E ec o he Si/B a io on he magne ic aniso opy dis ibu ion
o Fe73.5Si22.52
x
B
x
Cu1Nb3„
x
57,9,16…alloys along nanoc ys alliza ion
V. F anco, C. F. Conde, and A. Condea)
Depa amen o de Fı
´sica de la Ma e ia Condensada, Uni e sidad de Se illa, Apdo. 1065,
41080 Se illa, Spain
~Recei ed 20 Ma ch 1998; accep ed o publica ion 14 July 1998!
The e ec o he Si/B a io on he magne ic aniso opy dis ibu ion o Fe73.5Si22.52xBxCu1Nb3
(x57,9,16) alloys has been s udied. The in luence o isoch onal annealing on he hys e esis loop o
he h ee s udied alloys has been analyzed. They p esen wo minima in coe ci i y: he i s one can
be asc ibed o s uc u al elaxa ion, and he second one is ela ed o he a e aging o he
magne oc ys alline aniso opy, as p edic ed by he andom aniso opy model. The ela i e
impo ance o bo h minima changes wi h Si con en : he lowe he Si con en , he mo e e ec i e he
s uc u al elaxa ion and he less impo an he second minimum a e. The mean alue o he
magne ic aniso opy dis ibu ion p esen s a simila beha io , e idencing he g owing impo ance o
he magne oelas ic aniso opy o he elaxed amo phous samples as he Si con en is inc eased.
F om he e olu ion o he magne ic aniso opy dis ibu ion along nanoc ys alliza ion and he
mic os uc u al in o ma ion ob ained om ansmission elec on mic oscopy images, he beha io
o he coe ci i y minima wi h changes in Si con en can be asc ibed o a di e en deg ee in
compensa ion o magne oelas ic aniso opy due o he con ibu ions o di e en signs coming om
he nanoc ys als and he amo phous ma ix. © 1998 Ame ican Ins i u e o Physics.
@S0021-8979~98!04820-8#
I. INTRODUCTION
Since i s disco e y by Yoshizawa e al.,1 he amo phous
e omagne ic alloy known unde he ade name o
FINEMET has a ac ed much scien i ic in e es due o i s
ou s anding so magne ic p ope ies. This alloy, wi h a com-
posi ion o Fe73.5Si13.5B9Cu1Nb3, is sui able no only o
many echnological applica ions bu o pe o ming a consid-
e able amoun o undamen al s udies in magne ism. I s so -
es p ope ies a e ob ained a e nanoc ys alliza ion, while i
is cons i u ed o wo phases: an ul a ine g ained Fe,Si phase
which is embedded in he emaining amo phous ma ix.
The ole o he Cu and Nb elemen s in he achie emen
o his mic os uc u e and i s ela ionship wi h he magne ic
p ope ies has been ho oughly s udied.2,3 Coppe , which is
nonsoluble in bcc i on, seg ega es p io o o a he e y
beginning o nanoc ys alliza ion, o ming Cu- ich clus e s,
and he nuclea ion o Fe,Si g ains is hough o be mul iplied
by clus e ing. On he o he hand, he ejec ion o Nb a he
c ys al in e aces hinde s g ain g ow h. The e ec o he sub-
s i u ion o coppe by o he elemen s which a e soluble in
bcc Fe has also been epo ed,4,5 con i ming he impo ance
o he nonsolubili y o Cu o he achie emen o his special
mic os uc u e and i s cha ac e is ic magne ic p ope ies.
The e has also been some s udies de o ed o he co ela ion
o he magne ic p ope ies wi h he change o he Si/B a io
in he alloy,6–8 o wi h he Fe/Nb a io.9
Recen ly, changes in magne ic aniso opy dis ibu ion
upon hea ea men o a nanoc ys alline ma e ial ha e been
p esen ed.10 These esul s all ou o he capabili ies o He z-
e ’s model.11 In o de o explain he magne ic ha dening
which akes place a he e y beginning o nanoc ys alliza-
ion, a gene aliza ion o he andom aniso opy model has
been de eloped12 ha akes in o accoun he wo-phase cha -
ac e o nanoc ys alline ma e ials. This model can also ac-
coun o he e olu ion o he magne ic aniso opy dis ibu-
ion along he i s c ys alliza ion s age.
This wo k is de o ed o he i s c ys alliza ion s age o
Fe73.5Si22.52xBxCu1Nb3(x57,9,16) and o he in luence o
he Si/B a io on he magne ic p ope ies o he alloys along
nanoc ys alliza ion. Al hough FINEMET- ype alloys a e
usually ega ded as magne ically so e a e nanoc ys alliza-
ion han in he amo phous s a e, i can be easily seen ha
sligh changes in composi ion make his sen ence un ue.
The e has been p e ious s udies o he magne ic p ope ies o
hese alloys; howe e , o ou knowledge none o hem p e-
sen ed ei he he in luence o he changes in composi ion on
he dis ibu ion o magne ic aniso opy o i s e ec on he
ela i e impo ance o he coe ci i y minima ~be o e and a -
e nanoc ys alliza ion!. As magne ic aniso opy is exploi ed
in he design o mos magne ic ma e ials o comme cial in-
e es , i s dis ibu ion in he FINEMET- ype alloys is wo h
s udying. Also, his s udy will p o ide us wi h aluable in-
o ma ion abou he alloys hemsel es along nanoc ys alliza-
ion and will suppo an explana ion o he composi ional
dependence o he ela i e impo ance o he wo coe ci i y
minima.
a!Elec onic mail: [email p o ec ed]
JOURNAL OF APPLIED PHYSICS VOLUME 84, NUMBER 9 1 NOVEMBER 1998
51080021-8979/98/84(9)/5108/6/$15.00 © 1998 Ame ican Ins i u e o Physics
II. EXPERIMENT
Amo phous ibbons o alloys wi h nominal composi ions
o Fe73.5Si22.52xBxCu1Nb3(x57,9,16) we e p epa ed by a
mel -spinning echnique in he Ins i u e o Physics o he
Slo ak Academy o Sciences in B a isla a.
Specimens o ansmission elec on mic oscopy ~TEM!
s udies we e p epa ed by using a win-je elec ochemical
polishing de ice wi h a 10% pe chlo ic acid–ace ic acid so-
lu ion. TEM obse a ions we e pe o med on an Hi achi
H-800 elec on mic oscope wi h an accele a ing ol age o
200 kV.
Quasis a ic M-Hhys e esis loops we e measu ed a
oom empe a u e by using a compu e ized loop ace de el-
oped a ou labo a o y.13 Samples 10 cm long, 25
m
m hick,
and 1 cm wide we e p e iously submi ed o 1 h isoch onal
annealing in a halogen-lamp u nace unde a gon a mo-
sphe e.
The magne ic aniso opy dis ibu ion can be ob ained di-
ec ly om he magne iza ion cu e.14 To do ha , le us con-
side a dis ibu ion o uniaxial local easy axes wi h di e en
s eng h bu all di ec ed a igh angles wi h espec o he
applied ield. The aniso opy ield HKcan be de ined as he
ield necessa y o sa u a e a uniaxial pa icle a igh angles
o i s easy axis and is ela ed o he aniso opy cons an Kby
2K5
m
0MsHK,~1!
whe e Msis he sa u a ion magne iza ion. Thus, a egion o
such an ensemble o easy axes, cha ac e ized by an aniso -
opy ield HK, unde he in luence o a ield Hpe pendicula
o he easy axes will magne ize in he ollowing way:
M~H,HK!5Ms3~H/HK!
M~H,HK!5Ms
o H,HK,
o H.HK.~2!
Le us de ine P(HK) as he no malized p obabili y o
inding he alue HK o he aniso opy ield. By no maliza-
ion we mean ha he o al p obabili y o inding an aniso -
opy alue g ea e han ze o is uni y
E
0
`P~HK!dHK51. ~3!
Consequen ly, he mac oscopic educed magne iza ion
(m5M/Ms) o he whole ensemble will be
^
m~H!
&
5m 1
E
0
HP~HK!dHK1H
E
H
`P~HK!
HKdHK,~4!
whe e he i s e m is he educed emanence (m
5M /Ms), he second one accoun s o he sa u a ed e-
gions (H.HK), and he las one is he con ibu ion o he
unsa u a ed egions. Equa ions ~3!and ~4!assume alues o
he aniso opy ield up o `; howe e , his would co espond
o an in ini e aniso opy cons an . Al hough his alue would
be un easonable, he no maliza ion o he dis ibu ion im-
plies ha i ends o ze o as he ield ends o in ini y. I
should be no ed ha in he case whe e he e we e easy axes
pa allel o he applied magne ic ield, hey would con ibu e
equally o he emanence and ha e no in luence on he de-
ined dis ibu ion P(HK).
Unde hese assump ions, he dis ibu ion unc ion o he
pe pendicula magne ic aniso opy can be ob ained om he
second de i a i e o he educed magne iza ion
P~HK!52HKd2
^
m
&
dH2.~5!
In o de o apply his me hod o he samples s udied, we
ha e o ake in o accoun he wo possible o ien a ions o he
easy axes in he amo phous ibbons: due o he geome ical
and dimensional cha ac e is ics o he samples, hey ha e an
easy axis di ec ed along he ibbon axis. The o he possibili y
o easy axis is pe pendicula o he ibbon axis and is due o
he s esses p oduced in he manu ac u ing o he samples.
As he ield will be applied along he ibbon axis, he p e i-
ously desc ibed me hod will only p o ide in o ma ion abou
he easy axes ha a e no mal o he ibbon plane, as he o he
di ec ion con ibu es only o he emanence.
A e hea ea ing he ibbons in he ange o he i s
c ys alliza ion s age, Fe,Si cubic nanoc ys als appea in
hem. Howe e , due o hei small size (;10 nm) and o he
exchange coupling ac oss he emaining e omagne ic amo -
phous ma ix, he whole sample can s ill be ep esen ed as an
ensemble o easy axes di ec ed a igh angles o he ibbon
axis. The small alues o he emanence o sa u a ion mag-
ne iza ion a io10 suppo he assump ion o small de ia ions
o he easy axes wi h espec o he di ec ion conside ed. I
he easy axis o he aniso opy is no exac ly pe pendicula o
he di ec ion o he ield, one can expec o ind a b oadening
o he magne ic aniso opy dis ibu ion e en o a single
alue o HKand, in pa icula , a dec ease in he symme y o
he dis ibu ion, mainly in he high ield egion.
When he samples a e annealed in he ange o he sec-
ond c ys alliza ion s age, he appea ance o bo ide- ype
phases causes he emanence o inc ease ab up ly, indica ing
ha he assump ion o uniaxial easy axes wi h p e e ed o i-
en a ions is no longe alid. Mo eo e , o samples wi h
la ge hys e esis, he magne iza ion cu e is a ec ed by he
i e e sible magne iza ion p ocesses, so he me hod loses ac-
cu acy. The e o e, he p esen s udy is limi ed o he i s
c ys alliza ion s age.
III. RESULTS
A. Hys e esis loops
The in luence o he annealing empe a u e on he coe -
ci e ield o he alloys s udied is p esen ed in Fig. 1. I can be
seen ha o empe a u es below he c ys alliza ion onse ,
coe ci i y diminishes wi h annealing empe a u e o he
h ee alloys. This magne ic so ening in he ange o
empe a u e/ ime below he beginning o he iso he mal
nanoc ys alliza ion mus ha e i s o igin in s uc u al elax-
a ion, as in con en ional e omagne ic alloys. I is e i-
denced ha he s uc u al elaxa ion is mo e e ec i e as he
silicon con en is dec eased, as can be seen om he lowe
alues o he coe ci i y.
He nando’s model p edic s a magne ic ha dening when
annealing a sligh ly highe empe a u es, co esponding o
he ea ly s ages o nanoc ys alliza ion. This is due o he low
olume ac ion o he c ys alli es ha makes he a e age
5109J. Appl. Phys., Vol. 84, No. 9, 1 No embe 1998 F anco, Conde, and Conde
dis ance be ween g ains much longe han he exchange co -
ela ion leng h be ween hem ac oss he amo phous ma ix
and p e en s he magne oc ys alline aniso opy om being
a e aged ou . The i s inc ease o coe ci i y is well co e-
la ed wi h he onse o nanoc ys alliza ion, and is ha dly de-
ec able o he alloy con aining 16 a % Si, becoming mo e
impo an as he Si con en is dec eased. I is wo h no ing
ha he olume ac ions ha p o oke his ha dening a e so
small ha hey canno be de ec ed by x- ay di ac ion.
When he annealing empe a u e and consequen ly he
olume ac ion o he c ys alli es is inc eased, He nando’s
model ends o he andom aniso opy model and he e ec-
i e a e aging o he magne oc ys alline aniso opy should
cause he coe ci e ield o dec ease. This dec ease is mo e
p ominen in he alloy con aining 16 a % Si and i s magni-
ude loses impo ance as he Si con en dec eases, becoming
unde ec able in he alloy wi h 7 a % Si. I should be no ed
ha he in luence o he magne oelas ic aniso opy could ac-
coun o his e ec , as will be poin ed ou la e .
The inal ab up inc ease o coe ci i y is ela ed o g ain
g ow h and he appea ance o bo ide- ype phases a he be-
ginning o he second c ys alliza ion s age.
B. Magne ic aniso opy dis ibu ion
The e olu ion o magne ic aniso opy dis ibu ion wi h
annealing empe a u e o he h ee alloys s udied is p e-
sen ed in Figs. 2–4. Samples con aining 16 a % ~Fig. 2!and
9a %Si~Fig. 3!p esen a mo e simila beha io . Fo bo h
o hem, when he samples a e annealed a empe a u es be-
low he onse o nanoc ys alliza ion ~and consequen ly below
he inc ease o he coe ci e ield!, he dis ibu ion becomes
na owe due o s uc u al elaxa ion, which gi es ise o a
smoo hing o he in e nal s esses. When he onse o nano-
c ys alliza ion is eached, he dis ibu ion begins o shi o
highe aniso opy ields and un olds in a ious maxima. The
eme ging c ys alli es cause he in e nal s esses o no be he
only o igin o mac oscopic aniso opy, bu his now has a
new s uc u al componen . C ys alli es also de elop new
s esses in he sample, induced by hem in he amo phous
ma ix. These g ains canno be exchange coupled as he ex-
change co ela ion leng h h ough he amo phous ma ix is
oo sho o cause he dis an g ains o in e ac , and he e o e
aniso opy canno be a e aged ou . Unde his condi ion,
c ys alli es ac as pinning cen e s o he domain walls and,
consequen ly, magne ic ha dening akes place.
Fu he annealing causes he c ys allized olume ac ion
o inc ease and, consequen ly, he a e age dis ance be ween
g ains is educed. This ac makes he dis ibu ion e u n o
lowe aniso opy ields due o he p og essi e a e aging o
he magne oc ys alline aniso opy. I should be no ed, how-
e e , ha he magne ic aniso opy dis ibu ion o he nano-
c ys allized samples is di e en in bo h cases. The sample
which con ains mo e silicon, when he op imum a e aging o
aniso opy is eached (Ta5825 K), has a qui e na ow dis-
ibu ion ha consis s o only a single maximum. On he
o he hand, he sample con aining 9 a % Si can only each a
FIG. 1. Coe ci e ield dependence wi h Si con en s and annealing empe a-
u es.
FIG. 2. E olu ion o no malized magne ic aniso opy dis ibu ion wi h an-
nealing empe a u e o he alloy con aining 16 a % Si ~ e ical o se is
p o ided o make he igu e clea e !. Fo he sample annealed a 475 K he
do ed line is he ze o-le el line.
FIG. 3. E olu ion o no malized magne ic aniso opy dis ibu ion wi h an-
nealing empe a u e o he alloy con aining 9 a % Si. Fo he sample an-
nealed a 475 K he do ed line is he ze o-le el line.
5110 J. Appl. Phys., Vol. 84, No. 9, 1 No embe 1998 F anco, Conde, and Conde
dis ibu ion o magne ic aniso opy ha is wide and wi h a
mean alue ha is highe han ha o he elaxed amo phous
one.Finally, he beginning o he second c ys alliza ion s age
p o okes a new b oadening o he dis ibu ion, and he
maxima end o un old again. The ha dening o igina ed by
he appea ance o bo ide- ype phases causes an inc ease no
only in he coe ci i y o he samples, bu also in he ema-
nence o he sa u a ion magne iza ion a io. This ac causes
he me hod o lose accu acy, as could be seen by he loss o
symme y o he dis ibu ion when ob ained om he di e -
en b anches o he loop. Consequen ly, he second c ys alli-
za ion s age canno be s udied wi h his p ocedu e.
The sample con aining 7 a % Si ~Fig. 4!has, a i s
sigh , a di e en beha io . Al hough he s ess elaxa ion
p ocess has he e ec o na owing he dis ibu ion, once he
c ys alliza ion onse is eached, he dis ibu ion ends o g ow
wide and he mac oscopic aniso opy does no dec ease
along he i s c ys alliza ion s age.
IV. DISCUSSION
The i s i em wo h no ing is he con inuous b oadening
o he magne ic aniso opy dis ibu ion o he sample con-
aining 7 a % Si once he nanoc ys alliza ion begins. This
could seem o be in con adic ion wi h he andom aniso opy
model o i s gene aliza ion, p edic ing an a e aging o he
magne oc ys alline aniso opy when he olume ac ion in-
c eases. I should be emembe ed, howe e , ha he e a e
wo con ibu ions o he mac oscopic aniso opy a his s age:
he magne oc ys alline and he magne oelas ic. The o me
can be a e aged, as p edic ed by he models, bu he o e all
alue o he aniso opy indica es ha he e is an impo an
con ibu ion o he la e .
Ano he impo an poin is he change in dep h o he
coe ci i y minima ~be o e and a e nanoc ys alliza ion!wi h
Si con en . When he mean alue o he dis ibu ion is ep-
esen ed e sus he annealing empe a u e o he h ee
samples s udied ~Fig. 5!, i is obse ed ha he dep h o he
second minimum, which is he alue o
^
HK
&
co esponding
o he op imum a e aging a e nanoc ys alliza ion, is
smalle when he Si con en is dec eased, o be unde ec able
o 7 a %. On he o he hand, he beha io o he i s mini-
mum ~ elaxed amo phous!is he opposi e: i s dep h is in-
c eased as he Si con en is dec eased. This is in ag eemen
wi h he e ec obse ed in coe ci i y and, consequen ly, he
o igin mus be he same.
The plo o he magne ic aniso opy dis ibu ion o he
op imally elaxed amo phous samples ~Fig. 6!clea ly shows
ha no only is he mean alue inc eased, inc easing he Si
con en , bu he wid h o he dis ibu ion is also inc eased.
This e idences a g owing impo ance o he magne oelas ic
aniso opy when he Si con en is inc eased. I we assume
ha once he op imum elaxa ion is achie ed he h ee
samples p esen a simila alue o he esidual s esses, he
p esen ed cu es can indica e an inc ease o lso he amo -
phous phase, inc easing he Si con en . Some esul s epo ed
o simila composi ions (Fe74.5Nb2Cu1SixB22.52x) suppo
his assump ion o he e olu ion o ls.7
As he magne oelas ic aniso opy o he nanoc ys allized
sample is a combina ion o he posi i e con ibu ion o he
amo phous phase and he nega i e one o he nanoc ys als, i
is logical o hink ha he lowe he alue o lso he amo -
phous, he lowe he olume ac ion o c ys alli es ha
would be necessa y o compensa e bo h componen s. Conse-
quen ly, i would be expec ed ha he minimum alue o he
magne oelas ic aniso opy would be eached a lowe c ys-
allized olume ac ions when he Si con en o he sample
FIG. 4. E olu ion o no malized magne ic aniso opy dis ibu ion wi h an-
nealing empe a u e o he alloy con aining 7 a % Si. Fo he sample an-
nealed a 475 K he do ed line is he ze o-le el line.
FIG. 5. Dependence o he mean aniso opy ield wi h annealing empe a-
u e o he h ee alloys s udied.
FIG. 6. No malized magne ic aniso opy dis ibu ion o he op imally e-
laxed amo phous samples: ~squa es!16 a % Si annealed o 1ha 700K;
~ci cles!9 a % Si annealed o 1ha 700K;~ iangles!7 a % Si annealed
o 1ha 675K!.
5111J. Appl. Phys., Vol. 84, No. 9, 1 No embe 1998 F anco, Conde, and Conde
is dec eased. This is in good ag eemen wi h he composi-
ional dependence o he i s ha dening o he samples: o
samples wi h lowe Si con en , he compensa ion o magne-
oelas ic aniso opy occu s a lowe c ys allized ac ions and
a u he inc ease o he olume ac ion o he c ys alli es
p o okes an inc ease in magne oelas ic aniso opy being so
high ha he a e aging o he magne oc ys alline aniso opy
canno be de ec ed. ~I can only be seen annealing a 800 K
o 1 h, bu he dis ibu ion is so wide and la ha no claims
can be made.!When he Si con en is inc eased, he op imum
alue o he c ys allized ac ion o magne oelas ic compen-
sa ion is close o ha a which he magne oc ys alline an-
iso opy is a e aged and, consequen ly, he dec ease o he
mean alue o he dis ibu ion can be seen mo e easily and is
deepe as he Si con en inc eases.
To suppo hese s a emen s, he magne ic aniso opy
dis ibu ion o he samples annealed a 775 K o 1ha e
p esen ed in Fig. 7. These a e he same samples whose TEM
images a e displayed in Fig. 8. An inc easing olume ac-
ion o he c ys alli es is obse ed when he Si con en is
inc eased. Al hough he sample wi h 7 a % Si has a lowe
olume ac ion and a smalle mean g ain size, hey a e no
small enough o explain he huge wid h o he dis ibu ion.
The e o e we can claim ha by inc easing he Si con en
~which p o okes an inc ease in olume ac ion!, he dis i-
bu ion is na owed due o he di e en con ibu ions o he
magne oelas ic aniso opy.
I should be no ed ha he alue o lso he amo phous
phase e ol es along nanoc ys alliza ion due o he changes
in composi ion and his mus be aken in o accoun i quan-
i a i e calculus is pe o med. Mo eo e , changes in he Si
con en o he nanoc ys als will also in luence he alue o
hei magne os ic ion cons an .
V. CONCLUSIONS
Th oughou his wo k, he e olu ion o he magne ic an-
iso opy dis ibu ion o Fe73.5Si22.52xBxCu1Nb3(x57,9,16)
upon hea ea men has been p esen ed, along wi h coe ci -
i y da a and TEM obse a ions. The shape and mean alue o
he dis ibu ions a e well co ela ed wi h he mic os uc u al
and magne ic da a. The esul s ob ained all ou o he capa-
bili ies o He ze ’s model and ha e o be explained in ligh
o he gene alized andom aniso opy model.
The p esen ed esul s indica e ha his me hod is a pow-
e ul ool in he s udy o no only he s ess elaxa ion p o-
cess, bu o he e olu ion o he FINEMET- ype alloys along
he i s c ys alliza ion s age. By using he in o ma ion ha i
p o ides, he beha io o he wo coe ci i y minima ~be o e
and a e nanoc ys alliza ion!wi h changes in he Si con en
can be explained: i is asc ibed o he change in he compen-
sa ion o magne oelas ic aniso opy due o he con ibu ions
o di e en signs coming om he nanoc ys als and he
amo phous ma ix.
ACKNOWLEDGMENTS
The au ho s hank D . P. Duhaj and D . P. S ec ~Ins i u e
o Physics o B a isla a, Slo akia! o supplying he ibbons.
This wo k was suppo ed by he CICYT o he Spanish Go -
e nmen ~P ojec No. MAT95-0961-CO2-01!and by he PAI
o he Jun a de Andalucı
´a. One o he au ho s, V.F., is g a e-
ul o he Fundacio
´nCa
´
ma a o he Uni e si y o Se illa o
a esea ch ellowship.
1Y. Yoshizawa, S. Oguma, and K. Yamauchi, J. Appl. Phys. 64, 6044
~1988!.
2Y. Yoshizawa and K. Yamauchi, Ma e . T ans., JIM 31, 307 ~1990!.
3F. an Bouwelen, J. Sie sma, and A. an den Beukel, J. Non-C ys . Solids
156-158, 567 ~1993!.
4C. F. Conde, M. Milla
´n, and A. Conde, J. Non-C ys . Solids 232–234 346
~1998!.
5V. F anco, C. F. Conde, and A. Conde, in Non-c ys alline and Nanoscale
FIG. 7. No malized magne ic aniso opy dis ibu ion o he samples an-
nealed o 1ha 775K.
FIG. 8. TEM images o he samples annealed o 1ha 775K:~a!16 a %
Si; ~b!9 a % Si; ~c!7 a % Si sample.
5112 J. Appl. Phys., Vol. 84, No. 9, 1 No embe 1998 F anco, Conde, and Conde

Ma e ials, edi ed by J. Ri as and M. A. Lo
´pez-Quin ela ~Wo ld Scien i ic,
Singapo e, 1998!, p. 284.
6M. Mu
¨lle , N. Ma e n, and I. Illgen, Z. Me allkd. 82, 895 ~1991!.
7J. Zhi, K. He, L. Cheng, and Y. Fu, J. Magn. Magn. Ma e . 153, 315
~1996!.
8C. F. Conde, V. F anco, and A. Conde, in Rapidly Quenched and Me a-
s able Ma e ials, edi ed by P. Duhaj, P. M a ko, and P. S ec ~Else ie ,
Ams e dam, 1997!, p. 254.
9T. Kulik and A. He nando, J. Magn. Magn. Ma e . 160, 269 ~1996!.
10V. F anco, C. F. Conde, and A. Conde, J. Magn. Magn. Ma e . 185, 353
~1998!.
11G. He ze , IEEE T ans. Magn. 25, 3327 ~1989!.
12A. He nando, M. Va
´zquez, T. Kulik, and C. P ados, Phys. Re . B 51,
3581 ~1995!.
13V. F anco, J. Ramos-Ma os, and A. Conde, Re . Sci. Ins um. 67, 4167
~1996!.
14J. M. Ba andia an, M. Va
´zquez, A. He nando, J. Gonzalez, and G. Ri e o,
IEEE T ans. Magn. 25, 3330 ~1989!.
5113J. Appl. Phys., Vol. 84, No. 9, 1 No embe 1998 F anco, Conde, and Conde