Quantitatively reinterpreting atomic force microscopy via the data science paradigm

Abstract

From the invention of the atomic force microscope (AFM) in 1986, tremendous efforts have been put into developing this tool. The AFM has long been considered as one of the top choices to probe the nanoscale world with the ability to achieve nanoscale resolution imaging of surfaces under different environments. Advances in instrumentation combined with the exploitation of sophisticated data analysis methodologies are set to meet the demand for higher resolution forms of microscopy that allow for direct visualization, identification, nanometric or atomic defects and structure, and material phases. This combination is not arbitrary but responds to the necessity of employing algorithms to decouple and interpret the complex signals and contrast channels that result from both standard instrumentation and the extra complexity added by the instrumentation designed to increase throughput and enhance resolution and quantification. Starting with interpreting AFM data using single mode force spectroscopy method to explicating multiple channels acquired with advanced multifrequency methods, it has reached a point that resorting to big data approaches might provide broader understanding toward surface properties in the material science community. Finally, this thesis shows that it is possible to submit the data capturing complex physical phenomena like the tip-surface interaction in AFM to a specific question and obtain the answer regardless of the complexity and/or unknown factors of the phenomena.