Biomedical Engineering Reference
In-Depth Information
Fig. 3.21. Schematics of various implementations of MFM. A: lifting probe between topography
and MFM images. B: Bard method of lifting lever between scan lines. C:
z
set-point oscillation.
D: Hosaka method of moving probe close to surface, and recording MFM signal at various points for
each height.
(Figure 3.21A). This works well for flat samples, but is prone to problems of features from
the sample topography appearing in the MFM image, and also to problems from thermal
drift. As described by Bard [214], an improved method is to record the sample topography
first, then lift the probe, and measure the long-range forces while following the shape of
the topography, but at a certain 'lift height'. This is applicable to STM, EFM (see the
following section), or MFM. Typically, this is carried out in alternate scan lines, allowing
the topography data to be included in the second, magnetic scan line, meaning the probe
can stay approximately the same distance above the sample, even with changes in
topography (Figure 3.21B) [215]. It's also possible to change the
z
set-point while
scanning, meaning the probe will be constantly moving towards the sample to check the
topography, and then away again to register magnetic field information (Figure 3.21C).
Finally, in the method described by Hosaka [216], at each pixel the probe is lifted above
the surface, and the field is measured at several points as the probe is lowered again
(Figure 3.21D), to obtain a magnetic field gradient. The probe is then moved to the next
lateral point, lifted again, and so on. This method is probably the least prone to thermal
drift, but is rather slow to implement. Whichever method is used, lifting the tip from the
surface reduces resolution, and resolution in MFM is typically no greater than 30 nm
laterally [212].
For these lifting modes to work, it helps if there is little sample drift, or to have
linearized scanners. Typically, MFM is carried out in one of the dynamic modes, and
the magnetic effects on the cantilever are detected via phase shift, but they may also affect
the oscillation amplitude. Unfortunately, even at lift heights of several tens of nanometres
from the sample surface, short range forces other than magnetic interaction may affect the
cantilever oscillation, giving a false indication of magnetic contrast [217], an effect which
