Lawrence Livermore National Laboratory

Back to top

Back to top

Predicting Crystal Structure and Stability

Lawrence Livermore and Carnegie Mellon University have demonstrated that crystal structure prediction simulations could reduce the need for experimental, trial-and-error synthesis of energetic materials. Using machine learning to model ubiquitous polymorphism—how molecules pack together into a crystal structure—the researchers found that crystal properties could be predicted ab initio (without experimental input), an ability that could streamline compound selection and solid form development processes in energetic materials synthesis. Their work, featured in the August 16, 2023, edition of Crystal Growth and Design, examined the structures of commonly used energetic materials 2,6-diamino-3,5-dinitropyrazine-1-oxide (LLM-105), hexogen (RDX), and octogen (HMX).

The crystal structure prediction simulations use a physics-based algorithm that randomly generates crystal structures, optimizes them, and ranks them based on their predicted energy. A machine-learned model estimates the unit cell volume for the crystal based on its molecular structure. Then, the random structure generator provides a set of possible structures based on space group symmetries. The genetic algorithm GAtor optimizes these structures using evolutionary principles and feature selection to find the lowest energy structure.

This work has important applications for Livermore in the future. “The work impacts the Laboratory’s mission because the new method could be useful in the development of novel energetic materials,” says Brad Steele, a research lead. “We demonstrated that if we have knowledge of the molecular structure of an energetic material, we can predict its crystal structure and then a variety of other important material properties.”

Contact: Brad Steele (925) 422-9902 (

Scaling Solidification from Nanoseconds to Aeons

Researchers from Lawrence Livermore and Sandia national laboratories and the University of California at Davis have developed a scaling law to analyze the kinetics of high-pressure, rapid solidification of metastable liquids observed across a range of materials. Scaling laws illustrate the relationship between physical quantities that scale with one another over a defined interval, such as time or rate. The team analyzed experiments in which a dynamic-compression platform rapidly condensed liquids in nanoseconds—a feat under nonequilibrium conditions—and observed that solidification kinetics “renormalize” to match the quick timescale of these experiments. The researchers developed a scaling law to represent this compression-rate dependence across different experimental platforms, and their results are published in the September 7, 2023, edition of Physical Review Letters.

Previous theory, data on water’s solidification under dynamic compression, and experimentation with gallium informed the team’s finding. Deriving the scaling law with gallium and water, two very different materials, means it could apply to a broad range of materials with further refinement. “Understanding and controlling such properties is a necessary step toward enabling the long-term goal of engineering and synthesizing novel materials via high-pressure techniques,” says Livermore’s Jon Belof, a co-author of the paper.

Contact: Jon Belof (925) 424-3199 (

New Composite Improves Hydrogen Storage

Hydrogen is an attractive renewable energy carrier due to its high theoretical energy density and variety of sustainable production sources, but implementation challenges arise when storing and transporting it. A new approach stores hydrogen within a solid metal hydride based on lightweight magnesium or lithium. However, non-ideal thermodynamic properties and slow performance require high temperatures and hydrogen pressures for operation in the highest demand use cases. Lawrence Livermore scientists, alongside researchers at Lawrence Berkeley and Sandia national laboratories, discovered a way to improve light metal hydrides’ performance. Their research is featured on the cover of the September 22, 2023, issue of Advanced Materials Interfaces.

The team developed a composite based on magnesium and lithium metal amides, which were incorporated into a porous shell of reduced graphene oxide. Theoretical calculations indicated that a synergistic effect occurs when the composite is formed: The graphene oxide facilitates transfer of electrons to the surrounding lithium and hydrogen atoms at the three-way interface of the materials, leading to a weakening of the bonds that hydrogen forms with lithium and nitrogen. This result increases the rates of hydrogen regeneration and resorption, allowing for more efficient release of hydrogen gas from storage when needed. “Our result encourages future consideration of complex metal hydride materials to work as versatile hydrogen-storage media in high-energy-density applications,” said Livermore scientist Brandon Wood, one of the study’s authors.

Contact: Brandon Wood (925) 422-8391 (