Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “molecular data”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,117 records · Page 62
All sourcesNASA NTRSDOE OSTI
NASA NTRS1992

High molecular weight first generation PMR polyimides for 343 C applications

The effect of molecular weight on 343 C thermo-oxidative stability (TOS), mechanical properties, and processability, of the first generation PMR polyimides was studied. Graphite fiber reinforced PMR-15, PMR-30, PMR-50, and PMR-75 composites (corresponding to formulated molecular weights of 1500, 3000, 5000, and 7500, respectively) were fabricated using a simulated autoclave process. The data reveal that while alternate autoclave cure schedules are required for the high molecular weight resins, low void laminates can be fabricated which have significantly improved TDS over PMR-15, with only a small sacrifice in mechanical properties.

Malarik, D. C.↗
NASA NTRS2006

NASA Laboratory Astrophysics Workshop 2006 Introductory Remarks

NASA Laboratory Astrophysics Workshop 2006, is the fourth in a series of workshops held at four year intervals, to assess the laboratory needs of NASA's astrophysics missions - past, current and future. Investigators who need laboratory data to interpret their observations from space missions, theorists and modelers, experimentalists who produce the data, and scientists who compile databases have an opportunity to exchange ideas and understand each other's needs and limitations. The multi-wavelength character of these workshops allows cross-fertilization of ideas, raises awareness in the scientific community of the rapid advances in other fields, and the challenges it faces in prioritizing its laboratory needs in a tight budget environment. Currently, we are in the golden age of Space Astronomy, with three of NASA s Great Observatories, Hubble Space Telescope (HST), Chandra X-Ray Observatory (CXO), and Spitzer Space Telescope (SST), in operation and providing astronomers and opportunity to perform synergistic observations. In addition, the Far Ultraviolet Spectroscopic Explorer (FUSE), XMM-Newton, HETE-2, Galaxy Evolution Explorer (GALEX), INTEGRAL and Wilkinson Microwave Anisotropy Probe (WMAP), are operating in an extended phase, while Swift and Suzaku are in their prime phase of operations. The wealth of data from these missions is stretching the Laboratory Astrophysics program to its limits. Missions in the future, which also need such data include the James Webb Space Telescope (JWST), Space Interferometry Mission (SIM), Constellation-X (Con-X), Herschel, and Planck. The interpretation of spectroscopic data from these missions requires knowledge of atomic and molecular parameters such as transition probabilities, f-values, oscillator strengths, excitation cross sections, collision strengths, which have either to be measured in the laboratory by simulating space plasma and interactions therein, or by theoretical calculations and modeling. Once the laboratory data are obtained, a key step to making them available to the observer is the creation and maintenance of critically compiled databases. Other areas of study, that are important for understanding planet formation, and for detection of molecules that are indicators of life, are also supported by the Laboratory Astrophysics program. Some examples are: studies of ices and dust grains in a space environment; nature and evolution of interstellar carbon-rich dust; and polycyclic aromatic hydrocarbons. In addition, the program provides an opportunity for the investigation of novel ideas, such as simulating radiative shock instabilities in plasmas, in order to understand jets observed in space. A snapshot of the currently funded program, mission needs, and relevance of laboratory data to interpreting observations, will be obtained at this workshop through invited and contributed talks and poster papers. These will form the basis for discussions in splinter groups. The Science Organization Committee will integrate the results of the discussions into a coherent White Paper, which will provide guidance to NASA in structuring the Laboratory Astrophysics program in subsequent years, and also to the scientific community in submitting research proposals to NASA for funding.

Hasan, Hashima↗
NASA NTRS1994

Hydrogen and related materials at high density: Physics, chemistry and planetary implications

Recent studies of low-Z molecular materials including hydrogen to multimegabar pressures (less than 300 GPa) have uncovered a range of phenomena relevant to understanding the nature of the interiors of the outer planets and their satellites. Synchrotron x ray diffraction measurements (to 42 GPa) have been used to determine the crystal structure of the solid (hexagonal-close packed) and equation of state. Sound velocities in fluid and solid hydrogen (to 24 GPa) have been inverted to obtain elastic constants and aggregate bulk and shear moduli. In addition, an improved intermolecular potential has been determined which fits both static and shock-wave data. Use of the new potential for the molecular envelope of Jupiter suggests the need for major revisions of existing Jovian models or a reanalysis of reported free oscillations for the planet. Studies at higher pressures (greater than 100 GPa) reveal a sequence of pressure-induced symmetry-breaking transitions in molecular hydrogen, giving rise to three high-pressure phases (1, 2, and 3). Phase 1 is the rotationally disordered hcp phase which persists from low pressure to well above 100 GPa at high temperature (e.g., 300 K). Phase 2 is a low-temperature, high-pressure phase (transition at 100 GPa and 77 K in H2) with spectral features indicative of partial rotational ordering and crystallographic distortion. The transition to Phase 3 at 150 GPa is accompanied by a weakening of the molecular bond, gradual changes in orientational ordering, strong enhancement of the infrared intramolecular vibrational absorption, and strong intermolecular interactions similar to those of ambient-pressure network solids. Studies of the phase diagram reveal a triple point near 130 K and 160 GPa. Higher pressure measurements of vibrational spectra place a lower bound of approximately 250 GPa on the predicted transition pressure for dissociation of molecular hydrogen to form a monatomic metal.

Hemley, R. J.↗
DOE OSTI2020

Multivariate Linear Models for In Situ Monitoring of Low-Activity Nuclear Waste Using Infrared and Raman Spectroscopy - 20218

This paper presents the use of in situ monitoring tools for molecular analysis of non-radioactive species found in the supernatant liquid and dissolvable salt cake of nuclear waste at Hanford. We have selected two in situ spectroscopic techniques that do not require any sample preparation or consumables: infrared (in the form of attenuated total reflectance Fourier transform infrared or ATR-FTIR) and Raman spectroscopy. This work aims to evaluate their performance for liquid waste simulants representative of the Hanford DFLAW (Direct Feed Low-Activity Waste) feed. In addition, we have evaluated the performance of four multivariate linear models that extract concentrations from spectral measurements for a data set that was designed to include key molecular species found in liquid low-activity waste. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2026

Leveraging 13C-Labeling to Assign Molecular Formulas to Unknown Yeast Metabolites

Mass spectrometry analyses have identified tens of thousands of unknown small molecule-associated peaks in different biological specimens. Notably, even the simplest and best studied organisms like Escherichia coli and Saccharomyces cerevisiae yield thousands of unknown peaks. A key question is how many of these reflect actual novel endogenous metabolites. To explore this, Mahieu and Patti used complete 13 C -labeling in E. coli to credential peaks as biological. This reduced the number of unknowns by more than 90%. Here, we carry out similar uniform 13 C-labeling in the Baker’s yeast S. cerevisiae and two less-studied bioenergy-relevant yeasts Rhodotorula toruloides (lipid producer) and Issatchenkia orientalis (organic acid producer). Identification of unknown metabolite peaks and their molecular formulas is facilitated through software tailored for 13 C labeling data and resulting knowledge of carbon atom count. A classification model evaluates the plausibility of each candidate formula, with peaks lacking plausible candidate formulas unlikely to reflect metabolite molecular ions. This approach prioritizes about one hundred candidate abundant unknown metabolites with logical molecular formulas. Most of these are species-specific rather than conserved across yeasts, and more are found in the nonmodel yeasts than S. cerevisiae. Thus, 13 C-labeling data on unknown metabolites highlights the potential for discovering new metabolites and pathways in nonmodel yeasts.

Carbon↗
NASA NTRS1995

The 1995 Atmospheric Trace Molecule Spectroscopy (ATMOS) Linelist

The Atmospheric Trace Molecule Spectroscopy (ATMOS) experiment uses a Fourier transform spectrometer on board the space shuttle to record infrared solar occultation spectra of the atmosphere. Described are the current version of the molecular spectroscopic database used for the analysis of the data obtained during three Space Shuttle missions between 1992 and 1994.

molecular database↗
DOE OSTI2025

MS25: Materials Science-Focused Benchmark Data Set for Machine Learning Interatomic Potentials

Here, we present MS25, a benchmark data set for evaluating machine learning interatomic potentials (MLIPs) across diverse materials-relevant systems including MgO surfaces, liquid water, zeolites, a catalytic Pt surface reaction, high-entropy alloys (HEAs), and disordered Zr-oxides. Five MLIP architectures (MACE, NequIP, Allegro, MTP, and Torch-ANI) are trained and tested, focusing not only on traditional metrics (energies, forces, and stresses) but also explicitly validating derived physical observables such as lattice constants, volumes, and reaction barriers. We find that most models reach comparable accuracy on standard error metrics across the simple systems, although equivariant MLIPs offer 1.5–2× improvements over nonequivariant MLIPs in energy and force error for structurally complex or compositionally disordered environments such as HEAs and Zr–O systems. Our analysis highlights that low errors in energy and force predictions do not guarantee reliable observables, emphasizing the necessity of explicit validation. We demonstrate limitations in cross-framework transferability, as models trained on one zeolite framework (CHA) fail to reliably generalize to predictions of structurally distinct frameworks (e.g., MFI). Size-extensive tests show some dependence on system size for MgO, resulting from forced periodicity. The HEA and Zr–O data sets are identified as challenging tests for future benchmarks and MLIP model architecture developments as they show significant differentiation in error between MLIP architectures and are still relatively difficult at 1000 training images. Moving forward, we recommend that benchmarking efforts shift their focus from marginal accuracy improvements in energy and force errors toward identifying and understanding model failure modes, rigorously assessing transferability, and evaluating how their errors affect observable predictions. For researchers looking to choose an MLIP architecture, we suggest selecting equivariant MLIP architectures if the complexity of the system is a challenge. For simple materials problems, auxiliary features such as integration with molecular dynamics engines, trade-offs between computational data set generation cost vs MLIP inference speed, and framework integration may play a more important decision factor than small differences in error metrics that are unlikely to matter for production-level research.

chemical structure↗
DOE OSTI2021

AENET–LAMMPS and AENET–TINKER : Interfaces for accurate and efficient molecular dynamics simulations with machine learning potentials

Machine-learning potentials (MLPs) trained on data from quantum-mechanics based first-principles methods can approach the accuracy of the reference method at a fraction of the computational cost. To facilitate efficient MLP-based molecular dynamics and Monte Carlo simulations, an integration of the MLPs with sampling software is needed. Here, we develop two interfaces that link the atomic energy network (ænet) MLP package with the popular sampling packages TINKER and LAMMPS. The three packages, ænet, TINKER, and LAMMPS, are free and open-source software that enable, in combination, accurate simulations of large and complex systems with low computational cost that scales linearly with the number of atoms. Scaling tests show that the parallel efficiency of the ænet–TINKER interface is nearly optimal but is limited to shared-memory systems. The ænet–LAMMPS interface achieves excellent parallel efficiency on highly parallel distributed memory systems and benefits from the highly optimized neighbor list implemented in LAMMPS. We demonstrate the utility of the two MLP interfaces for two relevant example applications: the investigation of diffusion phenomena in liquid water and the equilibration of nanostructured amorphous battery materials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRSDate not supplied

Enabling Space Biological Knowledge Discovery Through Image and Video Data Sharing

Increased biomedical risks associated with deep space crewed missions (cis-Lunar, Mars transit/surface) require development of health countermeasures, novel ecosystem support, risk modeling, and fundamental space biological knowledge discovery. Molecular-omics, physiological-phenotypic-behavioral, and environmental-radiation telemetry data from space biological and health studies are needed for reuse by scientists to address these tasks. The data as well as space-relevant biospecimens are being made more findable, accessible, interoperable, and reusable through NASA’s Open Science Data Repository (OSDR). This new OSDR umbrella grouping includes NASA GeneLab, the NASA Ames Life Sciences Data Archive (ALSDA), and the NASA Biological Institutional Scientific Collection. The OSDR system design appropriately handles metadata and processed-tabular results from ALSDA studies collected from space experiments. But raw and processed ALSDA bioimage and video datasets require an expansion of OSDR’s data architecture to handle ingestion, curation, and egress. The academic-industry bioimaging field saw a scientific renaissance in the past several years through leveraging open-source software, international collaborations, machine learning, and other open science/programming approaches. As crewed missions and more biological experiments are on the deep space horizon, OSDR is embracing data stewardship through listening to feedback from subject matter experts and designing an expanded architecture which is appropriate for NASA’s goals to enable analysis and reuse of bioimaging and video data for the public science community.Discovery Through Image and Video Data Sharing

space biology↗
DOE OSTI2022

A Molecular Gas Ring Hidden in the Sombrero Galaxy

We present Herschel, ALMA, and MUSE observations of the molecular ring of Messier 104, also known as the Sombrero galaxy. These previously unpublished archival data shed new light on the content of the interstellar medium of M104. In particular, molecular hydrogen measured by CO emission and dust measured by far-infrared light are uniformly distributed along the ring. The ionized gas revealed by Hα and [C II] emission is distributed in knots along the ring. Despite being classified as an SAa galaxy, M104 displays features typical of early-type galaxies. We therefore compared its [C II] and dust emission to a sample of early-type galaxies observed with Herschel and SOFIA. The [C II]/FIR ratio of M104 is much lower than that of typical star-forming galaxies and is instead much more similar to that of early-type galaxies. By classifying regions using optical emission line diagnostics, we also find that regions classified as H II lie closer to star-forming galaxies in the [C II]/FIR diagram than those classified as low-ionization emission regions. The good match between [C II] and Hα emission, in conjunction with the lack of correlation between CO emission and star formation, suggests that there is very limited active star formation along the ring and that most of the [C II] emission is from ionized and neutral atomic gas rather than molecular gas. From the total intensity of the CO line, we estimate a molecular hydrogen mass of 0.9 × 10 9 M ⊙ , a value intermediate between those of early-type galaxies and the content of the molecular ring of our galaxy.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Deep learning of dynamically responsive chemical Hamiltonians with semiempirical quantum mechanics

Conventional machine-learning (ML) models in computational chemistry learn to directly predict molecular properties using quantum chemistry only for reference data. While these heuristic ML methods show quantum-level accuracy with speeds several orders of magnitude faster than traditional quantum chemistry methods, they suffer from poor extensibility and transferability; i.e., their accuracy degrades on large or new chemical systems. Incorporating quantum chemistry frameworks into the ML models directly solves this problem. Here we take the structure of semiempirical quantum mechanics (SEQM) methods to construct dynamically responsive Hamiltonians. SEQM methods use empirical parameters fitted to experimental properties to construct reduced-order Hamiltonians, facilitating much faster calculations than ab initio methods but with compromised accuracy. By replacing these static parameters with machine-learned dynamic values inferred from the local environment, we greatly improve the accuracy of the SEQM methods. Trained on molecular energies and atomic forces, these dynamically generated Hamiltonian parameters show a strong correlation with atomic hybridization and bonding. Trained with only about 60,000 small organic molecular conformers, the resulting model retains interpretability, extensibility, and transferability when testing on much larger chemical systems and predicting various molecular properties. Overall, this work demonstrates the virtues of incorporating physics-based descriptions with ML to develop models that are simultaneously accurate, transferable, and interpretable.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Roadmap on cosmic EUV and x-ray spectroscopy

Cosmic EUV/x-ray spectroscopists, including both solar and astrophysical analysts, have a wide range of high-resolution and high-sensitivity tools in use and a number of new facilities in development for launch. As this bandpass requires placing the spectrometer beyond the Earth’s atmosphere, each mission represents a major investment by a national space agency such as NASA, ESA, or JAXA, and more typically a collaboration between two or three. In general justifying new mission requires an improvement in capabilities of at least an order of magnitude, but the sensitivity of these existing missions are already taxing existing atomic data quantity and accuracy. This roadmap reviews the existing missions, showing how in a number of areas atomic data limits the science that can be performed. The missions that will be launched in the coming Decade will without doubt require both more and improved measurements of wavelengths and rates, along with theoretical calculations of collisional and radiative cross sections for a wide range of processes.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Magnetic, thermal, and electronic-transport properties of EuMg 2 Bi 2 single crystals

The trigonal compound EuMg 2 Bi 2 has recently been discussed in terms of its topological band properties. These are intertwined with its magnetic properties. In this paper detailed studies of the magnetic, thermal, and electronic-transport properties of EuMg 2 Bi 2 single crystals are presented. The Eu + 2 spins 7/2 in EuMg 2 Bi 2 exhibit an antiferromagnetic (AFM) transition at a temperature T N = 6.7 K, as previously reported. By analyzing the anisotropic magnetic susceptibility χ data below T N in terms of molecular-field theory (MFT), the AFM structure is inferred to be a c -axis helix, where the ordered moments in the hexagonal a b -plane layers are aligned ferromagnetically in the a b plane with a turn angle between the moments in adjacent moment planes along the c axis of ≈ 120 ° . An alternate but less likely magnetic structure is a planar structure with nearest-neighbor Eu spins aligned at ≈ 120 ° with respect to each other, where these ordered-moment layers are stacked along the c axis. The magnetic heat capacity exhibits a λ anomaly at T N with evidence of dynamic short-range magnetic fluctuations both above and below T N . The high- T limit of the magnetic entropy is close to the theoretical value for spins 7/2. The in-plane electrical resistivity ρ ( T ) data indicate metallic character with a mild and disorder-sensitive upturn below T min = 23 K. An anomalous rapid drop in ρ ( T ) on cooling below T N as found in zero field is replaced by a two-step decrease in magnetic fields. The ρ ( T ) measurements also reveal an additional transition below T N in applied fields of unknown origin that is not observed in the other measurements and may be associated with an incommensurate to commensurate AFM transition. The dependence of T N on the c -axis magnetic field H ⊥ was derived from the field-dependent χ ( T ) , C p ( T ) , and ρ ( T ) measurements. This T N ( H ⊥ ) was found to be consistent with the prediction of MFT for a c -axis helix with S = 7 / 2 and was used to generate a phase diagram in the H ⊥ – T plane.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Probing Molecular Packing of Amorphous Pharmaceutical Solids Using X-ray Atomic Pair Distribution Function and Solid-State NMR

The structural investigation of amorphous pharmaceuticals is of paramount importance in comprehending their physicochemical stability. However, it has remained a relatively underexplored realm primarily due to the limited availability of high-resolution analytical tools. Here in this study, we utilized the combined power of X-ray pair distribution functions (PDFs) and solid-state nuclear magnetic resonance (ssNMR) techniques to probe the molecular packing of amorphous posaconazole and its amorphous solid dispersion at the molecular level. Leveraging synchrotron X-ray PDF data and employing the empirical potential structure refinement (EPSR) methodology, we unraveled the existence of a rigid conformation and discerned short-range intermolecular C–F contacts within amorphous posaconazole. Encouragingly, our ssNMR 19 F– 13 C distance measurements offered corroborative evidence supporting these findings. Furthermore, employing principal component analysis on the X-ray PDF and ssNMR data sets enabled us to gain invaluable insights into the chemical nature of the intermolecular interactions governing the drug–polymer interplay. These outcomes not only furnish crucial structural insights facilitating the comprehension of the underlying mechanisms governing the physicochemical stability but also underscore the efficacy of synergistically harnessing X-ray PDF and ssNMR techniques, complemented by robust modeling strategies, to achieve a high-resolution exploration of amorphous structures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

2D Retrieval Frameworks for Hot Jupiter Phase Curves

Spectroscopic phase curves provide unique access to the three-dimensional properties of transiting exoplanet atmospheres. However, a modeling framework must be developed to deliver accurate inferences of atmospheric properties for these complex data sets. Here, we develop an approach to retrieve temperature structures and molecular abundances from phase curve spectra at any orbital phase. In the context of a representative hot Jupiter with a large day–night temperature contrast, we examine the biases in typical one-dimensional (1D) retrievals as a function of orbital phase/geometry, compared to two-dimensional (2D) models that appropriately capture the disk-integrated phase geometry. We guide our intuition by applying our new framework on a simulated Hubble Space Telescope (HST)+Spitzer phase curve data set in which the “truth” is known, followed by an application to the spectroscopic phase curve of the canonical hot Jupiter, WASP-43b. We also demonstrate the retrieval framework on simulated James Webb Space Telescope (JWST) phase curve observations. We apply our new geometric framework to a joint fit of all spectroscopic phases, assuming longitudinal molecular abundance homogeneity, resulting in an a factor of 2 improvement in abundances precision when compared to individual phase constraints. With a 1D retrieval model on simulated HST+Spitzer data, we find strongly biased molecular abundances for CH{sub 4} and CO{sub 2} at most orbital phases. With 2D, the day and night profiles retrieved from WASP-43b remain consistent throughout the orbit. JWST retrievals show that a 2D model is strongly favored at all orbital phases. Based on our new 2D retrieval implementation, we provide recommendations on when 1D models are appropriate and when more complex phase geometries involving multiple TP profiles are required to obtain an unbiased view of tidally locked planetary atmospheres.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Quasi-Classical Trajectory Calculation of Rate Constants Using an Ab Initio Trained Machine Learning Model (aML-MD) with Multifidelity Data

Machine learning (ML) provides a great opportunity for the construction of models with improved accuracy in classical molecular dynamics (MD). However, the accuracy of a ML trained model is limited by the quality and quantity of the training data. Generating large sets of accurate ab initio training data can require significant computational resources. Furthermore, inconsistent or incompatible data with different accuracies obtained using different methods may lead to biased or unreliable ML models that do not accurately represent the underlying physics. Recently, transfer learning showed its potential for avoiding these problems as well as for improving the accuracy, efficiency, and generalization of ML models using multifidelity data. In this work, ab initio trained ML-based MD (aML-MD) models are developed through transfer learning using DFT and multireference data from multiple sources with varying accuracy within the Deep Potential MD framework. Further, the accuracy of the force field is demonstrated by calculating rate constants for the H + HO 2 → H 2 + 3 O 2 reaction using quasi-classical trajectories. We show that the aML-MD model with transfer learning can accurately predict the rate constants while reducing the computational cost by more than five times compared to the use of more expensive quantum chemistry training data sets. Hence, the aML-MD model with transfer learning shows great potential in using multifidelity data to reduce the computational cost involved in generating the training set for these potentials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Liquid–Vapor Phase Equilibrium in Molten Aluminum Chloride (AlCl 3 ) Enabled by Machine Learning Interatomic Potentials

Molten salts are promising candidates in numerous clean energy applications, where knowledge of thermophysical properties and vapor pressure across their operating temperature ranges is critical for safe operations. Due to challenges in evaluating these properties using experimental methods, fast and scalable molecular simulations are essential to complement the experimental data. In this study, we developed machine learning interatomic potentials (MLIP) to study the AlCl 3 molten salt across varied thermodynamic conditions (T = 473–613 K and P = 2.7–23.4 bar), which allowed us to predict temperature-surface tension correlations and liquid–vapor phase diagram from direct simulations of two-phase coexistence in this molten salt. Two MLIP architectures, a Kernel-based potential and neural network interatomic potential (NNIP), were considered to benchmark their performance for AlCl 3 molten salt using experimental structure and density values. The NNIP potential employed in two-phase equilibrium simulations yields the critical temperature and critical density of AlCl 3 that are within 10 K (∼3%) and 0.03 g/cm 3 (∼7%) of the reported experimental values. An accurate correlation between temperature and viscosities is obtained as well. In doing so, we report that the inclusion of low-density configurations in their training is critical to more accurately represent the AlCl 3 system across a wide phase-space. The MLIP trained using PBE-D3 functional in the ab initio molecular dynamics (AIMD) simulations (120 atoms) also showed close agreement with experimentally determined molten salt structure comprising Al 2 Cl 6 dimers, as validated using Raman spectra and neutron structure factor. Furthermore, the PBE-D3 as well as its trained MLIP showed better liquid density and temperature correlation for AlCl 3 system when compared to several other density functionals explored in this work. Overall, the demonstrated approach to predict temperature correlations for liquid and vapor densities in this study can be employed to screen nuclear reactors-relevant compositions, helping to mitigate safety concerns.

Ab initio molecular dynamics↗
DOE OSTI2024

Elucidating the formation mechanisms of zeolites using data-driven modelling and in-situ characterization

Zeolites are the main solid catalysts used by the chemical industry. The use of zeolites in separations and as shape selective catalysts requires control of the width and connectivity of their pores. 235 distinct zeolite frameworks have been synthesized to date, of over 2 million that have been proposed. Recent work indicates that the limitation is in large part kinetic: new synthetic pathways are required to access new zeolites. Organic cations are used to direct the synthesis towards specific zeolites. However, the molecular mechanisms by which cations direct the nucleation towards specific zeolites is not known. Elucidating these mechanisms is key to realize new zeolites for catalysis and separations, and is the focus of this project. This project developed and implemented a synergistic, data-driven computational and experimental approach to resolve the molecular pathways of nucleation, growth, and polymorph selection of zeolites and the role of organic cations in directing their formation. The project developed computationally efficient and accurate models for the study of the nucleation and growth of pure silica zeolites in molecular simulations, using machine learning with data from experiments. Simulations with these models were integrated with scanning tunneling electron microscopy, computer vision, and deep learning to unveil the molecular pathways of formation of a zeolite. Of particular interest in this project was to elucidate the role of amorphous precursors in the nucleation of the zeolite. Previous experiments indicate that zeolites are born within non-crystalline aggregates in which the silicates and organic cations have local and medium range order similar to that of the zeolite. The organic cations that direct the formation of zeolites and those that direct the formation of ordered mesoporous silicas are similar. We hypothesized that the frustrated attraction that for large organic cations leads to the formation of stable mesophases that direct the synthesis of mesoporous silicas, could promote the formation of metastable mesophases that can assist in the nucleation and polymorph selection of zeolites. The simulations resolved how structure directing agents build crystalline order and showed that mesoscopic pre-ordering occurs is not required to facilitate the nucleation of zeolites, because the synthesis occurs at high driving forces, where the barriers for nucleation are negligible. This project unveiled that polymorph selection in zeolite synthesis occurs after nucleation, opening a distinct area of control through the kinetics of growth and not through nucleation barriers.

36 MATERIALS SCIENCE↗