Polymer Upcycling and Hard X-ray Sciences Projects Funded by the Joint Task Force Initiative

Machine Learning of Coherent Scattering from Dynamically Evolving Systems


PIs and Institutions: Paul Fenter and Irene Calvo Almazan, Argonne; Eric Jonas, UChicago

Description: The central thesis is to demonstrate that machine learning (ML) can bridge between imaging complex structures and probing structural dynamics for the case of temporally evolving structures. We seek to demonstrate this concept through analysis of simulated experiments for a simple model system. Specifically, our goal is to prove that the incorporation of temporal support and overlap constraints of a continuously-evolving system (i.e., in addition to the spatial support constraints), along with the unique capabilities of ML, can be used in conjunction with Q-space “over-sampling” to provide deeper insights into structural dynamics and to bridge between the (currently distinct) modes of analysis for CDI and XPCS.

Advanced control of strain gradients in quantum material systems via nanoscale Bragg ptychography


PIs and Institutions: Martin Holt, Alexander High, and F. Joseph Heremans, Argonne and UChicago

Description: The controllable creation and manipulation of crystallographic strain in quantum material systems is a transformative goal with potentially broad impact in engineered design of optical response and spin excitation levels with device-level consequences for information processing, communication, and nanoscale sensing. This is especially true in semiconductor material hosts for spin defect systems such as diamond, where near-defect strain, strain gradients, and structural dynamics all play a pivotal role in tuning quantum performance and optimizing coherence. While much progress has been made in understanding and exploiting mechanical coupling to optically active spin-defects in diamond, a common bottleneck is that the underlying structural mechanisms inducing a quantum response are often inferred through a model-dependent assessment of optical energy levels and are not observed directly. We propose to address and exploit this characterization gap by i) adapting synchrotron based 3D x-ray ptychographic strain imaging methods to directly observe strain and strain gradients in diamond membranes used as quantum material hosts, and ii) create process engineering tools that are uniquely based on this microscopic understanding through an inverse design approach to achieve predictive structural control of sparse optically active defects.

Data Driven Discovery of Regenerable Plastics through High-Throughput Experimentation


PIs and Institutions: David Kaphan, Magali Ferrandon, and Massimiliano Delfero, Argonne; Stuart Rowan, UChicago

Description: Living systems repeatedly demonstrate the importance of efficient, closed-loop utilization of (macro)molecular resources. For plastics, we are now seeing the repercussions of not having an effective closed-loop carbon economy. The EPA estimates that less than 10% of end-of-life plastic materials generated in the United States are recycled each year, and those not recycled are stranded in municipal and industrial landfills or combusted, leading to environmental contamination by microplastics and CO2 emissions. Following nature’s example of chemical circularity, we aim to design next-generation plastics with controllable methods of selective dismantlement. Our approach would allow for their deconstruction and would enable efficient regeneration of recycled materials with identical performance characteristics to virgin plastics through the incorporation of stimulus responsive cleavage sites referred to as “scissophores.” The inability to achieve this goal now is a major impediment to a truly circular economy for plastics.

Upcycling Plastics into Separating Membranes: A Sustainable Solution for Liquid Electrolyte Batteries


PIs and Institutions: Shrayesh Patel, Argonne; Lu Zhang, Argonne and UChicago

Description: In this proposal, we intend to establish a low-cost and high-efficiency processing system to upcycle used plastics to produce sustainable separating membranes for liquid electrolyte batteries. Our goals are: 1) to demonstrate utilizing a straightforward solvating, casting and drying process to convert used plastics to valuable membranes with tunable pore sizes and properties; 2) to utilize automated platforms to accelerate the discovery of low-cost solvating systems for dissolving a large range of plastics and build understanding of solvating principles; 3) to establish high throughput characterization apparatus and gather systematically measured properties of the membranes, such as pore size, conductivity, to provide a dataset for machine learning of such systems.