论文标题

一维量子物质中的压壁纠缠

Squashed entanglement in one-dimensional quantum matter

论文作者

Maiellaro, Alfonso, Romeo, Francesco, Citro, Roberta, Illuminati, Fabrizio

论文摘要

挤压纠缠及其通用上限,即量子条件相互信息,是双方量子相关性的忠实度量。因此,它们对量子系统的细粒结构敏感。在此观察结果的基础上,我们在量子多体系统的边缘之间介绍了量子条件互信息的概念。我们表明,该数量表征了明确的一维拓扑绝缘子和超导体,与前者中的钟形纠缠相等,而后者的半铃态纠缠则反映了两个系统中边缘模式的不同统计数据。在存在障碍或局部扰动的情况下,边缘到边缘的量子条件互信息是可靠的,即使在存在相互作用的情况下,系统大小将系统大小成倍收敛到量化的拓扑不变性,并且在琐碎阶段消失。因此,我们猜想它与受对称性保护的拓扑系统的整个地面相图中的边缘到边缘的紧缩纠缠相吻合,我们提供了支持该主张的一些分析证据。通过将它们与纠缠负性进行比较,我们进一步迹象表明,量子条件相互信息和壁板纠缠提供了一维量子物质中非本地相关模式的非常准确的表征。

Squashed entanglement and its universal upper bound, the quantum conditional mutual information, are faithful measures of bipartite quantum correlations defined in terms of multipartitions. As such, they are sensitive to the fine-grain structure of quantum systems. Building on this observation, we introduce the concept of quantum conditional mutual information between the edges of quantum many-body systems. We show that this quantity characterizes unambiguously one-dimensional topological insulators and superconductors, being equal to Bell-state entanglement in the former and to half Bell-state entanglement in the latter, mirroring the different statistics of the edge modes in the two systems. The edge-to-edge quantum conditional mutual information is robust in the presence of disorder or local perturbations, converges exponentially with the system size to a quantized topological invariant, even in the presence of interactions, and vanishes in the trivial phase. We thus conjecture that it coincides with the edge-to-edge squashed entanglement in the entire ground-state phase diagram of symmetry-protected topological systems, and we provide some analytical evidence supporting the claim. By comparing them with the entanglement negativity, we collect further indications that the quantum conditional mutual information and the squashed entanglement provide a very accurate characterization of nonlocal correlation patterns in one-dimensional quantum matter.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源