Paper Number
AR8 

Session
Applied Rheology and Rheology Methods

Title
High throughput microrheology: A path to rapid phase diagram and formulation mapping

Presentation Date and Time
October 11, 2021 (Monday) 2:20

Track / Room
Track 2 / Ballroom 7

Authors (Click on name to view author profile)

1. Luo, Yimin (University of California, Santa Barbara)
2. Gu, Mengyang (University of California, Santa Barbara, Department of Statistics and Applied Probability)
3. He, Yue (University of California, Santa Barbara, Department of Statistics and Applied Probability)
4. Edwards, Chelsea (University of California, Santa Barbara, Department of Chemical Engineering)
5. Helgeson, Matthew E. (University of California, Santa Barbara, Department of Chemical Engineering)
6. Valentine, Megan (University of California, Santa Barbara, Department of Mechanical Engineering)

Author and Affiliation Lines (in printed abstract book)
Yimin Luo¹, Mengyang Gu², Yue He², Chelsea Edwards¹, Matthew E. Helgeson¹ and Megan Valentine^3
1Department of Chemical Engineering, University of California, Santa Barbara, Santa Barbara, CA 93106; ²Department of Statistics and Applied Probability, University of California, Santa Barbara, Santa Barbara, CA 93106; ³Department of Mechanical Engineering, University of California, Santa Barbara, Santa Barbara, CA 93106

Speaker / Presenter 
Luo, Yimin

Keywords
experimental methods; computational methods; polymer solutions; rheology methods

Text of Abstract
Differential dynamic microscopy (DDM) combines the real-space resolution of microscopy and the ensemble sensitivity of scattering and has emerged as a promising tool to track the dynamic evolution of complex fluids. In DDM, image pair differences are ensemble averaged and analyzed in Fourier space through the image structure function D(q,?t), which encodes material functions including the intermediate scattering function and mean-squared displacement, without the need for particle tracking. The method boasts easy experimental setup, sensitivity to weakly scattering samples, applicability to scenarios where tracking is onerous or impossible, and offers a path to fully automated analysis. However, DDM analysis currently suffers from drawbacks associated with slow computation, lack of error quantification and robustness issues. We present a statistical approach by rigorously deriving the noise term, which arises due to imaging artifacts, and resampling through Gaussian Process Regression. As a result, a mere 1% of the original computation needs to be carried out at strategically selected design points, greatly improving the computational efficiency to near real time. We further extend DDM microrheology to more rigid materials where small displacement of probes limits the resolution of DDM relative to multiple particle tracking. These advances make possible truly automated microrheology characterization of material systems over a wide formulation space, which we demonstrate using a case study involving mixed polyelectrolyte solutions that undergo liquid-liquid phase separation (coacervation) at different polymer and salt concentrations. We show how coupling DDM microrheology with imaging in situ during phase separation enables identification of phase boundaries, partitioning of salt and polymer in the emerging phases, and rheological measurements of dense phase droplets. The high throughput measurements and analysis enabled by this method serve as a critical link for the materials discovery and inverse design.