CDS&E: MPATHS - Microscopic Pathway Analysis Toolkit for High-throughput Studies

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

Non-technical summary From the freezing of water into ice, to the ordering of proteins into arrays for analysis and drug design, to the assembly of nanoparticles into crystals that direct light, changes that occur in the structural arrangements of atoms, molecules, macromolecules and particles is ubiquitous in nature and in the synthesis and manufacturing of many materials and products. Understanding how these structural arrangements occur during a change from one phase of matter to another is critical to understanding self-organization processes in nature and to designing and making new materials that build themselves – predictively, reliably, and inexpensively -- from the bottom-up. By controlling the assembly process, novel materials that combine unique properties in important, unprecedented ways become possible, impacting everything from protective coatings and light-sensitive paints to materials that store and convert energy to stealth technologies and sports equipment. This project aims to develop and broadly disseminate powerful scientific software to aid researchers in tracking, analyzing, studying, and eventually engineering phase transition “pathways” that a system of material building blocks follows as it self-assembles and changes from one phase to another. The envisioned toolkit -- Microscopic Pathway Analysis Toolkit for High-Throughput Study (MPATHS) -- will be accessible, be easy to use, and exploit the fastest available computer architectures. It will make possible systematic, high-throughput studies of different types of phase transition pathways in ways that will make cross-system comparisons easy and will interoperate seamless with other open-source packages used by researchers who study phase transitions via simulations or experiments. Because MPATHS will reveal microscopic, mechanistic details of how order emerges from disorder during assembly processes, MPATHS will be of immediate interest to the nanoparticle and soft matter communities who can use it to study thermodynamic self-assembly of colloidal crystals and other complex structures, as well as swarming processes in active matter systems. MPATHS will also be of immediate and even broader interest to the materials, engineering, and chemistry communities interested in atomic and molecular self-assembly processes. An emphasis on accessibility to researchers outside of scientific computing fields will facilitate the adoption of MPATHS by a broader community. Technical summary Phase transitions in which the structure of a multi-particle system changes from one state to another are ubiquitous in nature and have been widely utilized for industrial purposes, such as the manufacturing of pharmaceuticals and the fabrication of materials, including metals, ceramics, plastics, nanocomposites and more. Therefore, understanding the mechanistic details of exactly how these structural transitions occur (i.e., the transition pathway) is crucial for a wide range of potential applications, including predicting and designing novel materials with desired properties as well as increasing yields or driving down costs for ones already in use. Hence, considerable research is devoted to the study of such phase transitions, especially from the point of view of thermodynamics and kinetics, including the study of such bulk quantities as free energy barriers to nucleation of ordered phases and nucleation rates. Such studies are informative but do not reveal the microscopic, particle-level details of how a particular sample changes from one state to another. For example, structural transitions such as crystallization, where a liquid solidifies into an ordered solid phase, are driven by a change in thermodynamic quantities such as temperature or pressure, which triggers changes in local structure involving a subset of particles (atoms, molecules, nanoparticles) that eventually spans the entire system. Microscopic details of how the local structure changes along the transition pathway controls the quality of the resulting crystal, as well as whether the resulting crystal is the thermodynamically preferred one or a metastable polymorph. However, despite the recent advances in computational power and experimental approaches, automated, system-agnostic tracking of structure evolution and detection of local and global changes in particle organization is largely nonexistent in materials and other fields of science and engineering, hindering the studies needed to predictively link processing parameters and particle attributes (such as shape and interaction patchiness in the case of nanoparticles) to final product. With such microscopic information in hand, researchers could precisely tailor not only processing parameters but also nanoparticle attributes to stabilize certain polymorphs over others, improve structural quality of the product, and optimize yield. This project will develop a generalized computational toolkit that will enable the systematic study and cross-system comparisons of structure evolution across a wide range of self-assembling materials. The system-agnostic nature of the toolkit will eliminate error- and bias-prone manual intervention, thereby accelerating the discovery, understanding, and engineering of pathways for multiple classes of materials. Using particle positions, orientations, and properties from either experimental or computational raw data as input, the Microscopic Pathway Analysis Toolkit for High-throughput Studies (MPATHS) will enable users to identify and label local structural motifs and track their development across the entire transition pathway. The project team will develop and package within MPATHS powerful and system-agnostic routines for finding neighbors, calculating per-particle order parameters and their cross-correlations, detecting local and global structural events using methods from information theory, and visualizing systems based on this information. MPATHS will enable expert workflows by non-experts and be system-agnostic so it can be used intuitively and with little-to-no computational expertise by researchers working on structural transitions across many disciplines. MPATHS will incorporate easy-to-use interfaces that abide by TRUE (transferable, reproducible, usable by others and extensible) principles, including a python scripting interface to facilitate the scripting of customized MPATHS analyses, a graphical user interface to foster accessibility, and a powerful command line interface. MPATHS will be made widely available and usable as either an offline or online tool to enable either post-analysis or on-the-fly control of simulations or experiments. MPATHS will permit the study of fundamental processes that only now are able to be studied due to advances in computing power that allow for detailed molecular simulations of complex structural transitions over long times with fine temporal detail, and exciting developments in in-situ electron microscopy that allow, for the first time, the visualization of dynamic nanoparticle rearrangements during the self-assembly of colloidal crystals of nanoparticles. MPATHS will be used to discover the microscopic processes driving two different structural transitions in simulated systems as exemplars: (i) assembly pathways of isostructural complex crystals from atoms and nanoparticles and emergence of ordered structures such as three-phase coexistence in active matter systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.