Preliminary Doctoral Defense of Francisco Almeida

Exchange springs and exchange bias studied withnuclear methods

Francisco Miguel Cruz Simoes de Almeida

Instituut voor Kern-en Stralingsfysica

May 24, 2011

Francisco Almeida (IKS) Doctoral dissertation May 24, 2011 1 / 25

Outline

1 Surface spin-flop in an Fe/Cr superlatticeInterlayer couplingSurface spin-flopPolarized Neutron Reflectometry

2 Exchange bias in Fe-Pt heterostructuresExchange biasFe-Pt alloysFePt/FePt3 bilayersOff-stoichiometric Fe-Pt film

3 Exchange bias through ion beam synthesisIon beam synthesisOxygen implanted cobalt

4 Conclusions


Surface spin-flop in an Fe/Cr superlattice


Fe/Cr superlattice

Multilayer consisting of 22 repetitions of an Fe/Cr bilayer structure

Fe layers 40 A thick.

Cr layers 11 A thick.

Fe

Cr

Fe

Cr

Fe

Cr

Fe

Cr


Fe/Cr superlattice

Multilayer consisting of 22 repetitions of an Fe/Cr bilayer structure

Fe layers 40 A thick and antiferromagnetically coupled.

Cr layers 11 A thick.


Interlayer coupling

Coupling between ferromagnetic layers separated by a spacer layer.

Bilinear Biquadratic


Magnetic response of an Fe/Cr superlattice

Easy axis

M/M

s0.0

0.2

0.4

0.6

0.8

1.0

25 mT

250 mT

[001]

[010]

H

a

Hard axis

µ0H (T)

M/M

s

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

0.2

0.4

0.6

0.8

1.0

25 mT

75 mT

[001

][010]

H

b

J. Meersschaut et al., PRB 73, 144428 (2006)


Magnetic response of an Fe/Cr superlattice

Easy axis

M/M

s0.0

0.2

0.4

0.6

0.8

1.0

25 mT

250 mT

[001]

[010]

H

a

Hard axis

µ0H (T)

M/M

s

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0

0.2

0.4

0.6

0.8

1.0

25 mT

75 mT

[001

][010]

H

b

J. Meersschaut et al., PRB 73, 144428 (2006)


Surface spin-flop

Transition (H(S)C ) that causes a top spin/layer to switch orientation before the

remaining spins/layers of an antiferromagnet. The irregularity propagates to the

middle of the spin chain, and leads to a bulk spin-flop (H(B)C ).

H(S)C : surface spin-flop field

H(B)C : bulk spin-flop field

How to visualize a surface spin-flop?


Surface spin-flop

Transition (H(S)C ) that causes a top spin/layer to switch orientation before the

remaining spins/layers of an antiferromagnet. The irregularity propagates to the

middle of the spin chain, and leads to a bulk spin-flop (H(B)C ).

H(S)C : surface spin-flop field

H(B)C : bulk spin-flop field

How to visualize a surface spin-flop?


Polarized Neutron Reflectometry

Similar to X-ray reflectometry.

Sensitive to structural periodicity.

Also sensitive to magnetic periodicity.

Experiment performed in the neutron reflectometer at the HelmholtzZentrum, Berlin.

Neutron wavelength λ = 4.66 A.


Polarized Neutron ReflectometryResults

R++

(a.u.)

10

10

10

10

10

-4

-3

-2

-1

0

5.0x10-5

1.0x10-4

1.5x10-4

Γ ( )-2

Å

0 200 400 600 800 1000Depth (Å)

0.00 0.04 0.08 0.12 0.16

QZ (

-1)Å

0 200 400 600 800 1000Depth (Å)

0.00 0.04 0.08 0.12 0.16

QZ (

-1)Å

0 200 400 600 800 1000

Depth (Å)

0.00 0.04 0.08 0.12 0.16

QZ (

-1)Å

0.00 0.04 0.08 0.12 0.16

QZ (

-1)Å

0 200 400 600 800 1000

Depth (Å)

μ H = 600 mT 0 μ H = 200 mT 0 μ H = 61.8 mT 0 μ H = 30 mT 0


Polarized Neutron ReflectometrySpins configuration: surface spin-flop transition

10-4

10-3

10-2

10-1

100

R+

+ (

a.u

.)

0.00 0.04 0.08 0.12 0.16

QZ (

-1)Å

10-4

10-3

10-2

10-1

100

R+

+ (

a.u

.)

10-4

10-3

10-2

10-1

100

R+

+ (

a.u

.)

10-4

10-3

10-2

10-1

100

R+

+ (

a.u

.)

[001

][010] Hμ H = 600 mT 0

μ H = 200 mT 0

μ H = 61.8 mT 0

μ H = 30 mT 0


Exchange bias in Fe-Pt heterostructures


Exchange biasExample of an oxidized cobalt thin film

Effects observed on the hysteresis loop upon field cooling

Loop shift and broadening

Training

-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

-1.0

-0.5

0.0

0.5

1.0

M/M

S (a

.u.)

0H (T)





Training

-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

-1.0

-0.5

0.0

0.5

1.0 room temperature 10 K

M/M

S (a

.u.)

0H (T)





Training

-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

-1.0

-0.5

0.0

0.5

1.0 room temperature 10 K (first) 10 K (second)

M/M

S (a

.u.)

0H (T)


Exchange biasMotivation

Origin of exchange bias notoriously difficult to identify.I Behavior very varied from system to system.I Many models of limited applicability.I Most applied techniques focus on the ferromagnet.

Looking into the antiferromagnet:I Local information from the antiferromagnet.I Mossbauer spectrometry of 57Fe.

Let us take advantage of Mossbauer spectrometry.












Fe-Pt alloys

FePt3 ( Antiferromagnetic, L12 )

FePt (d=100 Å)

MgO (110) (substrate)

Au (capping)

FePt (h)357

FePt ( Ferromagnetic, L10 )

(Fe: red spheres, Pt: white spheres)


Fe-Pt alloys

FePt3 ( Antiferromagnetic, L12 )

FePt (d=100 Å)

MgO (110) (substrate)

Au (capping)

FePt (h)357

FePt ( Ferromagnetic, L10 )

(Fe: red spheres, Pt: white spheres)


Magnetism of FePt3

a) T < 160 K: Q1 wave mode, uncompensated (110) surfaceb) T < 100 K: Q2 wave mode, compensated (110) surface

S. Maat et al., PRB 63, 134426 (2001)


FePt/FePt3 (110) bilayersMagnetic response

FePt semi-easy axis FePt hard axis

FePt

H

FePt3

-2.0 -1.0 0.0 1.0 2.0

-1.0

-0.5

0.0

0.5

1.0

M/M

S (

a.u

.)

0H (T)

SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS

μ

FePt

// HFePt3

-1.0 0.0 1.0

-1.0

-0.5

0.0

0.5

1.0

M/M

S (

a.u

.)

0H (T)

SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS

μ


FePt/FePt3 (110) bilayersExchange bias (100 A FePt/300 A FePt3)

Observed using a dual approach:

Volume magnetization information (magnetometry).

Local information (Mossbauer from the 57FePt3).

-100

0

100

200

300

400 HE

HC

-1.0 -0.5 0.0 0.5 1.0

-1.0

-0.5

0.0

0.5

1.0

M/M

S (

a.u

.)

0H (T)

μ H

,

μ H

(m

T)

00

CE

μ

0 50 100 150 200 250 300

T (K) v (mm/s)

yiel

d (a

.u.)

-1 0 1

630000

640000

650000


FePt/FePt3 (110) bilayersExchange bias (100 A FePt/300 A FePt3)

What we observe:

Very high roughness and/or intermixing.

Spin-flop coupling between the FePt and the FePt3.

-100

0

100

200

300

400 HE

HC

-1.0 -0.5 0.0 0.5 1.0

-1.0

-0.5

0.0

0.5

1.0

M/M

S (

a.u

.)

0H (T)

μ H

,

μ H

(m

T)

00

CE

μ

0 50 100 150 200 250 300

T (K) v (mm/s)

yiel

d (a

.u.)

-1 0 1

630000

640000

650000


Off-stoichiometric Fe-Pt film

What happens if we field cool a mixture of Fe0.33Pt0.67?

0 20 40 60 80 100 120 140

-30

-20

-10

0

0HE (m

T)

T (K)

SSSSSSSSSSSSSSS

v (mm/s)

Yie

ld (

a.u.

)

-5 0 5

615000

620000

625000


Off-stoichiometric Fe-Pt film

What happens if we field cool a mixture of Fe0.33Pt0.67?

0 20 40 60 80 100 120 140

-30

-20

-10

0

0HE (m

T)

T (K)

SSSSSSSSSSSSSSS

v (mm/s)

Y

ield

(a.

u.)

-5 0 5

615000

620000

625000


Exchange bias through ion beam synthesis


Ion beam synthesis

New method for obtaining exchange bias.

I Reactive implantation grows the antiferromagnet inside theferromagnet.

I Extremely granular structure.I Optimal contact surface.I Parametric control (e.g. fluence, energy).

Synthesis of CoxOy through implantation of 16O into Co.I MBE grown, capped cobalt thin films (thicknesses 600 A and 1000 A).


Ion beam synthesis

New method for obtaining exchange bias.I Reactive implantation grows the antiferromagnet inside the

ferromagnet.I Extremely granular structure.I Optimal contact surface.I Parametric control (e.g. fluence, energy).



Ion beam synthesis

New method for obtaining exchange bias.I Reactive implantation grows the antiferromagnet inside the

ferromagnet.I Extremely granular structure.I Optimal contact surface.I Parametric control (e.g. fluence, energy).



Exchange bias in oxygen implanted cobaltExample

Thin cobalt film implanted with 16O (E = 60 keV, Φ = 1016cm−2).

-0.4 -0.2 0.0 0.2 0.4

-1.0

-0.5

0.0

0.5

1.0 b)

M/M

S (a

.u.)

0H (T)

-1.0

-0.5

0.0

0.5

1.0 a)M

/M


ConclusionsSurface spin-flop in an Fe/Cr superlattice

Surface spin flop like transition detected in a biaxial Fe/Crsuperlattice, along the hard axis.

Demonstrated through Polarized Neutron Reflectometry.


ConclusionsExchange bias in Fe-Pt heterostructures

Exchange bias found in FePt/FePt3 bilayers.I FePt easy axis orthogonal to FePt3 easy axis (spin-flop coupling).I Local information extracted from the antiferromagnet via Mossbauer

spectrometry.

Exchange bias detected in a single layer of disorderedoff-stoichiometric Fe-Pt.


ConclusionsExchange bias in O implanted Co thin films

Exchange bias detected in oxygen implanted Co thin films.I New kind of exchange biased system.

F Antiferromagnet is embedded in the ferromagnet.F Heterogeneous effect.

I Requires a fluence of the order of at least Φ = 1016 cm−2.I The effect is sensitive to the fluence as well as the implantation energy.I Post implantation annealing enhances and homogenizes the obtained

effect.


Conclusion

Many new research opportunities regarding these new systems:

I Exchange bias between Fe-Pt alloys presents new challenges.F Evident direction for further research.F Mossbauer techniques give us more flexibility.

I Exchange bias obtained via ion implantation opens a new field ofresearch:

F Principle may be applied to many other systems besides Co/CoO.F Wealth of technological applications (e.g. patterned implantation).F Theoretical challenge: transition from 2D systems (bilayers) to 3D

systems (implanted films, fully granular structures).


Education

Preliminary Doctoral Defense of Francisco Almeida