Reconstruction of a Real Topographic Image distorted by Nonlinearities of a Vertical Scanner in Atomic Force Microscopy

Abstract

Atomic force microscope has become an important instrument for research and for inspection in many fields of industrial nano-fabrication. Especially, the industrial atomic force microscope requires the automation and the high accuracy in order to improve the efficiency of inspection and the reliability of results. In this dissertation we propose a new approach and post processing methods for the automation of atomic force microscope. First, we present a criterion for an automatic approach method in conventional atomic force microscope. The criterion can be used for decision making in determining mode transitions from fast motion to slow motion. It can be also used for finding the safe distance which is a boundary between the fast and slow motions. The criterion is calculated based on the average intensity of a Gaussian laser beam. The average intensity is the mean value of pixels inside a window which is a region of a visual image. The mode transition is made when the tip reaches a certain offset from the place where the average intensity reaches its maximum. We can remove or minimize the human intervention in the approach operation of atomic force microscope. Second, we develop new post processing methods for automation and improvement of the accuracy of scanned image: reconstruction and flattening methods. In the automatic scanning process, which requires fast scanning and high repeatability, the atomic force microscope must scan the sample surface immediately after completion of an approach operation. The scanned image may be badly distorted due to the rapid change that appears in the early stages of the creep effect of the vertical nano-scanner. We analyze the distorted scanned image caused by the creep of the vertical nano-scanner in the atomic force microscope operated in constant-force mode. The creep effect appears an illusory slope in the slow scanning direction. We present a method to measure the tilt angle of a sample and the creep factor of the vertical nano-scanner using two scanned images with no special tools. We can reconstruct a real topographic image based on the two scanned images, in which both the creep effect of the vertical nano-scanner and the slope effect of the sample have been eliminated. Additionally, we also present an advanced flattening method to flatten the scanned image that does not require the user be experienced. Under the assumption that the background of the sample surface is flat, the advanced flattening method is able to substantially reduce the artifacts caused by a conventional flattening method and the inspection time using a single scanned image. Furthermore, there is no need for the directional consideration of the flattening process or numerical creep model of the vertical nano-scanner. The above mentioned proposed methods can minimize or remove the human intervention from the atomic force microscope operation. Consequently, the proposed methods can be applied to the automation of the atomic force microscope operation. The effectiveness of the proposed improvements was confirmed through simulation and experiment results.