A INTRODUCTION

The most encouraging results in the porous silicon electroluminescence (EL) field are those obtained when the electrical bias is applied using a liquid electrode [1 3]. High external quantum yield, in the range of 0.1 to 1%, low applied voltages and voltage-induced spectral shift of the porous silicon electroluminescence [4] are the main attractive features of this junction. This 'wet' electroluminescence was first shown on anodically-polarised p-type porous silicon in contact with indifferent conducting liquid phase [1]. It was attributed to radiative recombination between holes supplied by the p-type substrate and electrons resulting from the oxidation of the Si-H and/or the Si-Si bonds on the porous silicon internal surface. The process via which minority carriers are injected is still the focus of on-going study [5 8]. Similar EL emission on p-type porous silicon was obtained by adding electron injectors to the indifferent solution. These species, such as the divalent methylviologen [9] ion or formic acid [10], oxidise through multistep electrochemical processes starting by hole capture and ending by electron injection into the porous silicon valence band. However, only a short-lived EL emission could be observed due to the simultaneous oxidation of the porous silicon layer which leads to the formation of a continuous oxide layer at the silicon/porous silicon interface. Consequently, the current flow through the porous layer is interrupted and the EL emission is suppressed [11]. Non-aqueous solvents [12] were also used in an attempt to suppress...