Skip to main content

Full text of "USPTO Patents Application 08479810"

See other formats


Mat, Res. Bull., Vol. 20, pp. 667-671, 1985. Printed in the USA. 
0025-5408/85 $3.00 + .00 Copyright (c) 1985 Pregamon Press Ltd. 



THE OXXCEN DEFECT PEROVSKITE BaLa^CUjO^ ^, A METALLIC CONDUCTOR 



C. Michel, L. Er-Rakho aad B. Rave a u 
Laboratoire de Cristaliographie, Chiaie et Physique des Sol ides, 0.A. 251 
ISMRa-Universite" de Caen, 14032 Caen Cedex, France 



(Received March 14, 1385; Hefereed) 



ABSTRACT 

A new oxygen defect pe rove kite BaLa^Ci^Op 4, characterised by a mixed 
valence of copper has been isolated ; the parameters of the tetragonal 
cell are closely related to tha t of the cubic perovskite; i-a • 8.644(4) i 
and c = 3.867(3) I "ftp- The X-ray diffraction study shows 
that the atoms are displaced from their ideal positions in the cubic 
cell, owing to the presence of ordered oxygen vacancies, the study of 
conductivity, magnetic susceptibility and the myoelectric power versus 
temperature shows that this oxide is a very good metallic conductor. 



INTROPUCTIOK 

Oxygen defect perovskites, have been more extensively studied these 
last years owing to their potential applications in catalysis, electrocata- 
lysis or as gauges (1-3). In this respect mixed valence copper oxides offer 
a wide field for investigation .: several perovskites (4) or perovskite^rela- 
ted structures have been isolated (5*6). these materials in which copper 
takes several coordinations simultaneously and a valence state intermediate 
between II and III can intercalate large amounts of oxygen according to the 
oxygen pressure and the tempera tare. Their electron transport properties 
ranging from semi-conductive to metallic (7) are closely correlated to the 
amount of intercalated oxygen. 

The present paper deals with a new oxygen defect perovskite 
aaLa4Cu50|3^4, which is like La^Ba^Cu^Oj^ (4) a mixed valence copper oxide 
but whose behavior is quite different* 

EXPERIMENTAL 

Synthesis 

Samples were prepared in platinum crucible and in air from appropriate 
airtures of dried oxides La 2 0 3 » CuO and carbonate BaO^. The mixtures were 
first heated a few hours at 9O0°C, ground and heated at 1000 °C during several 
hours. They were then ground again, and mixed with an organic binder, com- 
pressed into bars and then slowly heated up to IGOO'C. After 24 hours or 
»ore at I000*C» the bars were finally quenched to room temperature, The use 
of a binder was necessary to avoid that the compressed bars break before 

667 



BEST AVAILABLE COPY 



668 C. MICHEL, et al. Vol. 20. No. 6 

heating . In these conditions the compactness of bars was of about 80 X. 
Chemical analysis 

In order to determine the oxidation state of the transition metal ions, 
chemical analysis were carried out by iodotaetric titration using KI and by re- 
duction in a f low of 25 Z hydrogen id argon up to about 1000°C using a SETARAM 
raicrobalance for weight loss sieasurements. 

Structural analysis 

The cell parameters were determined from X-ray powder dif f ractograanas 
, registered with a Philips goniometer using Cu radiation. The space group 
was determined by electron diffraction using a JEOL 120CX electron microscope. 

Magnetic and electrical measurements 

The magnetic susceptibility was measured on powders by the Faraday me- 
thod in the range 6O-300K using a Cahn RC microbalance. 

The conductivity was measured by the four points method on sintered bars. 
It was calculated by measuring the intensity /voltage ratio between the points 
in each current: circulation direction in order to minimise the dissymetry ef- 
fect between the contacts* The Seebeck coefficient was measured on the same 
sintered bars bold between two Pt beads. 

Measurements were carried out up to 600K under an helium pressure of 
200 mbars for T < 290R and in air for T > 296K in order to avoid possible de- 
parture of oxygen* 

RESULTS AND DISCUSSION 

The scanning of the system La2°5" Ba ^~^ u0 for the compositions correspon- 
ding to the molar ratio (La + Ba) /Cu * 1 allowed us to isolate a perovskite 
for La/Ba «* 4. The X-ray diffraction pattern of this compounds presents be- 
sides the intense lines which can be indexed in a cubic perovskite cell, extra 
lines which are rather weak. This; feature is confirmed by the electron dif- 
fraction study, which shows superstructure reflections, leading to a tetra- 
gonal cell whose parameters are related to the cubic perovskite subcell (a ) 
as follows i P 

a a /5 c » a 

P P 

all the lines of the X-ray diffraction patterns can be then 
indexed with accuracy in the tetragonal system with a - 8.644(4) A and c .« 
3.867(3) L No reflection conditions are observed. The analysis of the oxy- 
gen content leads to the formulation BaLa^Cu^O^^ involving, the presence 
simultaneously of Cu(Il) and Cu(III) in spite of the presence of numerous oxy- 
gen vacancies (10.7 t) . The measure of the density by pyenometry in benzene 
at 25°c(d exo «• 7.05) confirms this composition for one mole per cell (d^j^ - 
7.03) . Thus it appears that the oxide BaLa^Cu^^sCu^^.gOji 4 j , p exhibits 
a great similarity with the oxygen defect perovskite Bail^CuTi^^Ou 111 ^^!^ 
previously described* However, this compound is very dif ferent fro® 
Ba^La-CugO}^^ from the point of view of the oxygen intercalation t no inter- 
calation or des intercalation of oxygen has been observed by annealing this pha- 
se at low temperature (400°C to 500 9 C) and under different oxygen pressures up 
to 1 bar contrary to Ba 31^.3(^1^0 14^5. In the same, way, no oxygen loss has been 
observed by TGA measurements for temperatures up to 650, 750 and 850°C and 
under oxygen pressure* of 0.02, 0.2 and 1 bar respectively. 

Taking into account the fact that the fundamental lines are indexed in a 
cubic perovskite cell and are strong with respect to the superstructure lines 
it was interesting to determine whether the metallic atoms were displaced 
from their ideal positions in the perovskite, or if the superstructure lines 



BEST AVAILABLE COPY 



d. 20, No. 6 



80 X. 



metal ions, 
I and by re- 
ag a SETABAM 



ctograans 
pace group 
microscope. 



a r ad ay me- 

intered bars 
the points 
syraetry ef- 
a the same 

ssure of 
ossible de- 



s correspotr 
erovskite 
e sent s bc- 

ceil, extra 
ctron dif- 
o a tetra- 
bcell (a p ) 



then 

and c • 
af the oxy- 
tesence 
umerous oxy- 
in benaene 

om 

: no inter- 
ing this pha~ 
pressured up 
oss has beeo 
SO°C and 

indexed in * 
cture lines 
displaced 
ture lines 



yd. 20, No. 6 



BaLa 4 Cu 5 0 13 



669 



gere only due to the ordering of oxygen vacancies . However, owing to the 
joall anoint of oxygen vacancies it was not likely to determine the distri- 
bution of the oxygen atoms by X-ray powder diffraction. Thus the structural 
study was undertaken for the composition La^BaC^Oj^ just to determine the 
positions of the atoms with respect to the cubic perovskite subcell. Eight 
space groups were possible, they were reduced to three PA, F4 and P4/m ta- 
king into account the analogy with the perovskite structure. Calculations 
cere carried out in the most symmetrical space group P4/a. Por a ranging 
frees 0 to 48*. 37 peaks i.e. 84 hkl were registered. The disparity between 
pkJcl and F hkl led us to introduce 139 hkl in the calculations. In the same, 
angle r nge 13 diffraction peaks (14 hkl) were indexed in the cubic perovski- 
te cell with a = 3.667 A, and used in a calculation with the atoms in the 
ideal positions of the cubic perovskite cell, involving only a refinement 
of the thermal factors B ; this first refinement led to a discrepancy' factor a « 
r « E|l obs o - I caic|/E I obs. of 0.066 with B(La, Ba) - 1.2 A 2 , B<Cu) « 2.6A' 
B(0) - 3.9 A 2 * The high B values let us think that the atoms were displaced 
front their ideal positions. A calculation carried out with all the iiite»si-v < 
ties in the P4/n space group and the same ideal positions and overall B - 1 A* 
(Table la), led to R « 0.35 in agreement with this point of view. Starting 
from these ideal positions, and assuming a statistical distribution of the 
oxygen vacancies in the oxides BaLa^CujO^ 4, the R factor was lowered to 
0.083, by refinement of the atomic parameters, the B factor being fixe at 
1 a?- Prom the final atomic parameters (Table Kb) it can be seen that se- 
veral atoms are displaced from their ideal positions in the cubic perovskite. 

TABLE I 

Atomic Parameters of BaLa^O^O^ $ (a) ideal positions (b) after refinement in 
the space group P4/m 



Atom 


Site 


X 


(a) 
V 


Z 




(b) 
Y 


Z 


Ba , La 


Kd) 


0.5 


0.5 


0.5 


0.5 


0.5 


0;5 


Ba, La 


4(k) 


0.1 


0.3 


0,5 


0.124(1) 


0.277(1) 


0.5 


Cu 


Ka) 


0.0 


o.d 


0.0 


0.0 


0.0 


0,0 


Cu 


Mi) 


0.4 


0.2 


0.0 


0.415(3) 


0.168(2) 


0.0 


0 


Kb) 


0.0 


0.0 


0.5 


0>0 


o.b 


0.5 


0 


2(e>^ 


0.0 


0.5 


0.0 


o.o 


0.5 


O.O 


0 


4(j) 


0.3 


0,4 


0.0 


0.261(7) 


0.384(8) 


0.0 


0 


4<j) 


0,2 


0.1 


0.0 


0.229(8) 


0.063(6) 


O.O 


0 


4(k) 


0.4 


0*2 


0.5 


0.428(10) 


0.155(6) 


0.5 



Portlier refinements, concerning the ordered distribution of oxygen 
in this structure, which is most probable, were not carried out due to the 
rather low content of oxygen vacancies, and the too small number of refle- 
ctions . 

This oxide is a very good conductor : its conductivity is about 
1.6 lo 3 (fi cm) - * at rooa temperature. * Figure I which represents the resistivi- 
ty P versus tempera tare, shows that this oxide exhibits a metallic conductivi- 
ty from 200 to 600K. The y value deduced from the equation p - p 0 O*Yt) 
(Y - 4.1 10-3 (T*) is very close to that of free electrons (y « 3.7 10" 3 C~ l ) . 

The molar magnetic susceptibility is very weak and nearly independent 
of temperature. This suggests a Paul! paramagnetism which is character is- 



BEST AVAILABLE COPY 



670 



C. MICHEL, et aK 



Vol. 20, No. 6 



A resistivity (a cm) x 10 4 



15 



10 



©o 



-100 0 100 200 

FIG. ! 

Resistivity plotted as a function of temperature 



300 t(X) 



tic of delocalized carriers. The Pauli susceptibility (8) calculated with 
nrVta * I arid for one carrier per Cu(III) (X^ = 5.3 l0~^ e.m.u) is however one 
order of magnitude lower than the experimental value ; xjj B 6 10~* e.m.u. The 
increasing of the Pauli susceptibility up to the experimental value needs 



10. This suggests a strongly correlated carriers gaz (degenerated 



spin polar on gaz) which was - introduced by Mott (9) to explain the magnetic 
susceptibility of LaCuOj and LaKi03 which are metals (10). At room tempera- 
ture, the Seebeck coefficient is also very weak and positive (ot - 9 uVK - *) 
and increases slightly with temperature (a$qoK * 18 urVK" 5 ) (Fig. 2) . This 



20 



10 - 



300 



400 



500 



TWO 



FIG. 2 



Evolution of the thermoelectric power as on function of absolute temperature. 



BEST AVAILABLE COPY