Mechanical-plowing-based high-speed patterning on hard material via advanced-control and ultrasonic probe vibration

Abstract

In this paper, we present a high-speed direct pattern fabrication on hard materials (e.g., a tungsten-coated quartz substrate) via mechanical plowing. Compared to other probe-based nanolithography techniques based on chemical- and/or physical-reactions (e.g., the Dip-pen technique), mechanical plowing is meritorious for its low cost, ease of process control, and capability of working with a wide variety of materials beyond conductive and/or soft materials. However, direct patterning on hard material faces two daunting challenges. First, the patterning throughput is ultimately hindered by the "writing" (plowing) speed, which, in turn, is limited by the adverse effects that can be excited/induced during high-speed, and/or large-range plowing, including the vibrational dynamics of the actuation system (the piezoelectric actuator, the cantilever, and the mechanical fixture connecting the cantilever to the actuator), the dynamic cross-axis coupling between different axes of motion, and the hysteresis and the drift effects related to the piezoelectric actuators. Secondly, it is very challenging to directly pattern on ultra-hard materials via plowing. Even with a diamond probe, the line depth of the pattern via continuous plowing on ultra-hard materials such as tungsten, is still rather small (<0.5 nm), particularly when the "writing" speed becomes high. To overcome these two challenges, we propose to utilize a novel iterative learning control technique to achieve precision tracking of the desired pattern during high-speed, large-range plowing, and introduce ultrasonic vibration of the probe in the normal (vertical) direction during the plowing process to enable direct patterning on ultra hard materials. The proposed approach was implemented to directly fabricate patterns on a mask with tungsten coating and quartz substrate. The experimental results demonstrated that a large-size pattern of four grooves (20 μm in length with 300 nm spacing between lines) can be fabricated at a high speed of ~5 mm/s, with the line width and the line depth at ~95 nm and 2 nm, respectively. A fine pattern of the word "NANO" is also fabricated at the speed of ~5 mm/s.

Control of a Magnetostrictive-Actuator-Based Micromachining System for Optimal High-Speed Microforming Process

“…The proposed MFDIIC method not only addresses this issue via iteration, but also achieves fast and smooth engagement by adjusting the transition coefficient. The superior tracking performance of the MFDIIC algorithm over feedback control has been clearly demonstrated in experiments [21].…”

“…In this paper, we propose a control framework that integrates a recently developed iterative control technique [21] with Kalman filtering and the optimal trajectory design for non-periodic tracking-transition switching [22]. As depicted in Fig.…”

“…The recently developed modeling-free differential-inversion iterative control (MFDIIC) approach [21] is utilized to track the aforementioned desired trajectory…”

“…This feature is particularly attractive for microforming as the dynamics of the large-range actuation system changes substantially when the workpieces are engaged, whereas it is practically infeasible to account for that change through remodeling each time-in order to maintain the tracking precision. Moreover, it has been demonstrated [21] that the aforementioned MFDIIC algorithm can compensate for both the hysteresis and the dynamics effects of hysteresis-dynamics systems including magnetostrictive actuator systems. We realize that techniques such as the optimal state-transition technique [28], the optimal output transition technique [29], or the input-shaping technique [30], might be employed for the output transition during the prewelding process.…”

“…First, a recently developed iterative control technique [21] is combined with the design of the transition trajectory to achieve rapid engagement of the workpieces without inducing post-engagement oscillations. Then, Kalman filtering is employed to rapidly detect and identify the prewelding to welding phase transition by accurately estimating the interaction force from the measured noisy data.…”