Optimal tuning of a Brownian information engine operating in a nonequilibrium steady state

Abstract

A Brownian information engine can induce directed motion of a Brownian particle in a single heat bath at constant temperature by extracting work from the information about the microscopic state of the particle, and serves as a model for artificial and biological submicron scale engines. Much of the experimental studies to date are limited to the realization of an information engine where the initial state of the system is in thermal equilibrium; however, most of the biological and artificial motors operate far from equilibrium. Here, we realize a cyclic information engine operating in a nonequilibrium steady state consisting of a Brownian particle in an optical trap and investigate the optimal operating conditions for maximum work, power, and efficiency. The performance of our information engine depends on the cycle period and the distance X-f that the trap center is shifted with respect to the reference distance x(m). We found that the extracted work increases with increasing tau and is maximum when tau reaches infinity and x(f )= 2x(m), while the extracted power is maximum at finite tau for x(f ) >= x(m), and when tau approaches zero for x(f ) < x(m). By measuring the steady-state information, we have also measured the efficiency at maximum power.