Energy dissipation measurements in frequency-modulated scanning probe microscopy

Local dissipation measurements by scanning probe microscopy have attracted increasing interest as a method for probing energy losses and hysteretic phenomena due to magnetic, electrical, and structural transformations at the tip-surface junction. One challenge of this technique is the lack of a standard for ensuring quantification of the dissipation signal. In the following, we explored magnetic dissipation imaging of an yttrium-iron garnet (YIG) sample, using a number of similar but not identical cantilever probes. Typical frequency-dependent dispersion of the actuator-probe assembly commonly approached +/- 1 part in 10(3) Hz(-1), much larger than the minimum detectable level of +/- 1 part in 10(5) Hz(-1). This cantilever-dependent behavior results in a strong crosstalk between the conservative (frequency) and dissipative channels. This crosstalk was very apparent in the YIG dissipation images and in fact should be an inherent feature of single-frequency heterodyne detection schemes. It may also be a common effect in other dissipation imaging, even down to the atomic level, and in particular may be a significant issue when there are correlations between the conservative and dissipative components. On the other hand, we present a simple method for correcting for this effect. This correction technique resulted in self-consistent results for the YIG dissipation measurements and would presumably be effective for other systems as well.