Restricted Research - Award List, Note/Discussion Page
Fiscal Year: 2023
68 University of North Texas (141956)
Principal Investigator: Du,Jincheng
Total Amount of Contract, Award, or Gift (Annual before 2011): $ 825,000
Exceeds $250,000 (Is it flagged?): Yes
Start and End Dates: 5/2/22 - 5/1/25
Restricted Research: YES
Academic Discipline: Materials Science & Engineer
Department, Center, School, or Institute: College of Engineering
Title of Contract, Award, or Gift: Novel Phosphate Waste Forms to Enable Efficient Dehalogenation and Immobilization of Salt Waste
Name of Granting or Contracting Agency/Entity:
Citrine Informatics
CFDA: 81.135
Program Title: none
Note:
The proposal research project will use state-of-the-art data science and physics-based simulation methods in conjunction with experiments to rapidly discover and develop novel phosphate waste forms. Accuracy of machine learning process-structure-property models can be a challenge in their use in discovery and design workflows, particularly for extrapolation to novel systems where training data is sparse, which is the case for alternative glass compositions (i.e., non-borosilicate and non-iron-phosphate). To ensure sufficient predictive capabilities to enable glass and ceramic waste form design, UNT will perform extensive simulations of relevant indicator properties for performance metrics, which will be closely coupled with machine learning models to improve extrapolation performance. Built upon recent development of interatomic potentials for aluminosilicate and borosilicate nuclear waste glasses, UNT will develop needed potential parameters for rare earth and transition metal phosphate, oxyhalides, and related compositions to enable simulations of halide containing phosphate and other novel waste compositions for MSR or reprocessing UNF. Density Functional Theory (DFT) and DFT-based ab initio molecular dynamics (MD) simulations will be used to provide input data for potential developments and potential validations. Once validated of the potentials, large scale and high throughput MD simulations of selected glass and ceramic systems can be performed to obtain the structures and properties of targeted waste form systems. Quantitative structureproperty relationship (QSPR) analysis based on the structural information of the crystals or glass waste forms will be performed for the structure-property relationship analysis. This approach has been proven effective in establishing correlations of initial dissolution rate and glass compositions of borosilicate nuclear waste glasses. The large amount of structure and property data from MD and related simulations can be used as an input database for machine learning of the compositions described in the previous section.
Discussion: No discussion notes