论文标题
磁场中二维电子系统的弹道到流动力过渡和集体模式
Ballistic-to-hydrodynamic transition and collective modes for two-dimensional electron systems in magnetic field
论文作者
论文摘要
二维电子系统中粘性流体动力电子流的最新证明对现有运输理论的有效性提出了严重的问题,包括弹道模型,碰撞引起的无碰撞流体动力学。尽管自由2D电子的流体力学到马力跨界的传输理论尚未确定,但对于磁场中的电子而言,这是不正确的。在这项工作中,我们开发了一个可解析的模型,描述了从弹道传输到流体动力传输的过渡,并改变了磁场中电子电子碰撞的强度。在此模型中,我们找到了2D电子高频非本地电导率张量的表达式。它在外部字段$ω$,回旋频率$ω_c$的频率和E-E碰撞频率$τ^{ - 1} _ {ee} $之间有效。我们使用所获得的表达来研究水动力至焊接跨界时2D磁性模式的转化。在真正的流体动力学制度中,$ωτ_{ee} \ ll 1 $,2DES支持一种单个磁性模式,该模式在Cyclotron Harmonics下未分裂。在弹道制度中,$ωτ_{ee} \ gg 1 $,等离子体分散在回旋子谐波上形成伯恩斯坦模式。动力学方程(“无碰撞流体动力学”)的形式的长波长膨胀预测,等离子体分散在$ω\2Ω_c$下的首次分裂。尽管如此,这种扩展仍无法预测真正的磁浮盘分散体的零和负组速度部分,为此需要完整的动力学模型。
The recent demonstrations of viscous hydrodynamic electron flow in two-dimensional electron systems poses serious questions to the validity of existing transport theories, including the ballistic model, the collision-induced and collisionless hydrodynamics. While the theories of transport at hydrodynamic-to-ballistic crossover for free 2d electrons are well established, the same is not true for electrons in magnetic fields. In this work, we develop an analytically solvable model describing the transition from ballistic to hydrodynamic transport with changing the strength of electron-electron collisions in magnetic fields. Within this model, we find an expression for the high-frequency non-local conductivity tensor of 2d electrons. It is valid at arbitrary relation between frequency of external field $ω$, the cyclotron frequency $ω_c$, and the frequency of e-e collisions $τ^{-1}_{ee}$. We use the obtained expression to study the transformation of 2d magnetoplasmon modes at hydrodynamic-to-ballistic crossover. In the true hydrodynamic regime, $ωτ_{ee} \ll 1$, the 2DES supports a single magnetoplasmon mode that is not split at cyclotron harmonics. In the ballistic regime, $ωτ_{ee} \gg 1$, the plasmon dispersion develops splittings at cyclotron harmonics, forming the Bernstein modes. A formal long-wavelength expansion of kinetic equations ("collisionless hydrodynamics") predicts the first splitting of plasmon dispersion at $ω\approx 2ω_c$. Still, such expansion fails to predict the zero and negative group velocity sections of true magnetoplasmon dispersion, for which the full kinetic model is required.