🔬 optics

Analytical Theory of Photon Tunneling and Near-Field Heat Transfer Between Dissimilar Materials

This paper derives a closed-form analytical framework for near-field radiative heat transfer between dissimilar materials, such as semiconductors and metals, by demonstrating that the dominant in-plane wave vector is an approximate average of those found in symmetric plasmonic-plasmonic and semiconductor-semiconductor cavities.

Original authors: Kartika N. Nimje, Mariano Pascale, Georgia T. Papadakis

Published 2026-07-14

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kartika N. Nimje, Mariano Pascale, Georgia T. Papadakis

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 two neighbors living in houses separated by a very thin, invisible fence. One neighbor is a "metal" type, full of free-roaming electrons that wiggle like a crowd at a concert (a plasmonic material). The other is a "semiconductor" type, like a solar panel or an LED, which only wakes up and absorbs energy if the music hits a specific high note (its bandgap).

Usually, when these two neighbors try to share heat through the fence, they do it by sending invisible waves that can't quite cross the gap unless they "tunnel" through. Scientists have long known that if both neighbors are the same type (metal-metal or semiconductor-semiconductor), they can predict exactly how much heat gets shared. But when they are different—like a metal talking to a semiconductor—it's been a messy puzzle with no simple formula to describe the conversation.

The Big Discovery
In this study, the researchers finally cracked the code. They derived a neat, closed-form mathematical recipe that explains exactly how heat tunnels between these mismatched neighbors. They found that the "dominant" way the heat travels isn't decided by just one side of the fence. Instead, it's like a team effort: the main path the heat takes is essentially the average of what would happen if the metal were talking to itself and what would happen if the semiconductor were talking to itself.

Think of it like a dance. If a metal dancer and a semiconductor dancer try to sync up, their best move isn't just the metal's move or the semiconductor's move; it's a perfect blend of both. The researchers showed that you can predict this "dance step" (the dominant wave vector) just by averaging the steps of two simpler, identical dance pairs.

The Rules of the Game
The paper is very clear about what doesn't work here. It argues against the old, simple idea that the heat transfer is just limited by the size of the gap (like saying the fence is so small that only waves of a certain size can fit). That old rule works for identical metal neighbors, but it fails completely for this metal-semiconductor mix. The paper explicitly states that the heat exchange is not a simple, broad flood; it is a spectrally narrow event. This means the heat only flows efficiently at very specific frequencies, and it depends heavily on how "lossy" (how much energy it wastes as heat) the materials are. If either material is perfect and loses no energy, the tunneling stops dead.

What They Actually Did
The team didn't just guess; they built a mathematical model based on the "Drude model" for the metal (which describes how electrons wiggle) and a "step-like" model for the semiconductor (which describes how it suddenly starts absorbing light at a certain energy). They tested their new formula against complex computer simulations (using something called fluctuational electrodynamics) for a specific setup: a 10-nanometer gap between Indium Tin Oxide (ITO) and Indium Arsenide (InAs).

The results were a match. Their simple formula predicted the heat flow and the specific "dance step" (the in-plane wave vector) almost perfectly compared to the heavy-duty computer simulations. They showed that by tweaking the metal's "plasma frequency" (how fast its electrons wiggle) or the semiconductor's absorption, you can tune exactly where the heat flows. For example, if you tune the metal so its resonance matches the semiconductor's "wake-up call" (the bandgap), the heat flows right where it's needed for solar cells. If you miss the tune, the heat flows to useless, higher energies.

The Bottom Line
This work doesn't claim to have solved every heat problem in the universe. It specifically provides a clear, analytical tool for understanding heat transfer between a metal and a semiconductor in a tiny gap. It proves that the behavior is a joint effort between the two materials, not just one or the other. This gives engineers a much better way to design things like super-efficient solar cells or cooling systems for tiny electronics, ensuring that the heat they exchange is exactly where and how they want it, rather than just hoping for the best.

Technical Summary: Analytical Theory of Photon Tunneling and Near-Field Heat Transfer Between Dissimilar Materials

Problem Statement
Near-field radiative heat transfer (NFRHT) allows heat exchange rates to exceed the blackbody limit through the coupling of evanescent waves and surface-polariton modes across nanoscale gaps. While this phenomenon underpins applications such as thermophotovoltaics (TPV), electroluminescent cooling, and thermal rectification, these systems typically involve asymmetric cavities composed of dissimilar materials—most notably a metallic (plasmonic) interface and a semiconducting interface. Despite the prevalence of this configuration, no closed-form analytical description exists for the near-field interaction between a semiconductor and a plasmonic metal. Existing analytical frameworks are largely restricted to symmetric systems (identical materials), while asymmetric cases generally require complex numerical integration over spatial and spectral frequencies.

Methodology
The authors derive a closed-form analytical expression for the photon-tunneling probability in an asymmetric cavity consisting of a semi-infinite plasmonic medium and a semi-infinite semiconductor separated by a vacuum gap of thickness $d$ .

Material Models: The plasmonic medium is modeled using the Drude model, characterized by plasma frequency ( $\omega_p$ ), optical loss ( $\gamma$ ), and background permittivity ( $\varepsilon_\infty$ ). The semiconductor is modeled as a dielectric with a step-like absorption profile, defined by a real permittivity ( $\varepsilon_r$ ) below the bandgap ( $\omega_g$ ) and a complex permittivity ( $\varepsilon_r + i\varepsilon_i$ ) above it.
Formalism: Building on the Polder–van Hove formalism, the authors focus on transverse-magnetic (TM) polarized evanescent modes, which dominate in local non-magnetic media. They apply a quasistatic approximation ( $k_0 d \ll 1$ and $\beta \gg 1$ ) to simplify the Fresnel reflection coefficients.
Derivation: The tunneling probability $\xi_{asy}$ is derived by expanding the denominator of the standard fluctuational-electrodynamics expression, $|1 - r_1 r_2 e^{-2\kappa_0 d}|^2$ , into three distinct terms representing Fabry–Pérot-like multiple reflections. This yields a closed-form expression for the spectral heat flux and the mode-resolved tunneling probability.

Key Contributions and Results

Closed-Form Tunneling Probability: The paper presents an explicit analytical formula (Eq. 5) for the photon-tunneling probability in an asymmetric plasmonic–semiconductor cavity. The numerator of this expression is proportional to the product of the metal's damping constant ( $\gamma$ ) and the semiconductor's absorption strength ( $\varepsilon_i$ ), confirming that both material losses are strictly required for heat exchange.
Decomposition of the Denominator: The denominator of the tunneling probability is decomposed into three terms: two exponentially weighted material-response terms and an interference term. This structure links material absorption directly to the background permittivity and plasma frequency.
Dominant In-Plane Wave Vector: A central result is the derivation of the dominant in-plane wave vector ( $\beta_{asy}$ ) that maximizes photon tunneling. The authors demonstrate that $\beta_{asy}$ is not determined solely by the plasmonic medium or a simple $1/d$ scaling. Instead, it is the arithmetic average of the dominant wave vectors of two symmetric reference systems: a plasmonic–plasmonic cavity and a semiconductor–semiconductor cavity (Eq. 11).
Validation: The analytical model is validated against full fluctuational-electrodynamics calculations (using a transfer-matrix method) for an Indium Tin Oxide (ITO)–vacuum–Indium Arsenide (InAs) system with a 10 nm gap. The analytical expression accurately reproduces the full numerical results in both frequency and wave vector domains, capturing the spectrally narrow and loss-dependent nature of the enhancement.
Parameter Sensitivity: Numerical sweeps illustrate how material parameters govern the heat transfer:
- Emitter Tuning: Adjusting the plasma frequency shifts the coupled resonance relative to the semiconductor bandgap. Optimal performance requires aligning the resonance with the bandgap to concentrate heat flux in the useful spectral region.
- Semiconductor Index: Varying the real part of the permittivity ( $\varepsilon_r$ ) alters the resonance frequency.
- Absorption: Increasing the imaginary part of the permittivity ( $\varepsilon_i$ ) strengthens the dissipative coupling (increasing flux) but broadens the resonance, thereby reducing spectral selectivity.

Significance and Claims
The authors state that the purpose of this work is not to replace full numerical computations but to provide a compact analytical framework that clarifies the specific roles of material properties—plasma frequency, optical loss, and semiconductor absorption—in near-field thermal exchange.

The primary significance lies in establishing that the near-field radiative exchange in asymmetric cavities is governed by the joint electromagnetic response of both interfaces, rather than the properties of a single material. This challenges the simplified view that heat transfer is dominated solely by the plasmonic emitter. Consequently, for applications like near-field TPV, the design objective is not merely to maximize total heat transfer, but to spectrally align the dominant tunneling contribution with the photovoltaic band edge. The derived framework allows for the prediction of spectral position, linewidth, and modal extent based on fundamental material parameters and gap size, offering a tool for optimizing asymmetric optoelectronic devices without resorting to complex numerical simulations.

More like this