Praseodymium valency from crystal structure in Pr-Ba-Cu-O and (Y-Pr)-Ba-Cu-O single crystals

HAL Id: jpa-00212437

https://hal.archives-ouvertes.fr/jpa-00212437

Submitted on 1 Jan 1990

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.

Praseodymium valency from crystal structure in

Pr-Ba-Cu-O and (Y-Pr)-Ba-Cu-O single crystals

G. Collin, P.A. Albouy, P. Monod, M. Ribault

To cite this version:

Praseodymium valency

from

_crystal

structure

in

Pr-Ba-Cu-O

and

_{(Y-Pr)-Ba-Cu-O single crystals}

G. Collin

_(a,

_b),

P. A.

_{Albouy (a),}

P. Monod

_(a)

and M. Ribault

_(a)

(a)

Laboratoire de

_Physique

des Solides

_(*),

Bât. 510, Université Paris-Sud, 91405

_Orsay,

France

(b)

C.P.C.M., DPhG, C.E.N.

_Saclay,

91191 Gif sur _Yvette, France

(Reçu

le 4 décembre 1989,

accepté

sous

forme définitive

le 23

février 1990)

Résumé. 2014 La substitution de Pr à Y conduit à des matériaux de formule

générale

(Y1-vPrv)(Ba2-x’Prx’)(Cu3-y~y)O6+x’/ 2-y+z,

avec une transition structurale autour de

x = v + x’ ~ 0,5. Pour x ~ _0,5 les cristaux sont

_{orthorhombiques,}

_type Y-Ba-Cu-O, mais avec une distortion atténuée. Pour x ~ _0,5 les cristaux sont

_{quadratiques,}

_type

_{La1,5Ba1,5Cu3O7 ± z,}

avec

le

_{tri-maclage caractéristique}

de cette

_phase.

La valence du Pr, de l’ordre de

_{3-3,2+ dépendant}

des conditions de

_{préparation,}

est déterminée

_d’après

les distances

_{interatomiques.}

Les cristaux

orthorhombiques

de

_{PrBa2Cu3-y~yOz}

présentent

un fort taux de défauts, y ~ 0,25, sur le site

Cu(1)

et sont semiconducteurs avec une loi d’activation en

T-1/4

attribuée aux fluctuations de valence du

praséodyme.

Abstract. 2014 The substitution of Pr to Y leads to materials with _a

general

formula

(Y1-vPrv)(Ba2-x’Prx’)(Cu3-y~y)O6+x’/ 2-y+z

and with a structural transition around x = _v + x’ ~ 0.5. For _{x ~} 0.5 the

crystals

_are orthorhombic, Y-Ba-Cu-O _type, but with _a lowered distortion. For v ~ 0.5 the

_crystals

are

_{tetragonal, La1.5Ba1.5Cu3O7 ± z}

_type, with the characteristic

tri-twinning

of this

phase.

The Pr

valency,

in the range 3-3.2+

depending

on

_preparation

conditions, is determined from interatomic distances. Orthorhombic

crystals

of

_PrBa2Cu3-

_y~yOz

prepared

at

_high

_temperatures exhibit a

_high

amount of

_{defects, y ~}

0.25 on the

_Cu(1)

site and are

semiconductors with a T-1/4 activation law attributed to the

praseodymium

valence fluctuation. Classification

Physics

Abstracts 61.10 - 61.70B - 74.70

1. Introduction.

Most of the Re-Ba-Cu-O materials are found to be

_{superconductive}

in the 90 K _range, even

with rare earth ions

_carrying

localized

_magnetic

moments

_except

for

_cerium,

_praseodymium

and

_terbium,

the three lanthanide

_ions,

which

_present

a marked

_tendency

to tetravalence and do not exhibit

_{high 7c superconductivity.}

_However,

the three mentioned elements do not

display

the same behaviour : Ce and Tb do not lead to the

_{required phase}

and

only

praseodymium

can substitute

_yttrium

and

_{gives isomorphic}

materials.

The

systems

(Y1 - v

Pr,)Ba2CU307 -,,

and

_{Y(Ba2-xPrx)Cu307:!:z}

have been

_investigated

on

powders by

several groups

[1-12]

and the end member «

_{PrBa2CU307 ±, »}

has been

_separately

described

_by

various authors

_{[11-20] .}

All agree with the occurrence of a

_{semiconducting}

character for v >

_0.5,

_{taking place}

after a

lowering

of

_{Tc in}

the 0 v 0.5 range. But in these

Pr-compounds

a further

complication

occurs due to the

_partial

substitution of Pr on the Ba

site,

as is the case with

light

_{rare earth}

(La-Nd),

substitution

_leading

to the

_general

formula

in which the

_{superconductivity}

can be

destroyed i) by

a

large

RE-Ba

_{(x ~ 0.5)}

substitution

level,

ii) by

an excess of defect on the

Cu(l)

site

_(y

>

0.15), iii)

as

_{usually, by}

an oxygen

deficiency (z

> -

0.2 )

and

iv) by

a

filling

of the

hybridized

_{copper-oxygen} bands

by

4f

electrons of Pr.

In this paper we

_investigate

these

_systems

on

_{single crystals,}

in order to determine which

parameter(s)

is

_(are)

_responsible

for the observed behaviour.

2.

_{Experimental.}

The materials are

_prepared

at 930 °

C,

from a mixture of rare earth oxide

_(prealably

reacted at

1 000 ° C in the case of mixed Y-Pr

_{materials), BaC03}

and CuO. After several

_grinding

and

firing

in the same conditions the

homogeneity

of the

_powder

is verified

_by

_X-ray

powder

diffraction.

_{Single crystals}

are obtained either

_by

_spontaneous

growth

from the

_powders

or

_by

flux

_(BaCu02

and

_CuO).

_Reannelings

in oxygen flux are carried out at 450-500 ° C for 12 h with a slow

_cooling

down to room

_temperature

in three

_days.

The reflections are collected on a four-circle

_{diffractometer,}

on a half-shere of

_reciprocal

space

(MoKa,

with a scan _{range oi} =

3°,

providing

the

required integration

of the different

twin

_samples

in the case of orthorhombic

_crystals).

After

_absorption

_correction,

the

independent

reflections are obtained as an _average of the

_equivalent

_reflections,

8 or 4 for the

tetragonal

and orthorhombic

_symmetry,

_{respectively.}

In all cases the whole set of reflections is used in the structural

_{calculations,}

_including

those with zero

_intensity.

3.

_Homogeneity

_{range and}

_crystal

structure determination.

On the

_powders

we have

_reproduced

the results

_{reported by previous}

_{authors ;}

the materials are

superconductors

_up to a substitution level v «

_0.5,

and semiconductors above.

However,

on

_{single crystals,}

a more subtle difference

_{appeared ;}

for v 0.5 the

_crystals

exhibit the orthorhombic

_YBa2CU3O7

_type,

whereas for

_{v > 0.5,}

_{they correspond}

to the

La1.5Ba1.5CU3O7.25

type.

Thus,

as the latter

_type

leads to

_{semiconducting}

_materials,

the

_change

observed around v = 0.5 does _not result

exclusively

from the electronic contribution of the

praseodymium

4f bands.

3.1 ORTHORHOMBIC CRYSTALS. - Under the usual conditions

(preparation

temperature

lower than 1 000

_{° C)}

and for v « 0.5 in the

_general

formula,

the

_crystals

exhibit a

_platelet

shape

and their lattice is

orthorhombic,

space groupe Pmmm

(Z

=

1 ),

hereafter referred _{to as}

type I,

with the usual

_{(a-b-c)-(b-a-c) twinning}

of

_{YBa2CU307 (a,}

b,

c the orthorhombic lattice

vectors),

but

_with

an orthorhombic distortion

decreasing

with

increasing

v.

In

_addition,

_{for crystals prepared by}

flux at rather

_high

temperatures

(starting

from

_{1 030° ),}

twinned orthorhombic

_crystals

with a

platelet shape

are also

obtained,

even for pure

praseodymium

materials. In order to

_investigate

the

_{changes occurring}

between

_yttrium

and

lanthanum,

orthorhombic

_{single crystals}

of

_gadolinium

and

_neodymium

materials were also

Table la. - Structural and thermal _parameters

_of

orthorhombic

_crystals.

Table Ib. -

Twinning

ratio =

_Vhkf/(Vhkf

+

_Vkhl).

Final formula :

For orthorhombic

_crystals

the cell _parameters are obtained from a

centering

routine on a

We determined the structure of a

_crystal

obtained from an initial

_{composition YO.7Pro.3}

(Y-Prl,

untreated),

of two _pure

_{praseodymium,}

Prl

_(untreated),

Pr2

_(oxygenated)

and of Gd and Nd

crystals (both untreated).

All these

_crystals

exhibit an

_unambiguous

orthorhombic

symmetry,

with a

_{twinning resulting}

in the

_{partial superposition}

of reflections issued from different domains at each node of the

_reciprocal

_{space and}

_clearly

observable on the

peak

profiles. Using

the

_{procedure previously}

described for this

_type

of

_{twinning [21],}

with the

_{corresponding}

derivatives

_including

an

_expansion

of

_temperature

factors up to the 4th rank in tensors, we obtained the final set of

_parameters

_reported

in table I. Our result confirms the

_regular

Y-Ba-Cu-O

_type

of these

_crystals

with

_i)

a _copper

_{deficiency, exclusively}

on the

_Cu(l)

site in-between two Ba

_units,

whereas the other sites are

_{fully occupied,}

_except,

as

usual,

the

0(4)-0(5)

sites and with

ii)

a

_{pronounced in-plane anisotropy}

of the Ba site with an

appreciable

anharmonic character

exclusively

observed for this site.

3.2 TETRAGONAL CRYSTALS. In contrast, for v >

_0.5,

the

_crystals

exhibit a cubic

shape

and are

tetragonal,

with the characteristic

tri-twinning (Fig. 1) already

observed for

lanthanum,

La(Ba2 - xLax)Cu307

± z

[22].

Indeed,

in

spite

of an

apparent

cubic

cell,

with

lattice constants

_{3 a * 3 a * 3 a,}

the diffraction

_patterns

of these

_{crystals correspond}

to a

twinning

of three distinct

_{tetragonal samples}

with lattice constants

_{respectively a}

* a * 3 _a, a * 3 a * _a, 3 a * 1I * a

(a

refers to the

_perovskite

substructure

_parameter)

because of the

Fig.

1. - Precession

photograph (MOÇ )

of the hk0

_{layer (equivalent}

to the

h0fand

_{0kf layers),}

of a

absence of reflections with less than two indexes = 3 n

(cf.

the

equivaieni

lanthanum

crystals

of Ref.

_[22]).

This

_tri-twinning

is characteristic of a

_high

level

of

Ba-R.E. substitution

(x

> 0.30 )

on the

_baryum

site. As

_yttrium,

under usual

condition,

does not

_{easily replace}

baryum,

this can

_only

be

_interpreted

as the _presence of

_large

amounts of Pr on the Ba site. In

all the cases the

_tetragonal

distortion is weak and cannot be

_appreciated

on the

_peak

_profiles

neither on

_precession

_photographs

nor on a counter whereas it should be in such twinned

crystals

with

_c/3

_appreciably

different from a

(in

the limit ouf 0.02

À

due to the

_mosaic).

The Pr-Ba substitution is

_easily

observed with

_{Pr-pure powders}

which shows that the

« stoichiometric » material

_{PrBa2CU307 ±,}

does not

_exist,

in our

_experimental

conditions and

corresponds

to a mixture with

_{BaCuO2 impurity.}

However,

from results on

_powders

with La

and

Pr,

we could show that this level of Ba-R.E. substitution is

_temperature

_dependent

and decreases with

_increasing

_preparation

_temperature.

Materials such as

R.E.

_{(Ba2 _ xR.E. x)CU30 Z’}

with low x,

0.20,

can be

_prepared

as

single phases, by quenching

from

_{T > 1 000 ° C,}

with a

_pronounced

orthorhombic distortion after

reannealing

in

02 (orthorhombic single crystals

examined in the

_previous

section are relevant of this

case),

whereas,

when

_prepared

at lower

_temperatures

or

slowly

cooled,

their diffraction

patterns

reveal a

_tetragonal

_symmetry

and exhibit

appreciable

amounts of

_BaCu02.

We may therefore

assign

a

_higher

Pr substitution level on the Ba site for

tetragonal

type

II

_{crystals (Pr3-4-5}

and

Y-Pr2-3)

than for orthorhombic

_type

1

_{crystals (Prl-2}

and

_Y-Prl),

in

_agreement

with the

preparation

conditions. This Pr-Ba substitution is furthermore

_responsible

for the observed

tri-twinning

because,

in

_regions

with a

_high

concentration of

_defects,

it

_changes

the

_packing

along

the a

_{(or b)}

axis into a c axis

_type

_packing.

But the number of 90° faults is small because

no

_{appreciable broadening}

of

_Bragg

reflections can be observed and because the sizes of the

integrated

volumes of the three

_types

of domains are _very different

_(see

Tab.

_{II) :}

a

_large

number of faults should lead to a

_{micro-twinning}

with

_twinning

ratio

_1/3,

_{1/3, 1/3.}

This _presence of

praseodymium

on the Ba site has also been mentioned

_by

Okai et al.

_[3],

Matsuda et al.

_[6],

Kinoshita et al.

[8]

who noted the presence of

BaCu02

in their materials with nominal

_composition

_{(Y1-’v Prv)Ba2Cu307:tz,}

and

_{by Sampathkumaran}

et al.

_[9]

and Suzuki et al.

_[10]

in materials

_{Y(Ba2-.J>rx)Cu307:tz.}

The

_crystal

of «

_PrBa2Cu307»

used

_by

Moran et al.

_[20],

was

_certainly

of this

_{type II,}

even if the

_tri-twinning

was not taken into

account

_by

the authors who mentioned that it was a

_tetragonal

small cube. The

_dispersion

of

values for lattice constants _{among the different authors is}

_surely

due to different levels of

Ba-Pr substitution

_according

to the

_preparation

conditions as is the case for the lanthanum

materials

_(cf.

Ref.

_[22]

for a

_review).

Under these conditions the

_{Tc lowering}

in the

v _range 0.0-0.5 cannot be

_{only interpreted}

as the effect of the

_particular

electronic

configuration

of Pr

ions,

because Pr-Ba substitution occurs even for lower Pr _content, with the

consequence of a

_lowering

of the orthorhombic distortion whithin this

_composition

range. For this

_{type II,}

five

_{single crystals}

were

_{examined ;}

two mixed-rare earth from initial flux

composition Yo.sPro.s(Y-Pr2),

reannealed in oxygen, and

YO.4Pro.6 (Y-Pr3),

untreated,

and three

_Pr-pure,

the first one obtained

_by

_spontaneous

_{crystallisation}

from the

_powder,

Pr3,

and

two others from

_flux,

Pr4 treated under

_partial

vacuum

(10-2

bar at

_{830 ° C)}

and Pr5 reannealed in an _oxygen flux. For the Gd material no such

_crystals

were found and in the Nd

one, in

spite

of a

possible

Ba-Nd substitution

_{[23-28], only}

a few

_crystals

with this

_shape

were

observed but much too

_tiny

to be

_{investigated.}

For these tri-twinned

_crystals,

a data collection

_using

a

_tetragonal

unit cell will

_provide

two

distinct

_types

of

reflections,

first from the

substructure, hkf with f

= 3

n, which

corresponds

to the

_{superposition}

of the diffracted intensities of the three

_samples,

and second from the

superstructure

hkQ

with £ =1=

3 n which are

_exclusively

issued from the

_{given sample}

retained

Table Ila. - Structural and thermal _parameters

_{of tetragonal crystals.}

Table IIb. -

Twinning

ratio.

with superstructure reflections

_Fhkl(t:#=

_{3 n)} calculated in the usual way and substructure reflections calculated as

A

_spontaneous

_equalization

of the two scale factors

_{corresponding}

to the two

_types

of reflections

_respectively

was obtained for the five

_crystals

_which,

after the

_isotypic

La substituted

_{crystals, definitely}

confirms the existence of a

_{systematic tetragonal tri-twinning}

of

these

type

II

_crystals.

Moreover,

as in lanthanum

_crystals,

both

Cu(l)

and

_Cu(2)

sites exhibit

_appreciably

anharmonic

_potential

_wells,

as revealed

by

the introduction of an

expansion

of the

temperature

factors up to the 4th rank in tensors

_{(Hermite polynomial expansion}

[29]).

This

can be related to the rare earth substitution on the barium site which introduces different

fluctuations in the

_{equilibrium positions}

for the copper atoms

- between ideal

Ba2 + -Cu-Ba2 +

and disturbed

Ba2 + -Cu- R.E 3+

_or

R.E3 + -Cu- R.E 3 +

configurations

around

_Cu(l)

- and between ideal

Ba2 +

-Cu-R. E 3+

and modified

R. E3+

-Cu-R. E 3+ packing

around

Cu(2).

This makes them

_appreciably

different from the

_tetragonal

_{YBa2Cu306 + E}

materials in which such a behaviour is not

_observed,

in

_spite

of

_closely

related structures. It should be

noted,

for these

_crystals,

that the

_tri-twinning

is not

_perfect,

with often a

_{predominant sample,}

we retained for the data

collection,

and that the anharmonic correction for copper

positions

is

smaller for mixed Y-Pr

_crystals

than for the

_Pr-pure

ones.

4. Results and discussion.

4.1 INTERATOMIC DISTANCES AND COPPER VACANCIES. It is difficult to

_appreciate

the level of Ba-Pr substitution with

_X-rays

because of the close number of electrons of these two

elements.

Moreover,

the different amount of vacancies on the

_Cu(l)

site in the two

_types

of _

crystals

makes it

impossible

to _compare the average

Ba(Pr)-Ox

distances which should be

shortened

_by

_{the presence of the lanthanide ion in the}

_{tetragonal samples.}

This results in

average distances in the range 2.85-2.90

À,

for which the fluctuation is

essentially

due to the

amount of oxygen

present

on the

_0(4-5)

sites.

An

_oustanding

feature is the amount of copper vacancies

increasing

from

yttrium

to

lanthanum in orthorhombic

_{crystals ;}

in the range 4-8 % for Y-materials

[19],

18 % for

Gd,

25 % for Nd and Pr and up to 30 % for La

_[20].

This is associated with an increase of the

in-plane

thermal vibration

_amplitude

and with a decrease of the

in-plane anisotropy

(U 11 ~ U22 )

except

for Y-Ba-Cu-O. This also results in a

lowering

of the _average

Cu(1)-O(1)

distance

_(along

_{c) and,}

as a _{consequence, in} an increase of the _average

Cu(2)-O(1)

distance

which in our orthorhombic

_crystals

is

_{systematically}

found to be

_{larger, ~}

2.30

_Á,

than that deduced from refinements on « stoichiometric »

_powders

or on less deficient

_type

II

_crystals

~ 2.20

Á.

We

_emphasize

the fact that all these

_crystals

have been treated in the same _{way ;} all

exhibit

_{platelet-like shape, --}

200 * 200 * 50

_{mm .}

with

_{equivalent absorption}

correction and

are refined

using

the same

_procedure

_including

the

twinning.

As

they

lead to different

amounts of vacancies

_with,

moreover, the mentioned

systematics,

this eliminates an intrinsic

artefact

due to the refinement

_procedure

and shall be attributed to a

_physical

reason.

_Indeed,

the

_synthesis

of

_{crystals requires}

a

_temperature

_{systematically higher}

from Y to La materials. This results in an enhancement of the copper

deficiency

on the

_Cu(l)

site

_{(exclusively).}

For

_large

_copper deficiencies such as those observed in La and Pr

_crystals,

_X-ray

diffuse

Table IIIa. Interatomic distances

(Â)

in orthorhombic

_crystals.

All standard deviations are

0.01

Â.

* _and **

one or both sites are

partially

vacant

_(cf.

Tabs. I and

_II).

Table IIIb. Interatomic distances

(Â)

in

tetragonal crystals.

an La

_{single crystal}

in

_{(Ref. [22]).}

The

_comparison

of

_figure

2a,

_{PrBa2CU~ 2.75Oz,}

and

_figure

_2b,

YBa2Cu= 2.95° z clearly

illustrates the difference between

_high

and low vacancy level in

agreement

with the results of refinement.

Moreover,

a further confirmation is

_{given by}

the level of vacancies in

_{tetragonal crystals}

(0 _ y , 0.15 % ), systematically

lower than in orthorhombic

_samples.

These

_crystals

are

prepared

at lower

temperatures,

as confirmed

_by

their

_higher

level of Ba-Pr substitution. The

case of the

_strictly

stoichiometric Pr3

_{crystal, directly}

issued from the

_powder

without

_melting,

definitely

confirms this

_point.

It should be noted that this concerns

_{exclusively crystals}

and that

_powders

are not or

only

slightly

copper deficient but are

_decomposed

at the

_high

_temperatures

used for the

preparation

of

_crystals.

The

_crystals

appear as

_{out-of-equilibrium high}

_temperature

quenched

Fig.

2. - X

ray diffuse

scattering

pattern

(MoK,, ) of a)

an’orthorhombic

Pr (Ba2 -

J>rx)Cu3 -

yOz

single

crystal (y - 0.25 )

and

b)

of an orthorhombic

YBa2Cu3 - yO

single crystal (y - 0.05 ) ;

incident beam

4.2 THE OXYDATION STATE OF PRESEODYMIUM. - In these materials the valence

(m + )

of

Pr,

which can _{vary from} 3+ to 4+ in

_oxides,

is still a controversial

_question.

Various authors _report values

_including

almost the whole variation range of the

praseodymium

oxidation states

_(Tab.

_IV).

Table IV.

_{- Praseodymium}

valence estimation

_{by different}

methods.

**

« Essentially single

phase

». The

impurity commonly

mentioned in ceramics is

_BaCu02.

*

« predominantly

trivalent ».

Our own _{measurements,} between 77 and 680 K

using

a

_Faraday

_balance,

on

_{semiconducting}

R.E.

_{(Ba2 - R.E.,,)2CU307 + xl 2}

_materials,

x =

0.15,

0.35 and 0.50

(strictly single phased

and

annealed)

and

_least-square

calculations

_{(after diamagnetic corrections)}

showed

_{that ;}

i)

for R.E. = Pr and Nd the

susceptibility

can be fitted

_by

_{a X = X 0 + C / ( T - 0 )}

law,

as

previously reported [6,

7,

12,

18],

at least above 150 K. At lower

_temperatures

the inverse

susceptibility

shows

_appreciable

curvature,

possibly

due to a

crystal-field splitting [17,

18].

In

order to eliminate all

_{possible problems,}

the moments were calculated between 290 and

680

K ;

ii)

the values

_{of X o .--}

1.5 x

10-3

emu, for Pr and Nd are

_{appreciably larger}

than the

_quasi

temperature

independent susceptibility, =

5 x

10-4

emu, obtained with R.E. = La which is a

better reference than the metallic

_{YBa2CU307 compound.}

These values of Xo are in the same

range as those

_{given by previous}

authors ;

iii)

the _average values of the

_magnetic

moments are

_{3.40 jL B}

for

Nd,

close to the theoretical value

_{3.62» B and = 2.9 IL B}

for Pr. This

_corresponds

to an _average valence close to

Pr 3.7+ (theoretical

values 3.58

_{IL Band}

2.54 _{IL B} for

Pr3 +

and

Pr4

_{+ ).}

From the interatomic distances it is

_possible

to

_get

a rather

_unambiguous

estimate of the Pr

oxidation state, because the conventional ionic radii

[32]

of

Pr3 +

_{(1.126 Å)}

and

Pr4 +

_(0.960

_{Â )}

are

_appreciably

different and should lead to

_{significantly}

_{distinct average}

Pr-Ox values in the 8-fold coordinated R.E.

_site,

with the limitation of

_possible

fluctuations in interatomic distances introduced

_by

the variable and unknown level of substitution of Pr on

the Ba site.

_However,

these corrections are small in front of the difference between

Pr3+

and

_{Pr4 + ,}

as the contribution

due

to the vacancies on the

Cu(1)

_site,

far away from the

Fig.

3. - Determination of the Pr

valency

from interatomic distances. The solid curve is

plotted

from the R.E.-Ox distances

_{of strictly}

3+ ions

_{(black circles),}

Nd, Gd

_(this

work Tab.

_III),

Ho, Y and La from

(Refs. [19]

and

_[20]).

The crosses _represent the ideal values for

Pr3 +

and

Pr4 +

ions. From the average

Pr-Ox distances

_{of crystals}

Prl-5, open circles, the Pr values of ionic radius are obtained. The

triangles

correspond

to the

points

dR.E.-Ox = fl (1 - v) * ry + v * rp,] in

(Y, -,,Pr,) (Ba2 - xprx)CU3 -,0-,,

single

crystals.

The different

_points

are localized on or _very close to the line, which constitutes a confirmation

of the result

_conceming

the deduced Pr values.

In

_figure

3 the different values of the average R.E.-Ox distances are

_plotted

as a function of

the ionic radii for

_{unambiguously}

trivalent

_ions,

_{Nd3+ ,}

Gd3 +

and

_including

the

La 3 +

and

Y3 +

values from

_{(Refs. [21] and [22]),}

obtained with the same conditions. We have also

given

the value of a

_crystal

with

_composition

_{HoBa2CU2.934(7)06.88 5 , R}

factor 2.2

%,

which

corresponds

to a value close to that

_{of yttrium (ionic}

radii 1.015

A and

1.019

À respectively).

The distances for the

_{praseodymium crystals reported}

on the curve lead to ionic radii from

1.094(7)

to

1.123(7) À:

i)

the

_{poorly oxygenated (Prl, type 1) crystal,}

exhibits the

_highest

R.E.-Ox distances

.2.493(4)

Á,

which

_corresponds

to valence

_{3.01(4) + ;}

ii)

after a low

_temperature

_reannealing

in _oxygen for

_{Pr2, type I,}

and

_{Pr5, type}

II

crystals,

the

distances,

2.482(4) Â

and

2.481(4) Á (valence

3 .10(4) +

and

_{3.11 (4) + )}

remain in the same

range as for non-annealed

_crystals

such as

_(Pr4,

_type

_II),

2.484(4) Â

and

_{3.09(4) + ;}

iii)

the Pr3

_crystal

exhibits the smallest

distances,

2.467(4)

Á.

This

_crystal,

which was

obtained from a

_{crystallisation}

in the

_powder

at a

_temperature

lower than the other

samples

prepared by

flux,

exhibits the

_highest

oxidation state,

3.20(4)

+ ,

close to the value

_{given by}

Chittipeddi et

al.

_[18]

and

_by

Moran et al.

_[25]

from

_{susceptibility}

measurements on

_powder

and

_crystal.

This indicates that a

_partial

Pr3 +

-

pr4 +

oxidation takes

place

at

_{relatively high}

temperatures,

around 900

_{° C,}

and that the

_preparation

conditions are a determinant

parameter

for the final oxidation state of

_praseodymium

as

_previously

mentioned

_by

Moreover,

taking

the deduced values of Pr ionic

_radius,

the

points

for the Y-Pr

_crystals

are

all in

_agreement

with the

_expected

values

_(Tab.

_V).

Another

_tetragonal

_type

II

_crystal,

with

composition

_{YO.54(1)Pr0.46(1)Ba(Pr)2Cu2.89(1)06.77(5)’}

(YPr2b, R

factor

_{2.4 %),}

which is also obtained

by

spontaneous

crystallisation

in the

_powder,

leads to the same value as the Pr3

crystal

for the Pr ionic

_radius,

_resulting

in a valence

_{3.23 (4)+}

_{(Tab. V).}

Table V.

_{- Experimental}

and calculated average R.E.-Ox distances in Y-Pr

crystals.

It should be noted that our results concern the

_praseodymium

atoms on the R.E. site

_only

and that we cannot deduce any information

concerning

the Pr ions on the Ba site.

These values deduced from structural data are in contradiction with most

_{susceptibility}

measurements on

_{powders, including}

_ours, which

_{systematically}

find a

_higher

amount of

Pr4 + ,

except

Chittipeddi et

al.

[18].

In contrast, then are in better

_agreement

with the results

of soft-X-Ray

absorbtion fine structure

_[4]

and XANES

[31].

This is

_probably

due to the fact that

_SXAFS,

XANES and structural determination lead to the intrinsic

_{properties, directly}

observed,

of the Pr

_ions,

whereas

_magnetic

measurements indicate

_only

_{the response of the} whole

_system

in a

_magnetic

field

_and,

as

_pointed

out

_{by Peng et}

al.

_[12],

it is

_important

to note

that a reduced value of the Pr moment can arise from

_{crystal-field}

effects. In contrast, a rather

precise

determination such as that deduced from bond

_length

leads to an

_unambiguous

answer

which accounts

_i)

for the difference observed between Pr

_{predominantly}

trivalent and which

gives

the

_regular

1-2-3

_phase

whereas Ce and

Tb,

more

spontaneously

tetravalent do _not, and

ii)

for the

_specific

behaviour of Pr as

_compared

to

_purely

trivalent rare

_earth,

due to a small

but noticeable amount of

_{hybrydisation}

of 4f orbitals which

_compensates

the holes on the

Cu-Oxygen

bands and suppresses the metallic character and

consequentely

the

_{superconductivity.}

However we

_emphazise

the fact

_that,

in addition to this fundamental

_mechanism,

Pr substitution on the Ba site is a further reason for

_suppressing

the metallic behaviour and

complicates

the

_{interpretation.}

Finally

we would like to mention that E.P.R.

_experiments

on Pr orthorhombic

_crystals

did

not reveal any

peak

of resonance between 6 and 300 K.

4.3 RESISTIVITY. - The

_resistivity

of orthorhombic

_{single crystals}

_{reannealed in oxygen} was

measured both in the a-b

_plane

and in the c direction between 300 and 80 K. Below this

temperature

the resistance becomes

_prohibitive

to _any reliable measurement.

_Typical

dimensions of the

_crystals

were 0.6 * 0.6 * 0.2 mm

3.

Gold wires were

_glued

on the

_{sample by}

silver

_bonding

adhesive

_(Dupont

4929 or

_6838).

At room

_temperature

the

_resistivity

in the a-b

_plane

and in the c direction is 4 fi.cm and 43 Q.cm

_{respectively.}

This

_anisotropy

does not

vary

significantly

in

decreasing

temperature.

A

_{semiconducting}

behaviour is observed on the

whole

_temperature

_range but this variation never fits an Arrhenius law

_{(Fig. 4).}

The accuracy

of our measurements does not allow us to

_distinguish

between 2-D or 3-D

_hopping

processes.

Fig.

4. -

Resistance versus T in the a-b

plane

of an orthorhombic

_{PrBa2Cu3 - yOz}

_{single crystal,}

reannealed in

_flowing

oxygen. Insert ; resistance versus

T- 1/4.

the 3-D

_hopping

law

_by

more than 1.5 %.

_Considering

the low value of the

_anisotropy

we

compare our results to the 3-D

_hopping

process : R oc

Ro exp [-

(To/T)1/4].

From

_sample

to

sample To

varies between 2.5 x

108

K to 4 x

108

K. This _range of values is three orders of

magnitude higher

than in

_La2Cu04

or in

_YBa2Cu306

or in the

_{LaBa2Cu3 -}

_YO,

_homologous

single crystals.

The last

_crystals

were also semiconductors but with a

_resistivity

which did not

exhibit such an activation law

_[20]

and were close to the

_{semiconducting YBa2Cu306 single}

crystals :

the copper vacancies

(

~ 30 % on the

_{Cu(l) site)}

forbid a

_{complete oxygenation,}

_up

to

_07,

and

_limiting

the

_composition

to

_{LaBa2Cu3 - y07 - Y}

_plays

the same role as the

introduction of _oxygen vacancies.

In

_praseodymium

_materials,

in addition to the

_{disappearance}

of the metallic behaviour due

to _copper

_vacancies,

a variable range

hopping

occurs which can be

specifically

attributed to

the _presence of Pr ions. This behaviour was

_previously

mentioned on « stoichiometric »

powders by

Matsuda et al.

[6]

who also

reported

a decrease of the Hall carrier number with

increasing

v in

_{YI - vPr vBa2Cu30 z}

and

_{by Chittipeddi et}

al.

_[18]

for v =

1,

at least between 300

and 80 K. This indicates that valence fluctuations of the Pr

_{ions, ~ 3.10-3.20+}

from our

determination,

in addition to

_{pair breaking}

_{[ 11,}

_12,

_18],

introduce fluctuations in the band

filling

responsible

for the occurrence of variable range

hopping.

5. Conclusion.

The characteristic behaviour of the

_{(Y-Pr)-Ba-Cu-O}

_system

exhibits a close

_analogy

with the

La-Ba-Cu-O

_system,

especiâlly

a marked

_tendency

of the Pr ion to substitute

_Ba,

_leading

to a

tetragonal

non-superconducting phase

and a

_high

amount of vacancies on the

_Cu(l)

site in

orthorhombic

_crystals.

In

_addition,

even for low Pr-Ba substitution

level,

the ~ 10-20 % of

tetravalent

_praseodymium

ions

prevent

the occurrence of a metallic

phase

which

gives

an

References

[1]

SODERHOLM _L., ZHANG K., HINKS D. G., BENO M. A., JORGENSEN J. D., SEGRE C. U. and SCHULLER I. K., Nature 328

₍₁₉₈₇₎

604.

[2]

LIANG J. _K., XU X. T., XIE S. S., RAO G. H., SHAO X. Y. and DUAN Z. L., Z.

_Phys.

B 69

₍₁₉₈₇₎

137.

[3]

OKAI B., KosuGE M., NOZAKI _H., TAKAHASI K. and OHTA _{M., Jpn} J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L41.

[4]

NEUKIRCH U., SIMMONS C. T., SLADECZEK P., LAUBSCHAT C., STREBEL O., KAINDL G. and

SARMA D. _D.,

_Europhys.

Lett. 5

₍₁₉₈₈₎

567.

[5] a)

GONÇALVES A. P., SANTOS I. C., LOPES E. B., HENRIQUES R. _T., ALMEIDA M. and FIGUEIREDO M. O.,

Phys.

Rev. B 37

₍₁₉₈₈₎

7476;

b) Physica

C 153-155

₍₁₉₈₈₎

910.

[6]

MATSUDA A., KINOSHITA K., ISHII _T., SHIBATA H., WATANABE T. and YAMADA T.,

Phys.

Rev. B

38

₍₁₉₈₈₎

2910.

[7] a)

DALICHAOUCH Y., TORIKACHVILI M. _S., EARLY E. _A., LEE B. _W., SEAMAN C. L., YANG K. N., ZHOU H. and MAPLE M. B., Solid State Commun. 9

₍₁₉₈₈₎

1001 ;

b) Physica

C 153-155

(1988)

858.

[8]

KINOSHITA K., MATSUDA A., SHIBATA H., ISHII T., WATANABE T. and YAMADA _{T., Jpn} J.

_Appl.

Phys.

27

₍₁₉₈₈₎

L1642.

[9]

SAMPATHKUMARAN E. V., SUZUKI, KOHN K., SHIBUYA _T., TOHDAKE A. and ISHIKAWA _{M., Jpn} J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L584.

[10]

SUZUKI A., SAMPATHKUMARAN E. V., KOHN K., SHIBUYA T., TOHDAKE A. and ISHIKAWA M., Jpn J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L792.

[11]

KEBEDE A., JEE C. S., SCHWEGLER J., CROW J. E., MIHALISIN _T., MYER G. _H., SALOMON R. _E.,

SCHLOTTMAN P., KURIC M. V., BLOOM S. H. and GUERTIN R. P.,

Phys.

Rev. B 40

₍₁₉₈₉₎

4453.

[12]

PENG J. L., KLAVINS P., SHELTON R. N., RADOUSKY H. B., HAHN P. A. and BERNARDEZ L.,

Phys.

Rev. B 40

₍₁₉₈₉₎

4517.

[13]

OSEROF S. B., VIER D. C., SMYTH J. F., SALLING C. R., SCHULTZ S., DALICHAOUCH Y., LEE B. W., MAPLE M. _B., FISK _Z., THOMPSON J. _D., SMITH J. L. and ZIRNGIEBL E., Novel

Superconductivity,

Eds. A. Wolf and W. Z. Kresin

_(Plenum

Press, New

_York)

1987, p. 679.

[14]

TARASCON J. M., GREENE L. _H., BAGLEY B. G., MCKINNON W. R., BARBOUX P. and HULL

G. _W., Novel

_{Superconductivity,}

Eds. A. Wolf and W. Z. Kresin

_(Plenum

Press, New

_York)

1987, p. 705.

[15]

MAPLE M. B., DALICHAOUCH Y., FERREIRA J. M., HAKE R. R., LAMBERT S. E., LEE B. W.,

NEUMEIER J. _J., TORIKACHVILI M. _S., YANG K. N. and ZHOU _H., Novel

_{Superconductivity,}

Eds. A. Wolf and W. Z. Kresin

(Plenum

Press, New

_York)

1987,

p. 839.

[16]

LEE S.-I., GOLBEN J. P., YI SONG, CHEN X.-D., MCMICHAEL R. D. and GAINES J. R., Novel

Superconductivity,

Eds. A. Wolf and W. Z. Kresin

(Plenum

Press, New

_York)

1987, p. 977.

[17]

TSURUMI S., IWATA T., TAJIMA Y. and HIKITA M.,

Jpn

J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L397.

[18] a)

CHITTIPEDI S., SONG _Y., COX D. L., GAINES J. R., GOLBEN J. P. and EPSTEIN A. J.,

Phys.

Rev.

B 37

₍₁₉₈₈₎

7454 ;

b)

CHITTIPEDI S., EPSTEIN A. J., SONG Y., GAINES J. R., FARNETH W. E. and MCCARRON III E. M.,

Synth.

Met. 29

₍₁₉₈₉₎

F651 ;

c)

CHITTIPEDI S., SONG Y., GAINES J. R., FARNETH W. E., MCCARRON III E. M. and EPSTEIN

A. J., J.

_{Phys. Colloq.}

France 49

₍₁₉₈₈₎

C8-2141.

[19]

LOPEZ-MORALES M. _E., RIOS-JARA D., TAGUENA-MARTINEZ J. and ESCUDERO _R.,

_Physica

C

153-155

₍₁₉₈₈₎

942.

[20]

MORAN E., AMADOR U., BARAHONA M., ALARIO-FRANCO M. A., VEGAS A. and RODRIGUEZ-CARVAJAL _J.,

_Physica

C 153-155

₍₁₉₈₈₎

423.

[21]

COLLIN G., AUDIER A. C. and COMES R., J.

_Phys.

France 49

₍₁₉₈₈₎

383.

[22]

COLLIN G., AUDIER A. C., ALBOUY P.-A., SENOUSSI S., COMES R., KONCZYKOWSKI M. and

[23]

IZUMI F., TAKEKAWA S., MATSUI Y., IYI _N., ASANO H., ISHIGAKI T. and WATANABE N.,

Jpn

J.

Appl. Phys.

26

₍₁₉₈₇₎

L1616.

[24]

TAKEKAWA _S., NOZAKI H., ISHIZAWA Y. and IYI N.,

Jpn

J.

_{Appl. Phys.}

26

₍₁₉₈₇₎

L2076.

[25]

NOZAKI H., TAKEKAWA S. and ISHIZAWA Y., Jpn J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L31.

[26]

TAKITA K., KATOH _H., AKINAGA H., NISHINO M., ISHIGAKI T. and ASANO H.,

Jpn

J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L57.

[27]

TAKITA K., AKINAGA H., KATOH H., ASANO H. and MASUDA K.,

Jpn

J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L67.

[28]

TSURUMI S., IWATA T., TAJIMA Y. and HIKITA M.,

Jpn

J.

_{Appl. Phys.}

27

₍₁₉₈₈₎

L80.

[29]

JOHNSON C. K. International Tables for

_{Crystallography}

_{(Kynoch Press)}

4

₍₁₉₇₄₎

311.

[30]

GUINIER A., Théorie et

_technique

de la

radiocristallographie (Masson

éditeur,

Paris)

1956.

[31]

ALP E. E., SODERHOLM L., SHENOY G. K., HIKKS D. G., VEAL B. W. and MONTANO P. A.,

Physica

B 150

₍₁₉₈₈₎

74.