The primary research objective of this proposal is to improve the temporal resolution of atomic force microscopy (AFM) through non-raster sampling schemes based on compressed sensing (CS). While AFM continues to be used heavily for the study of systems with nanometer-scale features, its temporal resolution limits its applicability to the study of dynamics. The research approach progresses from non-raster sampling of a single image, including robust time-optimal control techniques to move the tip of the microscope as rapidly as possible between measurement locations, to CS driven schemes for acquisition of image sequences. The methods developed will be implemented and tested on AFMs to demonstrate their capabilities.
If successful, the results of this research will extend the utility of AFMs by increasing the imaging rate while also decreasing the interaction with the sample, limiting any damage caused by the imaging process. While focused on AFMs, the techniques to be developed will be directly applicable to other scanning probe methodologies, such as scanning tunneling microscopy and near-field scanning optical microscopy, as well as more broadly to scenarios in which a short-range sensor is acquiring information in a large area. Examples of such scenarios include environmental monitoring by autonomous robots, large-scale data collection in ocean environments, and weather sampling. Further, the robust time-optimal control results will be broadly applicable in industries that include disk drives, tape drives, wafer scanning systems, electronic manufacturing, and more. Graduate and undergraduate students will benefit through participation in the research while outreach activities will engage middle and high school students in the Boston metro area, with a focus on students from low-income families, and in the Boulder metro area, with a focus on female students.
