AFM operating-drift detection and analyses based on automated sequential image processing

Nanomanipulation and nanoimaging with Atomic Force Microscopes (AFM) is a popular technique for nano-manufacturing. However, under ambient conditions without stringent environment control, nanomanipulation tasks are difficult to complete because various system drift can cause uncertainties of the spatial relationship between the AFM probe tip and the nano-entities to be manipulated. Researchers have speculated that thermal drift is one of the major causes of errors for nanomanipulation using AFM systems, but to this date, quantitative analyses of AFM drift phenomenon are almost non-existent. This paper gives a detection and analyses method for AFM operating-drift based on automated sequential image processing, which provides a quantitative understanding of the AFM drift phenomenon. Essentially, the drift of an AFM system can be measured by a Phase-Correlation Method among consecutively scanned images. In order to eliminate the effects of z-direction drifts in x, y displacements, a gradient calculation method is introduced. The influence of a PZT actuator's thermal expansion on overall system drift is also analyzed. The results showed that although the length of the PZT actuator's expansion is the greatest among all the main system components, it may not be the main cause of the overall system drift.