TY - JOUR

T1 - Modeling the hysteresis of a scanning probe microscope

A1 - Dirscherl,Kai

A1 - Garnæs,Jørgen

A1 - Nielsen,L.

A1 - Jørgensen,J. F.

A1 - Sørensen,Mads Peter

AU - Dirscherl,Kai

AU - Garnæs,Jørgen

AU - Nielsen,L.

AU - Jørgensen,J. F.

AU - Sørensen,Mads Peter

PB - American Institute of Physics

PY - 2000

Y1 - 2000

N2 - Most scanning probe microscopes use piezoelectric actuators in open loop configurations. Therefore a major problem related to these instruments is the image distortion due to the hysteresis effect of the piezo. In order to eliminate the distortions, cost effective software control based on a model for hysteresis can be applied to the scanner. We describe a new rate-independent model for the hysteresis of a piezo scanner. Two reference standards were used to determine the accuracy of the model; a one-dimensional grating with a period of 3.0 mum and a two-dimensional grating with 200 nm pitch. The structures were scanned for different scan ranges varying from 5 V peak to peak to 440 V peak to peak, so that 99% of the scanners' full motion range was covered. A least-squares fit of the experiments to the hysteresis model provided standard deviations per scan range of around 0.2%. This represents an uncertainty of 1 pixel. Since our model is based on a differential equation, it is flexible even to simulate arbitrary experimental conditions such as a sudden change in the offset

AB - Most scanning probe microscopes use piezoelectric actuators in open loop configurations. Therefore a major problem related to these instruments is the image distortion due to the hysteresis effect of the piezo. In order to eliminate the distortions, cost effective software control based on a model for hysteresis can be applied to the scanner. We describe a new rate-independent model for the hysteresis of a piezo scanner. Two reference standards were used to determine the accuracy of the model; a one-dimensional grating with a period of 3.0 mum and a two-dimensional grating with 200 nm pitch. The structures were scanned for different scan ranges varying from 5 V peak to peak to 440 V peak to peak, so that 99% of the scanners' full motion range was covered. A least-squares fit of the experiments to the hysteresis model provided standard deviations per scan range of around 0.2%. This represents an uncertainty of 1 pixel. Since our model is based on a differential equation, it is flexible even to simulate arbitrary experimental conditions such as a sudden change in the offset

U2 - 10.1116/1.591249

DO - 10.1116/1.591249

JO - Journal of Vacuum Science and Technology. Part B. Microelectronics and Nanometer Structures

JF - Journal of Vacuum Science and Technology. Part B. Microelectronics and Nanometer Structures

SN - 1071-1023

IS - 2

VL - 18

SP - 621

EP - 625

ER -