WetCleanSIM - Computational modeling for wet cleaning inside semiconductor nanostructures

Wet spin processing has become an important technique for producing the highly complex small scale surface structures on silicon substrates (wafers) in semiconductor industry. The working liquid is generally supplied here through a vertical jet and is further spread as thin film across the rotating wafer driven by the centrifugal forces. The ongoing miniaturization of the surface structures, high output targets, and low consumption of chemicals, makes it increasingly difficult to ensure the required high quality of the process. For mastering this challenge, one needs a reliable comprehensive description of the relevant multi-scale transport phenomena, ranging from the macro- and microscopic liquid film flow scales on the rotating substrate down to smallest molecular scales inside the nanometer-size channel structures (vias), which are fabricated by the etching process. The present PhD-project shall develop a computational model for predicting the convective/diffusive transport of reactive species on their way in the liquid carrier film to the entrance of the nanoscopic vias. Already existing numerical as well as semi-analytical models shall be particularly extended to couple realistically the macroscopic on-wafer transport of mass, momentum, and heat in the rotating thin film with nanoscopic molecular diffusive transfer mechanisms inside the via including chemical reactions. The computational investigations shall cover the basically axisymmetric case of central liquid jet dispense as well as the significantly more complex case with off-center liquid jet dispense, which typically produces a bow wave and eventually a de-wetted central region. Measured film thicknesses and chemical abrasion rates of surface material are used for model validation.