Quasi-Fermi Levels

In the realm of semiconductor physics, the concept of the Fermi level provides an elegant description of a system in thermal equilibrium. However, nearly every active semiconductor device—from the transistor in a computer to the LED in a display—operates by being intentionally pushed away from this placid state. When a semiconductor is excited by light or an applied voltage, the equilibrium model fails, creating a knowledge gap that must be bridged to understand and engineer modern technology. The concept of quasi-Fermi levels provides this crucial bridge. This article introduces this powerful theoretical framework for analyzing semiconductors in non-equilibrium conditions. The following chapters will first explore the fundamental principles behind quasi-Fermi levels, detailing how they emerge from the single Fermi level of equilibrium and how they redefine the laws governing carrier concentrations and currents. Subsequently, we will journey through the vast landscape of their practical applications, seeing how quasi-Fermi levels are the key to understanding the operation of transistors, solar cells, lasers, and even chemical processes at semiconductor surfaces.

To understand the dance of electrons in the modern world of semiconductors, from the chips in our phones to the solar panels on our roofs, we must first appreciate the profound peace of equilibrium. Then, and only then, can we grasp the beautiful complexity that arises when that peace is broken.

Imagine a vast, closed marketplace teeming with countless electrons, all seeking the lowest available energy state. In the quiet of ​​thermal equilibrium​​, when the semiconductor is left to itself in the dark at a constant temperature, a remarkable simplicity emerges. The entire system, across all its available energy states in both the conduction and valence bands, settles on a single, uniform "price" for adding or removing an electron. This universal price, a concept born from the fundamental drive of nature to maximize entropy, is the ​​Fermi level​​, denoted as $E_F$.

This single, unwavering Fermi level acts like a universal chemical potential. It dictates the probability that any given state is occupied. A profound consequence of this unity is the celebrated ​​law of mass action​​. In the common scenario where carrier concentrations are not excessively high (the non-degenerate limit), the product of the electron concentration ($n$) and the hole concentration ($p$) is a constant, determined only by the material's properties and the temperature:

where $n_i$ is the intrinsic carrier concentration. This elegant relationship, $np=n_i^2$, is the mathematical signature of thermal equilibrium, a direct result of every electron in the system marching to the beat of a single drum—the Fermi level $E_F$.

What happens when we shatter this stillness? What if we shine a bright light on the semiconductor, or apply a voltage across it? We are pumping energy into the system, kicking electrons from the valence band up into the conduction band, creating electron-hole pairs. The equilibrium concentrations are disturbed. Now, the product $np$ is greater than $n_i^2$. The old law is broken. The system is in a ​​non-equilibrium steady state​​.

Has the elegant order dissolved into pure chaos? Not at all. A new, more subtle order emerges. The key insight lies in comparing timescales. Within the conduction band, electrons scatter off each other and the crystal lattice at an incredible rate, thermalizing amongst themselves in femtoseconds. Likewise, holes in the valence band rapidly reach an internal thermal equilibrium. However, the process of an electron from the conduction band falling back down to annihilate a hole in the valence band—​​recombination​​—is a comparatively slow process, often taking nanoseconds or even microseconds.

This vast difference in timescales allows us to picture the system as a "tale of two cities." The conduction band is one city, and the valence band is another. Commerce and interaction within each city are lightning-fast, so each city establishes its own stable, internal economy. But travel between the cities is slow and restricted.

Because each population is internally thermalized, each can be described by its own chemical potential, its own "local price." We call these the ​​quasi-Fermi levels​​. The population of electrons in the conduction band is described by the ​​electron quasi-Fermi level​​, $F_n$. The population of holes in the valence band is described by the ​​hole quasi-Fermi level​​, $F_p$. The single, unified Fermi level of equilibrium has split into two.

Under the common non-degenerate approximation, the carrier concentrations are now given by:

Here, $N_C$ and $N_V$ are the effective densities of states for the conduction and valence bands, $E_C$ and $E_V$ are the band edge energies, and $k_B T$ is the thermal energy. These simple exponential forms are valid as long as the carrier concentrations are not so high as to become degenerate, a condition that means $F_n$ and $F_p$ remain well within the bandgap, far from their respective band edges.

This splitting of the Fermi level, $F_n - F_p$, is not just a mathematical convenience; it is the very measure of the system's departure from equilibrium. By taking the product of the new expressions for $n$ and $p$, we discover a beautiful generalization of the law of mass action:

The amount by which the $np$ product exceeds its equilibrium value, $n_i^2$, is directly set by the energy separation of the quasi-Fermi levels. A larger split means the system is further from equilibrium, bustling with more excess electron-hole pairs.

This is where the concept truly comes alive.

• In a ​​solar cell​​, the incoming photons create a flood of excess carriers, causing a large $np$ product. This generates a splitting, $F_n - F_p$. This energy separation across the device's terminals is precisely what we measure with a voltmeter as the ​​open-circuit voltage​​ ($V_{oc}$)! We are, quite literally, measuring the energy of the shattered equilibrium: $qV_{oc} = F_n(\text{n-contact}) - F_p(\text{p-contact})$.

• In a ​​Light-Emitting Diode (LED)​​, we do the reverse. We apply a forward voltage $V_F$ across the p-n junction. This external voltage directly imposes a quasi-Fermi level splitting, such that $F_n - F_p \approx qV_F$ in the active region of the device. This huge splitting forces the $np$ product to be enormous, driving a massive rate of recombination. As the electrons and holes recombine, they release their energy as photons—the light of the LED.

The story becomes even more profound when we consider how these quasi-Fermi levels vary in space. In a device, there are electric fields that push and pull on charges (​​drift​​), and there are concentration differences that cause charges to spread out (​​diffusion​​). For decades, these were treated as two separate terms in the current equation.

The quasi-Fermi level concept unifies them. The true, fundamental driving force on a population of carriers is the gradient, or spatial slope, of its electrochemical potential—its quasi-Fermi level. The separate drift and diffusion terms magically combine into a single, exquisitely simple expression. For electrons, the current density $J_n$ is:

and similarly for holes. This powerful relation tells us that current flows in response to a slope in the quasi-Fermi level. If a quasi-Fermi level is flat ($dF_n/dx = 0$), there is no net current for that carrier type, regardless of what electric fields or concentration gradients might be present. The gradient of the quasi-Fermi level is the hidden engine driving all charge transport in this regime.

Like any powerful model, the quasi-Fermi level concept has its limits—boundaries where its underlying assumptions of local, internal thermalization break down.

• ​​The Crush of Degeneracy:​​ If we inject an immense number of carriers, the states at the bottom of the bands can fill up. This is ​​degeneracy​​. The simple Boltzmann approximations fail, and we must use the full, more complex Fermi-Dirac integrals to relate carrier densities to the quasi-Fermi levels. The simple $np$ product law no longer holds, and even the relationship between diffusion and mobility (the Einstein relation) requires modification.

• ​​The Blur of the Ultrafast and Ultrasmall:​​ The very idea of a quasi-Fermi level hinges on carriers having enough time and space to scatter and thermalize. If we probe a device with an ultrafast laser pulse that is shorter than the scattering time, or if we look at a region of a device that is smaller than the average distance between collisions (the mean free path), the concept breaks down. In these ​​ballistic​​ or near-ballistic regimes, the carrier distribution is no longer a simple thermal one, and the quasi-Fermi level loses its meaning. To describe these frontiers of physics, we must return to more fundamental tools like the Boltzmann Transport Equation.

From the serene stillness of equilibrium to the vibrant, light-emitting life of a biased diode, the concept of quasi-Fermi levels provides a powerful and elegant framework. It unifies disparate phenomena, connects microscopic statistics to macroscopic measurements, and beautifully illustrates how nature maintains order even in the midst of non-equilibrium chaos.

Now that we have acquainted ourselves with the machinery of quasi-Fermi levels, we can ask the most important question: What are they good for? Why did we go to the trouble of inventing these new "Fermi levels" for systems out of equilibrium? The answer, you will be delighted to find, is that they are not merely a theorist's abstraction. They are the key that unlocks the inner workings of nearly every semiconductor device that powers our modern world. From the glowing screen you're looking at to the solar panels powering our future, quasi-Fermi levels provide the language to describe, predict, and invent. Let us take a journey through this landscape of applications, from the heart of electronics to the frontiers of chemistry and materials science.

At its core, electronics is about controlling the flow of charge. And in the world of semiconductors, the quasi-Fermi levels are the puppet masters pulling the strings. A gradient in a quasi-Fermi level is the driving force for current. If the quasi-Fermi level for electrons, $F_n$, is flat, the electrons are content; they have no net urge to move. But tilt it, and an electron current begins to flow, as if sliding down an energy hill.

Consider the simplest, most fundamental electronic component: the p-n junction diode. In equilibrium, a single, flat Fermi level stretches across the entire device. But what happens when we apply a forward voltage, $V_F$? We are, in effect, pumping energy into the system. This energy splits the single Fermi level into two. The electron quasi-Fermi level, $F_n$, is pushed up on the n-side, and the hole quasi-Fermi level, $F_p$, is pulled down on the p-side. Across the junction, their separation becomes directly related to the applied voltage, such that $F_n - F_p \approx qV_F$. This splitting has a dramatic consequence: it lowers the potential energy barrier that was holding the carriers back. It’s like opening the floodgates. A torrent of electrons from the n-side pours into the p-side, and a flood of holes from the p-side pours into the n-side. This massive flow is the forward current of the diode.

This same principle of "injection" by splitting quasi-Fermi levels is the engine of the bipolar junction transistor (BJT). In a BJT, a small voltage applied to the emitter-base junction creates a split between $F_n$ and $F_p$ within the thin base region. This split represents the injection of a huge number of minority carriers—say, electrons into a p-type base. These injected electrons find themselves in a region with a sloping electron quasi-Fermi level, and so they diffuse across the base to be collected at the other end. The genius of the transistor is that a tiny change in the base voltage causes a large change in this injected current, giving us amplification.

Field-effect transistors (MOSFETs), the building blocks of every computer chip, use a slightly different trick, but the language of quasi-Fermi levels is just as crucial. A voltage on the gate electrode creates an inversion layer—a thin channel of minority carriers at the surface. In this state, if there are no contacts to the channel, the system can be in equilibrium, with a single flat Fermi level describing both the channel and the bulk substrate. But the moment we apply a voltage between the source and drain contacts at either end of this channel, we create a lateral slope in the electron quasi-Fermi level, $F_n$. This gradient is the direct cause of the current that flows through the channel, turning the transistor "on". A beautiful contrast arises: in a simple MOS capacitor without current, we have one flat Fermi level. In an operating MOSFET with current, we have two distinct quasi-Fermi levels, $F_n$ and $F_p$, with the former being tilted to drive the device's current.

For high-power applications, engineers have designed clever devices like the Insulated Gate Bipolar Transistor (IGBT), which combines the best of both worlds. In its "on" state, it doesn't just inject one type of carrier; it injects both electrons and holes into a lightly-doped drift region. This dual injection is so intense—a condition we call high-level injection—that the region becomes flooded with an electron-hole plasma. In the language of quasi-Fermi levels, this means $F_n$ is pushed all the way up near the conduction band, and $F_p$ is shoved all the way down near the valence band. The region behaves almost like a metal, its resistance plummeting. This "conductivity modulation" allows the IGBT to handle enormous currents with minimal wasted heat, a feat perfectly visualized by the dramatic separation of its quasi-Fermi levels.

The relationship between light and semiconductors is a two-way street, and quasi-Fermi levels pave the entire road. When light shines on a semiconductor, it can be absorbed; when carriers recombine, light can be emitted.

Let's start with absorption. When a photon with enough energy strikes a semiconductor, it can lift an electron from the valence band to the conduction band, creating a free electron and a free hole. This act of creation pushes the system out of equilibrium. Even if no current flows, the presence of these excess carriers forces the Fermi level to split. The separation, $F_n - F_p$, is a direct measure of the intensity of the illumination. You could say this energy difference is the "electrochemical potential of the light" that has been stored in the material.

Now, how do we get useful work out of this stored potential? This is the job of a solar cell, or photodiode. A solar cell is essentially a p-n junction designed to be exposed to light. The light creates electron-hole pairs, which are separated by the junction's built-in electric field, creating a current. If we leave the circuit open, no current can flow, but a voltage appears across the terminals. This is the open-circuit voltage, $V_{oc}$. What is this voltage, fundamentally? It is the difference in the quasi-Fermi levels at the electrical contacts: $qV_{oc} = F_n(\text{n-contact}) - F_p(\text{p-contact})$. This is the maximum voltage the solar cell can possibly produce.

But a deeper look reveals wonderful subtleties. If you measure the quasi-Fermi level splitting inside the solar cell, you'll find it's not constant. Why? Because even at open circuit, there are internal currents—photogenerated carriers diffusing one way, and recombination currents flowing the other—that must perfectly cancel out. These internal flows require gradients in the quasi-Fermi levels. And what if you just shine light on a uniform block of semiconductor? You get a split in the quasi-Fermi levels, but you measure no voltage! The perfect symmetry of the block means no net voltage can develop. To harvest the energy, you need asymmetry, like a p-n junction, to separate the carriers. Real-world imperfections, like tiny short-circuits called "shunts," can also play spoiler by providing a path for carriers to recombine without doing useful work, effectively decoupling the large internal quasi-Fermi level splitting from the smaller voltage you measure outside.

Now for the reverse process: creating light. This is the magic of the Light-Emitting Diode (LED) and the semiconductor laser. To get light out, we must put energy in. We do this by applying a strong forward bias to a p-n junction, injecting a massive number of electrons and holes into the same region. This creates a large quasi-Fermi level separation. For an LED, this is enough: as these excess electrons and holes find each other and recombine, they release their energy as photons.

To make a laser, we have to push even harder. We must inject carriers so furiously that the quasi-Fermi level separation becomes greater than the bandgap energy itself: $F_n - F_p > E_g$. This is the legendary Bernard-Duraffourd condition for population inversion. Under this extreme condition, the material contains so much stored energy that an emitted photon is more likely to stimulate another electron-hole pair to emit an identical, coherent photon than it is to be reabsorbed. This is the genesis of stimulated emission, the process that gives laser light its extraordinary properties. The carrier populations can become so dense that they behave more like a quantum liquid than a dilute gas, a state known as degeneracy, but the core principle remains: the quasi-Fermi level splitting is the thermodynamic engine that drives the laser.

The power of the quasi-Fermi level concept extends far beyond traditional electronics, providing a bridge to chemistry, surface science, and even the manufacturing process itself.

Imagine you want to study the surface of a material. You might use a technique like Kelvin Probe Force Microscopy (KPFM), which measures the work function—the energy needed to pull an electron out of the material. But what does it measure if the material is a semiconductor being illuminated? The light creates non-equilibrium carriers, so the "work function" is no longer a simple equilibrium property. What the probe actually measures is the position of the electron quasi-Fermi level at the very surface. Under illumination, two things happen: the photogenerated carriers tend to flatten the natural band bending at the surface (an effect called surface photovoltage), and the electron population itself increases, raising $F_n$. Both effects combine to change the measured work function. This shows that a quasi-Fermi level is not just a theoretical tool; it's a real, measurable physical quantity that governs how a material's surface interacts with the outside world.

This connection to surface properties takes us directly into the realm of chemistry. Consider a p-type semiconductor electrode dipped in a chemical solution. We might want to use this setup to drive a chemical oxidation reaction—for example, to split water to produce oxygen, a key step in artificial photosynthesis. The "oxidizing power" of the semiconductor is determined by the energy of its holes. And the energy of this population of holes is given precisely by the hole quasi-Fermi level, $F_p$. In the dark, the holes may not have enough energy (i.e., $F_p$ is not at a sufficiently positive potential) to drive the reaction. But when we shine light on the electrode, we create more holes. This pushes $F_p$ to a more positive potential, making the holes "stronger" oxidizing agents. Suddenly, they have the driving force needed to pull electrons from the molecules in the solution and make the reaction go. The light-induced shift in the quasi-Fermi level is the key that unlocks the chemistry.

Finally, in a surprising and beautiful twist, quasi-Fermi levels are not just for describing how a finished device works, but also for understanding how it is made. During the manufacturing of a silicon chip, a crucial step is annealing—heating the wafer to high temperatures to activate implanted dopant atoms and repair crystal damage. A common method, Rapid Thermal Anneal (RTA), uses intense lamps. This light doesn't just provide heat; it also generates a massive number of non-equilibrium electron-hole pairs, creating split quasi-Fermi levels within the silicon wafer as it is being processed. The diffusion of dopant atoms through the crystal lattice is often mediated by point defects, like silicon atoms that have been knocked out of their proper place. The charge state of these defects ($+1, -1, 0$) dramatically affects how they move and interact with dopants. And what governs their charge state? The balance of electron and hole capture and emission. In this non-equilibrium, high-temperature environment, this balance is controlled not by a single Fermi level, but by $F_n$ and $F_p$. To accurately simulate and control the formation of the ultra-precise junctions in a modern microprocessor, one must account for the influence of the quasi-Fermi levels on the very atoms that are diffusing into place. The light used to build the device fundamentally alters its construction at the atomic scale, a process described perfectly by the language of quasi-Fermi levels.

From switching a transistor in a nanosecond to driving chemical reactions with sunlight to forging the very structure of a computer chip, the concept of the quasi-Fermi level proves itself to be an indispensable tool. It is the thread that unifies our understanding of the dynamic, energetic, and endlessly useful world of semiconductors out of equilibrium.