Modeling forces between the probe of atomic microscope and the scanning surface

Atomic force microscope (AFM) is usually used to study the properties and surface structure of nanoscale materials. AFMs have three major abilities: force measurement, imaging, and manipulation. In the force measurement, AFM can be used to measure the forces between the probe and the sample as a function of their mutual separation. AFM compared to scanning electron microscope has a single image scan size; also the scanning speed of AFM is also a limitation. AFM images can also be affected by nonlinearity, hysteresis, creep of the piezoelectric material, and cross talk between the x, y, and z axes that may require software enhancement and filtering. Due to the nature of AFM probes, they cannot normally measure steep walls or overhangs in surface. In this study, the force between the Probe of Atomic Microscope and the surface is simulated by using force measurement ability of AFM and artificial neural network. The experimental data are used for training of artificial neural networks. The best model was found to be a feed-forward backpropagation network, with Logsig, Tansig and Tansig transfer functions in successive layers, respectively, and 3 and 2 neurons in the first and second hidden layers. According to the results, the proposed neural network is well capable of modeling the behavior of AFM probes in noncontact mode.