arXiv subjects

Ming-Yu Guo

Publications and source records attributed to Ming-Yu Guo.

3 recordsLinked to original sources

arXiv2026

Stoichiometric cluster learning for few-shot property prediction of multi-ionic integrated energetic materials

Multi-ionic materials pose a distinct representational challenge in machine learning-driven materials design. Different from single-molecule or composition-based materials, their properties arise from how charged building blocks aggregate into specific assemblies. Here, we show how pretrained machine-learned interatomic potentials (MLIPs) can bypass full crystal-structure prediction and support pre-synthesis screening from stoichiometric ionic clusters using multi-ionic integrated explosives (MIXs) as a synthesis-facing example. This strategy combines a stoichiometric ionic-cluster representation, which represents each candidate material by a non-periodic, stoichiometry-preserved formula-unit cluster, with multi-task fine-tuning (MT-FT), which adapts a pretrained atomistic backbone while retaining the energy--force objective as physical regularization for the sparse detonation-velocity labels. With the pretrained backbone regularized by MT-FT, this surrogate provides a cross-validated screen across only 25 structurally curated perovskite-type energetic materials (PEMs) with experimentally derived Kamlet--Jacobs (K--J) detonation velocities. Representation probes show that the learned descriptors implicitly retain site-aware ionic organization, density information, and coarse packing compatibility, implying why non-periodic clusters can remain predictive before full crystal structures are known. The surrogate extends known PEMs chemistry to three newly synthesized ABX$_4$ materials with both unseen ABX$_4$ stoichiometry and an unseen ethylenediammonium B-site cation, yielding three-point concordance with K--J reference velocities and a mean absolute error (MAE) of 92~m$\cdot$s$^{-1}$ without retraining. Together, these results establish stoichiometry-preserved cluster learning as a synthesis-facing screening strategy for data-scarce multi-ionic materials.

cond-mat.mtrl-sci↗

arXiv2025

Unraveling Antagonistic Collision-Controlled Reactivity in Energetic Molecular Perovskites with Deep Potential Molecular Dynamics

The precise regulation of chemical decompositions in energetic materials, whether towards rapid ignition or stable endurance, requires atomic-scale principles governing reactivity, which remain elusive yet. Herein, we resolve this challenge through deep potential molecular dynamics (DPMD) simulations, uncovering a universal collision-control principle in energetic molecular perovskites, $\text{(H}_{2}\text{dabco)B(ClO}_{4}\text{)}_{3}$, where $\text{H}_{2}\text{dabco}^{2+}$ = 1,4-diazabicyclo[2.2.2]octane-1,4-diium, B = $\text{Na}^{+}$, $\text{K}^{+}$, $\text{Rb}^{+}$, $\text{NH}_{4}^{+}$ for DAP-1, DAP-2, DAP-3 and DAP-4, respectively. Atomic-scale simulation with Arrhenius fitting for over 100-ps trajectories reveals that increasing B-site ionic radius ($\text{Na}^{+} < \text{K}^{+} < \text{Rb}^{+}$) simultaneously reduces both activation energy $E_{a}$, which enhances reactivity, and pre-exponential factor $A$ which suppresses collision probabilities for hydrogen transfer between site $X$ and site $A$, with sharply opposing kinetic consequences. This duality well explains the peak stability and insensitivity in $\text{K}^{+}$-based DAP-2, which optimally balance thermal endurance and collision dissipation. For ammonium-based DAP-4, though the radius of $\text{NH}_{4}^{+}$ is close to $\text{K}^{+}$, the reactive B-site cation triggers proton transfer that promotes $\text{C-H}$ bond rupture. By linking static cation radii to dynamic $E_{a}$-ln$(A)$ coupling, we rationalize non-monotonic decomposition temperatures ($\text{K}^{+} > \text{NH}_{4}^{+} > \text{Rb}^{+} > \text{Na}^{+}$) macroscopic stability and establishes cornerstones for universal energetic material design.

cond-mat.mtrl-sci↗

arXiv2025

The OpenLAM Challenges

Inspired by the success of Large Language Models (LLMs), the development of Large Atom Models (LAMs) has gained significant momentum in scientific computation. Since 2022, the Deep Potential team has been actively pretraining LAMs and launched the OpenLAM Initiative to develop an open-source foundation model spanning the periodic table. A core objective is establishing comprehensive benchmarks for reliable LAM evaluation, addressing limitations in existing datasets. As a first step, the LAM Crystal Philately competition has collected over 19.8 million valid structures, including 1 million on the OpenLAM convex hull, driving advancements in generative modeling and materials science applications.

cs.LG↗