Observation of the Brownian motion of a small probe interacting with its environment is one of the main strategies to characterize soft matter. Essentially two counteracting forces govern the motion of the Brownian particle. First, the particle is driven by the rapid collisions with the surrounding solvent molecules, referred to as thermal noise. Second, the friction between the particle and the viscous solvent damps its motion. Conventionally, the thermal force is assumed to be random and characterized by a white noise spectrum. Friction is assumed to be given by the Stokes drag, implying that motion is overdamped. However, as the particle receives momentum from the fluctuating fluid molecules, it also displaces the fluid in its immediate vicinity. The entrained fluid acts back on the sphere and gives rise to long-range correlation. This hydrodynamic memory translates to thermal forces, which display a coloured noise spectrum. Even 100 years after Perrins pioneering experiments on Brownian motion, direct experimental observation of this colour has remained elusive. Here, we measure the spectrum of thermal noise by confining the Brownian fluctuations of a microsphere by a strong optical trap. We show that due to hydrodynamic correlations the power spectral density of the spheres positional fluctuations exhibits a resonant peak in strong contrast to overdamped systems. Furthermore, we demonstrate that peak amplification can be achieved through parametric excitation. In analogy to Microcantilever-based sensors our results demonstrate that the particle-fluid-trap system can be considered as a nanomechanical resonator, where the intrinsic hydrodynamic backflow enhances resonance. Therefore, instead of being a disturbance, details in thermal noise can be exploited for the development of new types of sensors and particle-based assays for lab-on-a-chip applications.
Download