Yufeng Xi | Searcharxiv

arXiv2025

Scalable and deterministic construction of moir\'e superlattice in 2D materials using stressor films

Moir\'e superlattice in two-dimensional (2D) materials provides a powerful platform to engineer emergent electronic states, yet the construction of moir\'e superlattices remains lab-scale, involving much trial and error and with little control. Here, we demonstrate the construction of a heterostrain-induced moir\'e superlattice in transition metal dichalcogenides using a scalable process that deterministically induces strain to 2D materials. By applying patterned thin-film stressors and probing the resulting structures with scanning transmission electron microscopy, we directly resolve the induced heterostrain, lattice deformations, and stacking variations that produce the moir\'e superlattice. We find that uniaxial and biaxial heterostrain give rise to distinct moir\'e patterns, including stripes and distorted hexagonal patterns. With this approach, we create in-plane polar distortions and thus in-plane polarization at the domain boundaries of the moir\'e superlattice in MoS$_2$. The deterministic and scalable construction of moir\'e patterns using a well-established scalable process opens opportunities to design new moir\'e geometries in 2D materials.

cond-mat.mtrl-sci↗

arXiv2025

OPENXRD: A Comprehensive Benchmark Framework for LLM/MLLM XRD Question Answering

We introduce OPENXRD, a comprehensive benchmarking framework for evaluating large language models (LLMs) and multimodal LLMs (MLLMs) in crystallography question answering. The framework measures context assimilation, or how models use fixed, domain-specific supporting information during inference. The framework includes 217 expert-curated X-ray diffraction (XRD) questions covering fundamental to advanced crystallographic concepts, each evaluated under closed-book (without context) and open-book (with context) conditions, where the latter includes concise reference passages generated by GPT-4.5 and refined by crystallography experts. We benchmark 74 state-of-the-art LLMs and MLLMs, including GPT-4, GPT-5, O-series, LLaVA, LLaMA, QWEN, Mistral, and Gemini families, to quantify how different architectures and scales assimilate external knowledge. Results show that mid-sized models (7B--70B parameters) gain the most from contextual materials, while very large models often show saturation or interference and the largest relative gains appear in small and mid-sized models. Expert-reviewed materials provide significantly higher improvements than AI-generated ones even when token counts are matched, confirming that content quality, not quantity, drives performance. OPENXRD offers a reproducible diagnostic benchmark for assessing reasoning, knowledge integration, and guidance sensitivity in scientific domains, and provides a foundation for future multimodal and retrieval-augmented crystallography systems.

cs.CL↗