Text

theses

This thesis presents the development of ultra-high-speed large-range dynamic mode imaging of atomic force microscope (AFM), through the extension of the adaptive multi-loop mode imaging technique to the eld-programmable gate array (FPGA) signal acquisition and processing hardware platform. High speed imaging is needed in atomic force microscope imaging to observe constantly changing processes such as chemical reaction on the sample surface, microorganism activity or macromolecules reactions. However, conventional imaging modes are too slow to observe such processes. Contact mode is faster than tapping mode but can damage the sample while tapping mode has less distortion but much slower. Such a speed limiation can be largely alleviated by the adaptive multi-loop mode (AMLM) technique which utilizes an online iterative feedforward controller to overcome the time delay of the z-feedback loop in tracking the topography. The goal of this thesis is to integrate the AMLM technique with the FPGA platform to further image the imaging speed by an order of mangitude without loss of quality and imaging range. Challenges in FPGA programming must be overcome to account for both the AFM system dynamics characteristics and the FPGA hardware speci cations. For instance, the (inverse) Fourier transform in the feedforward controller is replaced with time domain signals along with introducing the tapping amplitude error into the feedforward control and the tapping mode deflection decoupling technique. Moreover, a modeling-free inversion-based iterative feedforward control (MIIFC) approach is implemented for the x-axis piezo actuator control to track the high frequency triangular wave needed for high speed imaging. In the experiment a calibration reference sample is scanned to validate the control scheme and present the imaging results by comparing with the much slower conventional tapping image. The experimental results show that by using the AMLM imaging technique on FPGA board, the imaging speed increases from 5 Hz to 100 Hz while maintaining good imaging quality. However, although the main features of the sample topography are captured, many details are loss due to the time delay in the control system and high frequency noise.

ETD_8850

M.S.

Includes bibliographical references

by Jiarong Chen

doi:10.7282/T3JQ14F5

ETD graduate

The author owns the copyright to this work.