A method for the linearity improvement of the topology measurement of the atomic force microscopy (AFM) using 2D wavelet transform

Abstract

An Atomic Force Microscopy (AFM) is a powerful measurement instrument for measuring the mechanical, electrical, and magnetic properties in nano scale areas. Recently, the demand for Atomic Force Microscopy is increasing in industrial applications such as roughness measurements for display glass used in cellular phones and height measurements for LED patterned sapphire substrates (PSS) having large height variations that are approximately few micrometers.

The topology image of the Atomic Force Microscope is obtained by the controller output from a force control system that is proportional to the height of a measurement sample with under the assumption that no dynamics exist in a PZT driven Z nano scanner. Nonlinear characteristics of the PZT driven Z nano scanner such as the hysteresis and the creep effect can be ignored in small height variation samples. However, these nonlinear characteristics can’t be ignored when a sample has large height variation because this would cause inaccuracies in the measurement image. When large height variation sample is measured, the low control bandwidth and vibration problem are more appeared. compare to when the Atomic Force Microscope scan low height variation samples. higher bandwidth and a vibration reduction technique of the Atomic Force Microscope is required for large height variation sample which having high aspect ratio.

In order to avoid such inaccuracies, an additional displacement sensor, strain gauge sensor is used to directly measure the displacement of the PZT driven Z nano scanner. However, this approach also has a disadvantage in its relatively low precision. In order to obtain high precision data with good linearity, I propose an accuracy improvement method for the topology measurement of an Atomic Force Microscope using a 2D wavelet transform for large height variation samples. The advantages of the proposed method are experimentally validated by using topology images.