Stochastic lattice model for atomic layer deposition and area-selective deposition of metal oxides: Visualization and analysis of lateral overgrowth during area-selective deposition

Abstract

Although area-selective deposition (ASD) has developed to augment lithographic patterning of nanoscale device features, computational modeling of ASD remains limited. As pitch sizes shrink, the extent of lateral overgrowth at the feature edge becomes critical to ASD processing. We report a stochastic lattice model that describes atomic layer deposition (ALD) and ASD of Al2O3 using trimethylaluminum and water as an example system. The reactant/surface interactions are constrained such that the resulting ALD film properties, i.e., Al/O atom ratio, fraction of unreacted (blocked) –OH groups, fraction of “void” sites, and growth per cycle, are reasonably consistent with the experimental results for Al2O3 ALD. In the ASD model, the film nucleates in a localized “growth” region and extends laterally over a co-planar adjacent “nongrowth” region, thereby simulating lateral growth evolution. Under the “base ALD” condition, the extent of lateral growth is equivalent to vertical growth, and the contact angle between the film and the substrate is 90°. Introducing additional constraints on reactant/nongrowth surface interactions leads to changes in the extent, shape, and contact angle of the lateral growth, enabling insight into chemical and physical mechanisms that influence the shape and extent of lateral overgrowth. The 3D model visualizations are directly compared with example ASD results, demonstrating consistency between the model output and experiments. Comparing the mechanisms introduced to the model with the experimental ASD processes and conditions provides insight into the mechanisms that drive film shape evolution and lateral overgrowth, enhancing understanding of means to control lateral profile evolution during ASD.