730 740-0.004

0.000

0.004

0.0081%

Diff

. (a

.u.)

Photon Energy [eV]

diff

1.15

1.20

1.25

1.30

1.35

1.40

Inte

nsi

ty (

a.u

.)

norm1 norm2

Summary

We can do great measurements at SLS

Transparent IDs are essential! Very difficult to make an EPU transparent. Have to rely on Orbit feedback

For the measurement “no” difference between slow and no Orbit feedback

Critical time scale second – hour (10 Hz – 0.0001 Hz)Intensity variation 0.1% 0.5 μmenergy variation 1meV 1 μm

Slow Orbit feedback is not sufficient

Fast Orbit feedback is great