• Nenhum resultado encontrado

Braz. J. Phys. vol.32 número3

N/A
N/A
Protected

Academic year: 2018

Share "Braz. J. Phys. vol.32 número3"

Copied!
5
0
0

Texto

(1)

A Novel Perolation Theory for High

Temperature Superondutors

E.V.L.de Mello, E.S. Caixeiro, and J.L.Gonzalez

Departamento deFsia,UniversidadeFederalFluminense,

Av. Litor^anias/n,Niteroi,R.J.,24210-340, Brazil

Reeivedon28February,2002

Wepresent aperolationtheoryforthe high-T

oxides pseudogapandT

dependeneonthe hole

level. Thedopingdependentinhomogeneous hargestrutureis modeledbyadistributionwhih

mayrepresentthestripemorphologyandyieldaspatialdistributionofloalT(r). Thetemperature

onsetofspatial dependentsuperondutinggap is identiedwiththevanishingof thepseudogap

temperatureT

. Thetransitiontoasuperondutingstateorrespondstotheperolationthreshold

amongregions of dierent T

. Asaparadigm we use aHubbard Hamiltonianwith amean eld

approximation to yield a doping and temperature dependent superonduting d-wave gap. We

showherethat thisnewapproahreprodues thephasediagram, explainsandgivesnewinsights

onseveralexperimentalfeaturesofhigh-Toxides.

I Introdution

High-T

oxideshasbeendisoveredfteenyearsago[1℄

butmanyoftheirimportantpropertiesremainsnotwell

understood. Amongthese,thepseudogapphenomenon,

that is, a disrete struture of the energy spetrum

aboveT

,identiedbyseveraldierentexperiments[2℄,

hasitsnaturenotyetbeenlaried. Suhopenproblem

hasattratedalotofexperimentalandtheoretialeort

beauseitisgeneralbeliefthatitssolutionisrelatedto

theunderstandingofthesuperondutingfundamental

interation.

The evidene of suh energy gap above the

superonduting phase is learly demonstrated by

tunneling[3, 4℄ and angle-resolved photoemission

spetrosopy[5,6℄experiments. Intheresistivity

mea-surementsitspreseneisseenbyadereaseinthelinear

behaviorbelowT

[7,8℄andinthespeiheatasa

sup-pressioninthelinearoeÆientofthetemperature[9℄.

Thereisalsomountingexperimentalevidenesthat

the hole doped inhomogeneity into the CuO

2

planes,

ommontoalluprates,isdiretlyrelatedtothe

pseu-dogap phenomenon. In a given family, the

under-doped ompounds near the doping onset of

superon-dutivity havethe more inhomogeneousharge

distri-butions and the larger T

. As the doping level

in-reases, thesamples beome morehomogeneouswhile

T

dereases[10, 11, 12℄. For overdoped ompounds

T

disappears or beomes equal to T

. The

inhomo-studies of high temperature superondutors[13℄ but

only after the disover of the spin-harge stripes[14℄,

they havebeome a matter of systemati studies. In

the spin-harge stripes senario, regions of the CuO

2

planesarehole-rih(thestripes)andothersregionsare

hole-poorwhih llthespaebetweenthestripes.

Reently, magneti exitations as the vortex-like

Nernst eet have been reported above T

[15℄ and a

loalMeissnerstate,whihusuallyappearsonlyinthe

superonduting phase, has been seen as a preursor

tosuperondutivity[16℄. Suhinhomogeneous

diamag-neti domains developnear T

and growontinuously

asT dereasestowardsT

. NearT

,thedomainsappear

toperolate,aordingtoFig.3fromIguhietal.[16℄.

Basedonalltheseexperimentalndings,aboutthe

pseudogap phenomenon, loal harge inhomogeneities

andanon perolativeloalMeissner statebetweenT

andT

,weproposeanewsenariotoexplain thehigh

T

superondutorsphenomenology: duetothedoping

dependenthargeinhomogeneitiesinagivenompound

withaveragehargedensity

m

,there isadistribution

ofloal lusterswith spatialdependenthargedensity

(r),eahwithitssuperondutingtransition

tempera-tureT

(r). T

isthelargestofallT

(r). Asthe

temper-ature falls below T

, some lusters beome

superon-duting,buttheyaresurroundedbymetalliand/or

an-tiferromagnetiinsulating domains and, onsequently,

(2)

num-ature dereases, so the superonduting regions grow

and, eventually, at a temperature T

, they perolate

and beomeableto hold amarosopi dissipationless

urrent. ExatlyastheMeissner statedomainsshown

inFig.3ofIguhietal.[16℄. Similarideasweredisussed

byOvhinnikovat al[17℄. Theywereonernedmainly

with themirosopi mehanismwhih leadsto a

dis-tributionofT

(r)anditseetonthedensityofstate.

Inorderto showthat these ideasareableto make

quantitativepreditionsandreproduethehigh-T

ox-idesphasediagram,wehaveperformedalulationson

aHubbardmodelandagapequationisobtainedwithin

meaneldapproahtoderiveT

. Afterwards,using

theperolation approah, weshowthat wean

repro-duealsotheexperimentalvaluesfortheBi2212T

.

II The Charge Distribution

The presene of mirosopi harge inhomogeneities

distribution in the CuO

2

planes in a striped

ongu-rationiswelldoumented[14℄. Italsoimpliesinstrong

modiationsfortheloalstrutureastheloalCu O

bondlengthhangesitssizewiththeaverageloal

dop-inglevel[10,11,18℄.

The onsequene of these intrinsi mirosopi

harge inhomogeneities distribution in the CuO

2

planes, is theexistene oftwophases whih are

spon-taneously reated in the CuO

2

planes; regions whih

arehole-rihformthestripesandothersregionswhih

are hole-poor(the antiferromagneti insulatorregion)

and are reated between the harge-rih stripes (the

metalli region). Theexatlyformofthese harge

dis-tributions is not known and it is presently matter of

researh[10,11,18℄butweandrawsomeinsightsfrom

the STM/S results of Pan et al.[12℄. Based on these

STM/Sdata,wehavehosenaombinationofa

Pois-sonandaGaussiandistribution. Foragivenompound

with an average harge density

m

, the hole

distribu-tionP isafuntionoftheloalholedensity,P(;

m ).

It isdivided in twobranhes: Thelowdensity branh

representstheinsulatingregions, andthehigh density

one represents the metalli regions. As onerns the

superondutivity, only theproperties of thehole rih

branh are important sine the urrent ows through

themetalliregion. Suhdistributionmaybegivenby:

P() = (

)exp( (

)

2

=2(

)

2

)=

( 2

)(2 exp( (0:10) 2

=2(

)

2

)) (1)

The plus sign is for the hole-rih (

m ) for

m

,theminusto thehole-poorbranh(

=0:10)

for 0:10 and P() = 0 for 0:10 .

FromtheexperimentalSTM/SGaussianhistogram

dis-tribution for the loal gap[12℄ in an optimally doped

Bi

2 Sr

2 CaCu

2 O

8+x (

m

0:32)(seeFig.2dofRef.12),

we get that the half-width for the metalli branh is

abouthalfofthemeandopingvalue,thatis2

+

=0:16,

sinetheloalgapisproportionaltotheloalhole

den-sity. Furthermore

+

is relatedwith the degreeof

in-homogeneities, and weestimate the

+

values for the

other ompounds usingthe fat thatit dereaseswith

theholedensity.

ThedistributionP()for

m

=0:32and

+ =0:08

is shownin Fig.1a.. As a generalfeature for a family

ofompounds,basedalsointheexperimentalresultsof

Pan et al.[12 ℄, for samples with

m

0:5, theharge

distributionbeomesasimpleGaussianenteredat

m

with 0:04,whih reets thenon-existene of the

stripesphasesforoverdopedsamples.

Thevalues of for agivensample are hosen in

order that perolation in the hole-rih branh ours

exatlyatagivendensity

p

. ThusT

(

p

)isthe

maxi-mumtemperaturewhihthesystemanperolateand

whihweidentifyasequaltoT

(

m

). Aordingtothe

perolationtheory,the thresholdofperolationours

in asquarelattie when59%of thesitesorbondsare

lled[19℄. Thus, we nd the density where the

hole-rih branh perolates integrating R

P()d from

m

till the integral reahes the value of 0.59, where we

dene

p

. Below T

(

m

) the superonduting region

perolatesand, onsequently,it isableto holda

dissi-pationless superurrent. To estimate T

(

m

) weneed

to alulateT

asfuntionof.

III The Phase Diagram

To develop the dynamis of the hole-type arriers in

theCuO

2

planes, weadopt atwodimensionextended

HubbardHamiltonian inasquarelattieof lattie

pa-rametera

H =

X

ij t

ij

y

i

j +U

X

i n

i" n

i#

+ X

<ij> 0

V

ij

y

i

y

j 0

j 0

i

(2)

wheret

ij

isthehoppingintegralbetweensitesiandj;

U istheCoulombon-siteorrelatedrepulsionandV

ij is

theaphenomenologialattrativeinteration between

nearest-neighborsites.

UsingaBCS-typemean-eld approximationto

de-velop Eq.(2) in the momentum spae, one obtains

(3)

tempera-tures[20,21,22℄ k = X k 0 V kk 0 k 0 2E k 0 tanh E k 0 2k B T ; (3) with E k = p " 2 k + 2 k

; whih ontains the potential

V

kk

0. Forad-waveorder parameter,U issummedout

of the gap equation and the amplitude of the

attra-tive potential V is the only unknown variable whih

beomes anadjustable parameter[22, 23℄. "

k

is a

dis-persion relation. The hole density and the gap

equa-tion, are solved self-onsistently for a d-wave order

parameter[22,23℄.

InFig.1bweplotthetemperaturesofvanishinggap

T

asfuntion of, derivedwith theaboveequations.

Here weuseV = 0:150eVwhihreprodueswell the

experimentalvaluesofT

fortheBi2212system[7℄.

Thehoppingparametersarewithin15%tothosetaken

from ARPESmeasurements[24℄with anearest

neigh-borhopping t

ij

=0:12eV(t

ij

=0:147eVinref.24)and

further hopping parametersupto fth neighbor. The

partiular form oftheT

urvedepends onthevalues

ofthese hoppings. Notiethatthedensityofholeshas

afatorof2withrespettothevaluesgivenbyRef.[7℄,

butinagreementwithref.[21℄asitismoreappropriate

fortheBi2212system[25℄.

0.05

0.15

0.25

0.35

0.45

0.55

0.65

ρ

0

50

100

150

200

250

Temp (K)

T

co

(Raman)

T

co

(STS)

T

*

(

ρ)

T

*

(

χ

c

/

χ

ab

)

Experimental T

c

0

2

4

6

P(

ρ

)

ρ

m

=0.32

ρ

p

=0.48

a)

b)

T

*

T

c

Figure 1. 1a)The probability distributionfor theoptimal

ompoundwithm=0:32. 1b)Thealulatedonsetof

van-ishing gap T

(). The thik arrow shows T

(m) andthe

dot-dasharrowshowshowT

(

m

)isfound. The

experimen-talpointsandthesymbolsaretakingfromRef.[7 ℄

The T m and T m

diagrams are obtained

as follows: A ompound with an average holeontent

higherthanthe onsetofsuperondutivityforBi2212,

that is,

m

0:15, has its metalli branh with

lo-al densities(r)varyingfrom

m

tohighervalues. In

this range, T

is a dereasing funtion of .

There-fore for temperatures above T

(

m

), there is neither

any gap nor any superonduting region and T

( )

marks the appearane of a gap in the sample. That

iswhy T

(

m

) is theT

of thesystem. On the other

hand,T

(

m

)is alulatingintegrating R

P()dfrom

m

till theintegralreahesthevalueof0.59(the

inte-gratedregion is markedwith vertial lines in Fig.1a),

where we dene

p

. Only at T T

(

p

) (see Fig.1a

for

p

= 0:48) the ompound may hold a

superon-dutingurrentsinethesuperondutingregions

per-olatesthroughthesample. ThusT

(

p

)isthe

pero-lating threshold and it is equalto T

(

m

) as onean

seefollowingthedot-dashedlinein Fig.1aand1b).

Itisworthwhiletomentionthattheperolating

ap-proah ouldalsobeused withothersmethods to

al-ulatingT

m

,like,forinstane,theonewhihuses

theHubbard-Holstein Hamiltonian[27℄.

IV Disussion

Thefat that T

dereases ontinuouslywith in the

superonduting region(

m

>0:15),asseeninFig.1b,

isverysuggestiveanditisin agreementwiththeearly

ideasregardingaphononmediatedsuperonduting

in-terations: Materials whose vibrating atoms interat

stronglywiththeeletrons,andarepoormetals,should

beome superondutors at higher temperatures than

those good metals, whose atoms interat weakly with

eletrons[26℄. For uprates, asthe doping level of the

samplesinreases,itiswellknownthattheompounds

hange from verypoormetals in thenormal phase to

verygoodmetalswithtypialFermiliquidbehaviorfor

overdopedsamples. SineT

m

isadereasing

fun-tionofdopingforanyupratefamilybeyondtheonset

ofsuperondutivity,suhbehaviormaybeastrong

in-diationofthephononisuperondutinginteration.

There are several observations and measurements

that an be well explained within the perolating

ap-proah,wewill disusshereonlyafewexamples.

1-Harriset al.[28℄, throughARPES measurements,

havereportedtheanomalousbehaviorof

0 (

m )whih

dereasessteadily with thedoping

m

although T

in-reases by a fator of 2 for their underdoped

sam-ples. Inthe overdopedregion, sineT

also dereases,

thebehavior is theexpeted onventional

proportion-ality. It is well known that superondutors have a

onstant valuefor the ratio 2

0 =k

B T

, being3.75 for

usual isotropi order parameter and 4.18 for d

x 2

y 2

solution[29℄.

Atlowtemperature, sinethe superonduting

re-gionperolates throughdierent regions, eah with a

given

0

(r),tunnelingandARPESexperimentsdetet

thelargestgappresentintheompound. Consequently,

0 (

m

)mustbeorrelatedwiththeonsetofvanishing

gapT

(

m

)whih isthelargestsuperonduting

tem-perature in the sample, and not with T

(

m

). As we

show in Fig.2, orrelating the values for T

(

m ) with

(4)

ofHarriset al.[28℄on Dy BSCCO andexplainsthe

dierentenergysalespointedoutbyHarrisetal. and

severalothersauthors.

0.05

0.1

0.15

0.2

δ

0

5

10

15

20

25

30

0

(meV)

Bi

2

Sr

2

Ca

1−x

Dy

x

Cu

2

O

8−

δ

Theory

2.14kT

c

Figure2. Thezerotemperaturegap for9samplesas

mea-suredbyHarrisetal[28℄andouralulations.

2- The resistivity for underdoped and optimum

dopedompoundsdeviatesatT

fromthelinear

behav-iorand falls faster asthe temperature[2, 7℄ dereases.

This behavior an be understood by our model, with

theinreasingofsuperondutingluster numbersand

size, as the temperatures dereases below T

. Eah

superondutinglusterproduesashortiruitwhih

dereases the resistivitybelow thelinear behavior

be-tweenT

andT

.

3-Asmentioned in theintrodution, the existene

ofsuperondutinglusterbetweenT

andT

easily

ex-plainstheappearaneofloal diamagnetiorMeissner

domains,and,ifthereisatemperaturegradientinthe

sample, theloalux owsand produesthedynami

ux owstate[15℄.

4-Anotherimportantonsequeneisthat the

pair-ing mehanism must be investigated by experiments

performed mainly at T

. A suh experiment was

a-omplished byRubio Tempranoet al.[30℄, whih

mea-sured a large isotope eet assoiated with T

and

an almost negligible isotopi eet assoiated with T

in the slightly underdoped HoBa

2 Cu

4 O

8

ompound.

The results strongly support the fat that

eletron-phononinduedeetsarepresentinthe

superondut-ingmehanismassoiatedwithT

.

V Conlusions

We have demonstrated that the perolating approah

foraninhomogeneoushargedistributionontheCuO

2

planesprovidesnewphysialexplanationsformany

ex-perimentsperformedonhigh-T

oxidesandquantitative

Contrarytosomeurrenttrends,inwhihT

is

re-garded as aphase oherene temperature and the

ex-istene of a gap phase without oherene between T

and T

, in our approah, T

is a perolating

temper-ature for dierent regions whih, due to the

inhomo-geneities, possessdierentloal superonduting

tran-sition T

(r). Similarly, insteadof having a

superon-duting gap

s

and an exitation gap assoiated

with T

, we have a distribution of loally dependent

s (r).

Theperolationmethodintroduedhereanbe

ap-plied inanyuprateandyields alsoseveralnew

impli-ations whih will be disussed elsewhere, but one of

themostinterestingis thatoneansearh for

materi-alswith verylargeT

'sif abetter ontrol of theloal

dopinglevelisahieved.

Finanial support of CNPq and FAPERJ is

grate-fully aknowledged. JLG thanks CLAF for a

CLAF/CNPqpos-dotoralfellowship.

Referenes

[1℄ J.G.BednorzandK.A.Muller,Z.Phys.64,189(1986).

[2℄ T. Timusk and B. Statt, Rep. Prog. Phys. 62, 61

(1999).

[3℄ C.Renner,B.Revaz,J.-YGenoud,K.Kadowaki,and

O.Fisher,Phys.Rev.Lett.80,149(1998).

[4℄ M.Suzuki,T.Watanabe,andA.Matsuda,Phys.Rev.

Lett.82,5365 (1999).

[5℄ Z-X.ShenandD.S.Dassau,Phys.Rep.253,1(1995).

[6℄ H.Ding,T.Yokkoya,J.C. Campuzano,T.Takahashi,

M.Randeria,M.RNorman,T.Mohiku,K.Kadowaki,

andJ.Giapintzaki,Nature,382,511996

[7℄ M. Oda, N. Momono, and M. Ido, Superond. Si.

Tehnol.13,R139,(2000).

[8℄ H.Takagi, B.Batlogg,H.L.Kao,R.J.Cava,J.J.

Kra-jewski, and W.F. Pek, Phys. Rev. Lett. 62, 2975

(1992).

[9℄ J. Loram, K.A.Mirza, J.R. Cooper, and J.L. Tallon,

PhysiaC282-287, 1405(1997).

[10℄ S.J.L.Billinge,E.S.Bozin,M.Gutmann,andH.

Tak-agi, J. Superond., 2306 (2000), Proeedings of the

Conf. \Major Trends in Superondutivity in New

Milenium",andond-mat0005032.

[11℄ E.S.Bozin,G.H.Kwei,H.Takagi,andS.J.L.Billinge,

Phys.Rev.Lett.84,5856, (2000).

[12℄ S.H. Pan, J.P. ONeal, R.L. Badzey, C. Chamon, H.

Ding,J.R.Engelbreht,Z.Wang.H.Eisaki,S.Uhida,

A.K. Gupta, K.W. Ng, W.W. Hudson, K.M. Lang,

and J.C. Davis, Nature 413, 282 (2001) and

ond-mat/0107347.

[13℄ T.Egamiand S.J.L.Billinge, in\Physial Properties

of High-Temperatures Superondutors V" edited by

(5)

[14℄ J.M.Traquada,B.J.Sternlieb,J.D.Axe,Y.Nakamura,

andS.Uhida,Nature(London),375,561(1995).

[15℄ Z.A. Xu, N.P. Ong, Y. Wang, T. Kakeshita, and S.

Uhida,Nature406,486(2000).

[16℄ I.Iguhi,I.Yamaguhi,andA.Sugimoto,Nature,412,

420(2001).

[17℄ Yu.N.Ovhinnikov,S.A.Wolf,andV.Z.Krezin,Phys.

Rev.B63, 6452, (2001), andPhysiaC341-348, 103,

(2000).

[18℄ T.Egami, Pro.oftheNew3SCInternational

Confer-ene,tobepublishedinPhysiaC.

[19℄ D.F. Stauer and A.Aharony,Introdution to

Pero-lation Theory,Taylor&FranisEd.,London,1994.

[20℄ E.V.L.deMello,PhysiaC259,109(1996).

[21℄ G.G.N.Angilella,R.Pui,andF.Siringo,Phys.Rev.

B54,15471(1996).

[22℄ E.S. Caixeiro, and E.V.L. de Mello, Physia C 353,

103(2001).

[23℄ E.S.Caixeiro,andE.V.L.deMello,submittedtothe

J.Phys.CM.

[24℄ M.R.Norman,M.Randeria,H.Ding,J.C.Campuzano,

andA.FBellman,Phys.Rev.B52,615(1995).

[25℄ P.Konsin,N.Kristoel, andB.Sorkin,J.Phys.C.M.

10,6533(1998).

[26℄ B.T.Matthias,Superondutivity,SientiAmerian,

92,Novemberof1957.

[27℄ C. Di Castro, M. Grilli, and S. Caprara,

ond-mat/0109319.

[28℄ J.M. Harris, Z.H. Shen, P.J. White, D.S. Marshall,

M.C.Shabel,J.N.Ekstein,andI.Bozovi,Phys.Rev.

B54,R15665(1996).

[29℄ H.WonandK.Maki,Phys.Rev.B49,1397(1994).

[30℄ D.RubioTemprano,J.Mesot,S.Janssen,K.Conder,

A.Furrer,H.Mutka,andK.A.Muller,Phys.Rev.Lett.

84,1990(2000).

[31℄ M.T.D. Orlando, A.G Cunha, E.V.L. de Mello, H.

Belih, E. Baggio-Saitovih, A. Sin, X. Obradors, T.

Burghardt, and A. Eihler, Phys. Rev. B61, 15454

(2000).

[32℄ J.L.Gonzalez,M.T.D.Orlando,E.S.Yugue,E.V.L.de

Mello,andE.Baggio-Saitovih,Phys.Rev.B63,54516

Imagem

Figure 1. 1a) The probability distribution for the optimal
Figure 2. The zero temperature gap for 9 samples as mea-

Referências

Documentos relacionados

formation of single phase Pr123 with larger lattie pa-. rameter

provided lear evidene of a pair-breaking eld at the Ni site for non-superonduting ompounds.. This eld is not present in the superonduting ollinear AF DyNi2B2 C, however it

powder diratometry, sanning eletron mirosopy and d-magnetization.. Samples

most of the reported results have been obtained using aligned rystallites, -axis oriented thin lms.. and

Room temperature and 4.2 K M ossbauer spetra of the samples MgC(Ni1 xFex)3 with x = 0:01 and x = 0:5. Band struture alulations

YBCO blok in 1-D and mapping the trapped magneti eld in those bloks in 2-DC. I In

is proportional to the average grain-domain radius, but inverse proportional to the total length of.. grain-boundaries of the sample, a simple physial model has been provided and

transverse pinning of moving ommensurate vortex latties and smetis ours.. It is