INVESTIGACiÓN REVISTAMEXICANA DE FíSICA 44 141J77-.'R11 AGOSTO 1998

Oxygen behavior in Nd2-xCexCu04_6 observcd by rcsistivity rncasurerncnts

KA. Vargas, JE Diosa and B.-E. Mellander!"1Departamell/o de Física, U"h'ersidad del Valle

Apartado Aéreo 253ÓO. eoli. Colombia

Recibido el 14 de marzo de 1997; aceptado el 7 de mayo de 199X

We prcscnt in situ rcsistivily measurcmcnts 01' the high.lcmpcraturc phasc hehavior associalcd with oxygcll variation ol' the ceramic syslcm

Nd:l_rCCrCUÜ4_<i (NCCO) when anncalcd in ambient air. \Ve ohservcd a resistivity nnornaly <lrollnd 550 K in all deoxygcnated samples

preparcd hy rapid cooling ¡nto I¡quid N2. which is smallcr for highcr Ce contcnt. \Ve <turihule Ihis anomaly 10 oxygen in-diffusion, which

is Icss wilh incrcasing x. The implicatinns 01' Ihese rcsults on the role (JI"lhe ()xy~cn vJcancics to est<lhlish a metallic phase in NCCO arediscussed.

Key\wm1.f: T' -type cuprales; oxygen diffusion

Presentamos medidas de resistividad in situ del comportamiento de fases a altas temperaluras asociado con la variación de oxígeno delsistema cerámico Nd2-:l'CczCu04_J (NCCO) cuando se templa en Jire, Ohservamos una anomalía en la resistividad alrededor de 550 K entodas las muestras desoxigenadas preparadas por r;ípido enfriamiento en N2líquido, 1;1('ual es menor p;lra altos contenidos de Ce. Atribuimosla anomalía a 1;1difusión de oxígeno hacia el interior de la Illuestra. la cual es menor a rnediJa que crcce .". Se discuten las implicaciones deeste resuhado en el papel que juegan las vacancias de oxígeno en el establecimiento de la fase metálic¡\ ell NCCO.

Descriptores; Cupralos tipo T': difusi6n de oxígeno

PACS: Ilumbers: 6ó.30.-h, 74.72.J<; 74.62.Dh

1. Introduction

The T' -lype system Ndz_xCexCuO,_, can be prepared inlhe range of Ce content O :s .T :s 0.2 at ambient pressure,Although oxygenaled samples show semiconducling n-typebchavior, superconduclivity wilh lhe highest Te Ilear 24 Kappears after reduction annealing for.r = O.I~..{j.IS 11-3J.Various experimental works have been done [4-i] lo studythe inl1uence 01" the Ce- doping and the rcduction oj" oxy-gen content on establishing a metallic phase in NCCO. Itis cvident from those resulls that critical levels of carriersand oxygen vacancies <lre essenlial for the T'-type cupraleslO exhihit superconduetivily. Kawashima el al. 14] using lher-mogravimetry, de susceptibility and X-ray diffraclion showthat for x -= 0,15 or 0.16 a lesscr amount 01" oxygen defectswas needed lo realize supcrconduClivity compared wilh thecase ,1-" -= 0.1,,) suggesting that introduction of oxygen de-fects works lo add carriers (electrons) to Ihe system. Alp el

al. 151 using X-ray aosorption spectroseopy show Ihallhe Ceions add carricrs (eleetrons) lo the CUOl planes. Takayamael l/l. [G] using neulron powder dilfraetion suggests thal theoxygen defccls are inlroduced at least lo lhe sites within lheCUOl Jayer. Howevcr, lhe reduction process and the role ofoxygen defeets (vacaneies) thus created have not been ade-quatcly undcrstood so far [7].

In the present study, we prepared samplc with x -= O,O.OS, 0.12, 0.1G and invesligaled in delail lhe varialion ofoxygen conlenl in the tempcralUfe range 300-1 000 K hy in si-tu rcsislivity mcasuremcnts, We belicve, the rcsistivily chan-

ge, !::.¡>(T), belween lhe oxidized and reduced phases of NC-CO in Ihe normal slale may provide informalion on carrierdoping due Ihe occlIpancy on the oxygcn sites.

2. Experimental

The Ndl_,Ce,C"O.,_, samplcs (.r = O, O.OS, 0.12, andO.IG) were prep"red by solid stale reaetion [SI at 1223 K inai!" using Nd20:t. CC02 and CuO as starting reagents withseveral regrinding and refiring schcdules to improve hOTllo-gcneily. The powders were pressed into pcllets and thermallylreatcd in air. To gel samplcs with lhe minimum oxygen con-tent,lhey wereequally redueed al 1213 K in air for i~ hours,l!len quenchcd inlo liquid nitrogen. X-ray diffraction patternsshowed that alllhe s:unples were cssenlially single rhase andthe lines are indexed on the oasis ol' letragonal structure wilhIhe ialtice p"rameters "f pure Nd,CuO.¡ [9] (see Fig. 1).

Thc resislivily 01" all the samples was measured in airlIsing lhe four conlacl van del" Pauw method [ID) with a sinu-soidal applied current. Ohmic oehavior of lhe conlacls wereootained hy pressing !Ine platinum pins against lhe borderof the rmllt racc 01" lhe samples. which were disk-shape pe-lIets 01' 10 mm in di:uneter and 1.0 mm thickness. A Hewlclt-I'aekard Model4274A Muilifrequeney LCR meler conlrolledhy a HP900 compulcr was uscd lo Illcasure the rcsistance oflhe smnple for each four-point conflguralion rcquired hy thetechnique, at a freqllency 01' 1Khz and amplitude of 0.1 V.Our resistivity prooe was designed to allow air at amhientpressurc 10 llo\V around lhe whole sample at all limes. Thc

.17M KA. VARGAS. lE. DIOSA. AND B.-E. ~lELLAN[)ER

Nd2Cu044\ (room temperature)

FlnURE l. Powdcr X-ray ditfraction pattcrns ol' rcprcsentativcs;lmples of Nd2Ülo.l_J al room tempcralUrc «1) and al' dopcd sal11-pie Nd2_.rCC.rCuü.I_J wilh l" = 0.08 al 1073 K (h).

OOCO

0.004

0.=O.oca

700

x=O.06.1.a=O.436

x=016 0002,ia=O.1361

0001

OOCO

x=O.12t.a=0.1844 0.0024

0.0012

(b)

.~.

.\I .. \...

.\

1\.} ._--

/\x=O,06

600 800 5CX) seoTemperature (K)

(a)

0.0 9 • x=O.16

\0.06

0,03 ......--..~•.- •.- •.- •.0.12 . x=O.12

0.06

EOOBoe-"'-0.04:z..;;.~ 0.16o¡¡;Q)o:: 0.12

FIGURE 2. A sel of tirsl swcer hcating resistivily curves 01' seve-ral equally rcduced smnples of Ndl_.1 Ce".Cu04_<'i with varying Cecontcnts. ,l'. ;11a l'tlll.~t(lllt rale (Jf :2 K/lllin in amh¡cnt air: a) Tolalp(T) in Ihe 400--X()() K temperature range: b) resistivity changc.~p(T) nctwcel1 de reduced ami normal curves (sec Fig. 3) in the5<Xl--700 K tcmpcraturc rangc.

50

5045

45

40

40

35

35

30

30

29 [degree]

(b)

29 [degree]

25

25

(a)

Nd2_.Ce,CuO •.• (x=O.OB) (1073 K)

20

20

~ •

g ~.-:; I~r ~

""-o

o

BOO

BOO

<ñ'e 600~8~ 400

.~Q) 200e

<ñ'e 600~8~400.~Q) 200e

TAHLE I. Arca.6./1 ofthe peak shown in Fig. 2b as a function ofCeconcenlration

I"or two reprcscntative samplcs (Ce con tenis of.r = O and.r = O.lG) takcn 011Ihe first hcating run (data points markedhy solid upwanl triangles, Ihe sunscquent cooling run (poinlsmarkcl by sol id dO\vnward trianglcs). lhe second healing run(poinls marked hy opcn circles) and lhe second cooling run(points markcd by open squares). The data clearly show thatthe addition 01"Ce lllarked changcs in hoth lhe magnitude and(he lemperaturc dcpendence 01' the rcsiSlivity on lhe tirsL hc-.lling run (sec Fig. 2). As lile Icmpcrature was incrcased, Iheresistivity lirsl decreased. indicating a semiconducting bcha-vior. hut awulld 500 K it showcd an upward hend. Beyond

prohc was positiollcd in a tcmpcraturc-controllcú vertical tll-he furnacc. Constanl ramping modes (hcating ami cooling)uf anncaling wcrc uscd.

3. Reslllts amI disclIssion

The pml,.'(\cr X-fay diffraction pattcrns of reprcsentativc S,1Il1.

pies 01' Nd1Cuo.1 (;r = O) al room tcmpcraturc (a) and dopcdsample wilh.c = 0.08lakcn al 1073 K (h) are shown in Fig. 1.Bo(11 patterns are pcrfcct rcproduction 01' lhe powder pattcrnreponed oy Singh el al. [9), confinning that lhe samples are asingle phase with Ihe IClragonal structurc of pure Nd2Cuo.¡.

Figure 2a shows the high-tcmpcralurc rcsistivity in Ihe4()(}-XOO K tempcraturc range ol' several Nd2_xCexCuo.1_lisamples with varying Ce contents, which were previously al1-Ilcaled al 1213 K in air and quenched inlo liquid nilrogen.Thc dala wcrc takcn al a conslanl heating rate ol' 2 Klmin inairo from room tempcrature up lO 1000 K. Figure 2h showson an ampliticd scale the variation of Ihe resistivity wilh res-pecl lo the hase line 01' the observed peak anomaly. The area(~(I.) nI' (he peak as a funclion of Ce conccnlralion was tllenca1culalcd (scc l~lhle 1). Figure 3 shows lhe resistivity data

Ce cOl1centration (x)

O.IX)

O 08

0,1:20.16

Peak arca . .ó.u¡¡¡cm K)

498.1

04360.1844

0.1361

RCI'. ¡\1t'x. Fú, ..w (-1-) (199R) 377--3XO

OXYGEN HEHAVIOR IN Nd:l_:o;o.~:o;CllO.I_6 OHSERVED BY RESISTIVITY .\lEASUREI\IENTS

20

250 500 750 1000

Temperature (K)

FJ(iURE 3. Rcsistivity data, p(T). 01' two reprcsclltativcsNd:'!_.l,CeXCu04_6 samplcs with varying Ce conccntralions: a) ror./'= 0,1G and h) :1" = 0,0 al a constant rafe (JI' 2 K/min in amhicntairo On Ihe fina heating run (solid uptrianglcs). on lhe subSCqUClltwoling run (so/id downtrianglcs). on the second heating run (opencireles) ami the sccond cooling run (open squarcs).

Nd Ce CuO2., ',' 4.0;...-..-.- ..-...-.•

!• >«016;....-.,.-._.- ..-.-.•

I• >«012

. .-...-...-....-.I•

I• >«000

! •, • .-.-...-..~...!• >«00

as shown in rig. 4 rOl"various :1' conccntrations (lhe healingrale 01' 2 Klmin is sufficiently slow for DptDt ;:::;::O). Thcrcforc.from the resislivity changc. ~p(T). hetwcen lhe rcduced andnormal oxygcn contcnt phases of NCCO aboye room tem.peralure (Fig, 2h) \Ve cOllclude lha( samplcs wilh lower Ce-doping show higher (lxygcn in-dillusion in thc 500-550 KtClllperatme range. When Ihe sample is satuf<ltcd wilh oxy-gen al a particular Icmperature 01" lhe partial oxygcn pressure01' the gas phase (amhicnl air) is in cquilihrium with Ihe com-pound, Ihe resislivily foJlows lhe corrcsponding IcmperalurcvarialioJl rOl' lhe ,1' and lúlly oxidized phase.

f-rGUIH: 4, Temrcraturc dcrivative 01' lhe I1rsl healing resistivityd~lta, dp/dT 1',\'. T. rOl" v;uiotls ,1' conccllIr:1lions of NCCO

can

.(l4ll4lJ liD 1m 10:0

Temperalure (K)

.(l0>l

.(l00l4caro

~~E .(lCDt1oS. Q<XD•...:!2a."O .(lcm

Qaro

4. COIlc/usions

By cOlllparing in situ rcsistivity Illeasmemcnts p(T) in rhe sc-micollducling rhase 01" NCCO for various Ce concenlration,wc fOlllld Ihal the lempcrature variation of the resistivity asthc samplcs cxchangcs oxygcn wilh the amhienl airo showsCOllllllon !"eatmes ror lhese oxides. On healing a deoxygcna-lel! samplc in amoient air al a mol!eralc rate (2 K/min) the

x=0.16

x=0.00

0.30 o•..•(j

g.E'0.15 • 9u ••c: 8.- •.•.~t:;.~ ~OOO~uGa:~ 0.00 f-~~~~-~~-~_.!!1 40l/lQla:::

o

approximalely 550 K, lhe rcsislivily dccreased again, hUIahoye 700 K il levclcd lo a constant valuc. On lhe suhsc:quc:nteooling runs. the resislivity data did not show any <lnomalyaround 500 K: it followed lhe sarne li!le dow!llo approxima-lely 700 K. and helow lhis lemperature lhe resistivi(y dcpar-led from Ihe lirst he<lting run wilh a similar scmiconduclinghchavior. On suhsequenl lhermal cyclcs (healing ami coo-ling) p(T) followed Ihe ¡¡rsl cooling curve (see Fig. 3).

The ohservcd hehavior 01' lhe f1rst heating p(T) curves01" as-prcp~lrcd dcoxygcnatcd samplcs (reduccd <lt Ihe sametempcrature (1213 K) to sccurc identical oxygen detkienl)rOl' various Ce concenlrations c1early reveals same inlcrcs-ling characlerislics oI' lheir cxchange of oxygcn \vith lhe gasrhase. Thus, all lhe samplcs showed a scmiconducling hcha-vior hut lhe resislivity decreascd wilh Ce.doping indicatinglhat il creales carriers (elcctrons) by Ihe Cc.l+ subSlitulion onlhe Nd3+ sites (Notice Ihe draslic reduclion oI' Ihe resislivityscale as x increases in Figs. 2 and 3). The occurrcnce 01" Iheupward bcnd nf p(T) arnund 500 K marks Ihe dominancc nfIhe dccrcase 01' carricrs (so lhe rcsistivily gocs up) <lSa resullof in-diffusion 01' oxygcn inlO Ihe oxide. In uther wonJs, wcare claiming lhat lhe in-diflusion rale 01' oxygel1, dÓ/dl, en.lcring Ihe sample (grains) al constant hcaling rale, dT/r1I, isproponionallo the lemperalure derivative 01' Ihe curve p(T, t)

ReI'. Mex. Fú . .¡.¡ (4) (1998) 377-380

JS() RA VARGAS. lE. DIOSA. ANIlIl .•E MELLANDER

oxygcn l'Oncentration is prcscrved Up lO approximatcly 500 Kwhcrc oxygcn starts lo diffuse ioto those encrgctically avai-lable cmpty si tes of the corrcsponding oxide. By comparingd¡l/df for cqually rcJuccd samplcs with vaTinus Ce content.\Ve founJ that the rate of oxygen uptakc dccrcascs wilh in-crcasing .C OUT rcsults then reveal that cstablishing criticalIcvcls 01' oxygen-vacancy doping in NCCO rcquircs a goodcOlltrol of low partial oxygcn prcssures in cquilibriulll wilhthe sample maioly ahoye 500 K. Morcover, ;" siru rcsisli-vily cxpcrirnents whcn these oxides are thcfmally tfcatcd are

(/l). Dcpartmcnt 01"Physics. Chalmcrs University ofTcchnology. S.41296. Gothcnburg. Sweden.

1. Y. Tokura. H Takagi. and S. Uchida. Narure 337 (1989) 345.2. H. Takagi. S. Uchida. and Y. Tokura. Ph.rs. Rel'. Lell. 62(19R9)

1197.3. C.N. Picclulewski, K.S. Kirkpatrick. and T.o. Masol1. 1. Am.

('"mm. So(". 73 (191)0) 2141.

.1. T. Kawashima and E. Takayama-Muromachi. I'hv!>ica e 219(1994) 389.

shown lo he a simple 1001 ro 1ll0niWf thcir carricr doping duelo varialiolls 01' oxygcn con ten!.

Acknowled!(ments

Thc authors gratcfully acknowleJge the support 01"the eo-lomhian Research Agcncy, COLCIENCIAS. ami lhe Intcrna-tional Programs in lhe Physical Scicnccs. IPPS. 01' UppsalaUnivcrsity, Swcdcn. \Vc thank E. Ortií'. and F. Jurado ror so-me 01' the sampk prcparations.

G. E.E. Alp ('1 al .. Phy.\'. Ri'l'. U.tu (19R9) 2617.

G. E. Takayama-~1uromachi el (I/.. PhYJü'a C159 (1989) 634.

,. O.G. Singh i'/ al .. Solid Sal(' COllllnU/I. 100 (1996) 721.

8. W. J Zhll. P. Lin. ano Z. X. Zhao. Phv.\"ica C919() (1992) 2R5.

a. o. (L Singh. B. D. Paoalia. O. Prakash. K. Suba. A. V. Narlikarami L.e. Gupta. P/¡ysica e 219 (1994) 156.

10. L. 1. van oer Pauv.'. Philil's T('C/¡II. Re!'. 20 (195R/59) 220.

R,,'. Mex. Fis. 44 (4) (199R) 377-380