Biomedical Engineering Reference
In-Depth Information
When scanning the sample begins, it is useful to see a line-scan (a two-dimensional plot
of the signal the instrument is recording). Often, the height data, as well as the
z
-error
signal (in contact mode, the cantilever deflection) can be shown, and sometimes both
forward and reverse signals are shown. The function of the AFM software that displays
these signals is similar to an oscilloscope, so it is sometimes referred to as the oscilloscope
window. This can be extremely useful for optimization of scanning. As forwards and
backwards scanning lines measure (almost) the same parts of the sample, even when the
slow scan axis is enabled, the two height traces should coincide. Large differences
between forwards and backwards traces are an immediate indication that something is
not right with the scanning. There are a number of possible reasons for forwards and
backwards traces not matching but the most common reason is that imaging parameters
(gains, set-point, and scanning speed) are not yet optimized. An example of the signals
shown by the oscilloscope window is shown in Figure 4.6, illustrating the effect of
different feedback settings on the results obtained on a simple sample. For clarity, only
results from one direction are shown.
If the AFM probe is scanning over the sample in the normal raster motion (as shown
in Figure 2.22), the features in the oscilloscope window will of course keep changing.
It can be extremely helpful to have the probe scan in a line over the sample, without
moving in the slow scan axis. Usually, the software will have an option to do this, and
it is often the best way to adjust the scanning parameters as their effect on the scanning
can be seen directly without interference from changes in sample topography. It is
highly recommended that the line-scan option is used if difficulty arises in setting the
gains, etc. After optimization, then the slow scan axis movement can be re-enabled and
the image quality checked. When first learning to operate an AFM, it is helpful to scan
a test sample and see the effect of the feedback parameters on the height (
z
voltage)
and deflection signals. Such a sample has a very simple, reproducible topography
(usually a series of square pits or posts), so it is easy to see when the scanning
parameters are perfect. An image of such a sample, with the effect of varying the
feedback parameters is shown in Figure 4.7. Some useful test structures are discussed
in Appendix A.
The general procedure to use to adjust scanning parameters is as follows.
1. Increase feedback gains (PID values) step by step, observing for the start of feedback
oscillation (the fine-structured noise seen towards the bottom of Figure 4.7). Typic-
ally, the integral gain is increased first, and then the proportional gain adjusted in
approximately the same proportion.
2. When feedback oscillation occurs, reduce the gains again, until it disappears. The
optimal value is the highest gain setting you can use without adding feedback noise
to the image.
3. When gains are optimized, adjust the set-point. Ideally we would use the minimum
value to keep the probe on the surface, in order to reduce probe wear. However,
sometimes a greater force is required.
4. The gains may need to be optimized again to account for change in set-point.
5. Adjust scan speed if desired.
6. Gains and set-point may need adjusting once more to take account of change in
scanning speed.
