Directed Self-Assembly of Linear Nanostructures by Optimal Control of External Electrical Fields

TL;DR

This paper develops an optimal open-loop control strategy using electric potentials to guide charged particles into desired nanostructures, accounting for uncertainties in particle dynamics through two models.

Contribution

It introduces a novel optimal control framework for directed self-assembly of nanostructures using electric fields, applicable to both continuous and discrete particle models.

Findings

Control potentials can be optimized to reliably form desired patterns.

The approach is effective despite uncertainties in initial conditions and particle behavior.

Numerical examples demonstrate the feasibility of the control strategy.

Abstract

An optimal control strategy is developed to construct nanostructures of desired geometry along line segments by means of directed self-assembly of charged particles. Such a control strategy determines the electric potentials of a set of electrodes located at fixed points in the line segment. The particles move under the electric forces generated by these electrodes and by the interactions between the particles themselves to form a desired pattern eventually. Due to technology limitations, the particle positions cannot be measured during the course of control, so that the control is open-loop in nature. Such an open-loop control optimally changes the electrode potentials in time in order to create a desired pattern with the highest probability, despite the inherent uncertainty in the initial positions and the dynamical behaviors of the particles. Two models are proposed to describe the uncertain dynamics of the particles: a continuous model relying on a set of nonlinear stochastic differential equations, and a discrete Ising model consisting of a large dimensional continuous-time Markov chain. While the first model is more mathematically tractable, the second one more precisely describes particles at the nanometer scale. The control design procedure begins with the continuous model and identifies the structure of its stable equilibria, which is used later to propose a piecewise constant structure for the control and to demonstrate that the optimal value of each piece is independently obtained from a certain static optimization problem. It is shown next that the design procedure can be applied to the discrete model with only minor modifications. A numerical example of control design is presented.