Abstract

Scanning electron microscopes (SEM) are poorly characterized imaging instruments. The SEM optics (lens and aperture assembly) is designed to form a focused electron probe that scans the specimen to produce images. Therefore, the imaging efficacy in SEM strongly depends on the quality of the optics. However, there is no way to accurately characterize the optics in an uncorrected SEM (without multipole correctors) due to the lack of an exit wave. This work focuses on improving the imaging capabilities of an uncorrected SEM by understanding its optical performance using wave optical theory, simulation, and experiment. In this dissertation, we have developed two different simulations based on wave optical treatment of the electron beam to model the optical column and probe formation. Using one of the simulations for data generation, we have developed an aberration diagnostic method for the uncorrected SEM based on deep learning. Further, we have developed an experimental technique to perform non-interferometric phase retrieval of the electron probe. We have used the recovered phase information to visualize the point-spread function of the SEM optics (PSF_optics) for the first time. Finally, we have proposed an experiment based on electron vortex beams to improve phase retrieval. This work lays out the initial steps required to move towards "aberration-free" imaging in SEM without the use of multipole correctors.

Publication Date

4-24-2024

Document Type

Dissertation

Student Type

Graduate

Degree Name

Imaging Science (Ph.D.)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science

College

College of Science

Advisor

Richard Hailstone

Advisor/Committee Member

Nathan Cahill

Advisor/Committee Member

Roger Easton Jr.

Campus

RIT – Main Campus

Download

Share

COinS