Lateral metrology using scanning probe microscopes, 2D pitch standards and image processing

To obtain high accuracy we have improved a previously reported algorithm for calculation of the unit cell by sub-pixel Fourier analysis. We characterize the linearity by a new method where the positions of individual features associated with the unit cell are detected by a correlation technique and compared with predicted positions based on the calculated unit cell. The differences between the actual and predicted locations reflect the non-linearity. We are able to measure the non-linearity at sub-pixel level, i.e. smaller than 0.2% of the scan range for a 512×512 pixel image. Evaluation of the linearity has given us a tool that can confirm the unit cell measurements based on Fourier analysis to within 0.3% of the repeat distances. When analyzing a series of images acquired over a time frame of 6 months, we find a reproducibility of better than 0.7% for the calibration parameters. We demonstrate that we are able to measure angular distortions smaller than 0.2° and that corrections of small residual non-linearities have a positive influence on the accuracy of the pitch measurements. The developed software is made freely available.