Date of Award

1999

Document Type

Doctoral Thesis

Degree Name

Doctor of Philosophy

Department

Applied Physics and Instrumentation

First Advisor

Dr. Liam McDonnell

Abstract

The overall aim of this work was to develop computer based methodologies for the on-line scanning force microscopic detection of immunological binding events. To achieve this aim it was necessary to develop a rapid, high-resolution data acquisition system to digitise scanning force microscope (SFM) data on-line and then to develop an effective method for extracting the scanner induced background curvature from this data to reveal the nanoscale height binding events. Four techniques were investigated in terms of operational speed and performance accuracy; namely, polynomial interpolation; least squares approximation; real-time recursive least squares approximation; and wavelet approximation.

It was found that the method of least squares curve fitting will always provide an accurate determination of the scanner background provided that the polynomial is estimated using background data points only. The predictor-corrector type recursive least squares approximation method provides the fastest possible means of removing the scanner background, however the quality of fit is subject to the indeterminate nature of the occurrence of binding events.

A wavelet approximation method used with least squares curve fitting was developed to provide optimal fitting of the scanner background. The computational overhead of this method was compensated by fitting the polynomial to a lower resolution version of the scanner background. The use of wavelets is novel to SFM and the use of wavelets as a new visualisation tool in SPM is described. An extension of the technique to separate scanner curvature, antibody background and antigen foreground is also presented.

Finally a reliable method for the detection individual binding events of 3 nm in height was developed. The ability of the system to detect single ferritin molecules over 100 µm scan ranges for concentrations as low as 125 pg/ml has been demonstrated. In its present form the prototype appears to have a detection limit of 30 pg/ml for the ferritin assay.

The technology underpinning the new scanning force microscopic immunoassay (SFMIA) prototype has application within for example the semiconductor industry where there is a need for rapid methods to determine the sizes and locations of particulate contamination. The methods developed to remove scanner background from the SFM images have general application within SPM and could be deployed by instrument manufacturers to enhance the performance of their systems as well as by users seeking to optimise SFM images and the metrology data contained therein.

Access Level

info:eu-repo/semantics/openAccess

Included in

Physics Commons

Share

COinS