Manipulation of ellipsoidal nanoparticles considering roughness based on atomic force microscopy

The contribution in this paper is to investigate the manipulation of ellipsoidal nanoparticles by atomic force microscopy, taking into consideration of roughness. For the first phase of manipulation, roughness was investigated just for the substrate, but for the second phase, it begins by particle movement substrate and particle roughness is considered. The particle is of gold material; moreover, tip and substrate are made of silicon. Having examined the effective parameters of the contact mechanics, including the indentation depth and contact area for the Hertz, Jamari and Jeng–Wang models, two Jamari and Jeng–Wang contact models were used to consider the particle–probe contact and the particle–substrate contact respectively. By these two models, the first phase of manipulation process is simulated. At the end of this step, the force and the time, by which the particle starts moving, are detected. The next phase of manipulation begins with particle movement initiation and ends when it reaches the target point, which is called the second phase of manipulation. In order to simulate the second phase of manipulation, the Wilson model is used. Also, due to the fact that no surface is completely smooth and it has roughness that affects the amount of friction force and consequently manipulation relations, in order to bring the results closer to reality, in this article the Rumpf and Rabinovich models are used. In the first phase of manipulation for modelling the substrate roughness, Rumpf and Rabinovich models are used. Moreover, Rumpf model is used for modelling the particle and substrate roughness in the second phase of manipulation. Finally, to validate the results of the manipulation simulation process, they are compared with the results of the existing researches. In the first phase of the manipulation, based on the simulation results, the critical time and force error for the sliding mode were 0.94 and 1.1%, respectively. Meanwhile, the critical time and force error for rotation mode are 1.3 and 1.7%, respectively. For manipulation, taking roughness into account, the critical time and force error for sliding and rotation modes in the first phase of manipulation are about 6%. The error for the second phase, using Wilson model, considering roughness for the particle and the substrate when the particle slides on the substrate is 3.2% as well. As a result, there is a good match between the simulation and the experimental results for both manipulation phases.