Role catalytique de l’oxyde d’indium dans la réduction de l’oxygène

Laboratoire de génie électrique des Universités Paris VI et Paris XI, associé au CNRS, 33, avenue du Général-Leclerc, 92260 Fontenay-aux-Roses, France.

Résumé

L'oxyde, l'hydroxyde et le sous-hydroxyde d’indium sont les composés constitutifs d’un matériau pulvérulent obtenu par un traitement approprié d’un amalgame d’indium. Ce matériau catalyse la réduction électrochimique de l’oxygène. La détermination des isothermes de chimisorption des diifé- rents composés concernés conduit à attribuer le rôle catalytique au seul oxyde d’indium ; ce que confirment des essais électrochimiques directs. La faible teneur en oxyde d’indium du matériau considéré soulève une difficulté. Les auteurs proposent une explication.

Abstract

Indium amalgams are generally liquid at room temperature. By an appropriate treatment they can be transformed into a pulverulent material, each grain of which is constituted by a liquid core surrounded by an indium hydroxyde shell. By a thermical operation, we realise the evaporation of liquid cores while indium hydroxyde In(OH)₃ is transformed into InOOH and In₂O₃. Each grain of powder becomes a kind of empty conch, the constitutive material of which is a mixture of the three former compounds.

This powder which we have called « active material » is able to catalyse the electrochemical reduction of oxygen. We have tried to identify the compound which is responsible of this catalytic effect. Chemical adsorption analysis of oxygen at 25 °C shows that only indium oxyde is able to adsorb this gaz. These experiences are confirmed by electrochemical direct essays which indicate that oxyde is able to reduce oxygen while hydroxyde isn’t.

Then a difficulty appears because active material which can contain just a little percentage of indium oxyde and pure indium oxyde have the same catalytic effect. We have formulated the following assumption : indium oxyde would present an amphoteric character ; it would be a N semiconductor when isolated and a P semiconductor inside active material. The respective coverages of N and P types would allow to scratch out the difficulty. Our purpose is to verify this hypothesis during our next works.