Diffusion-based generative models with property guidance have emerged as a promising paradigm for inverse materials design by enabling the generation of crystalline materials with user-specified target properties. However, despite recent advances, the effectiveness of property guidance, its influence on crystallographic symmetry, and the physical viability of generated materials remain poorly understood. To address these questions, we develop a property-guided framework based on the lightweight diffusion model DiffCrysGen using parameter-efficient adapter fine-tuning and classifier-free guidance (CFG). The resulting framework enables efficient multi-property crystal generation while preserving the knowledge learned during unconditional pre-training. Using formation energy together with saturation magnetization and Vickers hardness as representative inverse-design tasks, we systematically investigate the influence of CFG across a broad range of guidance strengths. Increasing the guidance scale progressively steers the generated property distributions toward the prescribed targets while reducing the fraction of lowest-symmetry ($P1$) structures and increasing the proportion of higher-symmetry structures. To evaluate physical viability, generated structures are geometrically prescreened and subsequently validated using a machine-learning interatomic potential (MLIP)-based workflow comprising structural relaxation and thermodynamic, dynamical, and property-specific analyses. The framework identifies thermodynamically and dynamically stable magnetic and mechanically hard materials with overall success rates of 12.3\% and 3.9\%, respectively. These results establish property-guided DiffCrysGen as an efficient framework for inverse materials design while providing new insights into the role of classifier-free guidance in crystal generation.
A physics-guided generative AI framework for property-targeted porous media design, combining a property-aware variational autoencoder, a conditional latent diffusion model, and an independently trained differentiable structure-to-property surrogate is proposed.
Graph representations compactly encode polycrystalline microstructures while retaining grain topology and grain boundary information. We present a conditional graph diffusion framework for property-guided inverse design of dual-phase Ti-6Al-4V microstructures. An enhanced grain graph neural network (GNN) with grain boundary edge features, learnable node and edge embeddings, and multi-statistic pooling serves as a forward surrogate for stress prediction and candidate evaluation. The conditional diffusion model generates candidates through reverse diffusion under prescribed {\alpha}-phase volume fraction, elastic modulus, and yield-stress proxy targets. Across four target regimes and independently seeded starting sets, generated candidates consistently approach the prescribed properties, including a target outside the property envelope of the existing microstructures. Local crystallographic consistency is evaluated post-generation from deviations from the Burgers orientation relationship (BOR). BOR-aware ranking increases mean BOR consistency by up to 44.9% and 56.4% for the in- and out-of-envelope targets, respectively, while maintaining property alignment. Finite element validation of the five best candidates in each primary design case yields a maximum absolute relative error of 1.0% in their mean properties. In a representative benchmark, diffusion requires 32 candidate evaluations per input graph, compared with approximately 40,000 for random search and evolutionary optimization, and reduces runtime by approximately two orders of magnitude in the tested implementations. These results establish conditional grain-graph diffusion as an efficient framework for property-guided polycrystalline microstructure design.
Yuheng Zhou, Xiao Shang, Huicong Chen et al.· 0 citations
Polymer property prediction and inverse generative design targeting desired properties are two crucial tasks in machine learning-assisted polymer design. While the former has received considerable attention, there have been limited methods developed for the latter. Existing methods focus on single-property optimization in the generative process, whereas accurate prediction of macroscopic material behavior requires simultaneous control of multiple physical properties. In this paper, we provide a transformative framework for direct optimization of a large collection of polymer properties. We propose PolymerGPT, a decoder-based GPT model that incorporates up to 37 commonly used polymer properties into the generative process via learned conditioning prefixes. It also supports a scaffold condition that specifies a desired scaffold for predicted structures. Our experimental results demonstrate that PolymerGPT achieves exceptional performance for unconditional and conditional generation while maintaining high validity, uniqueness, and novelty. Conditioning on five key properties yields generated structures whose predicted values closely match all target properties simultaneously.
C. Pyle, A. Gadari, C. A. Figg et al.· 0 citations
ABSTRACT Recent advancements in artificial intelligence (AI)–based design strategies have expanded the ability to generate complex mechanical metamaterials across multiple length scales. However, achieving precise control of mechanical properties while preserving structural connectivity remains a major challenge, especially for functionally graded metamaterials with heterogeneous unit cell architectures. Here, a latent diffusion–based design framework is proposed for 3D graph metamaterials that enables stable generation and accurate inverse design in a discrete, topology‐aware latent space. By integrating vector‐quantized latent representations with a diffusion‐based generative process and mechanistic guidance, the framework effectively explores complex design spaces while steering generated structures toward target elastic properties. The proposed approach enables the generation of graph metamaterials ranging from repetitive lattices to functionally graded architectures with smooth mechanical transitions and robust connectivity. These results demonstrate that latent diffusion with mechanistic guidance provides a scalable alternative to conventional interpolation‐based or purely data‐driven generative models for mechanistically optimized metamaterial design.
Diffusion-based deep generative models have emerged as powerful tools for inverse materials design. Yet many existing approaches overlook essential chemical constraints, such as oxidation-state balance, which can lead to chemically invalid structures. Here we introduce 'crystal generator with valence-constrained design' (CrysVCD), a modular framework that integrates chemical rules directly into the generative process. CrysVCD first uses a transformer-based elemental language model to generate valence-balanced compositions, followed by a diffusion model to generate crystal structures. The valence constraint enables orders-of-magnitude more efficient chemical valence checking compared with pure data-driven approaches with post-screening. When fine-tuned on stability metrics, CrysVCD achieves 85% metastability (Ehull < 0.1 eV per atom) and 68% phonon stability. Moreover, CrysVCD supports conditional generation of functional materials, enabling discovery of candidates such as high thermal conductivity semiconductors and high dielectric constant (high-κ) materials. Designed as a general-purpose plugin, CrysVCD can be integrated into diverse generative pipelines to promote chemical validity, offering a reliable, scientifically grounded path for materials discovery.
: Designing two-dimensional anisotropic mechanical metamaterial unit cells from prescribed effective properties remains a challenging inverse problem, particularly when directional stiffness and material usage need to be controlled simultaneously. In this work, a data-driven conditional diffusion framework is developed for generating unit-cell structures with target effective elastic moduli and volume fractions. A structure–property database containing 57,000 binary unit-cell images is first established through a random target-property-driven inverse homogenization method. The effective elastic moduli in the x and y directions, together with the volume fraction, are used as conditional labels, denoted as ( E x , E y , V ). A conditional denoising diffusion probabilistic model (DDPM) with a U-Net denoising backbone, referred to as DDPM-UNet, is then trained to generate unit-cell layouts under prescribed property conditions. To improve the consistency between generated geometries and target properties, a convolutional neural network (CNN) surrogate model is introduced. The CNN provides a property-consistency loss during diffusion model fine-tuning and is also used as a fast evaluator for candidate screening during inference. Compared with DDPM-UNet, the DDPM-UNet-CNN framework produces more valid candidates under 5%, 10%, and 20% relative error thresholds. FEA-based re-homogenization further shows that the selected candidates generally approach the prescribed effective properties. The developed framework provides a practical generative route for identifying candidate anisotropic mechanical metamaterial unit cells under prescribed property conditions.
Hao Sun, Xiaohong Ding, Min Xiong et al.· Computers, Materials & C...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
