Diffusive and hydrodynamic magnetotransport around a density perturbation in a two-dimensional electron gas

This paper theoretically demonstrates that in a two-dimensional electron gas under a strong magnetic field, a density perturbation with a power-law tail creates a large "no-go" region of exponentially suppressed current and a rotated Landauer resistivity dipole outside it, with these effects being further modulated by electron viscosity.

The Gist

Imagine a crowded dance floor where everyone is moving in a coordinated line, representing electrons flowing through a flat, two-dimensional material like graphene. Now, imagine someone suddenly drops a giant, invisible boulder in the middle of the floor. This boulder represents a "density perturbation"—an area where the crowd of electrons is thinner or missing entirely.

This paper explores what happens to the flow of electrons when they encounter this "boulder," but with a twist: a very strong magnetic field is turned on.

Here is the breakdown of their findings using simple analogies:

1. The Magnetic "Twist"

Without a magnetic field, if you threw a ball at a wall, it would bounce back or slide along it. But with a strong magnetic field, the electrons behave differently. They don't just bounce; they start to spiral.

Think of the electrons as dancers who, when a magnetic field is applied, are forced to spin in tight circles as they try to move forward. When they hit the "boulder" (the empty spot), they don't just stop; they get trapped in a swirling vortex around the obstacle.

2. The "No-Go" Zone

The most surprising discovery is the size of the empty area around the obstacle.

• The Expectation: You might think the electrons would only avoid the physical size of the boulder.
• The Reality: The electrons avoid a much larger area. The authors call this the "No-Go" radius.

Imagine the boulder is the size of a basketball, but the electrons act as if there is a massive, invisible force field the size of a swimming pool around it. Inside this pool, the current is almost completely blocked. The stronger the magnetic field gets, the bigger this invisible "No-Go" pool becomes.

3. The Shape of the Obstacle Matters

The paper looks at two types of "boulders":

• The Hard Wall: A sudden, sharp drop in electron density (like a cliff).
• The Gentle Slope: A gradual thinning out of electrons (like a hill that slowly fades away).

They found that if the slope is gentle (mathematically described by a "power-law tail"), the "No-Go" zone is even larger and the way the current spirals around it is different than if the wall were sharp. It's like how water flows differently around a smooth, rounded rock versus a jagged, sharp cliff.

4. The "Landauer Dipole" (The Wake)

When water flows around a rock in a river, it leaves a wake behind it. In this electron world, the "wake" is called the Landauer resistivity dipole.

• Without Magnetism: The wake points straight back, like a boat's wake.
• With Magnetism: The wake gets twisted. The authors found that the angle of this twist depends on how gentle or sharp the "boulder" is. If the density drops off gently, the wake twists at a specific, predictable angle that is different from the sharp-wall case.

5. The "Viscous" Effect (The Honey Analogy)

The paper also considers what happens if the electrons act more like a thick fluid (like honey) rather than individual particles. This happens when electrons bump into each other very frequently.

• The Result: If the fluid is thick enough (high viscosity), the "No-Go" zone grows much faster as you increase the magnetic field.
• The Scale: In this thick-fluid scenario, the size of the disturbance is set by something called the Gurzhi length. Think of this as a "reach" of the fluid's stickiness. The "No-Go" zone is tiny compared to this reach, but the reach itself is huge compared to the actual size of the obstacle.

Summary

In short, the authors used math to show that in a strong magnetic field, a small empty spot in a 2D electron gas acts like a giant, invisible magnet that repels current from a very large area. The current doesn't just go around it; it spirals in a complex pattern. The size of this repelled area and the angle of the spiral depend on how "smooth" the empty spot is and whether the electrons are flowing like individual particles or a thick, sticky fluid.

These findings help scientists interpret images taken by high-tech microscopes that try to "see" how electricity moves through materials like graphene, allowing them to understand the hidden rules of electron flow.

Technical Summary

1. Problem Statement

The paper investigates the behavior of electrical current flow around a local density inhomogeneity (specifically a circular depletion or accumulation region) in a two-dimensional electron gas (2DEG) subjected to a strong magnetic field ($B$).

• Context: Recent scanning tunneling potentiometry (STP) experiments in graphene have revealed that current does not simply flow around a density depletion (acting as an obstacle) but forms a twisted, spiral pattern. The current is suppressed in a region much larger than the physical size of the depletion.
• Gap in Knowledge: Previous theoretical models often assumed a "hard-wall" boundary condition (an abrupt step in electron density). However, realistic density perturbations (e.g., induced by local gating) often exhibit smooth, gradual transitions. The authors aim to understand how the power-law tail of a smooth density profile affects the current distribution, the size of the current-suppressed region, and the orientation of the resulting resistivity dipole.
• Regimes: The study covers both the diffusive regime (dominated by electron-impurity scattering) and the hydrodynamic regime (dominated by electron-electron scattering and viscosity).

2. Methodology

The authors employ a combination of analytical modeling, asymptotic matching, and numerical verification.

• Modeling the Perturbation: Instead of a step-function density, they model the electron density $n(r)$ as approaching the background density $n_0$ via a power-law tail:
  $$\frac{n(r)}{n_0} \simeq 1 - \left(\frac{a}{r}\right)^\beta \quad (\text{for depletion})$$
  where $a$ is the geometric size of the perturbation, $r$ is the radial distance, and $\beta > 2$ is the decay exponent.
• Diffusive Regime:
  • They solve the continuity equation for the electrochemical potential $\Phi$ using Ohm's law with position-dependent conductivity tensors ($\sigma_{xx}, \sigma_{xy}$).
  • The problem is mapped to an advection-diffusion equation where the "flow velocity" is proportional to the gradient of the Hall conductivity.
  • They derive exact solutions using Gauss hypergeometric functions and approximate solutions using the WKB method and modified Bessel functions.
• Hydrodynamic Regime:
  • They incorporate electron viscosity ($\nu$) into the transport equations, leading to a fourth-order partial differential equation for the stream function $\Psi$.
  • They introduce the Gurzhi length ($l_G = \sqrt{\nu \tau_{mr}}$), which characterizes the scale of viscous effects.
  • They use asymptotic matching to connect the near-field solution (near the "no-go" radius) with the far-field solution (Landauer dipole).

3. Key Contributions and Results

A. The "No-Go" Radius ($R_{no}$)

The most significant finding is the existence of a "no-go" region where current is exponentially suppressed, which is much larger than the geometric size $a$ of the perturbation.

• Diffusive Scaling: In the diffusive regime, the no-go radius scales as:
  $$R_{no} \sim a \left( \frac{1}{\alpha} \right)^{1/\beta} \propto B^{(1-\chi)/\beta}$$
  where $\alpha = \sigma_{xx}/\sigma_{xy} \ll 1$ is the inverse Hall angle. This implies that even a weak density depletion can repel current over a macroscopic distance that grows with the magnetic field.
• Hydrodynamic Scaling: In the viscous regime (where $l_G \gg a$), the scaling changes significantly. The no-go radius grows much faster with the magnetic field:
  $$R_{no} \propto B^{1/(\beta-2)}$$
  This indicates that viscosity dramatically enhances the repulsion of current from the inhomogeneity.

B. Spiral Current Patterns

Inside the no-go region, the current and electrochemical potential exhibit spiraling patterns.

• The equipotential lines and current streamlines twist as they approach and recede from the obstacle.
• The direction of the spiral depends on whether the density is depleted or accumulated and differs between the diffusive and hydrodynamic regimes.

C. The Landauer Resistivity Dipole

Outside the no-go region, the potential correction behaves as a dipole (decaying as $1/r$).

• Rotation Angle ($\theta_L$): The dipole is rotated relative to the average electric field.
  • Hard-wall limit ($\beta \to \infty$): $\theta_L \approx 2\theta_H$ (where $\theta_H$ is the Hall angle).
  • Smooth profile ($\beta$ finite): The rotation angle is reduced:
    $$\theta_L \approx 2\theta_H - \frac{\pi}{\beta}$$
  • Hydrodynamic case: The angle is slightly larger than $2\theta_H$ due to logarithmic corrections involving the Gurzhi length.
• Dipole Size ($R_L$):
  • In the diffusive regime, $R_L \sim R_{no}$.
  • In the hydrodynamic regime, $R_L$ is set by the Gurzhi length ($l_G$), which is typically much larger than $R_{no}$. Specifically, $R_L \approx l_G / \sqrt{\ln(l_G/R_{no})}$.

D. Density Accumulation

The authors also analyze density accumulation (positive perturbation). They find that the sign of the power-law correction reverses, causing the Landauer dipole to rotate in the opposite direction compared to the depletion case.

4. Significance and Implications

• Interpretation of Experiments: The results provide a theoretical framework for interpreting recent STP images in graphene and other 2D systems. The observation of a large "no-go" region and spiral currents can be used to distinguish between diffusive and hydrodynamic transport.
• Viscosity Measurement: Since the scaling of the no-go radius with the magnetic field differs between the diffusive ($B^{1/\beta}$) and hydrodynamic ($B^{1/(\beta-2)}$) regimes, experimentalists can track the growth of this radius to estimate electron viscosity.
• Beyond Hard-Wall Models: The paper demonstrates that the "hard-wall" approximation is insufficient for realistic density profiles. The power-law exponent $\beta$ of the density tail is a critical parameter that quantitatively alters the transport characteristics (rotation angle and suppression radius).
• Stokes Paradox Resolution: In the hydrodynamic limit, the authors show how the "Stokes paradox" (the breakdown of dipole solutions in 2D viscous flow) is resolved by the finite size of the no-go region and the presence of the magnetic field, leading to a well-defined dipole size determined by the Gurzhi length.

Conclusion

Parashar and Fogler demonstrate that in strong magnetic fields, the interplay between density inhomogeneities and transport mechanisms creates a macroscopic "shadow" where current is blocked. The size of this shadow and the orientation of the resulting resistivity dipole are highly sensitive to the smoothness of the density profile ($\beta$) and the electron viscosity, offering new diagnostic tools for characterizing 2D electron systems like graphene.