(367c) High-Throughput Microrheology of Synthetic and Biomolecular Polyelectrolytes Using Differential Dynamic Microscopy

Rapid formulation of complex fluids and biological materials often demands rapid, accurate analysis and improved measurement methods for the rheology of scarce and microscopically heterogeneous materials. Although microrheology holds promise to address this challenge, current analysis approaches involving multiple particle tracking (MPT) limit throughput and place barriers on automation. Alternatively, differential dynamic microscopy (DDM) has emerged in recent years as a promising new tool to track how materials systems evolve in space and time. By representing image fluctuations in the Fourier domain, DDM connects real-space image data to dynamical properties accessed in dynamic light scattering, and was recently extended to the estimation of mean-squared displacements, providing access to microrheology measurements without the need to track probes directly. By comparison to MPT, DDM microrheology is therefore accessible to a number of different experimental conditions where MPT is difficult or impossible. Most importantly, a lack of user-defined inputs provides a path to fully automated microrheology analysis using DDM. However, in its current state, DDM analysis suffers from several drawbacks, including slow computation, lack of error quantification and limited robustness.

Here, we present a new strategy to overcome these limitations and enable automated, high-throughput microrheology. A statistical approach to estimate the contribution to the DDM signal from fluctuating noise enables more accurate determination of mean-squared displacements. By coupling this noise estimator with Gaussian process regression, we demonstrate improved computational efficiency of DDM analysis through downsampling of the image correlation data while retaining accuracy to as little as 1% data utilization. The resulting analysis workflow stabilizes large fluctuations inherent to limited data sampling and preserves error quantification from the acquired imaging data, while accelerating computation to near real-time analysis. We demonstrate the utility of this new high-throughput DDM microrheology framework involving two experimental studies. The first involves oppositely charged polyelectrolytes that undergo complex coacervation. We demonstrate how DDM microrheology can be used systematically to characterize the viscoelastic properties of the dense coacervate phase across a range of macro-ion and salt concentrations, and use spatially-resolved high-throughput microscopy to track morphology evolution of the resulting phase separation. The second study involves high-throughput viscometry of self-assembling proteins that form discrete clusters due to a combination of short-range attractive and long-range repulsive intermolecular interactions. Understanding how protein clustering influences viscosity across a range of physicochemical factors ultimately provides a mechanistic model for connecting protein assembly and colligative properties in vitro with its native function in vivo. Overall, these case studies highlight how high-throughput microrheology analysis enabled by DDM paves the way for rapid screening and formulation of biomolecular materials, and serves as a critical link for material discovery.