🔬 mesoscale physics

Non-Hermitian impurity scattering in graphene: Boltzmann transport and thermoelectric response

This paper investigates charge and thermoelectric transport in monolayer graphene with dilute non-Hermitian impurities, demonstrating that while gain enhances carrier lifetime and thermoelectric performance, absorption suppresses them, thereby establishing non-Hermitian scattering as a tunable mechanism for controlling graphene's energy-dependent relaxation and thermoelectric response.

Original authors: Juan A. Cañas, Daniel A. Bonilla, A. Martín-Ruiz

Published 2026-09-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Juan A. Cañas, Daniel A. Bonilla, A. Martín-Ruiz

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a sheet of carbon atoms, just one layer thick, where electricity and heat flow with a speed and efficiency that seems almost too good to be true. This material, known as graphene, is not just a laboratory curiosity; it is a two-dimensional platform where tiny particles called electrons behave like massless waves, zipping through the material at a constant, high speed. Scientists have long been fascinated by how to control this flow, particularly for converting waste heat into electricity, a process known as thermoelectricity. To do this, researchers usually introduce tiny imperfections, or impurities, into the graphene. These impurities act like speed bumps, scattering the electrons and changing how they carry energy. In the standard view of physics, these interactions are perfectly balanced: an electron hits a bump and bounces off, but the total number of electrons remains the same, and energy is conserved.

However, a new study explores what happens when this balance is broken. The researchers investigated a scenario where the impurities do not just scatter electrons but actively exchange them with their surroundings. Some impurities act like tiny drains, absorbing electrons and removing them from the flow, while others act like pumps, injecting new electrons into the system. This concept, known as non-Hermitian physics, moves beyond the traditional rules of conservation to describe open systems where particles can appear or disappear. The question driving this research was simple yet profound: if you have a sheet of graphene filled with these active drains and pumps, how does the flow of electricity and heat change? The answer reveals a surprising way to tune the material's ability to turn heat into power.

The researchers, working from the National Autonomous University of Mexico, built a theoretical model to simulate this behavior. They imagined a single layer of graphene populated with a sparse, random distribution of circular impurities. Unlike ordinary obstacles, these impurities were modeled with a complex potential, meaning they had a real part that acted like a standard barrier and an imaginary part that described the gain or loss of particles. To solve the problem, they used a precise mathematical approach to calculate exactly how an electron wave would scatter off these circular regions. They found that the scattering process splits into two distinct effects: one that changes the direction of the electron's momentum, and another that changes the total number of electrons passing through.

To make sense of this in a real-world transport scenario, where a steady flow of electricity is required, the team introduced a crucial concept: an external reservoir. In their model, if the impurities are absorbing electrons, this external source must constantly inject new ones to keep the system stable. Conversely, if the impurities are pumping electrons in, the reservoir must act as a drain to prevent the system from overflowing. This setup allowed them to define a new, effective time scale for how long an electron travels before its path is disrupted. This time scale is governed by both the usual bouncing off obstacles and the new process of gaining or losing particles.

The results of their calculations showed a clear and distinct difference between the two types of impurities. When the impurities acted as absorbers, removing electrons from the flow, the overall ability of the graphene to conduct electricity and heat decreased. The electrons were simply being removed from the channel faster than they could travel. However, the opposite happened when the impurities acted as amplifiers, adding electrons to the flow. In this case, the effective lifetime of the carriers increased, leading to a boost in both electrical and thermal conductivity. The presence of these active sources made the material more efficient at moving charge and heat.

Perhaps the most significant finding concerned the material's ability to generate electricity from a temperature difference, a property measured by the Seebeck coefficient. The study found that while absorbing impurities reduced this thermoelectric power, the active, amplifying impurities significantly enhanced it. By increasing the lifetime of the electrons and altering how they respond to energy changes, the gain mechanism made the material much better at converting heat into voltage. This enhancement also improved the electronic figure of merit, a key metric for the efficiency of thermoelectric devices. The researchers noted that this behavior stands in contrast to other studies involving coherent quantum transport, where loss sometimes improved efficiency, highlighting that the rules change depending on whether the transport is a smooth, wave-like flow or a diffusive, scattering process.

The team also examined how these effects play out at different temperatures and energy levels. They found that the standard laws of physics, which usually hold true for metals at low temperatures, remained largely intact, with the ratio of heat to electrical conduction staying close to expected values. The non-Hermitian effects primarily acted as a fine-tuning mechanism, modifying the small corrections that appear at higher temperatures. Importantly, the model showed that these active impurities could resolve a theoretical problem that often arises in graphene physics near the point where the material has no charge carriers. In standard models, the conductivity can become infinite or undefined at this point, but the inclusion of particle loss or gain naturally stabilizes the system, keeping the conductivity finite and well-behaved.

Ultimately, this work suggests that non-Hermitian scattering provides a new, powerful tool for controlling the properties of graphene. By adjusting the balance between particle loss and gain, it is possible to tailor the material's response to heat and electricity without necessarily changing the size or density of the impurities. While the current study focuses on the electronic contribution and assumes a stable environment maintained by an external reservoir, the findings point toward a future where the thermoelectric performance of graphene could be significantly improved. The ability to use gain to boost efficiency offers a fresh perspective on how to design materials for energy conversion, moving beyond simple obstacles to active, dynamic control of the flow of energy.

Technical Summary: Non-Hermitian Impurity Scattering in Graphene

Problem Statement
The paper investigates charge and thermoelectric transport in monolayer graphene containing a dilute distribution of finite-range non-Hermitian scattering centers. While standard transport theory typically assumes Hermitian potentials where particle number is conserved, realistic defects or engineered environments can involve local particle exchange (gain or loss) with an external reservoir. The authors address how such non-Hermitian effects—modeled as complex circular potentials—modify the energy-dependent relaxation of carriers and, consequently, macroscopic transport coefficients like electrical conductivity, thermal conductivity, and the thermoelectric figure of merit. A key challenge is formulating a consistent semiclassical description where local particle non-conservation is balanced by an external reservoir to maintain a stationary state.

Methodology
The study employs a multi-step theoretical framework:

Microscopic Scattering: The authors solve the Dirac scattering problem exactly for a single circular complex potential $V(r) = (U + iW)\Theta(a-r)$ using a partial-wave approach. Here, $U$ represents the Hermitian potential depth, and $W$ represents the non-Hermitian strength ( $W<0$ for absorption/loss, $W>0$ for amplification/gain). This yields a non-unitary scattering matrix $S_{m_j}$ , from which they derive the transport cross-section ( $\sigma_{tr}$ ), scattering cross-section ( $\sigma_{sc}$ ), and absorption cross-section ( $\sigma_{abs}$ ).
Semiclassical Transport: To connect microscopic scattering to macroscopic transport, they formulate a modified Boltzmann equation. Recognizing that local loss/gain violates particle conservation, they introduce a phenomenological external reservoir that compensates for the net flux exchange. This leads to an effective relaxation time $\tau_{eff}$ governed by both elastic momentum scattering and non-Hermitian flux exchange: $1/\tau_{eff} = n_{imp}v_F (\sigma_{tr} + \sigma_{abs})$ .
Thermoelectric Coefficients: Using the full energy-dependent $\tau_{eff}(E)$ , they calculate the Onsager coefficients via exact numerical integration (avoiding the Sommerfeld expansion where it breaks down due to strong energy dependence). From these, they derive the DC electrical conductivity ( $\sigma_{DC}$ ), electronic thermal conductivity ( $\kappa_{el}$ ), Seebeck coefficient ( $S$ ), Lorenz ratio ( $L/L_0$ ), and the electronic thermoelectric figure of merit ( $ZT_{el}$ ).

Key Contributions

Exact Non-Unitary Scattering Formalism: The paper provides an exact analytical derivation of scattering cross-sections for Dirac fermions in the presence of complex potentials, explicitly separating elastic momentum redistribution from net carrier exchange.
Effective Relaxation Time Model: It proposes a kinetic model where the non-Hermitian flux exchange acts as an additional relaxation channel (for loss) or a negative scattering mechanism (for gain), modifying the effective carrier lifetime.
Resolution of Low-Energy Divergence: The study demonstrates that the inclusion of absorptive impurities ( $W<0$ ) naturally suppresses the unphysical divergence of the transport time and conductivity near the Dirac point, a known issue in standard semiclassical treatments of graphene with impurities.
Contrast with Coherent Transport: The work highlights a fundamental dichotomy between coherent transport (Landauer formalism) and diffusive transport (Boltzmann formalism) regarding non-Hermitian effects, specifically showing opposite trends for the thermoelectric figure of merit.

Results

Scattering Cross-Sections: In the absorptive regime ( $W<0$ ), the absorption cross-section is positive and dominates at low energies, drastically increasing the total scattering rate. In the gain regime ( $W>0$ ), $\sigma_{abs}$ is negative; if the gain is too strong at low energies, the total cross-section becomes negative, signaling a breakdown of the steady-state Boltzmann description.
Conductivities: Weak gain ( $W>0$ ) increases the effective carrier lifetime, thereby enhancing both electrical and electronic thermal conductivities. Conversely, absorption ( $W<0$ ) suppresses these conductivities.
Lorenz Ratio: The Lorenz ratio remains close to the Sommerfeld value ( $L_0$ ) at low temperatures. Non-Hermiticity primarily modifies the finite-temperature corrections rather than inducing a qualitative change in the transport regime.
Thermoelectric Response:
- Seebeck Coefficient: Gain significantly enhances the magnitude of the Seebeck coefficient, while loss suppresses it. This is attributed to the gain-induced steepening of the energy dependence of the relaxation time near the Fermi level.
- Figure of Merit ( $ZT_{el}$ ): Active gain increases the electronic thermoelectric figure of merit, whereas loss decreases it. The maximum calculated $ZT_{el}$ is approximately 0.1.
Comparison with Coherent Transport: The paper notes that these results contrast with recent findings for coherent transport through a single non-Hermitian barrier, where loss was found to enhance $ZT$ . This underscores that thermoelectric consequences in non-Hermitian systems depend critically on whether the physics is governed by coherent transmission or the energy dependence of a diffusive relaxation time.

Significance and Claims
The authors claim that non-Hermitian scattering provides an "additional mechanism" for controlling the energy dependence of carrier relaxation and the resulting thermoelectric response in graphene. They position their work as an extension of the standard impurity-scattering framework to situations where defects exchange particles with their environment.

The paper is modest in its scope, explicitly stating that:

The model is phenomenological; it does not self-consistently determine the microscopic mechanism of particle exchange or the specific conditions for reservoir stability.
The results are restricted to the weak non-Hermitian limit ( $|W| \ll U$ ), dilute disorder, and finite electron doping (away from charge neutrality).
The study considers only the electronic contribution to thermoelectric performance, excluding phonon thermal conductivity and the energetic cost of maintaining the external reservoir.
While the electronic $ZT$ is enhanced by gain, the total efficiency in real devices remains limited by lattice thermal conductivity, suggesting that non-Hermitian impurity engineering should be combined with other strategies (like phonon scattering reduction) for practical applications.

Ultimately, the paper establishes that complex, non-unitary scattering potentials offer a new degree of freedom for tuning graphene's thermoelectric properties, distinct from conventional control via impurity density or size.

Technical Summary: Non-Hermitian Impurity Scattering in Graphene

More like this