Influence of surface morphology to laser cleaning efficiency

Abstract

Laser cleaning is an effective cleaning method that can be widely used in microelectronic industry. Mechanisms of laser induced removal of particulate contaminants from solid substrates, such as Si wafer, are of great concern. The previous works on laser cleaning are mostly based on perfectly flat surface model, while few of them are objected to the influence of surface morphology. In the IC process, however, the Si surface morphology will change dramatically after few steps of processing, therefore its influence to laser cleaning efficiency is inevitable. In this experiment, the cleaning efficiencies of 2.5 and 1.0 μm spherical silica particles from Si with different surface morphologies are investigated. These surfaces were achieved by anisotropic etching of Si wafers in KOH solvent, with etching times raging from 1 minute to 30 minutes. Atomic force microscope (AFM) observation showed the Rp-v roughness of thus-processed Si surfaces ranges from 1.6 nm to 70 nm, and rms roughness ranges from 0.1 nm to 9.0 nm. For both kinds of particles, the cleaning efficiencies decrease with increasing surface roughness. The possible reason is that the rough surface may buffer the sudden thermal expansion of the substrate, making the particles more difficult to be removed.