发表机构
Institute for Advanced Study in Physics, Zhejiang University; Department of Applied Physics, Stanford University; Google Quantum AI（浙江大学物理高等研究院; 斯坦福大学应用物理系; 谷歌量子人工智能）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出莫尔异质双层系统的扭转角无序模型，通过复制技巧发现其与费米子耦合非厄米无序的对偶模型存在关联，为研究扭转角无序提供新视角并揭示其与常规无序的本质差异。
AI 中文摘要
我们提出并研究了一种莫尔异质双层系统模型，其中扭转角仅在有限范围内保持均匀且定义明确，在更大距离处则发生随机形变。此时局部扭转角服从某一概率分布，该分布会惩罚围绕其均值的大幅波动。在复制技巧（replica trick）框架内，我们证明无序平均会产生副本间符号交替的非平庸关联，该关联仅在每个局域畴的边界附近及特定方向上达到最大。我们进一步表明，对于低能莫尔能带，在费米子与非厄米无序耦合的对偶模型中可产生完全相同的关联，该模型能动态规避安德森局域化。这种对偶性为研究莫尔系统中的扭转角无序提供了新视角，并揭示了扭转角无序与常规无序之间的若干本质差异。
英文摘要
We propose and study a model of moire heterobilayer systems where the twist angle remains uniform and well defined only within a finite range, but deforms randomly at larger distances. The local twist angle is then subject to a certain probability distribution that penalizes large fluctuations around its mean value. Within the framework of the replica trick, we show that disorder averaging produces a nontrivial correlation with alternating sign between replicas, maximized only close the boundary of each local domain and along certain directions. We show that for the low-energy moire bands, exactly the same correlation can be generated in a dual model where fermions are coupled to non-Hermitian disorder, which can evade Anderson localization dynamically. This duality provides a new perspective to investigate twist-angle disorder in moire systems, and unveils some of the essential differences between twist-angle disorder and conventional disorder.
CommentsComments welcome
