System and method for deconvoluting the effect of topography on scanning probe microscopy measurements

A method for creating an image reflecting a physical property of a surface of a sample from data obtained using a physical property scanning probe microscope, comprising:

• (a) training a neural network to separate topographic effects from physical property effects on said data;
• (b) scanning the surface of the sample using a physical property scanning probe microscope to obtain a topographic data set and an apparent physical property data set, the values of which reflect both physical property effects and topographic effects;
• (c) applying the neural network to separate the topographic effects from the physical effects on the values of said apparent physical property data set;
• (d) generating an image reflecting variations in the physical property of the surface of the sample.

An apparatus for obtaining images reflecting a physical property of a surface of a sample comprising:

• (a) a physical property scanning probe microscope;
• and
• (b) a neural network,
• wherein the neural network has been trained to separate topographic effects from physical property effects;
• and wherein the neural network is used to separate the topographic effects from the physical effects to generate an image reflecting variations in the physical property of the surface of the sample.

A thermal scanning probe microscope comprising:

• (a) a cantilever arm;
• (b) a thermal probe extending from said cantilever arm;
• (c) means for scanning the thermal probe across the surface of a sample, and for recording thermal and topographic data as a function of position on the sample to obtain a data set reflecting thermal and topographic properties of the surface of the sample;
• (d) a neural network for separating topographic effects from physical effects in the data set;
• and
• (e) means for displaying an image reflecting the thermal properties of the sample.

A method for using a neural network to deconvolute the effect of topography from a thermal conductivity measurement made by a thermal scanning probe microscope comprising the steps of:

• in a training phase of using the neural network:
• (a) obtaining a sample of a material having a smooth surface;
• (b) determining a true thermal conductivity of the sample;
• (c) obtaining another sample of the material, the another sample having a rough surface;
• (d) selecting a point on the rough surface;
• (e) determining a local topography at the selected point;
• (f) measuring a thermal conductivity at the selected point using said thermal scanning probe microscope;
• and
• (g) storing the selected point, the local topography, the true thermal conductivity of the material and the measured thermal conductivity at the selected point in a table;
• (h) training the neural network using the table;
• and in the measurement phase of using the neural network:
• (i) obtaining a sample to be tested;
• (j) selecting a point on the surface of the sample to be tested;
• (k) determining a local topography around the point selected on the sample to be tested;
• (l) measuring a thermal conductivity at the point selected on the sample to be tested using a thermal scanning probe microscope;
• and
• (m) applying the local topography around the point selected on the sample to be tested and the measured thermal conductivity at the point selected on the sample to be tested to the neural network to deconvolute the effects of topography from thermal conductivity.

A method for deconvoluting the effects of topography from a thermal conductivity measurement made by a thermal scanning probe microscope, comprising the steps of:

• (a) selecting a point on the surface of a sample of a material for which the true thermal conductivity is known;
• (b) determining a local topography around the selected point;
• (c) measuring a thermal conductivity at the selected point using a thermal scanning probe microscope;
• (d) using the local topography, the measured thermal conductivity and the true thermal conductivity to determine the effect of topography on the measured thermal conductivity;
• (e) storing the determination of the effect of topography on the measured thermal conductivity;
• (f) selecting a point on the surface of a sample of a test material;
• (g) determining a local topography around the selected point on the surface of the sample of the test material;
• (h) measuring a thermal conductivity at the selected point on the surface of the sample of the test material using a thermal scanning probe microscope;
• and
• (i) deconvoluting the effect of topography on the measured thermal conductivity at the selected point on the surface of the sample of the test material using the stored determination of the effect of topography on the measured thermal conductivity, the determined local topography around the selected point on the surface of the sample of the test material and the measured thermal conductivity at the selected point on the surface of the sample of the test material.

A system for deconvoluting the effects of topography from a thermal conductivity measurement, comprising the steps of:

• means for selecting a point on the surface of a sample of a material having a true thermal conductivity;
• means for determining a local topography around the selected point;
• means for measuring a thermal conductivity at the selected point using a thermal scanning probe microscope;
• means for using the local topography, the measured thermal conductivity and the true thermal conductivity to determine the effect of topography on the measured thermal conductivity;
• means for storing the determination of the effect of topography on the measured thermal conductivity;
• means for selecting a point on the surface of a sample of a test material;
• means for determining a local topography around the selected point on the surface of the sample of the test material;
• means for measuring a thermal conductivity at the selected point on the surface of the sample of the test material using a thermal scanning probe microscope;
• and
• means for deconvoluting the effect of topography on the measured thermal conductivity at the selected point on the surface of the sample of the test material using the stored determination of the effect of topography on the measured thermal conductivity, the determined local topography around the selected point on the surface of the sample of the test material and the measured thermal conductivity at the selected point on the surface of the sample of the test material.