Applications of Spin-Polarized Electrons inCondensed Matter Physics

Dan Pierce

Electron Physics Group, NIST

http://physics.nist.gov/epg

Supported in part by the Office of Naval Research

Principle of GaAs Polarized e- SourceOptical excitation of spin polarized electrons

For positive helicity σ+ light, the theoretical polarization of the electrons photoexcited by bandgap radiation is

N NN NP ↑ ↓

↑ ↓

−+= 1-3

1+3= -50%

Polarization is easily and rapidly reversed by switching helicity of light

=

Pierce and Meier, PR B 13,5484-5500 (1976)

Negative Electron Affinity (NEA) GaAsSurface treatment gives world’s best photoemitter

Source characteristics:P ~ 30% for bulk waferI = 20 µA/mWcontinuous or pulsed operationlow initial energy and energy spreadhigh figure of merit = P2I

Charlie’s Hippie Days

Spin Polarized Electron Gun

Spin polarized e- scattering apparatus

I IA I I

↑↑ ↑↓

↑↑ ↑↓

−= + M∝Measure

Pierce, Celotta, Wang, Unertl, Galejs, Kuyatt, Mielczarek, RSI 51, 478-499 (1980)

Polarized e- Gun for SLAC Parity Violation Experiment

Beam polarization, 35% to 45% Beam current, 1 mA cw to 15 A peak

Complete Polarized e- Injection Systemfor SLAC Parity Violation Experiment

Technical Support in the Early Days

Development of Measurement Techniques Using Spin-Polarized Electrons

Spin-polarized electron gun (NEA GaAs)Spin polarized electron scattering

Spin polarized low energy electron diffraction (SPLEED)

Spin polarized inverse photoemission spectroscopy (SPIPES)

Spin polarized electron energy loss spectroscopy (SPEELS)

Spin polarized low energy electron microscopy (SPLEEM)

Electron spin polarization analyzerSpin polarized photoemission

Spin polarized Auger electron spectroscopy

Spin polarized secondary electron emission

Scanning Electron Microscopy with Polarization Analysis (SEMPA)

Surface Sensitivity

Surface sensitive due to short electron mean free paths

Inelastic mean free path especially short in transition metals with many d holes

Spin dependent attenuation length in ferromagnets at low energy

Pappas et al, PRL 66, 504 (1991)

Spin Polarized Electron Gun

Spin polarized e- scattering apparatus

I IA I I

↑↑ ↑↓

↑↑ ↑↓

−= + M∝Measure

Pierce, Celotta, Wang, Unertl, Galejs, Kuyatt, Mielczarek, RSI 51, 478-499 (1980)

Spin Polarized Low Energy Electron Diffraction (SPLEED)

First surface magnetization measurement with SPLEED

A=A(E,θ,H,T)

Test for true magnetic effect by measuring field and temperature dependence

A(H

) A(T

)

I IAAAB,

θ

Celotta, Pierce, Wang, Bader, Felcher, PRL 43, 728 (1979)

Spin-Polarized Electron Scattering

Pierce, Celotta, Unguris, Siegmann, PR B 26, 2566 (1982)

M TM BT( )( ) ..../0 1 3 2= - + Theory:

BBsurface

bulk=2 Experiment: BB

sb>2

Results explained by extending theory to include weaker exchange coupling of surface to bulk, JS⊥<JB

Thermal excitation of spin waves

Measure low temperature surface magnetization

Ni40Fe40B20

Bs/Bb=3

Spin-Polarized Electron ScatteringDetermine surface critical exponent

A lot of theory but few experiments formagnetic critical behavior at surfaces

Aex ∝ tβ, where t=[1-(T/TC)]. Find β≅0.78 for both Ni(110) and Ni(100) surfaces, independent of E and θ

Neutron measurements give β=0.38 for bulk

Alvarado, Campagna, Hopster, PRL 48, 51 (1982)

Spin Polarized Inverse Photoemission SpectroscopySpin dependent bandstructure of unfilled states

hw = ±97 035. . eV

A Pn nn n=-+

F

HGGG

I

KJJJA BA B

10cosa Large minority spin peak just above

Fermi level from the d-holes responsible for ferromagnetism in Ni

Energy, angle (k), and spin-resolved inverse photoemission

Photon detector sensitivity

Measure

Unguris, Seiler, Celotta, Pierce, Johnson, Smith, PRL 49, 1047 (1982)

Running with Friends

Bay to Breakers

Stopping to read while others catch up

Tough

Happy on top

of the mountain

A Sinclair Grand Canyon

Expedition

Determined

You’re not tryin’ if your not bleedin’!

Principled

Two things I can’t tolerate:

Incompetence and Cold Coffee

Married in 1983!

How ‘bout them legs

Spin Polarized Secondary EmissionHigh spin polarization where have high secondary intensity

Fe81.5B14.5Si4

nB nV PFe 2.2 8 0.28

Co 1.7 9 0.19

Ni 0.055 10 0.055

M n nB=- -A Bm ( )

Valence electron polarization

# Bohr magnetons# Valence electrons

BV

n n nP nn n↑ ↓

↑ ↓

−= = =+

Unguris, Pierce, Galejs, Celotta, PRL 49, 72 (1982)

Scanning Electron Microscopywith Polarization Analysis (SEMPA)

High resolution magnetization imaging

Magnetic Specimen

Incident Electrons FromScanning Electron Microscope

MagnetizationM= -µB (N↑ - N↓)

Spin-PolarizedSecondary Electrons

Spin-Polarization Analyzer

P =N↑ - N↓

N↑ + N↓

PolarizationAnalyzers

In-Plane

Out-of-Plane

MetalEvaporators

RHEED Screen

Secondary &Backscatter

ElectronDetectors

AugerElectronEnergy

Analyzer

Ion GunSample

ElectronSource

1 10 100 10000.01

0.1

1

10

100

1000

Beam diameter (nm)

Acqu

isiti

on ti

me

(min

) 10 keV20 keV

New Field Emission

Old LaB6

NIST SEMPA Apparatus

Input Optics

Drift TubeAnode

MCPG2

G1

E1

E2

Au evaporatorAu & C Targets

150

eV e

-

Low Energy Diffuse Scattering Analyzer

A

C

D B X

Y

End View

4 cm

SEMPA Example – Iron CrystalMeasure the topography and two magnetization components(simultaneously but separately):

MyMxIntensity

Directionθ = tan-1(Mx/My)

MagnitudeM = (Mx

2+ My2)1/2

Then derive the magnetization direction and magnitude:

Cobalt Vector Magnetization

Out-of-plane M

In-plane M

M contour

SEMPA – MFM ComparisonHard disk test sample with Au patterned overlay

SEMPAIntensity M AFM MFM

350 nm

Interlayer Exchange CouplingCr wedge schematic

Cr Wedge

Fe Whisker

Fe Film

0 - 20 nm

∼ 1-2 nm

∼ 500 µm

M

Unguris, Celotta, Pierce, PRL 67, 140 (1991)

Fe/Cr/Fe Oscillatory Exchange CouplingPrecise measurement of coupling periods

0 10 20 30 40 50 60 70 80

-0.2

0.0

0.2

Cr Thickness (layers)

P(Fe

)

-0.04

0.0

0.04

P(C

r)

RHEED

M

5.73 ±0.05

2.48 ±0.05

Interlayer Measured period(layers)

Cr0.144 nm/layer

Ag0.204 nm/layer

Au0.204 nm/layer

2.105 ±0.00512 ±1

2.37 ±0.07

8.6 ±0.3

Leadership

Vision

We’ll get there if you can keep up

with me

Between a rock and a hard place

Magnetic Depth Profiling Co/Cu MultilayersMagnetoresistance and domain structure

Current through magnetic multilayer depends on magnetization alignment,

Mi

but real domain structures can be complex:

Co

Ion milled crater(2 keV Ar+)

Cu Top Co layer

2nd Co layer

Top layer magnetization predominantlyantiparallel to 2nd layer

Borchers, Dura, Unguris, Tulchinsky, Kelley, Majkrzak, Hsu, Loloee, Pratt, Bass, PRL 82, 2796 (1999)

Interlayer Domain CorrelationsQuantify layer-by-layer angle distributions

Angle images:

1st layer 2nd layer

0 90 1800.00.20.40.60.81.0

FM AFM

P(∆φ

)

∆φ (°)

Correlation distribution:

Magnetic materials

Air Side Wheel Side

Transformer core ferromagnetic glass(Allied Signal)

Magnetic NanostructuresFe “nanowires”

213 nm

M

FeCr

Fe evaporated at grazing incidence on Cr lines fabricated by laser focused atom deposition

Tulchinsky, Kelley, McClelland, Gupta, Celotta, JVST A16, (1998)

Ready to retire???

Acknowledgements

Spin-polarized electron work at NBS/NISTMany contributors,

especially Bob Celotta and John Unguris

Pictures of CharlieJoel Falk

Jim Murray

Roger Route