The Role of Configurational Entropy in Real-Time Mass Sensing

We present a theoretical framework for determining the mass deposited on a mechanical resonator subject to a flux of incoming particles of a single species. We consider the specific example of a vibrating nanostring and infer the history of mass deposition events from the frequency shifts in real time using a numerical optimization alogrithm that correctly compensates for the configurational entropy. Our approach is tested against simulated data and is shown to perform well. We comment on its applicability to common inverse problems in physics, such as the extraction of spectral functions from imaginary-time Green's functions, that also suffer from configurational entropy effects.