1.Physics Department, Hong Kong University of Science and Technology, Hong Kong, China 2.School of Physics and Technology, Wuhan University, Wuhan 430072, China
Fund Project:Project supported by the Hong Kong Research Grants Council, China (Grant Nos. 16304717, AoE/P-02/12)
Received Date:21 February 2020
Accepted Date:04 March 2020
Available Online:20 March 2020
Published Online:05 August 2020
Abstract:Transmission line is a common kind of one-dimensional waveguide. In addition to being widely used in engineering, the transmission lines can be used in proof-of-principle experiments in basic scientific research. For example, the wave equations governing the transmission line and quantum wire are equivalent, so transmission lines are widely used in the research of quantum graphs. The transmission line network equations are similar to the equations of zero-energy tight binding model, so the transmission line network can also be used to study some physical properties predicted by the theories based on tight binding model, and examples include Anderson localization, band dispersions, topological properties, etc. According to the transmission line network equations, we review some applications of transmission lines in the research fields mentioned above. We will discuss Anderson localization in one-, two-, and three-dimensional networks, the band structures of periodic and quasiperiodic networks, and the angular moment-dependent topological transport in transmission line network. We introduce the methods and results in detail to show the potential of transmission lines in basic scientific research. Keywords:transmission line/ tight binding model/ topological material/ Anderson localization
(4)式是一个关于各个节点的方程, 被称为网络方程或者耦合方程, 大部分关于传输线网络的计算也都基于该方程. 其方程形式和能量为零的紧束缚模型十分相似, 第一项可以看作是节点的本地能量, 第二项则可视为不同节点之间的耦合项. 首先考虑一个图1(a)中的开放网络. 该网络有两个节点A和B, 由两条长度l1的电缆连接而成. 另外在输入和输出线上分别增加了额外端点“in”和“out”, 分别距离A, B点l2长度. 对于节点A, B, 分别写下其网络方程: 图 1 传输线网络中透反射谱和能带的计算 (a) 有输入输出的开放性传输线网络; (b) 一维的周期性传输线网络; (c) 二维的周期性传输线网络; (d) 三维的周期性传输线网络; (e)?(g) 分别对应(b)?(d)中三种网络的能带图 Figure1. Transmission/reflection spectra and band structures in transmission line network: (a) The transmission line network with input and output (open boundary); (b) one-dimensional (1D) periodic transmission line network; (c) two-dimensional (2D) periodic transmission line network; (d) three-dimensional (3D) periodic transmission line network; (e)?(g) correspond to the band structures of the networks shown in (b)?(d) respectively.
3.用传输线网络研究安德森局域化安德森局域化是一个研究了几十年的经典问题, 其描述的现象是无序引入的多重散射会导致原本向远处传导的波被局域在一个范围里. 本节就对一维、二维、三维传输线网络中安德森局域化的一些研究工作进行回顾介绍[9,10]. 首先介绍一维网络的研究[9], 研究的对象如图2(a)所示. N个节点沿一个方向一字排开, 相邻节点之间用若干根电缆进行连接, 而无序性可以通过不同的电缆数目(配位数)和长度来引入. 当所有电缆取一样的长度l, 则(4)式变形为 图 2 一维和二维传输线网络中安德森局域化的研究[9] (a) 一维的随机传输线网络, 通过配位数和电缆长度的变化来引入随机性; (b) 二维的随机传输线网络, 通过在正方格子网络中随机移除结点之间的电缆来引入随机性 Figure2. Anderson localization in 1D and 2D transmission line networks[9]: (a) 1D random transmission line network, the randomness is introduced by changing the coordination numbers and cable lengths; (b) 2D random transmission line network, the randomness is introduced by removing the cables randomly in the square lattice network.
其中a, b为不依赖于p, $\theta $的系数. 数值结果偏离这个形式比较大, 这也许是因为精确度不够造成的. 当$p \to 1$, 局域长度$\xi $会变得非常大, 如果想得到更精确的结果, 所需计算的样品长度也需要随之增大, 会非常耗时, 所以文献[9]没有做进一步计算. 为了进一步确认电缆网络模型的普适性类别, Zhang和Sheng[9]计算了$\xi (M)/M$和$\xi /M$的关系曲线. 根据有限尺寸标度理论[25], 假如电缆网络模型和无序紧束缚模型同属一个普适性类别, 则$\xi (M)/M$是以$\xi /M$为单一变量的函数, 和其他任何参数都无关. 计算结果显示, 当M分别取值M = 32, 64和128时, 改变p和$\theta $的值得到基本重合的三条曲线, 证明网络模型和紧束缚安德森模型同属于一个普适性类别[9]. 在另一项研究中, Zhang等[10]在三维网络中实验观测到了安德森局域化, 其研究的三维网络见图3(a), 网络在z方向为AB层交替堆叠的结构, 相邻层之间重叠的节点直接用0.5 m单根电缆相连. A和B层的结构如图3(a)所示, 水平方向的直线代表单根0.5 m电缆, 垂直方向的直线代表两根长度分别为0.5和2 m电缆连接(如虚线框所示). 在该研究中, x和y方向分别有5个和6个节点, z方向共有ABABA五层, 所以样品尺寸为5 × 6 × 5. 第一层左下角的节点坐标标注为(x, y, z) = (1, 1, 1), 其他位置坐标依次类推. 首先, 他们数值计算并实验测量了该有序网络的透射频谱(见图3(b)左图), 信号输入点为(x, y, z) = (1, 4, 3), 输出点为(x, y, z) = (5, 3, 3), 这里计算中考虑进了真实电缆的损耗, 实验和模拟结果符合得很好. 从透射谱中可以看到一个41—60 MHz的带隙(通过能带计算, 将这个带隙更加精确地确定在了44—59 MHz). 在这个带隙中52 MHz的位置上有一个尖峰, 对应着表面态. 图 3 在三维传输线网络中实验观测安德森局域态[10] (a) 样品结构示意图, 样品为AB层交替堆叠的结构; (b) 左图为(a)所示的样品测量得到的透射频谱, 圆点虚线和实线分别为实验测量和模拟结果; 右图为引入缺陷后得到的透射频谱, 圆点虚线和实线分别为实验和模拟结果, 细虚线为不考虑传输线损耗的模拟结果; (c) 引入随机性后的透射频谱, 圆点虚线和实线分别为实验和模拟结果, 细虚线为不考虑损耗的模拟结果; (d) 在不同的样品尺寸下计算IPR来评测态的延展程度; (e) 当频率为46.66 MHz时的散射波函数强度场图; (f), (g) 分别为实验测量和模拟得到的安德森局域态的强度场图, (g)图中还给出了该随机样品的电缆连接情况 Figure3. Experimental observation of Anderson localization in 3D transmission line network[10]: (a) The structure of the sample (it is the AB alternatively stacked structure); (b) the left panel is the transmission spectrum measured for the sample in (a), the circles and solid curves are the measured and calculated results respectively; the right panel is the transmission spectrum after introducing a defect, the circles and solid curves are the measured and calculated results respectively, the dotted curve is the calculated result without dissipation; (c) the transmission spectrum after introducing randomness, where the circles and solid curves are the measured and calculated results, and the dotted curve is the calculated result without considering dissipation; (d) inverse participation ratios (IPRs) for different sample scales; (e) the field intensity pattern of the scattered wave function at 46.66 MHz; (f) and (g) are measured and calculated field intensity pattern of localized states respectively, (g) also shows the connections of the random network sample.
5.传输线网络的拓扑传输近些年来拓扑材料成为了研究热点, 而传输线网络可灵活连接的特点为研究拓扑材料提供了很好的研究平台. 通过将电缆连接成具有特定拓扑性质的网络结构, 就可以直接在这个网络中观测到拓扑传输. 这一节将介绍一项这方面的研究[16]. 首先介绍该研究中所用的紧束缚模型[16]. 图6(a)所示为具有轨道角动量涡旋的元原子构成的六角蜂巢晶格结构. 最简化的模型下面, 每个元原子可由三个离散化的节点表示(图6(b)), 这里要求: 节点上的波函数幅值相等, 彼此之间的相位差可取$2{\text{π}}/3$, 0, $ - 2{\text{π}}/3$, 分别对应m = 1, m = 0, m = –1三个轨道角动量. 在图6(b)中离散模型的基础上, Jiang等又通过图6(c)所示的模型来引入轨道角动量和波矢之间的耦合. 一个元原子中的三个节点被分别置放在三层中的同一个亚晶格A或者B上. 例如图中A1, A2, A3为一个位于A亚晶格上的元原子的三个节点. 图6(c)中蓝线表示了层内节点之间的耦合, 黄线表示了不同层节点之间的耦合, 并且第一层被要求以同样的连接方式连回第三层. 随后, 用紧束缚模型可以在$K( {0, 4{\text{π}}/3\sqrt 3 a})$和 $K' \left( {0, - 4{\text{π}}/3\sqrt 3 a} \right)$点展开得到有效哈密顿量: 图 6 具有拓扑性质的传输线网络[16] (a) 由具有轨道角动量涡旋的元原子构成六角蜂巢晶格结构; (b) 将(a)中的元原子离散化为三个节点; (c) 在(b)的基础上引入轨道角动量和波矢之间的耦合, 这里只画出了晶格中的一个六边形; (d), (e) 计算(c)所示的传输线网络能带, 分别对应m = 0和m = 1; (f), (g)分别是m = 1情况下沿着x和y方向的投影能带, 红色和蓝色分别代表位于两个相对边界的边界态 Figure6. Transmission line network possessing topological properties[16]: (a) A hexagonal ring formed by the meta-atoms which possess angular momentum; (b) discretizing the meta-atom into three nodes; (c) shows how to introduce the coupling between the angular momentum and wave vector to the model in (b). Here we only show a hexagon of the honeycomb lattice; (d), (e) calculated band structures of the model in (c) for the m = 0 and m = 1 sectors respectively; (f), (g) projected band structures along x and y directions for m = 1 modes, the red and blue curves represent the edge states at the opposite boundaries respectively.
其中$P = \displaystyle\sum\nolimits_{f \leqslant {f_{\rm{c}}}} {\left| {{u_f}} \right\rangle \left\langle {{u_f}} \right|} $是将截止频率${f_{\rm{c}}}$以下的所有本征态$\left| {{u_f}} \right\rangle $全部累加起来得到的投影算符, 而${P_{ij}} = \left\langle {{x_i}} \right.\left| P \right.\left| {{x_j}} \right\rangle $为实空间波函数在位置xi和xj之间的空间关系. 计算局域陈数前, 先选取一块计算区域, 并且将其分成A, B, C三个子区域, 分别对应图8(a)所示的红、蓝、绿三个区域. 最后(17)式的计算结果标记的是计算区域中心处的局域陈数. 需要注意这个计算区域的尺寸需要大小合适, 如果选取得过大或过小都会导致计算结果变为零. 图 8 有限样品中局域陈数的计算[16] (a) 计算局域陈数的方法; (b) 从左至右分别为能级图、局域陈数和截止频率的关系、局域陈数场图, 研究的样品尺寸如右图所示; (c) 和(b)图相同, 但所计算的样品尺寸和实验样品尺寸相同 Figure8. Calculation of local Chern number in the finite samples[16]: (a) The computation domain for local Chern number calculations; (b) from left to right: energy level, the relation between local Chern number and cutoff frequency, local Chern number patterns. The size of sample can be seen in the right panel; (c) the same as (b), but the sample size equals to the size in experiments.