© ICSPI Corp. 2007-2021
nGauge AFM User Manual 2.0
74
9.1.4.1.
Calibration Grating Selection
At a minimum, the calibration gratings should span the expected data range. In other words:
One calibration grating with a height lower than the lowest expected data result, and one
calibration grating with a height higher than the highest expected data result should be used. If
possible, other gratings which are within the expected data range should be used to increase
accuracy.
Gratings are available in standard sizes: 20 nm, 110 nm, 520 nm and 1500 nm. As an example,
if you are scanning 50-nm tall features for example, you should collect calibration data at 20 nm
and at 110 nm and use these measurements to correct your 50-nm tall feature heights.
9.1.4.2.
Create a Linear or Polynomial Correction Function
Follow the same post-processing steps as with single-point calibration: Level, align rows, then
level again. Measure the step height with your preferred method.
Create a table of the measured step heights and the known step heights and add in (0,0):
Measured step height (nm)
Known step height (nm)
0
0
26
20
126
110
580
520
Fit a linear or second-order polynomial to the data. This can be done in Excel, Matlab, Python,
or other scientific computing tools. Since we are applying the function to “measured step height”
data, the measured step height is the independent variable (x) and the known step height is the
dependent variable (y).
An example of a linear fit to the data above in Excel:
