Understanding H isotope adsorption and absorption of Al-alloys using modeling and experiments (LDRD: #165724)

Donald K. Ward, Xiaowang Zhou, Richard A. Karnesky, Robert Kolasinski, Michael S. Foster, Konrad Thürmer, Paul Chao, Ethan N. Epperly, Jonathan A. Zimmerman, Bryan M. Wong, Ryan B. Sills2015 · DOI

TL;DR

This paper presents a multiscale modeling framework, combining DFT, MD, and DD, to understand hydrogen isotope interactions with Al-Cu alloys. This work develops a novel Al-Cu-H interatomic potential and experimental validation, crucial for predicting the long-term performance of materials in hydrogen environments.

What Problem Does This Paper Address?

Hydrogen embrittlement is a critical issue for materials used in hydrogen storage and containment, particularly for stainless steels. While aluminum alloys offer better resistance due to low hydrogen solubility, their long-term performance can still be affected by hydrogen interactions with surfaces and internal defects like dislocations and precipitates. Predicting component lifetime requires a fundamental understanding of these interactions, which is challenging due to the vast time and length scales involved. This paper aims to bridge this gap by developing a multiscale modeling approach and validating it with experiments to provide insights into hydrogen behavior in Al-Cu alloys.

Methods

The study employs a hierarchical, multiscale modeling approach:

  • Density Functional Theory (DFT): Used to calculate fundamental properties like cohesive energies and atomic volumes, serving as a high-fidelity reference for developing interatomic potentials.
  • Bond-Order Potential (BOP): A novel, high-fidelity ternary BOP for the Al-Cu-H system was developed. This potential is parameterized using DFT data and validated through stringent molecular dynamics (MD) simulations, including vapor deposition and chemical reaction simulations.
  • Molecular Dynamics (MD): Used to simulate atomic-scale phenomena, including surface adsorption, growth simulations under irradiation, and calculating dislocation core energies and precipitate misfit strains.
  • Dislocation Dynamics (DD): A mesoscale model that uses inputs from atomistic simulations to study the collective behavior of dislocations. The authors developed a new formulation for DD that can handle arbitrarily shaped precipitates and improved collision detection algorithms.
  • Experimental Techniques:
    • Low Energy Ion Scattering (LEIS) and Direct Recoil Spectrometry (DRS): Used to study hydrogen adsorption on Al surfaces and determine binding configurations.
  • Thermal Desorption Spectroscopy (TDS): Employed to measure hydrogen trapping energies in Al-Cu alloys.
  • Electrical Conductivity and Differential Scanning Calorimetry (DSC): Used to track precipitate evolution and dissolution in Al-Cu alloys.

Datasets

  • DFT Calculations: Extensive DFT calculations were performed to generate a library of cohesive energies, geometric parameters, and defect energies for various Al, Cu, H, and their binary/ternary compound structures. This data served as the primary dataset for parameterizing the BOP.
  • Experimental Data:
    • Surface concentration data from LEIS/DRS experiments on Al surfaces.
  • TDS spectra of deuterium-charged Al-Cu alloys.
  • Electrical conductivity and DSC data for Al-Cu alloys.
  • Literature data for pure Al and Al-Cu alloys were used for comparison and validation.

Key Insights and Ideas

  • A Novel Al-Cu-H Potential: The development of a high-fidelity ternary BOP is a significant achievement, enabling accurate simulations of H interactions with Al-Cu systems across various scales.
  • Capturing Complex Chemistry: The BOP successfully captures challenging aspects like the H₂ → 2H dissociation reaction and phase separations involving H₂, which are often difficult for simpler potentials.
  • Dislocation Core Energy Anomalies: Simulations revealed that dislocation core energies in aluminum do not monotonically increase with character angle, suggesting a more complex behavior than previously assumed and necessitating a revised understanding.
  • Precipitate Influence on Hydrogen Trapping: The study identified a distinct precipitate trapping site in Al-Cu alloys with a binding energy of 18.4 kJ/mol, indicating that precipitates play a crucial role in hydrogen retention.
  • Multiscale Validation: The framework successfully links atomistic insights (DFT, MD) to mesoscale phenomena (DD) and experimental observations, providing a robust approach for studying material behavior.

Key Results

  • A new Al-Cu-H ternary bond-order potential was developed, which meets six critical criteria for accurately describing Al-Cu-H systems.
  • MD simulations confirmed the BOP's ability to predict low-energy structures and capture phase separation phenomena.
  • Atomistic simulations determined misfit strains and equilibrium configurations of dislocations interacting with GP-I zones, providing essential parameters for DD simulations.
  • LEIS/DRS experiments identified specific binding sites for hydrogen on Al surfaces, with a discrepancy noted between experimental and computational predictions for the (100) surface.
  • TDS experiments revealed a precipitate trapping peak in Al-Cu alloys with a binding energy of 18.4 kJ/mol, suggesting that these alloys retain more hydrogen than pure Al.
  • The developed DD formulation successfully incorporated arbitrarily shaped precipitates and improved collision detection, enabling large-scale simulations of precipitation-strengthened alloys.

Limitations and Future Work

  • Dislocation Core Energy Model: While a model was developed to describe the non-monotonic behavior of dislocation core energies, further analysis is needed for complete understanding.
  • Discrepancy in Surface Binding: A difference between experimental and computational predictions for hydrogen binding sites on the Al(100) surface remains unexplained and requires further investigation.
  • Precipitate Cutting Model: Differences between MD and DD results for precipitate cutting suggest the need for further refinement of the cutting model, potentially by adjusting the precipitate aspect ratio.
  • Future Work: The authors propose several avenues for future research, including:
  • MD simulations of surface diffusion on Al-Cu systems to validate models against experiments.
  • Expanding the study to include other alloy species present in AA2219.
  • Investigating hydrogen interaction with oxygen-covered surfaces using a combination of experimental techniques.
  • Benchmarking experimental results with tritium experiments.
  • Further development of the multiscale modeling framework to address long-term hydrogen exposure effects.

Summary updated

Abstract

Current austenitic stainless steel storage reservoirs for hydrogen isotopes (e.g. deuterium and tritium) have performance and operational life-limiting interactions (e.g. embrittlement) with H-isotopes. Aluminum alloys (e.g.AA2219), alternatively, have very low H-isotope solubilities, suggesting high resistance towards aging vulnerabilities. This report summarizes the work performed during the life of the Lab Directed Research and Development in the Nuclear Weapons investment area (165724), and provides invaluable modeling and experimental insights into the interactions of H isotopes with surfaces and bulk AlCu-alloys. The modeling work establishes and builds a multi-scale framework which includes: a density functional theory informed bond-order potential for classical molecular dynamics (MD), and subsequent use of MD simulations to inform defect level dislocation dynamics models. Furthermore, low energy ion scattering and thermal desorption spectroscopy experiments are performed to validate these models and add greater physical understanding to them.
BibTeX
@misc{ward2015understanding,
  author = {Donald K. Ward and Xiaowang Zhou and Richard A. Karnesky and Robert Kolasinski and Michael S. Foster and Konrad Thürmer and Paul Chao and Ethan N. Epperly and Jonathan A. Zimmerman and Bryan M. Wong and Ryan B. Sills},
  title = {{Understanding H isotope adsorption and absorption of Al-alloys using modeling and experiments (LDRD: \#165724)}},
  year = {2015},
  doi = {10.2172/1222665},
}

Loading related content...

Figure