A INTRODUCTION

Silicon substrates are high quality single crystals, nearly defect free, with a very low volume density of impurities and a controlled amount of dopants. The surface may also be prepared flat on the atomic scale by standard methods used in microelectronics [1]. The formation of unidirectional pores [2] on such ideal electrode materials is therefore intriguing since corrosion pitting is generally associated with material defects and leakage from a passive layer. A recent attempt to treat the two questions in parallel [3] led to the conclusion that porous silicon formation arises from a specific process. Thus, the purpose of this Datareview is to examine the possible mechanisms responsible for pore initiation and propagation.

Since the early studies of Ulhir [4] and Turner [5] and later by Canham [6] porous silicon has mainly been obtained by anodisation in aqueous or ethanoic HF solution [2, 7]. Recent reports show, however, that organic solvents such as acetonitrile and DMF may also be used [8 10], A typical J-V curve is shown in FIGURE 1 for a diluted HF aqueous solution [11, 12], Pore formation occurs in the initial rising part of the curve for 0 < V < V _ep, with V _ep the potential of the small sharp peak. This peak, also called the electropolishing peak, has an intensity J _ep which depends mostly on the solution composition and little on the substrate [12]. For V > V _ep