• Nenhum resultado encontrado

Braz. J. Phys. vol.30 número4

N/A
N/A
Protected

Academic year: 2018

Share "Braz. J. Phys. vol.30 número4"

Copied!
4
0
0

Texto

(1)

The 3d Ising Spin Glass

PerNordblad

Department ofMaterialsSiene,

UppsalaUniversity,

Box534, SE-75121Uppsala,Sweden

Reeivedon5August,2000

Experimentallyderivedphysialpropertiesof3dIsingspinglassesare disussedandomparedto

whatis antiipatedfromtheoryandsimulationsofmodelsystems. It isfoundthatthe spinglass

experienes a zero-eld phase transition to a low-temperature spin-glass phase, that there is no

phasetransition inamagnetieldandthat nonequilibriumandhaoseets oftendominatethe

dynamisofthelow-temperatureglassyphase. These resultsare inagreementwithndingsfrom

MonteCarlosimulationsandwithpropertiesderivedfromthedropletmodel.

I Introdution

3dspinglassesaredesribedbytheEdwards-Anderson

(EA)model:

H = 1

2 X

i;j J

ij S

i S

j

(1)

whereJ

ij

arerandominterationswithzeromeanthat

an be hosen only as nearest neighbor interations

andS

i andS

j

arespinsonaregularthreedimensional

lattie [1℄. This model Hamiltonian is realized bythe

Fe

0:5 Mn

0:5 TiO

3

ompound. The magneti struture

ofthetwoparentompounds ofthis spinglass system

isantiferromagnetiatlowtemperatures;thestruture

maybedesribedbyahexagonallattiewiththespins

direted along the -axis of the rystals. In FeTiO

3

the magneti atoms are ferromagnetially oupled in

thehexagonalplanesbutadjaentlayersare

antiferro-magnetiallyoupled. InMnTiO

3

,ontheotherhand,

neighboring Mn atoms are antiferromagnetially

ou-pled within the layers. Mixing iron and manganese

ausesrandomnessandfrustration,andataboutequal

onentrationof FeandMnatoms,thissystemalmost

ideally reprodues a short range 3d EA spin glass[2℄.

Results from experiments onFe

0:5 Mn

0:5 TiO

3

will be

used to disuss stati and dynami saling analyses

thatsuggesttheexisteneofanitetemperaturephase

transition in zero eld but no phase transition in a

magneti eld. Nonequilibrium dynamis and aging

phenomena [3℄will be exemplied with representative

datafromother3dspinglassmaterials.

II Phase Transition

Isthereaspinglassphaseinreal3dIsingspinglasses?

This keyquestionmaysound trivialto resolve

experi-mentally, but hasturned out to requireextensiveand

areful studies on good model systems to nd a

reli-ableanswer. Weherereallstatianddynamisaling

results on one and the same Fe

0:5 Mn

0:5 TiO

3 single

rystalline spinglass material [4, 5℄. Fig. 1shows the

inandoutofphaseomponentsofthesuseptibilityof

this material in awide frequeny range[5℄, and Fig. 2

shows results from a suseptibility measurements in

superposed d magneti elds [6℄. Someessential

fea-turesofspinglassdynamisanbeseeninthesegures,

twoofwhihare: (i)thereiswideregioninwhih

onse-quenesofaslowingdownofthedynamisareevident

in the a-suseptibility; (ii)there is asubstantial

sup-pressionofthesuseptibilitywitheld.

Theobservedfrequenydependene ofthe

suseptibil-ityallows to safelydistinguisha ritialslowing-down

behavior:

0 =t

z

(2)

where t is the redued temperature ((T T

g )=T

g ),

the relaxation time,

0

the mirosopi spin ip time

and z the dynami ritialexponent. More and less

sophistiated dynami saling analyses of the data of

Fig.1allnallyyieldz=100.5. [5℄

Approahingthe ritialtemperaturethe

susepti-bility,asisimpliedfromFig.2,beomesmoreandmore

eld dependent. This non-linear suseptibility an be

aurately measuredfrom,e.g.,thedependene of the

a suseptibility on aninreasing superposed d eld.

Employing suh data in stati saling analyses yields

(2)

Figure1. A-suseptibilityofFe

0:5 Mn

0:5 TiO

3

at dierent

frequeniesfromref.[5 ℄.

Figure 2. Asuseptibility of Fe0:5Mn0:5TiO3 at 125Hz

indierentsuperposedd-eldsvaryingfromtoptobottom

Theonlusionfromthesedynamiandstati

sal-ingresultsisthatthe3dspinglassdevelopstowardsan

equilibriumlowtemperaturephaseasthetemperature

isloweredtowardsT

g .

A more ontroversial problem is whether or not

thereisaspinglassphaseinanappliedmagnetield.

Mean eld solutions of the innite range spin glass

modelsaythatthereisaphasetransitioninamagneti

eld,whereasthedropletmodelfor3dIsingspinglasses

saysno. AnalytialworkandsomereentMonteCarlo

(MC)simulations onshort range 3dIsing spinglasses

[7,8℄supporttheanswerno,whereasotherMCworkis

interpretedtofavoranin-eldphasetransitioninsuh

systems[9℄.

What do experiments suggest? Fig. 3 shows

on-stant relaxation time ontours for our model system

derivedfrom a suseptibility datain applied d

mag-neti elds [6℄. To extrat information on whether a

phase transition is antiipated or not on ooling in a

eld, we have analyzed the slowing down of the

dy-namisalongthetwolines drawninFig.3. Theresult

ofvariousdynami salinganalysesofthedataisthat

theslowingdownisbestdesribedbythefollowingtwo

salingfuntions:

ln(

0 )/T

1

f

(3)

forH =onstantand

ln(

0 )/T

(1+H )

f

(4)

for H=T =onstant, where T

f

is the freezing

temper-ature orresponding to eah relaxation time and, in

termsofthedropletmodel,isanenergybarrier,

H

governsthe eld dependene of theorrelation length

and is thebarrier exponent. Theresultimpliesthat

there is no phase transition in a magneti eld for a

realisti3dIsingspin glass. Theexponent

H

=0.65,

yielding =0.8using aderived valueof the exponent

H

=0.8. Experimentsona3dIsingmodelsystemthus

supportthesenariothatthereisaphasetransitionin

zero eld and no phase transition in a nite eld, in

aordwiththedropletmodelpredition.

III Nonequilibrium dynamis

In this setion the nonequilibrium nature of the spin

glassphaseisdisussed. Mostexperimentsonthiskind

ofdynamis havebeenperformedon3dspinglass

sys-tems of less lear Ising nature than Fe

0:5 Mn

0:5 TiO

3

and we have therefore hosen to exemplify

nonequi-librium dynamis bymeasurements onother, but still

harateristi, spin glass systems. The observed

be-haviorishoweveralsoharateristiof3d Isingmodel

(3)

0

5

10

15

20

25

0

1

2

3

T [K]

H [T]

1700 Hz

0.017 Hz

Figure 3. Constant relaxation time ontours for

Fe

0:5 Mn

0:5 TiO

3

in anH-T diagram. The frequenes are

0.017,0.17,1.7,17,170and1700Hzandtheorresponding

relaxationtimest=1=! [6 ℄.

An aging phenomenon in amagneti system is

re-vealed bya nonstationaryresponse to the appliation

of an external magneti eld, i.e., the response

fun-tion depends on the waiting time (t

w

) spent at

on-stanttemperaturebeforethemagnetield isapplied.

Fig.4showsresultsfromzeroeld ooled(ZFC)

mag-netizationmeasurementsona3dspinglassfordierent

waiting times; (Fig. 4a)m(t) vs. log(t) and(Fig. 4b)

the orresponding relaxation rate (S(t) = 1=h m(t)

ln(t) ).

Astrikinghangeoftherelaxationwitht

w

isobserved

and theoureneof amaximum in S(t) at an

obser-vation time of order the waiting time is harateristi

forspinglasses. Inana-suseptibilityexperiment,the

observation time is onstant t = 1=! and at low

fre-queniesaginganbeobservedasaspontaneousdeay

of both the in-phase and the out-of-phase omponent

ofthesuseptibilitywithtimeatonstanttemperature.

Thisbehaviorhasbeenamplydemonstratedinboth2d

and3dspinglassesandalso ours inother frustrated

magnetisystems[11℄.

The omplexity of the nonequilibrium dynamis

of spin glasses is further illustrated in Fig. 5, where

the low frequeny out-of-phase omponent of the

a-suseptibilityisplotted vs. temperature. Thedierent

urveshavebeenreordedonontinuousheatingofthe

sample after dierent ooling proedures: ontinuous

ooling (referene), ooling with one long stop at 40

K(single)and oolingwithlongstopsat twodierent

temperatures, 40 and 50K(double) [12℄. Thedips in

the urves our at the temperatures where the

sam-ple hasbeenkeptat onstanttemperatureduring the

otherwiseontinuousoolingproess. Tworuialspin

glasspropertiesareillustratedbythisgure: (i)haos

- equilibratingthe sample at onespei temperature

yields a spin onguration that appears random at a

dierent temperature; the system isrejuvinated when

thetemperatureishangedand(ii)memory -the

equi-libration that has been imposed on the system at a

tionandisretrievedwhenthattemperatureisreovered

onheating. Theombinedmemoryandhaosbehavior

hasbeenfoundto distinguishaspinglassphasefrom,

e.g., a frustrated ferromagneti phase [13℄ in whih a

rejuvination ours both on heating and ooling, i.e.,

thereisnomemoryeet.

Figure 4. (a) Magnetisation vs. log(t) and (b)the

orre-spondingrelaxation rate(b)measuredafterdierent

wait-ingtimes,t

w

foranamorphousmetalli3dspinglass.

Figure 5. Low frequeny 00

(T), measured on a Cu(Mn)

spin glass on ontinuous heating after ontinuous ooling

(ref), after one halt at 40 K (single) and after halting at

twodierenttemperatures,40and50K,(double).

(4)

temperature experiment. Fig. 6shows the ZFC, eld

ooled (FC) and thermoremanent (TRM)

magnetiza-tion measured on aAg(Mn) spin glass by ooling the

sampletoalowtemperatureinonstanteldandthen

applying/keeping/removing (ZFC/FC/TRM) a weak

magnetieldandreordingthemagnetizationon

on-tinuous heating, (open symbols) [14℄. The urves

marked by losed symbols have been reorded in the

same way, the only dierene being that the sample

waskeptataonstanttemperatureT

h

=27Kforsome

hoursduringooling. Thereisamarkeddip(bump)in

the ZFC(TRM) urvein the region aroundT

h . This

behavioragainillustrateshaosandmemoryinthespin

glass phase. (This simpleexperimentanbemade on

anyhighresolutionmagnetometersystemtodistinguish

aspinglassphasefromotherdisorderedphasesin less

wellunderstooddisorderedsystems).

20

25

30

35

40

45

0

1

2

3

4

5

6

T(K)

M (arb. units)

ZFC

FC

TRM

Figure6. ZFC,FCandthermoremanent(TRM)measured

onheatingafteroolingtheAg(Mn)spinglassto20Kand

then measuring on heating. Theurvesmarked with full

symbolsrepresentasamplesubjetedtoalongstopat

on-stanttemperature(27K)andeldduringooling[14 ℄.

IV Disussion and onlusions

In this paper we have reapitulated somekey

experi-mental propertiesof 3dspinglasses. Theobserved

be-haviorsupports many features inherentto thedroplet

model of short-range 3d spin glasses. The questions

regarding thephase transition are answeredin aord

with this model: there exists azero eld phase

tran-sition, but there is no phase transition in a magneti

eld.

Nonequilibrium dynamis, whih in the droplet

model is related to the growth of spin glass ordered

withthedropletpiture. Thememoryphenomenon is

afeaturethatperhapsanbeinorporatedinadroplet

piture[15℄ormayrequirealternativeapproahes[12℄.

There are urrently major eorts to understand the

nonequilibrium and aging phenomena that disordered

andfrustratedphysialsystemspossess[16℄.

Aknowledgments

FinanialsupportfromtheSwedishNaturalSiene

ResearhCounil(NFR) isaknowledged.

Referenes

[1℄ ForreentreviewsonspinglassesseeSpinGlasses and

RandomFields,A.P.Young,Ed.,(WorldSienti,

Sin-gapore,1998).

[2℄ H.ArugaandA.ItoJ.Phys.So.Jpn.62,4488(1993).

[3℄ L. Lundgren,P.Svedlindh, P.Nordblad, andO.

Bek-man,Phys.Rev.Lett.51,911(1983).

[4℄ K.Gunnarsson,P.Svedlindh,P.Nordblad,L.Lundgren,

H.Aruga,andA.Ito,Phys.Rev.B43,8199 (1991).

[5℄ K.Gunnarsson,P.Svedlindh,P.Nordblad,L.Lundgren,

H.Aruga,andA.Ito,Phys.Rev.Lett.61,754(1988).

[6℄ J.Mattsson,T.Jonsson,P.Nordblad,H.ArugaKatori,

andA.Ito,Phys.Rev.Lett.74,4305(1995).

[7℄ Seee.g.,M.Palassini,andA.P.Young,Phys.Rev.Lett.

83,5216 (1999); M.Palassini, andA. P.Young,

ond-mat/0002134,andreferenestherein.

[8℄ Seee.g.,M. A.Moore, H.Bokil, andB.Drossel, Phys.

Rev.Lett. 81, 4252 (1998). B. Drossel, H. Bokil, and

M.A.Moore,ond-mat/0007315,andreferenestherein.

[9℄ Seee.g.,E.Marinari,G.Parisi,andJ.J.Ruiz-Lorenzo,

inSpin Glasses andRandom Fields,A.P.Young,Ed.,

(WorldSienti,Singapore,1998),p.59,andreferenes

therein.

[10℄ P.Jonssonetal.,unpublished.

[11℄ P. Nordblad, in Dynamial Properties of

Unonven-tionalMagneti Systems, A. T.Skjeltorpand D.

Sher-rington.Eds.,(Kluwer,1998),p.343.

[12℄ K.Jonason,E.Vinent,J.Hammann,J.P.Bouhaud,

andP.Nordblad,Phys.Rev.Lett.81,3243 (1998).

[13℄ E.Vinent,V.Dupuis,J.Hammann,M.Alba,andJ.

P.Bouhaud,Europhys.Lett.50,674(2000).

[14℄ R.Mathieu,D. N.H.Nam,P.Jonsson,andP.

Nord-blad,ond-mat/0007070.

[15℄ T.Jonsson,K.Jonason, P.Jonsson,and P.Nordblad,

Phys.Rev.B59,8770 (1999).

[16℄ J.P.Bouhaud,L.F.Cugliandolo,J.Kurhan,andM.

Mezard M., inSpinGlasses and Random Fields,A. P.

Imagem

Figure 1. A-suseptibility of F e 0:5 M n 0:5 T iO 3 at dierent
Fig. 4 shows results from zero eld ooled (ZFC) mag-
Figure 6. ZFC, FC and thermoremanent (TRM) measured

Referências

Documentos relacionados

ritial behavior is haraterized at high magneti onentrations..

body transverse Ising model with random bonds and elds. The ouplings and the magneti elds are

transition in suh a model, and study the orresponding ritial temperature as the spin value ande. the

the master equation on the phase spae, and the evolution of loal and global spin autoorrelation.. funtions, in terms of independent relaxation modes, whih are eigenvetors of the

We propose an improvement of a Monte Carlo method designed to treat the Ising model in a

universality lass at the ritial point marking

We alulate the spatial and temperature dependenes of various spin orre-.. lation funtions, as well as the wave-vetor dependene of the spin suseptibility for

relations has been applied to obtain the time evolution of a non-ommuting spin