No abstract

No abstract

Process simulation based on a two dimensionaldiffusion model for the reaction ofthe hexamethyldisilazan (HMDS) within the resist considering the crosslinks between photo active component (PAC) and resin and the following dry development is outlined. Results of silylation and etching simulation are discussed. Furthermore the influence of the resin composition and the conditions of silylation will be presented for an ORWO-resist. 1.IntroductionWafer stepper application can be extended by using a surface imaging process such as the DESIRE-Process [1]. The original pattern are transfered into the surface layer of a thick planarising resist layer. This surface layer modified by the incident ligth is used for the formation of an etching mask. The complete resist structure is obtained by dry development with anisotropic etching using this mask (dry development). In the DESIRE-Process the etching mask is produced by diffusion of a silicon-organic compound (most HMDS) and by the reaction of this compound with the resin selectivly within the exposed resist areas. Process simulation is of advantage for a further process optimization and its later application. We simulated numerically this process and derived necessary input parameters. The outlined process simulation is based on a two dimensional diffusion equation for diffusion and reaction of the HMDS within the resin. Beginning with a photolithography simulation programme [5] the PAC-concentration distribution after exposure was calculated using Dill's ABC-parameters [6]. The etchiiig simulator based on a cell-removal-model is applicated for processing the silylated structures. The used parameters for simulation were derived by experiments. ExperimentsDifferent samples of a basic ORWO-photoresist GS 7010 (binder: novolac, PAC: modified naphthoquinonediazide) was used for all experiments. The samples differ in the way of resiil manufacturing and therefore in the number of the hydroxyl groups as an important parameter for the silylation process. The sample S2 in the presented study is that with the lower number of hydroxyl groups in the spincoated resist (arround 2.0 x 1021 cm3). For the sample Si we determinded around 3.4 x 1021 cm3 hydroxyl groups. Resist was spincoated in conventional manner (layer thickness 1.8um at 5000 rpm, soft bake for 50 s by 110°C on hot plate). For exposure with the g-line wafer stepper AUR (Carl-Zeiss Jena, numerical aperture NA = 0.3) a radiation dose of around 80 mJ/cm2 is necessary. This dose is around 50%lower than those necessary for wet development. For process application in device manufacturing and process optimization the exposed resist is baked and silylated on a hot plate in a prime module of a commercial wafer track. Experiments are performed on an experimental equipment "Syll". In this equipment it is possible, to do experiments with higher concentration of the HMDS by heating the bubbler in a thermostat and with gas mixtures (e.g. HMDS, NH3 and other). The progress of silylation is controlled by reflection measurements using a He-...

No abstract