Graphite Outshines Diamond
Researchers from Lawrence Livermore, the University of California at Davis, and George Washington University used a machine-learning model to simulate if carbon crystallizes as graphite or diamond to build from previous experiments. A paper published in Nature Communications on July 9, 2025, demonstrated that graphite can spontaneously form under conditions in which diamond is more stable and expected to form.
The extreme conditions required for carbon crystallization and the rapidity of the transition are difficult to achieve in a laboratory setting. The team trained their machine-learning model on density functional theory, a quantum mechanical model used in condensed matter physics. The results showed that at high pressures, diamond would form. However, at lower pressures, graphite formed even when conditions were still ideal for diamond formation. These unexpected results are explained by graphite’s metastability and lower interfacial free energy, which favor its formation over diamond.
Knowing what separates graphite from diamond and liquid molten carbon enables the modeling of planets and supports fusion ignition at Livermore’s National Ignition Facility (NIF). “In NIF experiments, diamond is commonly used as the surface of the target capsule. During the initial phase of implosion, the diamond is driven to ablate as evenly as possible,” says Livermore scientist Margaret Berrens. “Understanding carbon crystallization near the graphite–diamond–liquid triple point is essential to ensure the transition remains controlled rather than chaotic to obtain high-yield shots.”
Contact: Margaret Berrens (berrens1 [at] llnl.gov (berrens1[at]llnl[dot]gov)).
New Material Bends, Bounces, and Absorbs
Livermore scientists and their collaborators created a new class of programmable soft materials that can absorb impacts, stiffen or soften, and change shape depending on their architecture and environmental conditions. In a study published in Advanced Materials on June 23, 2025, Livermore and partners from Harvard University, the California Institute of Technology, Sandia National Laboratories, and Oregon State University 3D-printed polymers called liquid crystal elastomers (LCEs) into responsive lattice structures. When placed under stress, the lattices behaved differently from conventional materials whose properties are cemented during manufacturing, making them more adaptable.
During the 3D-printing process, researchers aligned the molecular structure of the LCEs as they were deposited to orient the molecules, enabling researchers to encode behaviors in how the material might shrink or expand in different conditions. To better test and track these behaviors, the team created computer models to test and track material behaviors.
Based on the initial findings of the LCE lattices’ durability and flexibility, the team is studying dynamic applications, such as body armor and biomedical devices that move with the body. “What excites me most is the unprecedented level of control we now have—from the molecular scale up to the macroscopic structure—enabling us to design materials that respond and adapt to their environment,” says Livermore engineer Rodrigo Telles. “This opens new possibilities for engineering materials with tunable mechanical properties.”
Contact: Elaine Lee (925) 422-4939 (lee1040 [at] llnl.gov (lee1040[at]llnl[dot]gov)).
Microscope Takes 3D Ghost Images of Nanoparticles
Livermore researchers developed the first 3D quantum ghost imaging microscope that can study microscopic environments with limited light. Ghost imaging uses entangled photons to reveal an object’s silhouette without directly viewing it, minimizing light exposure and reducing photo damage to sensitive samples. Before this research, published in Optica on July 20, 2025, quantum ghost imaging had been limited to two dimensions.
Researchers created the 3D image by separating photon pairs that were entangled or linked together in space and time. Taking different paths after hitting a mirror, the signal photon interacts with the sample, and the idler photon goes directly to a camera-like detector. Coordinates and an impact time for each pixel are associated with both the ghost image, created by idler photons, and the standard image, created by signal photons. Once photon pairs are reidentified based on the timestamp, their coordinates are assembled to form the 3D image.
This scientific advancement changes the scale at which 3D ghost imaging can take place. “This microscope is the first of its kind,” says Livermore scientist Ted Laurence. “We are achieving three spatial dimensions of information at the micrometer scale.” The team plans to apply this new technology to enable the high-speed tracking of the movements of cells in relation to each other.
Contact: Ted Laurence (925) 422-1788 (laurence2 [at] llnl.gov (laurence2[at]llnl[dot]gov)).
