ABSTRACT
The adhesion and removal of small particles on substrates is extremely important in many industrial applications [1, 2]. In some cases adhesion is desired (e.g.,
copiers, filtration units, coatings), but for many applications particles are considered to be contaminants that must be removed in order to improve process yield and performance. Removal of particles is generally performed in a liquid medium. This is due to the fact that in a liquid the adhesion force between the particles and the surface is significantly lower than in vacuo or air and, hence, removal is likely to be easier. A consequence of working in a liquid medium, however, is that the cleaning liquid must be removed from the substrate. This is done in a rinsing step where the process liquid is replaced with another, more dilute, rinsing liquid. The latter is eventually removed from the substrate in a drying step. Although much effort is typically spent on the optimization of the actual cleaning step, little effort is put into the rinsing step. Often one simply assumes the rinsing step to be particleneutral, i.e., neither adding or removing extra particles. We shall show that this is not necessarily the case. For critical applications — where final particle levels on the surface must be kept minimal — it is, therefore, crucial that the rinsing step is optimized together with the cleaning step. One of these critical applications is the manufacturing of integrated circuits. Because the typical dimension of their smallest components is of the order of 70 nm, unwanted particles — even as small as a few tens of nanometers — can result in device failure.
