Observation of magnetization distribution in a correlated spin glass system : amorphous Tb-Co magnetic films

HAL Id: jpa-00212383

https://hal.archives-ouvertes.fr/jpa-00212383

Submitted on 1 Jan 1990

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Observation of magnetization distribution in a correlated

spin glass system : amorphous Tb-Co magnetic films

M. Schlenker, J. Pelissier, B. Barbara, J.P. Guigay, G. Fillion, R.H. Geiss, A.

Liénard, B. Blanchard

To cite this version:

483

Observation of

_{magnetization}

distribution in

a

correlated

spin

glass

system :

amorphous

Tb-Co

_magnetic

films

M. Schlenker

_(1),

J. Pelissier

_(2),

B. Barbara

_(1),

J. P.

_Guigay

_(1),

G. Fillion

_(1),

R. H. Geiss

_(3),

A. Liénard

₍₁₎

and B. Blanchard

₍₄₎

(1)

Laboratoire Louis Néel du CNRS, associé à l’Université

_Joseph

Fourier, B.P. _166, 38042

Grenoble, France

(2)

Département

de

_{Métallurgie, CEN-G,}

B.P. _85, 38041

_Grenoble,

France

(3)

IBM Almaden Research

Laboratory,

650

Harry

Road, San Jose, CA 95120-6099, U.S.A.

(4)

S.E.A.,

Département

de _{Chimie, CEN-G,} B.P. 85, 38041

Grenoble,

France

(Reçu

le 14

_septembre

_1989,

_accepté

le 23 octobre

₁₉₈₉₎

Résumé. 2014 La distribution

spatiale

de l’aimantation a été _{observée par}

_{microscopie électronique}

en transmission sur des couches

_amorphes

de

_{TbCo3 à anisotropie}

aléatoire obtenues par

pulvérisation

cathodique,

à

_température

ambiante ainsi

qu’à

hautes et basses

températures.

Les couches brutes de

_{préparation, d’épaisseur}

8 x

10-8

m environ, montrent à l’ambiante une

structure

_magnétique

désordonnée dont l’échelle

_{caractéristique, 10-7}

m, est

_compatible

avec le modèle de

Imry

et _{Ma pour} un verre de

_spins

corrélés. Cette structure _{n’est pas affectée par le}

refroidissement, mais elle se

_simplifie

de

_façon

irréversible et

_progressive

lors du

_chauffage.

Le contraste des

_parois

de domaines est _inversé, sans modification des

positions

des

parois,

à la traversée de la

température

de

_{compensation,}

voisine de 270 K. Dans des

_fragments

amincis, la structure en domaines est

_{qualitativement}

_{différente par suite de la perte} de terbium dans les

régions

superficielles.

Abstract. 2014 The

magnetization

distribution in

_{sputtered random-anisotropy TbCo3 amorphous}

films was observed

_by

transmission electron

microscopy

at room, low and

_high

temperature. In the as-grown films, about 8 x

10-8

m _thick, a disordered structure on a scale of

10-7 m,

consistent

with the

_Imry

and Ma

_picture

for a correlated

_{spin glass,}

is observed at room _temperature. This structure is unaffected

_{by cooling}

but is

irreversibly

and

_{progressively}

_simplified

on

_heating.

Domain wall contrast is reversed, with no

_change

in wall

_positions,

when

_{going through}

the

compensation

temperature of about 270 K. In thinned

fragments,

the domain structure is

qualitatively

different, due to terbium loss near the surfaces.

J.

_Phys.

France 51

₍₁₉₉₀₎

483-492 1ef MARS 1990,

Classification

Physics

Abstracts

61.16D - 75.50K - 75.60 - 75.70

Introduction.

Random-anisotropy

magnets

have attracted considerable attention since the

_pioneering

work of

_Imry

and Ma

_[1]

and of

Aharony

and

_Pytte

_[2],

who

_predicted

the formation of small clusters with correlated

_magnetic

moment directions when random

_anisotropy,

characterized

by

D, competes

with

_exchange

J ;

a transition to standard thin film

_{ferromagnetism}

with

ripple

and domain structures is of course

_expected

for low

_anisotropy.

_Chudnovsky

and Serota

_[3]

as well as

_Aharony

and

_Pytte

[4]

showed that the

superimposition

of a uniform

uniaxial

_anisotropy

can

bring

about such a transition for D > 0. The effect of the unavoidable

dipolar coupling

was considered

_by

Cullen

_[5].

The case of weak random

_anisotropy

was

discussed

_{by Chudnovsky,}

Saslow and Serota

_[6],

who introduced the

_concept

of the

correlated

_{spin glass}

_(CSG).

Saslow

_[7]

_investigated

the effect of

_applied

_magnetic

field. An

excellent review was

_{recently published by Chudnovsky}

_[8].

Domain observations have been

_{abundantly performed}

on thin

_films,

both

_crystalline

and

amorphous,

of interest for

_magnetic

and

_{magneto-optical}

_storage

_applications

and in

particular

in

_amorphous

terbium-iron,

a

_system

_apparently

similar to the one we are

investigating [9-12].

Their relevance for

_{applications,}

_however,

_usually

_implies

_{the presence of}

high perpendicular anisotropy,

and observations focused on films

_featuring

it.

Although

neutron

_scattering

measurements

_{by Rhyne}

_[13]

_yielded

a wealth of information

about the overall

temperature

dependence

of the

_{magnetization}

_{fluctuations,}

no direct

microscopic

observation was

_yet

_made,

to the best of our

_knowledge,

on the behavior of materials with

_{anisotropy predominantly}

local.

The aim of the

_present

_study

was to

_directly

observe the

_{magnetization}

_{distribution,}

in a

system

with random

_anisotropy

dominant over coherent

_anisotropy,

as a function of

temperature.

The

_amorphous

terbium-cobalt

_system,

_regarded

as a

_{typical sperimagnet}

with local

_anisotropy,

seemed

_appropriate.

The fact that the Curie

_temperature

is well above room

temperature

[14]

makes observations

easier,

and

_cooling

could be

_expected

to increase the effect of local

_anisotropy

on the disordered terbium sublattice.

Experimental.

Thin

films,

about 8 x

10- 8

m

thick,

were

_{prepared by}

DC triode

_magnetron

_sputtering,

at

normal

incidence,

from a

_{TbCo 1.8}

_target,

in argon at 1 _{mTorr pressure, with} a bias of 42

_V,

a

voltage

of 1 500 V and an ion current of 0.12

_A,

onto cleaved rocksalt substrates mounted on a

liquid-nitrogen

cooled

rotating

holder

using

the

equipment

developed by

Rebouillat. A

vacuum of 5 x

10- 7

Torr was maintained in the

_sputtering

unit for 20 hours before

_sputtering

was started and the

target

was cleaned

_{by pre-sputtering}

for 15 minutes on a

_{dummy sample}

holder before

_making

the

_samples.

When thick

_{amorphous samples}

were

_prepared

earlier from the same

_target

for bulk

_magnetic

and neutron measurements, their

_composition

was found to be

_{ThCol.98 [15] ;}

as will appear in the

results,

there is evidence that the

_composition

of the films

_investigated

here is different.

The films were then floated off the

substrate,

and observed

_by

conventional transmission electron

_{microscopy, mostly}

in the defocused Lorentz

mode,

but also in standard observation as well as _{normal and very}

_small-angle

diffraction modes.

Parts of the

_experiments

were

_performed

on a 100 kV

_Philips

301

_{microscope, using}

a

commercial

_{liquid-nitrogen}

_{stage ;}

there the

_samples

had to be thinned down

_{by ion-milling}

with argon

ions,

typically

with a

_voltage

of 5.5 kV and ion current about 0.2 mA for a few minutes. No

_quantitative

measure of the thickness was then

available,

and

only comparisons

between

_regions

of different thickness can be made. The bulk of the observations were

performed

on the unthinned films at the 1 MV

_facility

of

_Département

de

_Métallurgie

at

CEN-G, Grenoble,

operating

at 800

_kV,

_using

a

_{liquid-helium}

cold

_stage

and a hot

stage.

In Lorentz

_(defocused)

mode

_observation,

the

_objective

lens was switched off to avoid

submitting

the film to its

_magnetic

field. The amount of

_defocusing

was controlled

_by

the

current in the intermediate lens. All the features

_expected

for

_magnetic

domain wall

observation,

viz. contrast inversion when

_moving

from overfocus to

_underfocus,

and

disappearance

at the focused

_position,

were encountered and found very

_reproducible

at all

485

Neither of these

_microscopes

had ultra

_high

vacuum ; the vacuum in the 1 MV Grenoble instrument is estimated to be 2 x

10-7

Torr.

Since the films were

_unprotected,

oxidation was

_expected,

and

_Secondary

Ion Mass

Spectroscopy

(SIMS)

was

_performed

to ascertain its extent as well as the actual

_composition

of the

_films,

_using

6 keV xenon ions on the Cameca-Riber instrument of SEA at CEN-G. We

determined the concentration

_profiles

of

terbium, cobalt,

oxygen, chlorine and

hydrogen,

as well as the

_degree

of oxidation as revealed

_by

the ratio between the ion intensities collected with and without oxygen

background ;

the terbium/cobalt concentrations were normalized

_by

comparison

with

_crystalline

_{TbCo2 kindly}

_{provided by}

D.

_Gignoux.

Magnetic

measurements were

_performed

on a

_{SQUID magnetometer.}

The

_specimen

was a coupon about 1

cm 2

, floated off the

rocksalt,

and

supported

on an aluminum

grid.

The

integrity

of the film was

_roughly

_retained,

_although

a few holes

_appeared

due to surface tension effects

_during

_drying.

Results.

Because the

_experiments

were

spread

over a rather

long

time,

and oxidation

_effects

_proved

non

negligible,

we indicate in

brackets,

_for

each

_experiment,

the number

of days

that had

elapsed

since

specimen

preparation.

Figures

1,

2 and 4a

_show,

at room

_temperature,

the

_{striking variety}

of

configurations

obtained with

_varying

film thickness.

_Figure

1

_{[166 d]}

shows a

_{specimen heavily}

thinned

_by

ion-milling ;

the domain walls and

_ripple

are reminiscent of

classical,

low

_anisotropy

materials,

and indeed the film behaves like a soft

_magnetic

material : when the

_specimen

was

slightly

tilted with

_respect

to the

_optical

axis,

wall movement, due to the

_in-plane

_component

(a

few

_Oersted)

of the remanent

_magnetic

field of the unused

_objective

lens,

was

_{observed ;}

this behavior was not observed on thicker

_specimens

at room or low

_temperature.

_Figure

2

[71 d]

shows an ion-milled

_sample

with a

_quite

_{inhomogeneous}

thickness as evidenced from the differences in

_density

on the

_{original plate,}

which were

_{deliberately dodged}

out under the

enlarger

to allow details to be

_{visible ;}

the transition to a feather-like

_{configuration}

is clear in the thicker

part.

Figure

3

_[1013

_d]

is

_typical

of the unthinned

_specimen

in the «

_virgin

»

condition,

i.e. before

any

heating

above room

_temperature

was

_performed.

It was taken under the conditions where the best resolution can be

_expected,

viz. with low

_{defocusing (for}

details to be

_observable),

Fig.

1. -

Magnetic

structure as seen

_by

Lorentz electron

_microscopy,

at room _temperature, on thin area

Fig.

2. -

Region

with

_highly

_{inhomogeneous}

thickness, differences in

_{density deliberately dodged}

out

on

_{enlarging :}

the _top

_right

_part is much thinner than the bottom left _{part ;} 100 kV.

and at low

temperature,

where the

_{magnetization}

is

_largest

_(for

contrast to

_occur).

It shows rather

_anisotropic

domains,

with an

_elongation

direction

_{roughly along}

the direction of

creases in the

_film,

visible as broad _{gray lines} on the

_edge,

which were

_parallel

to the bars of

the

_grid

on which the film was

_{deposited, probably}

due to surface-tension induced stress that

occurred when the film was

_picked

_{up with the}

_grid

from the beaker on which it floated.

The characteristic distances are

_obviously

on the order of

10- 7

m in width and

10-6

m in

length.

Very small-angle

diffraction in the electron

_microscope

from a selected area

_including

a wall is a standard way of

_displaying

the distribution of the

_in-plane

_component

of

magnetization :

for normal

_{ferromagnetic}

films it

_provides

two fine

_spots

with a

_separation

proportional

to the

_product

of the

_in-plane

_spontaneous

_{magnetization}

times the film thickness. We

observed,

in the thinned

_specimens,

broad

_spots

_{with very small}

_separation,

implying large

fluctuations in the

_in-plane

_component

of

_{magnetization.}

_Tilting

the

_specimen

with

_respect

to be beam did not increase the

_spot

_{separation, indicating}

that in this case the situation is not one with

_{magnetization}

almost

_{perpendicular}

to the film. The same kind of observation could not be carried out for thicker

_specimens

in the

_{corresponding}

_virgin

state

because the domains were then too small and it was

impossible

to have a

_single

wall in the beam.

On

_cooling

the thinnest

_regions

down to 115

K,

the walls remained unaltered in

_shape

and

position,

and no contrast reversal was observed.

487

-Fig.

3. - Low-defocus observation of domains

at 34 K in unthinned film. 800 kV.

room

_temperature

and about 20

_K,

_{but very detailed examination revealed that the} contrast

of each wall was reversed. The

_{corresponding pictures}

are not shown here because this

contrast _{reversal is very difficult} to see on such a fine

_scale,

_{and appears in much clearer form} below.

Figure

4

_[1016

_d]

consists of

_pictures

taken with constant

_{underfocusing}

at various

temperatures.

Due to the variation in

_{magnetization,}

contrast

_{changes appreciably.}

It shows the

_effect,

_starting

from room

_temperature

(Fig. 4a),

of

_heating

to about 500 K

_{(Fig. 4b),}

then

cooling again

to room

_temperature

(Fig. 4c)

and to 34 K

_{(Fig. 4d)}

an unthinned

_sample.

The mottled

_image,

characteristic of

_closely

knit walls at this

_{defocusing, prevails}

at room

temperature

in the initial condition.

On

_raising

the

_temperature,

a distinct

_change

to a structure with more

_{broadly separated}

walls is observed. This structure remains when the

_sample

is

_brought

back to room

temperature.

Observation of the successive

_pictures

suggests,

and the video

_recording

confirms,

that this transition occurs

_{irreversibly through}

the sudden

_{disappearance}

of more

and more wall

_segments,

while

practically

no

_{rearrangement}

of the

_remaining

_segments

is

seen.

On

_{cooling again,}

the walls remain

_{practically unchanged.}

Contrast reversal for each wall is now

_clearly

observed

_although

the

_defocusing

is constant.

_Actually

the contrast at low

temperature

is

_{disturbingly high,}

because the

_spontaneous

_{magnetization}

is now much

_larger

than at room

_temperature.

Hence the

_picture

at 157 K

_{(Fig. 4e)}

is easier to _{compare with the}

Fig.

4. - Effect of

temperature

cycling

on unthinned

_specimen,

observed at 800 kV under constant

underfocusing. a)

Room température ;

b)

T = 500 K ;

c)

_room

température :

d)

T = 34 K ;

e)

T = 157 K.

Figure

5

_{[1015 d]}

shows the same behavior in more

_spectacular

form : the

_sample

_(a

different one than for

_Fig.

₄₎

_{had been heated up} to 690

_K,

and as a result the structure was further

_{simplified :}

_{the domains look very much} more « normal », and white walls at room

temperature

become black at low

_temperature

at constant

_defocusing.

It can be noted that the diffraction

_diagram

of the

_sample

that was

_cycled

several times above room

_temperature

and up to 690 K showed some

_{changes, probably}

indicative of structural

_relaxation,

with

_respect

to the

_{virgin amorphous specimen,}

but no clear indication of

_{crystallization.}

489

Fig.

5. -

Contrast reversal, at constant

_defocusing,

associated with

_compensation

_point

in unthinned

specimen

that was _{heated up} to 690 K.

_a)

T = 345 K ;

b)

T = 30 K.

at 230 K in

_TbCo3.8

_by

McGuire and Hartmann

_{[6] (there}

is however a

_discrepancy

with the

compensation

temperature

shown as about 50 °C for

_TbCo3

_by

Shieh,

Yamasaki and

_Kryder

[17]

from

_{magneto-optical}

and

_spontaneous

Hall effect

_{measurements).}

When the

_specimen

is

gradually

warmed back to room

_temperature,

contrast

_decreases,

_disappears

around 260 K in these unthinned

films,

and reappears in reversed form. We note that in the thinnest ion-milled

fragments

no contrast

reversal,

hence no

_{compensation point,}

was observed

_{[156 d]}

between

room

_temperature

and 125 K.

The SIMS

_analysis

_{[1358 d]}

showed that the films have a

_{composition corresponding}

to

TbCo3,

with oxidation of about 10 % for terbium and 5 % for

cobalt,

except

in

_regions

within about 1 x

10- 8

m from the

surfaces,

where the terbium concentration

drops

and the oxidation

rate climbs. This

_composition

is in

_good

agreement

with the

_compensation

_temperature

observed.

After an unthinned

_sample

was

_{visibly recrystallized,}

as evidenced both

by

the diffraction

diagram

and

_by

dark-field observations of the

_{crystallites,}

but also

_possibly

further

oxidized,

Switching

the

_(magnetic)

_objective

lens on

_temporarily

had no visible effect on the domain

structure in the unthinned

_specimens.

To determine the extent of

_{perpendicular anisotropy,}

measurements were

_performed

[1.8

x

103

_d]

_]

on a

_{SQUID magnetometer}

_equipped

with a

_sample

rotation

_possibility

and

detection of the

_magnetic

moment

_components

_along

two

_{perpendicular}

directions. The

application

of

_magnetic

field

_{perpendicular}

to the film

_appeared

unwise,

both because of the

chance of the film

_moving

and of the effect of

_eddy

currents in the aluminum

_{plate holding}

the

film,

and because of the

_{expected difficulty}

in

_correcting

for

_{demagnetizing}

field in this

complicated

situation. The evolution of the direction in space of the remanent

_magnetic

moment, induced

by

the

_temporary

_application

of a

_magnetic

field in

_plane,

was monitored as a function of

_temperature

around the

compensation point,

where the

demagnetizing

field

vanishes. A rotation of the

_magnetic

moment from

_in-plane

towards the

_{perpendicular}

direction was observed

extremely

close to the

_{compensation point.}

Since this transition occurs

when the uniform

_anisotropy

is of the order of the

_{demagnetizing}

field,

the conclusion was

that

_{perpendicular anisotropy}

is

_{negligible compared}

to the local random

_anisotropy

in these films.

Discussion.

The Tb-Co

system

is

_slightly

different from the random

_anisotropy

_magnet

considered in theoretical work because it is a

_sperimagnet

in which the cobalt moments are

expected

to be

largely aligned

due to the

_high

Co-Co

_exchange,

and

_only

the terbium moments will be induced

_by

random

_anisotropy

to take on orientations

_spread

over a

_large

_part

of a

half-sphere

relative to the Co moment direction

_[18].

Furthermore,

as the SIMS

analysis

showed,

our films could

_definitely

be considered neither as

_homogeneous

nor as free of oxidation : the terbium losses from the latter result in the formation of a

_highly

enriched cobalt-like

_layer

on the

_surfaces,

i.e. on either side of the film

of

_composition

close to

_TbCo3.

In

fact,

careful

_{SQUID magnetometer}

measurements

performed

on the unthinned

_samples

with

_magnetic

field

_in-plane

showed

_complicated

hysteresis loops

including crossing points,

reminiscent of the results of simulations for

multilayers involving

a

_{random-anisotropy}

film sandwiched between two soft

_{ferromagnetic}

ones

_[19].

The observation of

_{qualitatively}

different domains with soft

_magnetic

behavior at room

temperature

and no

_{compensation point}

on the thinnest

_{specimen fragments}

confirms that their

_{magnetization}

is cobalt-dominated. This is borne out

_by

the SIMS

_profiles,

since the

observed

_region

is near one of the surfaces.

The

_{magnetization}

distribution in the thicker films

_{is, however,}

_strikingly

disordered. From these observations

alone,

the choice would remain of either

_assigning

this to _{the presence of}

high

perpendicular

anisotropy,

with fluctuations around the direction

_{perpendicular}

to the film

_{(bubble-related domains),}

or to the effect of random

anisotropy

(Imry

and Ma-like

domains).

The

_magnetic

measurements make it clear that this is not a

_{perpendicular-anisotropy}

491

where

_Ra

_{is the range of the}

_short-range

structural

order, K

the

_anisotropy

constant and A the

_exchange

constant, for a

_speromagnet.

For our

_sample,

the ratio

_D/J

could be evaluated from

_{magnetization}

curves and should

roughly

be of the order of 0.1

_[15].

Another

_approach

to the evaluation

_{of J/D}

would be to note that in rare earth-cobalt

_alloys

the

_following

orders of

_magnitude

are

_usually

valid :

Since the dominant

_exchange

interaction is

_Jc.-c.

and the relevant

_anisotropy

_energy

D - JRE-C. -

DRE

because

_{non-colinearity}

of the Co moments with the

_half-sphere

axis of the RE moments is _{involved in rotation processes,} we

_get

_{DIJ - JRE-Co/jCo-Co -}

0.1.

The value for

_R

_agrees

_quite

well with the observed size of the fine structure seen most

clearly

in

_figure

3. This indicates that this structure

_corresponds

to the

_spatial

fluctuations associated with

_Imry

and Ma

_domains,

thus observed

_{directly probably}

for the first time. Their scale in this

_{low-anisotropy}

_system

is rather

_large,

which is

_why

we can see them with the

_{comparatively}

modest resolution we have in the defocused

mode,

and

_beyond

the range of

small-angle scattering experiments.

We note that for such sizes

_dipolar

fields cannot be

_neglected,

and

_they

should contribute both to the size and the

_shape

of the domains

_together

with the

_exchange

and

_anisotropy

energies

[6, 20].

The

_coupled

effect of a coherent

uniaxial,

probably

stress-induced,

anisotropy

and of the

dipolar

interaction is

_certainly

the

_origin

of the

_{anisotropic shape}

of the domains in

_figure

3. The irreversible but

_{gradual change}

of the

_{magnetic configuration}

on

_heating

above room

temperature

is

_likely

to be associated with relaxation to a condition associated with reduced random local

_anisotropy.

The fact that the

_{magnetization}

distribution,

whether

_simplified

or

not, does not

_change

when

_lowering

the

_temperature

_points

to effective

_pinning.

Two

possibilities

can be invoked to

_explain

the

_changes.

One is

that,

as the

temperature

is

raised,

the random

_anisotropy

decreases and the coherent

_anisotropy

becomes

dominant,

leading

to

ferromagnetic

domains with Bloch walls. The

_{magnetic configuration}

could then still be

thought

of as

_{being globally}

locked

_through

_{the presence of}

_{topological potential}

walls

separating valleys

on a

_mesoscopic

scale

_{[21, 22].}

The second

_possibility

is

_simply

that

_gradual

oxidation leads to increased terbium losses

_[23-25],

hence to a decrease in

_DIJ

which is

_highly

irreversible.

Along

with this first observation of

_Imry

and Ma

domains,

the reversal of domain wall

contrast across the

_compensation

_temperature

is

_visually

a _very

_spectacular

result. It

_implies

that there is no

_{rearrangement}

of the domains when the

_{macroscopic magnetization changes}

drastically

and in

_particular

_goes to zero. The

_dipolar

effects are thus

_{obviously negligible}

in

comparison

with the

_pinning

mechanism.

Acknowledgments.

The authors are

_happy

to thank R. H. Wade for valuable discussions and

_help,

I. B. Puchalska for advice on substrate

_preparation,

D. Givord and B.

_Dieny

for

_helpful

suggestions,

and to

_acknowledge

an

_{extremely illuminating}

short discussion with E. M.

References

[1]

IMRY Y., MA S.-K.,

Phys.

Rev. Lett. 35

₍₁₉₇₅₎

1399-1401.

[2]

AHARONY A., PYTTE E.,

Phys.

Rev. B 27

₍₁₉₈₃₎

5872-5874.

[3]

CHUDNOVSKY E. M., SEROTA R. A., J.

_Phys.

C 16

₍₁₉₈₃₎

4181-4190.

[4]

AHARONY A., PYTTE E.,

Phys.

Rev. Lett. 45

₍₁₉₈₀₎

1583-1586.

[5]

CULLEN J., J.

_Appl.

_Phys.

61

₍₁₉₈₇₎

4413-4415.

[6]

CHUDNOVSKY E. M., SASLOW W. M., SEROTA R. A.,

Phys.

Rev. B 33

₍₁₉₈₆₎

251-261.

[7]

SASLOW W. M.,

Phys.

Rev. B 35

₍₁₉₈₇₎

3454-3460.

[8]

CHUDNOVSKY E. M., J.

_Appl.

Phys.

64

₍₁₉₈₈₎

5770-5775.

[9]

GRUNDY P. J., LACEY E. T. M., WRIGHT C. D., JMMM 54-57

₍₁₉₈₆₎

227-228.

[10]

SUZUKI T., JMMM 31-34

₍₁₉₈₃₎

1009-1010.

[11]

TAKENO Y., SUWABE M., SAKURAI T., GOTO K.,

Jpn

J.

_Appl.

_Phys.

25

₍₁₉₈₆₎

L657-L660.

[12]

SUITS J.

C.,

GEISS R. H., LIN C. J., RUGAR D., BELL A. E., J.

Appl. Phys.

61

₍₁₉₈₇₎

3509-3513.

[13]

RHYNE J. J., IEEE Trans. magn. MAG-21

(1985)

1990-1995.

[14]

LEE J., HEIMAN N., AIP

_Conf.

Proc. 24

₍₁₉₇₅₎

108-109.

[15]

GIVORD D., LIÉNARD A., REBOUILLAT J. P., SCHWEIZER J.,

Physica

130B

₍₁₉₈₅₎

351-354.

[16]

MCGUIRE T. R., HARTMANN M., IEEE Trans.

_Magn.

MAG-22

₍₁₉₈₆₎

1224-1226.

[17]

SHIEH H.-P. D., YAMASAKI J., KRYDER M. _H., IEEE Trans.

Magn.

MAG-23

₍₁₉₈₇₎

3208-3210.

[18]

COEY J. M. D., CHAPPERT J., REBOUILLAT J. P., WANG T. S.,

Phys.

Rev. Lett. 36

₍₁₉₇₆₎

1061-1064.

[19]

DIÉNY _B., GIVORD D., N’DJAKA J. M., ALAMEDA J., to be

_published

in Proc. Conf.

_Magn.

and

Magnet.

Mater., Boston

_{(Nov. 1989).}

[20]

DIENY B., BARBARA B., J.

_Phys.

France 46

₍₁₉₈₅₎

293-302.

[21]

HENLEY C. L., Ann.

_Phys.

N. Y. 156

₍₁₉₈₄₎

368-411.

[22]

MOLHO P., PORTESEIL J. L., SOUCHE Y., GOUZERH J., LEVY J. C. S., J.

_Appl.

Phys.

61

₍₁₉₈₇₎

4188-4193.

[23]

GRUNDY P. J., LACEY E. T. M., WRIGHT C. D., IEEE Trans.

Magn.

MAG-23

₍₁₉₈₇₎

2632-2634.

[24]

HARTMANN M., KLAHN S., WITTER K., IEEE Trans.

_Magn.

MAG-23

₍₁₉₈₇₎

2946-2948.

[25]

MARINERO E. E., MILLER, D. C., BELL A. E., GUPTA A., PAYNE R. N., NOTARYS H., IEEE