Soft X-Ray Studies of Surfaces, Interfaces and Thin Films:

From Spectroscopy to Ultrafast Nanoscale Movies

Joachim StöhrSLAC, Stanford University http://www-ssrl.slac.stanford.edu/stohr

Work supported by the DOE Office of Basic Energy Science

Overview of my talk

The Power of Soft X-rays

Polarized X-Ray Absorption Spectroscopy Liquid crystal alignment on surfaces

X-Ray Spectro-Microscopy Ferromagnetic alignment on an antiferromagnetic surface

Time Dependent X-Ray Spectro-Microscopy Switching of magnetic nano-structures with spin currents

A Glimpse of the Future

What are soft x-rays anyway?

VUV Hard X-Rays

30 eV 3000 eV

Soft X-Rays

100 eV~ 10 nm

1000 eV ~1 nm

Opening the soft x-ray region – late 1970s

Stanford Synchrotron Radiation Lab

500 eV 800 eV

Oxygen SEXAFS oxidized Al surface 12/ 5/1977

Phot

on fl

ux

Photon energy (eV)

O K-edge

Grasshopper monochromator

Spectroscopy in the important region 280 – 1000 eV became possible

Tunable x-rays offer atom specific valence shell information through guided transitions

Element specificity, Chemical specificity, Valence properties

magnetic multilayerpolymer

Polarized x-ray absorption determines charge and spin orientation

Antiferromagnetic order

Orientational order of bonds

Ferromagnetic order

Liquid crystal alignment on rubbed polymer surfaces…discovered in 1907…

Use of soft x-rays to solve a 100 year old puzzle

Note LC “pretilt” out of plane

A $30 billion world-wide business

Alignment is basis of liquid crystal displays

X-ray diffraction on polyimide suggests epitaxy-like nucleation Oldest model assumes micro grooves in polymer surface

Conventional models of alignment mechanism

Models cannot explain LC “pretilt” angle up from plane

A key observation in 1998: Directional ion beam irradiated polymers also align liquid crystals

Pretilt direction is opposite !

• LCs align on a-carbon surface layer, not on polymer substrate• Is LC alignment due to bond orientation on substrate surface?

X-ray spectroscopy of ion beam modified polymer surface

referencesample

Do not need polymers at all ! start with a-Carbon – align with ion beam

• Rubbing and ion beam create molecular level orientational order• Highest resolution displays today use ion beam aligned carbon films

Nature 411, 56 (2001); Science 292, 2299 (2001)

Polarization Dependent Imaging with X-Rays

Oriented molecular regions

Antiferromagnetic regions

Ferromagnetic regions

Tackling a 50 year old mystery with x-rays: “Exchange bias”

How can a “neutral“ antiferromanet bias a ferromagnet?

Effect remained a puzzle ever since its discovery in 1956

Conventional techniques could not study the all-important interface

Key modern magnetic building blocks are based on fixed (“pinned”) ferromagnetic reference layers

does not turn in external fieldspinned by antiferromagnet

turns in external fields: “0” or “1” bitsReference layer

X-Rays reveal interfacial coupling of FM and AFM domains

Ni edge – use linear polarization – antiferromagnetic domains

Co edge – use circular polarization – ferromagnetic domains

H. Ohldag et al., PRL 86, 2878 (2001)

[010]

2m870 8740

5

10

15

Elec

tron

Yie

ld

Photon Energy(eV)

NiOXMLD

776 778

0

4

8

Elec

tron

Yie

ldPhoton Energy (eV)

CoXMCD

780

2nm

X-Rays-in / Electrons-out: A way to study thin film interfaces

pureCo/NiOpure

pureCo/NiOpure

Interface is mixed CoNiOx layer - is it magnetic?

Images of the Ferromagnet-Antiferromagnet Interface

Ohldag et al., PRL 87, 247201 (2001)

Interface layer contains ferromagnetic NiOx - is it coupled to AFM NiO?

Exchange bias model

• A thin interfacial diffusion layer (1–2 layers) of CoNiOx is formed• Interface layer contains ferromagnetic Ni spins from modified NiO• About 95% of interfacial Ni spins rotate with FM (not pinned)• Only < 5% of interfacial Ni spins are pinned to bulk NiO• This tiny fraction is the origin of exchange bias

Ohldag et al PRL 91, 017203 (2003)

What have we learned so far ?

• Interface effects play import role in modern nanoscale materials

• Suble interface properties can lead to important phenomena

• Soft x-rays are powerful tool to reveal interface-specific effects elemental specificity chemical specificity magnetic specificity orientational specificity nanoscale spatial resolution

The new frontier: dynamics or “the need for speed”

The ultrafast technology gap

Drivers of Modern Magnetism Research: Smaller and Faster

Fundamental Timescales

Operational Timescales

The goal

Bunch spacing 2 ns

Bunch width ~ 50 ps

Time Resolution:Pulsed X-Rays from Electron Storage Ring

beam line

pulsed 50 ps x-rays

State-of-the art ultrafast electronics :Y. Acremann et al., Rev. Sci. Instr. 78, 14702 (2007).J. P. Strachan et al., Rev. Sci. Instr. 78, 54703 (2007).

From reading to writing informationSuggested by J. Slonczewski & L. Berger in 1996

“spin torque switching” – no external magnetic field !

Verified by:F.J. Albert, J.A. Katine, R.A. Buhrman, D. Ralph, Appl. Phys. Lett. 77, 3809 (2000)

free

fixed

Time-Resolved Scanning Transmission X-Ray Microscopy

Detector leads for current pulses

2 nm magnetic layer

buried in 250nm of metals

current

~100 nm

Y. Acremann et al., Phys. Rev. Lett. 96, 217202 (2006)

X-ray image

5m

100nm

Spin Torque Switching: 180nm x 110nm x 2 nm nanostructure of CoFe

switch backcurrent pulse

switch

Y. Acremann et al., Phys. Rev. Lett. 96, 217202 (2006)J. P. Strachan et al., Phys. Rev. Lett. 100, 247201 (2008)

+_

200ps 400ps 600ps 800ps

t=0

100 nm

Vortices are important on all length scales

~ 50nm

100 km

100,000 light years= 1018 km

Hurricane

Milky Way

Nano-element

A Glimpse of the Future

X-ray snap shots on the fundamental time scales of motion of

atoms, electrons and spins

….femtoseconds and faster….

The Light FantasticThe Light Fantastic Birth of the X-Ray Laser …..and a New Era of Science

The End