Orthorhombicity and oxygen uptake by YBa2Cu3O6 + x

HAL Id: jpa-00211008

https://hal.archives-ouvertes.fr/jpa-00211008

Submitted on 1 Jan 1989

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.

Orthorhombicity and oxygen uptake by YBa2Cu3O6 + x

Michel Ain, Jean-Marc Delrieu, Alain Menelle, George Parette, Jocelyne Jegoudez

To cite this version:

Michel Ain, Jean-Marc Delrieu, Alain Menelle, George Parette, Jocelyne Jegoudez. Orthorhombic- ity and oxygen uptake by YBa2Cu3O6 + x. Journal de Physique, 1989, 50 (12), pp.1455-1461.

�10.1051/jphys:0198900500120145500�. �jpa-00211008�

1455

Orthorhombicity ^{and oxygen} uptake by YBa2Cu3O6 + x

Michel Ain (1), Jean-Marc Delrieu (1), Alain Menelle (2), George ^Parette (2)

and Jocelyne Jegoudez (3)

(1) Service de Physique ^{du Solide et de Résonance} Magnétique, Direction de la Physique, CEN-Saclay, 91191 Gif-sur-Yvette Cedex, ^France

(2) Laboratoire Léon Brillouin (CEA-CNRS) CEN-Saclay, 91191 Gif-sur-Yvette Cedex, ^France

(3) Chimie des Solides, Université de Paris-Sud, 91405 Orsay, ^France (Reçu ^{le 6} janvier 1989, accepté

forme définitive ^{le 8}

1989)

Résumé.

Nous

mesuré, ^à différents taux d’oxygène, ^les paramètres de maille de

YBa2Cu3O6+x ^{entre 24 et 260 K à} l’aide de neutrons froids. L’affinement de

résultats, présenté

^la figure 1,

_permet pas de confirmer les

de rayons-X, ^récemment publiées [1], ^montrant

anomalie de l’orthorhombicite b - a,

voisinage ^{de la} température

de transition supraconductrice. ^Il

aussi, ^été mis

évidence que le volume de la ^maille

cristallographique

contractait par absorption d’oxygène.

Abstract.

We have measured, ^at different oxygen contents, the cell parameters ^of

YBa2Cu3O6+x between 24 and 260 K using ^{cold neutrons.} The results of profile refinement of

high resolution data is shown in figure 1 ; ^{it does not} confirm the X-ray finding, recently reported [1], showing

cusp in the orthorhombic splitting b 2014 a, around the superconducting ^transition temperature. It has also been evidenced that the volume of the crystallographic cell is contracted

by oxygen uptake.

J. Phys. ^{France 50} (1989) ^{1455-1461 15 JUIN 1989,}

Classification

Physics ^Abstracts

74.70 - 61.12

Introduction.

It is a major ^{issue, to} examine the structural transformations, ^{at the onset of} superconductivity

in YBa2CU307 ; ^{to the extent that} coupling ^between pairs of carriers and the lattice could

provide information on the underlying microscopic mechanism of the superconductivity ⁱⁿ high 7c superconducting ^oxides. Unfortunately, the lattice strain observable in the vicinity ^of

the superconducting transition temperature, ^{is so weak} that this work brings ^no confirmation of the observations reported by ^{Horn et al.} [1], ^and reproduced ^on the insert of figure 1, showing ^{a maximum at} T,, ^{on the} plot ^{of the} parameter

2 (b - a )/ (b ⁺ a ) defining ^the

orthorhombic strain of the structure.

This is not the first attempt ^to reproduce the results of Horn et al. [1], by ^neutron diffraction, ^{and to our} knowledge, nearly ^all attempts have been unsuccessful. David et al. [2]

have seen no evidence of anomalous behavior of _{a ;} on the other hand, François et ^al. [3]

measured an 06,91 sample ^{and observed a} bump ^{in the} temperature dependence ^of

o,, they concluded that it was in agreement ^{with the} X-rays measurements. And finally, ^{in a}

first account of this work reporting ^{on a} YBa2CU306.88 sample ^we did in fact observed, ^near

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:0198900500120145500

1456

Fig. ^1.

^The orthorhombicity parameter o-

temperature ^for YBa2Cu306.98 ^and YBa2Cu306.99. ^An

error

bar is drawn

single point ^{at 83} K ; ^the

nearly ^constant

^the temperature range. The insert is the result of X-rays diffraction from reference [1], showing

different behaviour for

a versus

T.

90 K, ^a little ondulation smaller than the error bar but we concluded that it was not a

confirmation of the behavior published in reference [1].

Concerning ^the afore-mentioned account, ^{it was} suggested by ^the referee, ^{to remake the} experiment ^{on a} powder ^{that was as close as} possible ^to YBa2CU307.

Expérimental settings.

Abiding by ^these prescriptions, ^{a new} sample ^was prepared ^{from a} mixture of BaCo3, Y203, ^{and CuO} powders, ⁱⁿ stochiometric proportion according ^to the formula YBa2CU307.

The powders ^were gently grounded together ^and pressed by ^hand (to preserve ^a certain

porosity) into three cylindrically shaped samples ^{of 22 mm} height ^{and 18 mm} in diameter and

fired at 950 °C, ^{in air. The} preceding process ^was then repeated ^{and the} powder ^{was fired}

again for 24 hours at 950 °C, ^{and then} slowly cooled down to 450 °C. After 6 hours at 450 °C it

1457

was then brought ^{to room} temperature. The oxygen ^{content was} found to be 06.96, ^as

determined by the reiinement of the neutron diffraction pattern ^{recorded at 83} K ; ^{this same} sample ^{was then} replaced ^{in the oven, at 450 °C under} circulating oxygen, for 16 hours ; ^{and a}

new refinement of the oxygen occupancy ^was performed, using ^{the same} technique ^as above,

and it gave O6.9g. ^A last heat treatment under oxygen for 100 hours, resulted in a

06.99 sample.

Horn et al. [1], ^{did not measure} the oxygen ^content of their sample. ^{But we can estimate}

that it falls between 06.95 ^and 07 after 2 hours spent betweerr400 and 500 °C in flowing

oxygen.

The neutron diffraction data were collected by ^a position ^sensitive detector, with incident

wave-length ^{of À}

4.757 Â obtained from a vertically ^bent monochromator, this value being slightly beyond ^{the cut off} wave-length ^of aluminium at À

4.7 Â. The incident beam was

filtered by ^cold berrylium. ^The position sensitive detector (a bank of 400 cells subtending ^an angle ^{of 80} degrees, filled with BF3) ^{had to be} positionned twice in order to scan the whole

accessible sample’s diffraction pattern.

The sample ^was mounted in ^an aluminium cell in thermal contact with the cold finger ^{of a}

closed cycle displex cryogenerator. Neutron diffraction patterns ^{were recorded on all three} stages ^of oxygenation ^{of the} sample, ^as already ^{said. At one} single temperature ^of

83 K for the 06.96, ^{at several} temperatures between 40 and 130 K for the 06.98 ; ^{and last, the} 06.99 ^was measured between 24.1 K and 260 K by steps of 5 then 10 and finally

20 K ; in the range 72.5-115 K, ^the steps ^{were narrowed to} 2.5 K. The spectrometer, ^{as well as} the temperature regulator, ^{were driven} by ^a desktop computer, according ^{to the} following procedure : ^the temperature ^was stabilized to a drift not exceeding ^0.03 K/min ; ^afterwards

the detector was positionned ^{at his} low-angle location for a one hour counting, ^{then it was}

moved to its high angle position for another one hour counting, ^{and so on.}

Treatment of the spectra.

Profile refinement of the spectra ^were performed, ^on basis of space group Pmmm, using ^the

Rietveld program [4] ^modified by ^Hewat [5]. ^The spectra ^{collected at 83} K, ^were used, ^as said, ^to determine the oxygen ^content after each heat treatment. Oxygen occupancy of sites 04 and 05 (see Fig. 2) ^{were allowed to} vary in these particular ^runs of the refinement program. As ^a result, ^{we obtained an} unambiguous convergence of site 05 occupancy towards 0 %, and site 04 occupancy towards respectively ⁹⁶ %, ^{98 % and 99 %} (± ³ %) ^after

each of the three heat treatments the sample has gone through. ^{The exact} final formula of ^our

compound being ^then, YBa2CU306.99. The transition temperature ^is expected ^{to be around}

92 K, ^as indicated by figure 3 of reference [6].

In the subsequent refinements, only ¹⁷ parameters ^{were allowed to} vary, (two ^{less than} above, namely the oxygen occupancies ^{of site} 04 ^and 05, ^{that were held} constant) they ^were

distributed as follows : 5 for the

atomic coordinates of all atoms in the unit cell, excepting yttrium and those oxygen and copper atoms, ^which positions ^are imposed by ^the symmetry ^of the space group (see Fig. 2) ; 1 overall scale factor ; ^{3 for the} expression of the FWHM of the diffraction peaks ; ^{1 for the zero} point ^{of the} detector ; ^{3 for the} unit cell constant ; 3 anisotropic temperature parameters for oxygen 04 ; ^and only ¹ parameter for all the individual isotropic temperature factors, ^{that was used as} follows. We assumed that the kinetic energies Ec ^{of all atoms} excepting 04, ^were equal ; ^then starting from the fundamental relation :

where Ud is the vibration amplitude ^{of atom} d, Md, ^{the mass, and} ud ^the velocity amplitude.

1458

Fig. ^2.

^{The structure of the} orthorohombic cell, with space group Pmmm- Oxygene ^sites

^numbered

from 1 to 5. Yttrium atom is at the center of the cell, 04, 05 and Cul lie in the basal plane. ^The empty site 05 is represented by

square. Also shown, ^{the two} unequivalent sites of copper Cul and CuII. The so-called CU02 planes squeeze the yttrium ^ions.

Noting ^{that an} individual isotropic temperature ^{factor is} proportional ^to ud, ^i.e. Bd

au2d, ^we

write :

The preceding ^relation obviously shows that Bd ^is inversely proportional ^to Md. This feature has been derived here ^on physical grounds ^{but it can be} proved formally [7]. ^{As a}

consequence of relation (2) ^the unique ^variable parameter for the individual isotropic temperature ^{factor was} attributed to oxygens 01, 02, ⁰³ (see Fig. 2) and all the other factors

were deduced from it by multiplication with the ratio of the atomic masses such as, for

example :

We mention that due ^to the large wavelength, ^our diffraction pattern ^{is not} significantly

reduced by ^the Debye-Waller ^{factor at} large angles. ^These simple adjustments ^{of the} temperature ^factors gave ^{us, at} 82 K :

With all these settings, typical values for the different refinement factors that evaluate the

quality ^{of the} fit , ^{were :}

Re ^{is the} expected value for the refinement and ought ^{to be} nearly equal ^to

RWp, ^the weighted profile refinement factor, ^{and to} Ri ^the Bragg refinement factor.

RP ^{is the profile} refinement factor. Ri ^{is a} criterion for the diffraction pattern corresponding ^to

a given structural model ; ^while R,p ^{is a criterion for the} peak shapes being well described as

1459

well as for the goodness ^{of the} determination of the lattice parameters. ^{In the} present ^{case all} these parameters ^{have the same order of} magnitude indicating ^a very satisfactory refinement.

Since we merely ^{aimed to make a} measurement of the dimensionless parameter

Q ⁼

2 (b - a)/ (b ⁺ a ) ^versus temperature, it is admitted that the accuracy is sufficient. The

regular evolution of the lattice constants a, b, ^c gathered in table I confirms this.

Table I. - The parameters of the orthorhombic cell, ^at different temperatures. ^{The variance} of

the measure are between parentheses.

1460

Analysis of the data.

Figure ¹ represent the variation of the orthorhombicity parameter ^with temperature ^for samples 06.98 ^and °6.99. ^A simple calculation of the error bar, using ^the figures ⁱⁿ parentheses

in table I, gave ^{Aa = ±} 1 E-4 at 82.03 K, ^a value that is nearly independant ^{of the} temperature. ^{This error} is related to the counting statistic and does not include systematic

errors on the lattice parameters, that could proceed ^{from the uneven} representation ^of peaks

associated with each lattice constant. In that regard, ^the most accurate estimation among all the refined parameters, ought ^to be the lattice constant c, which is over-represented ^{in the}

diffraction pattern, because its value is nearly three times larger ^{than the two others.}

References [2, 3] that has been performed ^{at shorter} wavelength (1.05 ^{Â in Ref.} [3]) ^are expected ^to be much less sensitive to this effect.

We note that c decreases as the oxygen ^content increases, ^a feature that is ^now well- established. Furthermore the lattice constant c is ^more directly ^{related to} the oxygen ^content than is 6 ; indeed, _apart ^{from the} systematic ^errors mentioned in the preceding paragraph,

the statistical error on

is 0.6 % whereas it falls to 0.006 % ^{on c.}

Table II lists the parameters ^a, b, ^c and the cell volume, ^{at 83 K versus} oxygen

stochiometry. ^An important feature is evidenced in table II namely, the volume of the cell decreases with increasing oxygen ^content. Otherwise, one can see that the lattice constant c is

one of the lowest ever published ^{in the} litterature, Capponi et ^al. [8] ^{found a} lower value of c = 11.6329 (5) ^{À at 75 K for an} 07, but Horn et al. [1] ^{measured a} higher ^{one :}

c

11.6385 around 80 K, ^as estimated in figure 2 of reference [1], ^{in that} respect ^{we conclude} that our oxygen ^content is approximately ^{the same as} in reference [1].

Surprisingly, ^{our values for a are} lower than the corresponding ^ones in references [1], contradicting the above conclusion deduced from the value of parameter ^{c. This} discrepancy

is simply ^{due to} the fact that Horn et al. extracted their lattice parameters by fitting appropriately ^chosen peaks ^to Lorentzian line-shapes, ^{whereas we fitted our whole} spectrum thus favouring parameter ^{c, as} explained above. In fact in an early presentation of this work

[9], ^{we fitted some isolated} peaks ^{of our} spectrum, ^to Gaussian line shapes ^{and we obtained} higher ^values for a (T) than those in reference [1]. Nevertheless the o- (T) ^{curve was still}

similar in shape ^{to those in} figure 1, ^{but with a} greater dispersion ^{of the} experimental points.

This confirms that the discrepancy ^is simply ^{due to} systematic ^error in the evaluation of the different lattice constant, attributable to the fitting ^methods.

Table II.

The parameters o f ^the orthorhombic cell at 83 K, ^{at three} di fferent oxygenation stages. ^{The variance} of ^the measure are between parentheses. ^Note that the volume of ^{the cell,}

decreases with increasing oxygen ^content.

Discussion.

Remembering that the conduction in these superconducting materials, is attributable to holes

more or less confined in the CU02 planes (see Fig. 2), ^and admitting ^as in reference [10] ^that

1461

the appearance of ^a hole produces ^{a local} reorganization ^of the ion lattice charge distribution which requires ^a significant energy ; then the tendency ^for pairing would evolve from the fact that it costs less energy ^to add a second hole in the vicinity of the first _one, rather than to create the separate ^local charge reorganisations required ^{for two} isolated holes. But if the two directions of the copper plane ^{are not} equivalent, ^{then an} alignment ^{of the} paired holes will be favored if their axis is parallel ^to the easy direction. This effect being quite capable ^of disturbing ^the regular evolution of lattice constant a and b, ^with temperature. Although ^the suggestion ^{of a} bump ^{can be seen} in the data of figure ^{1, the scatter} of the data is much smaller than the statistical errors.

Conclusion.

The accuracy achieved in the present work, performed ^{on a} YBa2Cu306,99 sample, ^excludes

the existence of a maximum near Tc

⁹³ K, ^{on the curve of the} orthorhombicity parameter

versus temperature. ^The pending question ^is why, ^do X-rays display ^{such an} anomaly ^{on the} orthorhombicity parameter o-. ^{It could be} possible that this effect is somehow related to

superficial strains in the grains ^{of the} sample, ^{to which} X-ray diffraction is more sensitive than neutrons.

References

[1] ^{HORN P. M., KEANE D. T., HELD G. A.,} JORDAN-SWEET J. L., KAISER D. L. and HOLTZBZERG F. and RICE T. M., Phys. Rev. Lett. 59 (1987) ^2772.

[2] DAVID W. I. F., EDWARDS P. P., ^{HARRISON M.} R., JONES R. and WILSON C. C., Nature 331

(1988) ^245.

[3] FRANÇOIS M., ^JUNOD A., ^YVON K., ^{HEWAT A.} W., CAPPONI J. J., STROBEL P., ^MAREZIO M. and FISHER P., Solid State Commun. 66 (1988) ^1117-1125.

[4] RIETVELD H. M., ^J. Appl. Cryst. ² (1969) ^65.

[5] ^{HEWAT A.} W., ^Harwell report, AERE-57350 (1973).

[6] ^MONOD P., RIBAULT M., ^D’YVOIRE F., JEGOUDEZ J., COLLIN G. and REVCOLEVSHI A., J. Phys.

France 48 (1987) ^1369.

[7] Squires, Introduction to the theory of Thermal Neutron Scattering (Cambridge University Press) expression (3.74).

[8] CAPPONI J. J., CHAILLOUT C. , ^{HEWAT A.} W., ^LEJAY P. , ^MAREZIO M., ^NGUYEN N. , ^RAVEAU B., SOUBEYROUX J. L., THOLENCE J. L. and TOURNIER R., Europhys. ^{Lett. 3} (1987) ^1301.

[9] Tables Rondes du L.L.B (Novembre 1987).

[10] ^{HIRSCH J.} E., ^TANG S., LOH Jr E. and SCALAPINO D. J., Phys. Rev. Lett. 60 (1988) ^1668.