Modeling and boundary force control of microcantilevers utilized in atomic force microscopy for cellular imaging and characterization.

摘要

This dissertation undertakes the theoretical and experimental developments microcantilevers utilized in Atomic Force Microscopy (AFM) with applications to cellular imaging and characterization. The capability of revealing the inhomogeneties or interior of ultra-small materials has been of most interest to many researchers. However, the fundamental concept of signal and image formation remains unexplored and not fully understood. For his, a semi-empirical nonlinear force model is proposed to show that virtual frequency generation, regarded as the simplest synthesized subsurface probe, occurs optimally when the force is tuned to the van der Waals form. This is the first-time observation of a novel theoretical dynamic multi-frequency force microscopy that has not been already reported.;Owing to the broad applications of microcantilevers in the nanoscale imaging and microscopic techniques, there is an essential feeling to study and propose a comprehensive model of such systems. Therefore, in the theoretical part of this dissertation, a distributed-parameters representation modeling of the microcantilever along with a general interaction force comprising of two attractive and repulsive components with general amplitude and power terms is studied. This model is investigated in a general 2D Cartesian coordinate to consider the motions of the probe with a tip mass. There is an excitation at the microcantilever's base such that the end of the beam is subject to the proposed general force. These forces are very sensitive to the amplitude and power terms of these parts; on the other hand, atomic intermolecular force is a function of the distance such that this distance itself is also a function of the interaction force that will result in a nonlinear implicit equation. From a parametric study in the probe-sample excitation, it is shown that the predicted behavior of the generated difference-frequency oscillation amplitude agrees well with experimental measurements.;Following the proposed Euler-Bernoulli model, a more comprehensive model is developed by modeling the probe dynamics and including the effects of the rotary inertia and shear deformation under the same proposed tip-sample interaction force. An extensive comparative study between the Euler-Bernoulli and Timoshenko beam assumptions is conducted for different conditions including different base-excitation amplitudes and higher modes. The results underline that the comprehensive Timoshenko model unveils the effects of the nonlinear interaction force better than the Euler-Bernoulli beam model.;In addition to extensive modeling efforts on the microcantilever and its interaction with sample, an adaptive control framework is developed in order to make the microcantilever's tip follow a desired trajectory. This trajectory can further be considered as an important path acquired by the path planning techniques to manipulate the nanoparticles. There is a base excitation considered for this model and can be considered as an input force control to excite the probe by taking advantage of flexibility of the cantilever despite its complexity and under existence of the external nonlinear interaction forces between the tip and sample's surface.;When building such complicated controller on top of the proposed comprehensive model, the results could be extended to study a macro-micro hybrid rigid-flexible model of a microrobot to mimic the realistic behavior of the MM3ARTM microrobot. The MM3ARTM microrobot is equipped with a piezoresistive layer which functions as a force sensor and is capable of measuring very slight forces as small as micro to nano-Newton. Two types of controllers are investigated for the case of the tip force control. Lyapunov-based PD and robust adaptive controllers are developed for this purpose and their performances and stabilities are compared.;In the experimental part, a platform for performing the automated nanomanipulation and real-time cellular imaging is developed by integrating a microrobot, digital signal processor platform (dSPACERTM), computer, and a state-of-the-art light microscope. The closed-loop boundary force control framework is additionally developed for the autonomous in-situ applications. Since the incoming and outgoing signals of the piezoresistive microrobot are in the form of the electrical voltage and the string commands (ASCII code), respectively, an intuitive programming code for interfacing the MATLAB and dSPACE RTM has been written for the online quasi-data acquisition. As a result, the height of the corneal cell has been obtained and additionally, the microcantilever's tip force has been automatically controlled by taking advantage of the proposed control framework.