Mesoscopic current patterns and orbital magnetism induced by dc-voltages

纳米科学的很大一部分是关于电荷密度对外加电场的响应。在这个项目中，我们在原子尺度上分析了这种电荷响应的局部空间结构。我们的准备工作使用了我们自制的从头算输运模拟程序AITRANSS。该软件包允许计算功能化有机材料中的复杂电流模式，例如带有氢吸附剂的石墨烯薄片。这些电流模式显示出令人惊讶和强大的特征，例如本地环电流的幅度可以超过平均传输电流一个数量级。为了在从头算水平上更系统地研究这种影响，人们需要以自洽的方式考虑感应电场和磁场对电流流动的影响。因此，应该发展一种采用当前密度泛函理论(CDFT)的输运形式。这就是我们在这个项目中建议做的：在其主要研究线中，CDFT将在我们的AITRANSS包中实现。为了突出第一个定量估计，我们将忽略对Kohn-Sham规范势的交换关联修正。这相当于包含了电流诱导的洛伦兹力，但忽略了通过相互作用效应实现的(非平衡)重整化。我们展望了这种新颖的、基于CDFT的输运形式论在表面科学和介观物理的几个子领域中的丰硕应用。首先，我们将研究电流诱导的洛伦兹力是否可以变得足够大，以便它们可以用于分子磁体的可控开关。其次，我们将研究流线和旋涡在介观样品中对热产生的作用。第三，我们想要研究感应磁场的统计性质，例如，为了预测电流模式的局部性质如何进入全局(粗粒化)可观测。第四，研究非常规材料中的局部电流模式是非常有趣的，例如拓扑物质。事实上，拓扑绝缘体表现出反常的介电响应，一种量子化的双各向异性，这也应该在局部电流模式中表现出来。最后，我们提到，这项提案中的技术发展也应该与材料科学相关。也就是说，我们在这里建议计算的(与时间相关的)电场的磁响应也引起了最近在光子学和超材料领域中被认为非常重要的双各向异性。我们的研究有助于从从头算起计算这种双各向异性。

A very large part of the nano-sciences is concerned with the response of the charge density to externally applied electric fields. In this project we analyze the local spatial structure of this charge response on the atomistic scale. Our preparative work employs our homemade ab-initio transport simulation package AITRANSS. The package allows to calculate complicated current patterns in functionalized organic materials, such as graphene flakes with hydrogen adsorbants. These current patterns exhibit surprising and strong features such as local ring currents with an amplitude that can exceed the average transport current by an order of magnitude. In order to investigate such effects more systematically on an ab-initio level, one needs to take the effect of the induced electric and magnetic fields on the current flow into account in a self-consistent manner. Thus, a transport formalism should be developed that employs the current density functional theory (CDFT). This is what we propose to do in this project: In its main research line the CDFT is to be implemented in our package AITRANSS. To fascilitate a first quantitative estimate we will ignore exchange-correlation corrections to the Kohn-Sham gauge potential. This amounts to including current-induced Lorentz-forces but neglecting (non-equilibrium) renormalizations through interaction-effects. We envision fruitful applications of this novel, CDFT-based transport formalism in several subfields of surface sciences and mesoscopic physics. First, we will investigate whether the current induced Lorentz forces can become sufficiently sizable so that they may be used for controllable switching of molecular magnets. Second, we will investigate the role of streamlines and eddies for the heat generation in the mesoscopic sample. Third, we would like to study the statistical properties of induced magnetic fields, for instance, in order to predict how the local properties of current patterns enter global (coarse-graining) observables. Forth, it is highly interesting to investigate local current patterns in unconventional materials, such as topological matter. In fact, topological insulators exhibit an anomalous dielectric response, a quantized bianisotropy, that should manifest itself also in local current patterns. Finally, we mention that the technological developments made in this proposal should be relevant also in the materials sciences. Namely, the magnetic response to (time dependent) electric fields, which we propose to calculate here, also gives rise to those bianisotropies that have recently been recognized as very important in the field of photonics and meta-materials. Our research can help paving the way for the calculation of such bianisotropies from ab-initio.