Search

Recent advances in deep learning generative models (GMs) have created high capabilities in accessing and assessing complex high-dimensional data, allowing superior efficiency in navigating vast material configuration space in search of viable structures. Coupling such capabilities with physically significant data to construct trained models for materials discovery is crucial to moving this emerging field forward. Here, we present a universal GM for crystal structure prediction (CSP) via a conditional crystal diffusion variational autoencoder (Cond-CDVAE) approach, which is tailored to allow user-defined material and physical parameters such as composition and pressure. This model is trained on an expansive dataset containing over 670,000 local minimum structures, including a rich spectrum of high-pressure structures, along with ambient-pressure structures in Materials Project database. We demonstrate that the Cond-CDVAE model can generate physically plausible structures with high fidelity under diverse pressure conditions without necessitating local optimization, accurately predicting 59.3% of the 3,547 unseen ambient-pressure experimental structures within 800 structure samplings, with the accuracy rate climbing to 83.2% for structures comprising fewer than 20 atoms per unit cell. These results meet or exceed those achieved via conventional CSP methods based on global optimization. The present findings showcase substantial potential of GMs in the realm of CSP.

Lithium (Li) is a prototypical simple metal at ambient conditions, but exhibits remarkable changes in structural and electronic properties under compression. There has been intense debate about the structure of dense Li, and recent experiments offered fresh evidence for new yet undetermined crystalline phases near the enigmatic melting minimum region in the pressure-temperature phase diagram of Li. Here, we report on an extensive exploration of the energy landscape of Li using an advanced crystal structure search method combined with a machine learning approach, which greatly expands the scale of structure search, leading to the prediction of four complex Li crystal phases containing up to 192 atoms in the unit cell that are energetically competitive with known Li structures. These findings provide a viable solution to the newly observed yet unidentified crystalline phases of Li, and showcase the predictive power of the global structure search method for discovering complex crystal structures in conjunction with accurate machine-learning potentials., Comment: 6 pages, 4 figures

Crystal structure prediction has been a subject of topical interest, but remains a substantial challenge, especially for complex structures as it deals with the global minimization of the extremely rugged high-dimensional potential energy surface. In this manuscript, a symmetry-orientated divide-and-conquer scheme was proposed to construct a symmetry tree graph, where the entire search space is decomposed into a finite number of symmetry-dependent subspaces. An artificial intelligence-based symmetry selection strategy was subsequently devised to select the low-lying subspaces with high symmetries for global exploration and in-depth exploitation. Our approach can significantly simplify the problem of crystal structure prediction by avoiding exploration of the most complex P1 subspace on the entire search space and have the advantage of preserving the crystal symmetry during structure evolution, making it well suitable for predicting the complex crystal structures. The effectiveness of the method has been validated by successful prediction of the candidate structures of binary Lennard-Jones mixtures and high-pressure phase of ice, containing more than one hundred atoms in the simulation cell. The work, therefore, opens up an opportunity towards achieving the long-sought goal for crystal structure prediction of complex systems.

The geological conundrum of sulfur excess refers to the finding that predicted amounts of sulfur, in the form of SO2, discharged in volcanic eruptions much exceeds the sulfur available for degassing from the erupted magma. Exploring the source of the excess sulfur has been the subject of considerable interest. Here, from a systematic computational investigation of sulfur-oxygen compounds under pressure, a hitherto unknown S3O4 compound containing a mixture of sulfur oxidation states +II and +IV emerges and is predicted to be stabilized above a pressure of 79 GPa. We predict that S3O4 can be produced via multiple redox reactions involving subducted S-bearing minerals (e.g., sulfates and sulfides) at high pressure conditions relevant to the deep lower mantle, and conversely be decomposed into SO2 and S at shallow depths of Earth. Therefore, S3O4 can be considered as a key intermediate compound to promote the decomposition of sulfates to release SO2, which offers an alternative source of the excess sulfur released during explosive eruptions. These findings provide a possible resolution to the geological paradox of excess sulfur degassing and a viable mechanism for the understanding of S exchange between surface and the lower mantle for the deep sulfur cycle., Comment: 18 pages, 12 figures

While many Particle Swarm Optimization (PSO) algorithms only use fitness to assess the performance of particles, in this work, we adopt Surprisingly Popular Algorithm (SPA) as a complementary metric in addition to fitness. Consequently, particles that are not widely known also have the opportunity to be selected as the learning exemplars. In addition, we propose a Euclidean distance-based adaptive topology to cooperate with SPA, where each particle only connects to k number of particles with the shortest Euclidean distance during each iteration. We also introduce the adaptive topology into heterogeneous populations to better solve large-scale problems. Specifically, the exploration sub-population better preserves the diversity of the population while the exploitation sub-population achieves fast convergence. Therefore, large-scale problems can be solved in a collaborative manner to elevate the overall performance. To evaluate the performance of our method, we conduct extensive experiments on various optimization problems, including three benchmark suites and two real-world optimization problems. The results demonstrate that our Euclidean distance-based adaptive topology outperforms the other widely adopted topologies and further suggest that our method performs significantly better than state-of-the-art PSO variants on small, medium, and large-scale problems.