Skip to Main Content
Atomic force microscope (AFM) based nanomanipulation systems are generally slow, not repeatable and imprecise due to a lack of control on the success of limited attempts in the literature. To improve the amount of control, reliability and precision of such systems, this work proposes an automated nanomanipulation method. Spherical gold nanoparticles with 100 nm diameter are positioned mechanically on a flat mica substrate by contact manipulation by the AFM probe tip to a desired position autonomously. The most significant issue of the manipulation operation is the lack of real-time visual feedback. This issue is solved by developing a robust algorithm for particle center detection and using the AFM cantilever deflection (force) signals to detect contact losses in real-time and to repeat the manipulation again until the target location is reached. Using these solutions, an automated AFM manipulation system is developed and a statistical study is made, where gold nanoparticles are positioned for 50 times to random target positions in different directions and pushing distances. 86% of all the particles could be successfully positioned to the target positions with an accuracy less than 100 nm. Unsuccessful positioning operations are due to the particle sticking to either the tip (8%) or the substrate (6%). Additionally, performance of the successful manipulations are investigated on 60 manipulation operations in 12 different directions and 5 different distances. The metrics used to quantify performance are the final particle position error and the average manipulation speed.