Non-Equilibrium Semiconductors

In the world of materials, semiconductors occupy a unique position, forming the bedrock of modern electronics. In their most tranquil state, known as thermal equilibrium, their behavior is elegantly described by a single parameter: the Fermi level. However, most of the fascinating and useful things semiconductors do—from emitting light in an LED to generating power in a solar cell—happen precisely when we force them out of this placid state. This raises a critical question: how do we understand and describe a semiconductor that is actively being energized? The rules of equilibrium are no longer sufficient.

This article bridges that knowledge gap by introducing the powerful physics of non-equilibrium semiconductors. It provides a comprehensive guide to understanding what happens when a semiconductor is disturbed by an external energy source like light or a voltage. The reader will journey from the balanced world of equilibrium to the dynamic realm of the non-equilibrium steady state. First, in the "Principles and Mechanisms" chapter, we will deconstruct the concept of thermal equilibrium and detailed balance, then see how external excitation breaks this balance, necessitating the crucial concept of quasi-Fermi levels. Following this, the "Applications and Interdisciplinary Connections" chapter will demonstrate how this theoretical framework is not just an abstraction but the very engine driving optoelectronics, thermoelectrics, and even cutting-edge research in plasmonics and ultrafast physics.

Imagine a semiconductor as a bustling city of charge carriers—nimble electrons and their counterparts, holes. In a world of perfect tranquility, kept in a dark room at a constant temperature, this city is in a state of profound balance. This is the world of ​​thermal equilibrium​​, a state not of stillness, but of dynamic stability where every microscopic event is perfectly counteracted by its reverse. This is the principle of ​​detailed balance​​. In this placid state, the entire population of electrons and holes, no matter their location or energy, is governed by a single, unifying parameter: the ​​Fermi level​​, denoted as $E_F$.

Think of the Fermi level as the universal "sea level" for electrons in the material. It's a constant energy value that permeates the entire crystal. The probability of finding an electron in a state with energy $E$ is dictated by how high that state is relative to this sea level. States far above $E_F$ are mostly empty, while states far below are mostly full. This single parameter, $E_F$, elegantly determines the concentration of both electrons ($n$) in the high-energy "conduction band" and holes ($p$) in the lower-energy "valence band."

In this equilibrium world, a beautiful and powerful treaty governs the relationship between electrons and holes: the ​​law of mass action​​. It states that the product of their concentrations is a constant, $np = n_i^2$. Here, $n_i$ is the ​​intrinsic carrier concentration​​, a fundamental property of the semiconductor material that depends only on its band structure and temperature. This law is not an independent axiom but a direct consequence of detailed balance. At equilibrium, the rate at which electron-hole pairs are thermally generated is exactly matched by the rate at which they recombine.

It's crucial to understand that this thermodynamic law works hand-in-hand with another fundamental principle: ​​charge neutrality​​. While the law of mass action fixes the product $np$, charge neutrality dictates that the total positive charge (from holes and ionized donor atoms) must balance the total negative charge (from electrons and ionized acceptor atoms). Together, these two independent rules—one from thermodynamics, one from electrostatics—uniquely determine the individual values of $n$ and $p$, and thus the position of the Fermi level for a given semiconductor at equilibrium.

Now, let's disrupt this peaceful equilibrium. Let's shine a light on our semiconductor. If the photons in the light have enough energy (more than the semiconductor's band gap, $E_g$), they can be absorbed, kicking an electron from the valence band up to the conduction band, leaving a hole behind. We are actively creating new electron-hole pairs.

This external pumping of carriers breaks the delicate dance of detailed balance. The generation rate of pairs is now the sum of the old thermal generation rate plus this new optical generation rate. To find a new balance, the system must increase its recombination rate to match this higher total generation. The system settles into a ​​non-equilibrium steady state​​, where the total number of carriers is constant, but the underlying forward and reverse processes are no longer individually balanced. The city is no longer in a state of quiet commerce; it's now a city during a festival, with a constant influx of new arrivals.

How do we describe this new, more energetic state? The single Fermi level, the universal sea level of equilibrium, is no longer sufficient. The electron and hole populations have been "inflated" and are no longer in equilibrium with each other.

However, a wonderful simplification occurs. While electrons and holes are not in equilibrium with each other, the electrons in the conduction band collide among themselves and with the crystal lattice so frequently that they quickly establish a state of internal equilibrium. The same is true for the holes in the valence band. It's as if the electrons and holes have divorced and now live as two separate families, each maintaining its own internal household rules.

Each of these internally thermalized populations can be described by its own chemical potential, its own "sea level." We call these the ​​quasi-Fermi levels​​: an electron quasi-Fermi level, $E_{Fn}$, and a hole quasi-Fermi level, $E_{Fp}$. The electron concentration $n$ is now determined by the position of $E_{Fn}$ relative to the conduction band, and the hole concentration $p$ is determined by the position of $E_{Fp}$ relative to the valence band. The system is still at a single lattice temperature $T$, but it is described by two distinct chemical potentials.

What happens to the law of mass action in this new regime? By writing down the expressions for $n$ and $p$ in terms of their respective quasi-Fermi levels, we can compute their product. The result is a simple and profound generalization of the old law:

This equation is the heart of non-equilibrium semiconductor physics. It tells us that the product $np$ is no longer a fixed constant. Instead, it is magnified above its equilibrium value, $n_i^2$, by a factor that depends exponentially on the ​​quasi-Fermi level splitting​​, $\Delta E = E_{Fn} - E_{Fp}$. This splitting is the ultimate measure of how far the system has been pushed from equilibrium. If the external drive is turned off, the excess carriers recombine, the splitting $\Delta E$ collapses to zero, and we gracefully return to the equilibrium law, $np = n_i^2$.

The effect is dramatic. At room temperature ($T = 300 \, \text{K}$), the thermal energy $k_B T$ is about $0.026 \, \text{eV}$. A modest quasi-Fermi level splitting of just $0.18 \, \text{eV}$ causes the exponential factor to be $\exp(0.18/0.026) \approx 1057$. The $np$ product is boosted by over a thousand times! If we drive the system harder to achieve a splitting of $0.538 \, \text{eV}$, the $np$ product explodes to be more than $10^9$ times its equilibrium value. This is not just a theoretical curiosity; we can measure this effect. The light emitted by a semiconductor, called photoluminescence, is proportional to the recombination rate, which in turn is proportional to the $np$ product. By measuring the increase in brightness under illumination, we can directly calculate the quasi-Fermi level splitting.

Furthermore, the quantity $(np - n_i^2)$ acts as the thermodynamic "driving force" for net recombination. Any process, whether radiative or non-radiative, will have a net rate proportional to this difference. A positive splitting means $np > n_i^2$, driving net recombination that attempts to bring the system back to equilibrium.

The elegance of this framework truly shines when we consider situations that seem hopelessly complex. Imagine the region near the surface of a semiconductor, or within a p-n junction. Here, electric fields cause the energy bands to bend, meaning the band edge energies $E_C(z)$ and $E_V(z)$ change with position $z$. As a result, the individual concentrations of electrons, $n(z)$, and holes, $p(z)$, can vary by many orders of magnitude over just a few nanometers.

One might expect their product, $n(z)p(z)$, to be an equally complicated function of position. But if we make the reasonable assumption that within this active region, the quasi-Fermi levels $E_{Fn}$ and $E_{Fp}$ are nearly flat (constant with position), something magical happens. When we compute the product $n(z)p(z)$ using the generalized law of mass action, the position-dependent terms from the bending bands perfectly cancel out. We are left with an astonishingly simple result:

Even as $n(z)$ and $p(z)$ individually fluctuate wildly, their product remains steadfastly constant across the entire region. This powerful insight simplifies the analysis of nearly all semiconductor devices, revealing a beautiful unity in the underlying physics. It's a prime example of how a good physical concept can cut through apparent complexity to reveal an elegant, simple truth.

This entire theoretical structure is not merely an intellectual pursuit; it is the very foundation of modern optoelectronics.

A ​​Light-Emitting Diode (LED)​​ is a device engineered to exploit this principle. By applying a forward voltage to a p-n junction, we directly impose a large quasi-Fermi level splitting, $E_{Fn} - E_{Fp} \approx qV$, where $V$ is the applied voltage. This creates a massive $np$ product in the junction, leading to a furious rate of recombination. The device is designed so that this recombination is primarily radiative, releasing the excess energy as photons of light. The brightness of your LED screen is a direct manifestation of the non-equilibrium $np$ product.

A ​​solar cell​​ is the reverse. Sunlight creates the quasi-Fermi level splitting for us, free of charge. This splitting, $\Delta E = E_{Fn} - E_{Fp}$, appears as a measurable voltage across the device terminals—the open-circuit voltage. When we connect the solar cell to a circuit, we allow the "excited" population of carriers to flow out, do work (powering your calculator or your home), and in doing so, move back toward equilibrium.

The principles even extend to situations where the temperature itself is not uniform. In a thermoelectric generator, a temperature gradient $\nabla T$ can drive carriers and create a spatial variation in the quasi-Fermi levels, generating a voltage. A complete description requires us to use a local, temperature-dependent version of all our parameters, such as $n_i(T(x))$, showing the robustness and adaptability of the quasi-Fermi level concept. From lighting our world to powering it with the sun, the physics of non-equilibrium semiconductors is a testament to how pushing a system out of balance can lead to some of the most useful and beautiful phenomena in science.

We have spent some time understanding the rather abstract idea of a non-equilibrium semiconductor, where the placid, unified Fermi level of equilibrium splits into two distinct quasi-Fermi levels, one for electrons ($E_{Fn}$) and one for holes ($E_{Fp}$). Now, you might be tempted to think this is just a clever piece of theoretical bookkeeping, a physicist's trick for dealing with a messy situation. But nothing could be further from the truth. This schism in the Fermi sea, this energetic separation $E_{Fn} - E_{Fp}$, is not a subtle correction; it is the very engine that drives a vast swath of modern technology. To see how, we must leave the quiet world of equilibrium and venture into the dynamic realm where these principles come to life.

Perhaps the most direct and beautiful manifestations of non-equilibrium physics are in devices that interact with light. Let’s start with something you see every day: the Light-Emitting Diode, or LED. How does it work? We take a p-n junction, made from a special type of material called a direct bandgap semiconductor, and we apply a forward voltage. This applied voltage acts against the junction's natural built-in potential, lowering the energy barrier that normally keeps electrons on the n-side and holes on the p-side.

With the barrier lowered, the party starts. Electrons are injected from the n-side into the p-side, and holes are injected from the p-side into the n-side. The device is flooded with excess minority carriers, a classic non-equilibrium situation. It is here that the quasi-Fermi levels make their grand entrance. This injection of carriers is precisely what separates $E_{Fn}$ and $E_{Fp}$. The applied voltage provides the "pump" that pushes the electron population to a higher electrochemical potential than the hole population. Nature, abhorring such an imbalance, seeks to restore equilibrium. An injected electron finds itself in a sea of holes and, in a direct bandgap material, the most efficient way to relax is for the electron to fall from the conduction band back into an empty state in the valence band (a hole), releasing its excess energy. This energy, which is approximately equal to the material's bandgap $E_g$, is emitted as a photon—a particle of light!. The color of the light is determined by the bandgap of the semiconductor. A larger bandgap gives a higher energy photon, like blue or violet light, while a smaller bandgap gives red or infrared light. The brightness, in turn, depends directly on the rate of this recombination, which is governed by the concentration of injected carriers. A higher forward voltage leads to a larger split between the quasi-Fermi levels, a greater density of injected carriers, and thus, a brighter light.

Now, what if we run the process in reverse? Instead of supplying energy to get light, can we use light to get energy? Absolutely! This is the principle of the solar cell. When a photon with energy greater than the bandgap strikes a p-n junction, it can excite an electron from the valence band to the conduction band, creating an electron-hole pair. The built-in electric field of the junction then swoops in and separates this pair before they can recombine, pulling the electron to the n-side and the hole to the p-side. This process continuously pumps charge carriers, building up an excess population of electrons on the n-side and holes on the p-side.

This accumulation of charge is, once again, a non-equilibrium state. And what describes the energy of this state? The quasi-Fermi levels! The light has forced the electron quasi-Fermi level $E_{Fn}$ on the n-side to a higher energy and the hole quasi-Fermi level $E_{Fp}$ on the p-side to a lower energy. If we connect a voltmeter across the illuminated cell under open-circuit conditions (no current flowing), what we measure is precisely the potential corresponding to this energy separation, $qV_{oc} = E_{Fn} - E_{Fp}$. The photovoltage is the macroscopic manifestation of the microscopic, light-induced splitting of the Fermi level. It's a marvelous symmetry: in an LED, we use a voltage to split the Fermi levels and create light; in a solar cell, light splits the Fermi levels and creates a voltage.

The story, however, has more subtle and fascinating chapters. When a high-energy blue photon strikes a silicon solar cell, it has far more energy than the bandgap $E_g$ requires to create an electron-hole pair. What happens to this surplus energy, $E_{photon} - E_g$? It is given to the newly created electron and hole as kinetic energy. These carriers are "hot"—they are rocketing through the crystal lattice at high speed. Before we can even collect them to generate current, they collide with the atoms of the lattice, shedding their excess kinetic energy in tiny, discrete packets of vibrational energy called phonons. This process, known as thermalization, is incredibly fast, often occurring in less than a picosecond. The energy is lost as heat, warming up the solar cell but contributing nothing to the electrical output. This thermalization is one of the single biggest limits on the efficiency of conventional solar cells.

This leads to a tantalizing question: could we somehow collect these hot carriers before they cool down? Doing so would allow us to capture that excess energy and dramatically boost solar cell efficiency. This is the dream of "hot-carrier solar cells." The main obstacle is the incredible speed of cooling. To make such a device, we need to understand, and perhaps control, the cooling process.

The cooling rate depends critically on the material. In polar semiconductors (like gallium arsenide or metal-halide perovskites), hot electrons interact very strongly with a particular type of lattice vibration, the longitudinal optical (LO) phonon. This provides a very efficient cooling channel. In nonpolar covalent materials like silicon, the interactions are different and generally weaker. At very high light intensities, an even stranger thing can happen. The hot electrons can emit LO phonons so rapidly that the phonons themselves don't have time to decay and dissipate their energy. This creates a non-equilibrium population of "hot phonons," which can then be re-absorbed by the electrons, effectively slowing down the net cooling rate. This traffic jam of energy is known as the "hot-phonon bottleneck," a beautiful and complex dance between the non-equilibrium electron and phonon systems. Understanding these intricate details is at the forefront of materials physics, as scientists search for ways to keep carriers hot just a little bit longer.

The profound consequences of non-equilibrium states are not confined to optoelectronics. They stretch out to touch upon a surprising range of scientific fields.

Consider the field of thermoelectrics, which deals with converting heat directly into electricity. If you create a temperature gradient across a semiconductor, charge carriers will naturally diffuse from the hot end to the cold end, creating a voltage—the Seebeck effect. But something else is happening, too. The temperature gradient also creates a net flow of phonons, a "phonon wind" blowing from hot to cold. These phonons, though chargeless, carry momentum. Through momentum-conserving collisions, this phonon wind can literally drag the charge carriers along with it. Under open-circuit conditions, an extra electric field must build up to counteract this drag. This adds an extra component to the Seebeck voltage, an effect known as "phonon drag." To optimize a thermoelectric material, one seeks to maximize this effect while suppressing the lattice's ability to conduct heat. It is a subtle non-equilibrium interplay between electrons and the lattice, where a heat flow drives a momentum flow that, in turn, drives a charge flow.

The story gets even more exciting at the nanoscale, at the interface of materials science, chemistry, and nanophotonics. Imagine a tiny gold nanoparticle sitting on the surface of a semiconductor like titanium dioxide. When light of the right color shines on the nanoparticle, it can excite a collective oscillation of its free electrons, a phenomenon called a localized surface plasmon. This plasmon resonance acts as a powerful nanoscale antenna, concentrating the light energy. The plasmon decays incredibly quickly, not by emitting light, but by generating high-energy, non-equilibrium "hot" electrons within the metal itself. These hot electrons, born in the metal, can have enough energy to leap over the energy barrier (the Schottky barrier) at the metal-semiconductor interface and inject themselves into the semiconductor. Once in the semiconductor, they can be used to generate a current or, even more interestingly, to drive chemical reactions on the surface, a field known as plasmonic photocatalysis. Here, the non-equilibrium carriers are not even born in the semiconductor, but are fired into it from a neighboring metal antenna!

Finally, let's consider the sheer speed of these non-equilibrium processes. What happens if we hit a semiconductor with an extremely intense, but ultrashort, laser pulse? We can generate an enormous density of electron-hole pairs in a flash—on the order of femtoseconds. If the density of these free carriers becomes high enough, the semiconductor can momentarily behave like a metal. Its optical properties, such as its reflectivity, can change dramatically. As these carriers then rapidly recombine, the material reverts back to its semiconducting state. This gives us a way to create an "ultrafast optical switch," using one pulse of light to control the path of another, on timescales a million times faster than conventional electronics.

From the humble LED in your lamp to the grand challenge of next-generation energy conversion, the guiding principle is the same. By pushing a semiconductor away from its equilibrium slumber—with a voltage, with light, with heat, or with a combination thereof—we create a dynamic state of tension described by quasi-Fermi levels. And in the relentless drive of the system to resolve this tension, we find the power to generate light, harvest energy, and control the very properties of matter at the frontiers of science and technology.