Predicting the failure of ultrathin porous membranes in bulge tests

Silicon nanomembranes are thin nanoporous films that are frequently used as separation tools for nanoparticles and biological materials. In such applications, increased differential pressure across the nanomembranes directly increases process throughput. Therefore, a predictive tool governing the macroscale failure of the porous thin films is fundamentally important in application areas where high differential pressures are desired. Although the deflections and stresses of the nanomembranes can be reliably predicted, a straightforward and prognostic failure model has yet to be outlined. In this publication, a brittle macroscale failure model is established and validated with experimental results. Theoretical agreement with experiments within 10% accuracy offers reliable failure predictions for square membrane dimensions from 60 μm to 1.5 mm through over 100 trials. The methodology relies on an effective fracture toughness from previously published work that is incorporated through Griffith's law. These developments will be useful in the selection of nanomembranes for particular applications and will help guide the design of materials with improved strength. The model should also prove useful for high-volume, in-line processing and inspection of nanomembranes as their role becomes more prominent in industry.