2017 Theses Doctoral
Analytical Imaging for Complex Materials
Systems known as complex materials have key attributes that contribute to their designation as "complex''. For example, evolving dynamical properties add complexity, as is observed in supercooled liquids. Polymers and proteins are structurally complex as they can fold in different conformations, with these different conformations affecting different biological functions or physical properties. Complex materials generally have interesting macroscopic properties that are difficult to predict from their microscopic (molecular) constituents. The connection between microscopic features and macroscopic properties in these materials has been a subject of study for decades, yet the ability to wield strong predictive power in these materials remains elusive.
Imaging can provide information in both space and time necessary to understand how the microscopic details in these materials yield the observed macroscopic properties. Moreover, time-sequenced imaging allows one to understand how these properties might evolve. To image these details however, requires high resolution in both time and space. Unfortunately, obtaining images and extracting information from these images becomes quite difficult as the length scales of interest become small, especially when signal-to-background ratios are low.
My dissertation work addresses the challenge of extracting such information by developing quantitative methods to circumvent obstacles related to obfuscation from low signal as well as the diffraction limit, with high-throughput and high-resolution. I apply these techniques towards the study of the following complex materials via imaging: (1) single molecules rotating in a supercooled liquid (2) conjugated polymers containing multiple emitters within the diffraction limit (3) solvent vapor annealing mediated conjugated polymer aggregation and (4) collagen gels during their formation and perturbation.
The first chapter describes how the local dynamics in a supercooled liquid may be assessed by the rotations of single-molecule probes. To resolve rotations however, requires splitting the fluorescence signal from single-molecules into orthogonal polarizations thereby reducing the already low signal-to-background ratio (SBR) expected from single-molecule experiments. The data is further complicated by instances of photoblinking and out-of-plane rotation, when only background signal is present. A convenient method for excluding background signal was developed via a Monte Carlo simulation for discriminating points in the trajectory composed only of background signal that is robust to SBR. These simulations also showed an SBR dependance for the accuracy of the values extracted from rotational autocorrelation functions. This method was used experimentally to directly demonstrate ergodicity in supercooled liquids.
The next chapter focuses on conjugated polymers, which display a complex relationship between chain conformation and photophysics. This conformational complexity exists even at the single-chain level, obscuring the understanding of how excitons behave in the bulk, such as in a device. Understanding this relationship however, is difficult as conformation and photophysics are hard to access in operando not only because a single conjugated polymer chain is smaller than the diffraction limit, but also because a fluorescing conjugated polymer emits light from many locations. The overall conformation is typically assessed using polarization modulation measurements, which only provide mesoscale information about a chain's conformation. A super-resolution method was developed to map the distribution of emitters and trace out single-chain conformation. The extracted radii of gyration for these single-chains matched well with polymer theory.
Chapter three describes the development of experimental and image analytical tools to bridge single-chain studies of photophysics in conjugated polymer, to the photophysics that might be observed in a conjugated polymer device where chain-chain contacts and high levels of local ordering may be present. It has been shown previously that solvent vapor annealing can be used to prepare conjugated polymer aggregates of various levels of internal ordering. However, solvent vapor annealing is a process that is difficult to control and difficult to evaluate. Therefore, a first-of-its-kind apparatus was constructed that can generate and deliver solvent vapor in a controlled fashion to swell polymer films while monitoring both film dynamics by fluorescence imaging as well as swelling extent via a quartz crystal microbalance. Fluorescent images acquired during aggregation showed heterogeneous diffusion among aggregates, possibly indicating heterogeneous sizing. Fluorescent characterization of presumably differently sized aggregates indicates a possible emergent quenching phenomenon in a bulk conjugated polymer material.
The final chapter of this dissertation describes an effort to characterize dynamics in collagen gels. Collagen gels form through a complex sol-gel process precipitated by nucleation and growth fibrillogenesis. To probe the long length scales of interest here, single-molecule methods were not practical. Instead, an optical flow algorithm was explored to detect key physical events in the evolving system. One effort aims to characterize the dynamics of the early gelation process. In particular, the optical flow measurements provide a high-resolution measure for the moment at which the sol-gel transition occurs. Another application involves the use of optical flow to observe distortions in the collagen network while it is undergoing strain stiffening. Preliminary studies show that at critical strain, local breaks in the gel propagate throughout the gel until the gel completely loses its ability to sustain stress.
In general, the ability to quantify details about complex materials from imaging data can be quite a complex endeavor itself, requiring awareness of the physical phenomena of interest, how said phenomena manifests optically, and the use and development of appropriate algorithms. As described in this dissertation, the proper use and/or development of the proper image analysis methods allow for extraction of key information from dense data.
- Hoang_columbia_0054D_13720.pdf binary/octet-stream 5.1 MB Download File
More About This Work
- Academic Units
- Thesis Advisors
- Kaufman, Laura J.
- Ph.D., Columbia University
- Published Here
- February 1, 2017