Date of Award




Document Type


Degree Name

Doctor of Philosophy (PhD)


Department of Nanoscale Science and Engineering


Nanoscale Sciences

Content Description

1 online resources (xvi, 120 pages) : illustrations (some color)

Dissertation/Thesis Chair

Bradley L Thiel

Committee Members

Serge R Oktyabrsky, John G Hartley, András E Vladár, Milos Toth


cd metrology, low vacuum SEM, noise, optimization, signal, Scanning electron microscopy, Vacuum technology

Subject Categories

Nanoscience and Nanotechnology


Low vacuum scanning electron microscopy (LVSEM) is proposed and evaluated for next generation Critical Dimension (CD) metrology. Its ability to control charging artifacts and hydrocarbon contamination in order to obtain high signal-to-noise ratio, high resolution image data from insulating materials make the technology an excellent match for the increased use of high-k dielectrics and shrinking feature size in the semiconductor industry. The presence of a gas in the LVSEM chamber means that the probe characteristics and secondary electron amplification, detection, and signal-to-noise ratio differ significantly from conventional high vacuum tools. In order for low vacuum CD approaches to be viable, all of the processes must be understood and described to the degree of accuracy currently available on high vacuum systems. Consequently, the focus of this thesis is to determine an analytic form of the signal-to-noise ratio for two detector configurations: the simplified steady-state cascade system operating in the well defined Townsend's discharge region, and the high resolution, low vacuum immersion lens secondary electron detector, for which the physical amplification process has not been studied in the past. A physically realistic and ultimately predictive model, which could potentially be incorporated in CD simulation codes such as NIST's MONSEL, is developed. Its effectiveness is verified with experimental data acquired as a function of gas pressure for all important operating parameters, such as electron beam energy and current, detector bias, cascade distance, and gas type, and its capability for optimization of the imaging conditions is discussed. Noise characteristics are also analyzed using Monte Carlo gain histograms and pure statistical methods.