(127j) Simulating Chemically Fueled Molecular Motors

Molecular motors are at the heart of life. These diminutive engines drive muscle contraction, molecular transport, cellular motion, and other important life processes by transducing chemical fuel into mechanical work. Traditional computational approaches to studying the mechanism of such motors rely on projecting the dynamics into lower dimensions or probing the dynamics of chemically bound and unbound states separately. I will discuss efforts to instead simulate a motor’s complete dynamics in the presence of a chemical driving force.

I will present a model and methodology inspired by a catenane-based artificial molecular motor. The model consists of two interlocked rings, the smaller of which diffuses around the larger. Removable blocking groups gate this natural diffusion, guiding it in a preferred direction with a mechanism known as an “information ratchet.” Tetrahedral clusters—which serve as fuel—decompose at catalytic sites on the large ring to create these blocking groups. Meanwhile, grand canonical Monte Carlo moves at the boundary constantly inject tetrahedral clusters and remove the decomposition products, creating a nonequilibrium steady state. The model relies only on a single, time-independent, classical potential energy surface with no external forces or torques driving the motor. Rather, directed motion is caused only by the motor transducing free energy from the fuel, which is continually available in the nonequilibrium steady state.

Simulations of this model reveal the tradeoffs between motor accuracy, velocity, and efficiency and demonstrate the motor’s response to a range of operating conditions. These simulations are a playground for exploring how particle interactions affect motor performance and serve as a design tool for exploring different motor configurations as well. Some of these explorations have already revealed interesting behavior, such as a reversal of motor direction depending on the spacing between catalytic sites. Simulation data is also used to parameterize coarse-grained Markov state models, connecting to the framework of stochastic thermodynamics. Furthermore, by determining entropy production rates from the simulations, I evaluate the motor’s performance with thermodynamic uncertainty relationships, which bound the system’s precision in relationship to the entropy produced. With this approach, I can directly connect experimental studies to recent thermodynamic theories, creating a design tool for future artificial molecular motors and revealing how close they come to theoretical limits.