Single-Electron Tunneling

In the macroscopic world, electric current is perceived as a continuous, fluid-like flow of charge. However, as we shrink our perspective to the nanoscale, the fundamental graininess of electricity becomes undeniable: charge is carried by individual electrons. This simple fact opens up a new realm of physics and technology centered on a profound question: what are the rules that govern the transport of electrons one by one, and how can we harness this control? The phenomenon of single-electron tunneling provides the answer, revealing a rich interplay between classical electrostatics and quantum mechanics.

This article delves into this fascinating topic. The first section, "Principles and Mechanisms", will unpack the foundational concepts of charging energy, Coulomb blockade, and the quantum conditions required to observe these effects. We will explore higher-order processes like cotunneling and the crucial influence of the device's environment. The second section, "Applications and Interdisciplinary Connections", will then showcase how these principles become powerful tools, enabling breakthroughs in fields ranging from quantum metrology and spintronics to single-molecule chemistry and the study of exotic materials.

Imagine an island, infinitesimally small, floating between two vast continents, a source and a drain of travelers. These travelers are, of course, electrons. In our everyday world, electrons flow like a continuous river of charge. But on the scale of our tiny island—a nanoscale piece of metal or a quantum dot—the rules change. The discreteness of the electron, its fundamental indivisibility, becomes the star of the show. Here, we will explore the principles that govern an electron's journey onto and off this island, a phenomenon known as single-electron tunneling.

Let's begin with a simple, classical idea. Our island is a capacitor. To put charge on any capacitor, you have to do work. The energy stored in a capacitor with charge $Q$ and capacitance $C$ is given by the familiar formula $U = Q^2/(2C)$. Now, what happens when a single electron, with its indivisible charge $-e$, attempts to hop onto a neutral island?

Initially, the island is neutral, so its charge is $Q_i = 0$ and its stored energy is zero. When one electron arrives, the island's charge becomes $Q_f = -e$. The energy changes to $U_f = (-e)^2 / (2C_\Sigma)$, where $C_\Sigma$ is the total capacitance of the island, accounting for its connections to the whole universe—the source, the drain, and a nearby gate electrode. The energy cost to make this happen, the price of admission for this single electron, is what we call the ​​charging energy​​, $E_C$.

This energy acts like a toll. If an incoming electron doesn't have enough energy to pay it, it simply can't get on the island. This refusal of entry is the essence of ​​Coulomb Blockade​​. It's a blockade born from the electrostatic repulsion of a single electron's charge.

For this blockade to be more than a theoretical curiosity, it must stand up to the constant jostling and chaos of the thermal world. Electrons in the leads are not sitting still; they are buzzing with thermal energy on the order of $k_B T$, where $k_B$ is the Boltzmann constant and $T$ is the temperature. If this thermal energy is comparable to the charging energy, electrons can easily borrow enough energy from their heated surroundings to overcome the toll. The blockade would be washed away in a sea of thermal fluctuations.

Therefore, for the Coulomb blockade to be robustly observed, a clear condition must be met: the charging energy must be substantially greater than the thermal energy, $E_C \gg k_B T$. This simple inequality has profound consequences for engineering. To make $E_C$ large, the capacitance $C_\Sigma$ must be incredibly small. This is why single-electron devices are nanoscale objects. To see the blockade at the frigid temperatures of a dilution refrigerator, around 20 millikelvin, the total capacitance of the island must be smaller than a few femtofarads ($10^{-15}$ F), a testament to the marvels of modern nanofabrication.

Making the island tiny and the system frigidly cold seems like enough to build our electronic turnstile. But the electron is a quantum creature, and it obeys stranger rules. One of the most famous is Heisenberg's uncertainty principle. In one of its forms, it tells us that a state that exists for only a short time $\Delta t$ has an inherent uncertainty in its energy, $\Delta E$, of about $\hbar/\Delta t$ (where $\hbar$ is the reduced Planck constant).

Imagine an electron tunneling onto our island. It has a certain lifetime there before it tunnels off again. This lifetime gives the island's charged state an energy broadening, or "fuzziness". If this quantum fuzziness, $\Delta E$, becomes as large as the charging energy $E_C$ itself, the very concept of a well-defined energy toll collapses. The blockade is smeared out not by heat, but by quantum uncertainty.

To preserve the blockade, we must ensure the electron's state on the island is well-defined. This means its lifetime must be long, and consequently, the tunneling process must be slow. The "slowness" of tunneling is controlled by the opacity of the tunnel barriers that separate the island from the leads. In electronics, opacity is just another word for high resistance. This brings us to the second pillar of single-electron tunneling: the resistance of the tunnel junctions, $R_T$, must be large.

How large? The universe provides a natural scale for resistance in the quantum world: the ​​resistance quantum​​, $R_Q = h/e^2$, which is approximately $25.8 \, \text{k}\Omega$. To prevent quantum fluctuations from destroying the blockade, the tunnel resistance must be significantly larger than this value: $R_T \gg R_Q$.

This condition can be understood in a couple of beautiful ways. From the uncertainty principle perspective, a large $R_T$ implies a low tunneling rate, $\Gamma$. This long lifetime for the charged state ensures its energy broadening, $\hbar\Gamma$, is much smaller than the charging energy $E_C$. From the perspective of quantum transport, a large resistance implies that the probability of an electron transmitting through the barrier is very small. The electron's wavefunction is mostly reflected, preventing it from "leaking out" and delocalizing across the leads. This keeps the electron's charge firmly localized on the island, a prerequisite for the blockade to even make sense.

With our two golden rules, $E_C \gg k_B T$ and $R_T \gg R_Q$, we have successfully built a region of impregnable blockade, a "Coulomb diamond" in the map of experimental parameters where current should be zero. But quantum mechanics is the master of loopholes.

Even when an electron lacks the energy to pay the toll and stay on the island, it can make a fleeting, ghostly visit. Imagine an electron from the source lead tunneling onto the island. This is energetically forbidden. But if, at the same instant, an electron from the island tunnels off to the drain lead, the island's charge number is only changed for an infinitesimal, "virtual" moment. This coherent, two-electron shuffle is a second-order quantum process called ​​cotunneling​​.

This process is a direct consequence of the full quantum mechanical description of the system, encapsulated in its Hamiltonian. The total Hamiltonian, $H = H_{\text{leads}} + H_{\text{island}} + H_C + H_T$, contains the charging energy term $H_C = E_C(N - n_g)^2$, which creates the blockade, and the tunneling term $H_T$, which describes electrons hopping. Cotunneling is what happens when we consider the action of $H_T$ to second order.

The amplitude for this process involves the intermediate virtual state where the island is temporarily charged. Quantum perturbation theory tells us this amplitude is suppressed by the energy cost of that virtual state, which is on the order of $E_C$. The rate is proportional to the square of this already small amplitude. So, while cotunneling provides a "leakage" current through the blockade, it is exceedingly small.

This leakage current has its own rich inner life, which we can reveal by changing the temperature.

• ​​Inelastic Cotunneling​​: If the tunneling electron can give some of its energy to the island, it can leave behind an excitation, like creating an electron-hole pair. This is inelastic cotunneling. At finite temperature or bias voltage, there is energy available for such processes. The resulting current shows a characteristic quadratic dependence on temperature ($I \propto T^2$) or voltage ($I \propto V^2$).
• ​​Elastic Cotunneling​​: If the island is left in its ground state after the event, the process is elastic. This is the only form of cotunneling possible at absolute zero temperature and infinitesimal bias. It provides a tiny, temperature-independent floor to the conductance.

Imagine watching the current as we cool the device. At high temperatures, electrons have enough thermal energy to hop on and off sequentially, and current flows readily. As we cool, this sequential current "freezes out" exponentially. We then enter a regime dominated by inelastic cotunneling, where the conductance gently falls as $T^2$. Finally, at the lowest temperatures, even these inelastic processes are frozen out, and we are left on the flat, temperature-independent plateau of pure elastic cotunneling—a direct view into a subtle quantum tunneling process.

So far, we have only discussed the average current. But the flow of discrete electrons is not perfectly smooth; it has a rhythm. The fluctuations in the current, known as ​​shot noise​​, can tell us even more about the nature of the electron's journey.

Think of raindrops falling on a tin roof. If they fall randomly and independently, like a Poisson process, we hear a steady, characterless hiss. This corresponds to the "full" shot noise, with a magnitude $S_I(0) = 2eI$.

Now consider our single-electron transistor in the sequential tunneling regime. An electron tunnels onto the island. Due to the Coulomb blockade, it blocks any other electron from following until it has tunneled out. The electrons are forced to form an orderly queue! They are no longer independent; their motion is anti-correlated. This regularity reduces the randomness—the "hiss"—of the current. The noise becomes ​​sub-Poissonian​​, with a magnitude less than $2eI$. The measured noise is a direct signature of electrons interacting and "waiting their turn".

What about the cotunneling regime? Elastic cotunneling is a single, coherent quantum event that moves an electron from source to drain. If we assume each of these quantum events is statistically independent of the others, then the transport process is once again like random raindrops. The process is Poissonian, and the noise should return to the full value of $S_I(0) = 2eI$. By simply listening to the rhythm of the current, we can distinguish between the one-by-one, correlated march of sequential tunneling and the burst-like, independent events of cotunneling.

We come to a final, profound point. Our tiny island is not truly isolated. It is connected to a macroscopic world of wires, amplifiers, and measurement devices. This external circuit constitutes an electromagnetic ​​environment​​. And in quantum mechanics, you can never ignore the observer—or their apparatus.

When an electron tunnels, it causes a sudden jolt to the electric field. This jolt propagates out into the surrounding circuit, creating electromagnetic waves (photons). The environment can absorb this energy, or conversely, the fluctuations in the environment (even vacuum fluctuations) can give energy to the tunneling electron. This exchange of energy between the electron and its environment is the basis of ​​Dynamical Coulomb Blockade​​.

The theory that describes this, called ​​P(E) theory​​, tells us the probability $P(E)$ that a tunneling electron exchanges an energy $E$ with its environment. If the environment is resistive—if it's good at dissipating energy—it becomes very difficult for an electron to tunnel without losing a little bit of energy to this dissipation. The probability of a perfectly energy-conserving, "elastic" tunneling event ($E=0$) can be driven to zero!

This creates an even more severe blockade. The conductance at very low bias voltage is no longer a small constant, but is suppressed to zero according to a power law, $G(V) \propto V^\alpha$. The exponent $\alpha$ depends on the resistance of the environment. This means that the very act of wiring up the device to measure it fundamentally changes the transport properties. The edges of the Coulomb diamonds, once sharp, become smeared and rounded. It is a beautiful and humbling reminder that in the quantum world, no system is an island, entire of itself. Its dance is always choreographed in concert with the wider universe to which it is coupled.

After our journey through the fundamental principles of single-electron tunneling, one might be left with the impression of a rather abstract and delicate phenomenon, confined to the pristine environment of a low-temperature physics laboratory. But nothing could be further from the truth! This is where the story truly comes alive. The simple, almost naive idea that charge comes in discrete packets of $e$ and that adding one more to a small object costs energy, turns out to be a key that unlocks a vast and beautiful landscape of applications, from the most practical technologies to the most profound explorations of the quantum world. Single-electron tunneling is not just a curiosity; it is a tool—a quantum scalpel of exquisite precision.

What is the most fundamental task in electronics? Counting electrons. For centuries, we have defined our unit of electric current, the Ampere, based on classical forces between wires. But what if we could define it directly by its very nature—a flow of a certain number of electrons per second?

This is precisely the goal of metrological single-electron pumps. Imagine a tiny island connected to source and drain electrodes by two tunnel junctions, controlled by a rapidly oscillating gate voltage. The process is like a quantum turnstile. In the first half of a cycle, the gate voltage is adjusted to lower the energy of the island, making it favorable for exactly one electron to tunnel from the source onto the island. Then, in the second half, the gate voltage is changed to raise the island's energy, coaxing that same electron to tunnel off onto the drain. Repeat this process at a frequency $f$, and you get a perfectly quantized current: $I = ef$. One electron per cycle, as reliable as clockwork.

Of course, reality is a bit more complicated, and this is where the art of quantum engineering comes in. The designer of such a device faces a delicate balancing act. The tunnel barriers must be opaque enough to prevent electrons from leaking through when they're not supposed to (a process driven by thermal energy), and to suppress higher-order quantum "cheating" like an electron tunneling through both barriers at once. Yet, the barriers must be transparent enough to allow the electron to reliably hop on and off during the brief windows of time the turnstile is open. If the frequency $f$ is too high, or the barriers too opaque, the electron might miss its chance to tunnel, leading to an error. The optimal design often involves making the two tunnel junctions as symmetric as possible to maximize the current for a given tunneling rate. To achieve the incredible accuracy required for metrology—say, an error of less than one part in a million—engineers must carefully navigate these trade-offs, suppressing thermal errors by going to very low temperatures and designing the gate voltage pulses to create a large energy penalty for any unwanted tunneling events, all while trying to push the frequency as high as possible for a usable current. This technology, born from the principles of Coulomb blockade, is at the heart of the modern quantum definition of the Ampere.

The single-electron transistor (SET) is not just a switch or a pump; it can also be an extraordinarily sensitive spectrometer. A quantum dot, that tiny island of semiconductor, is often called an "artificial atom" because, like a real atom, its electrons can only occupy a discrete set of energy levels. How can we map out this secret energy ladder?

The answer lies in a beautiful technique called inelastic electron tunneling spectroscopy (IETS). Imagine an electron in the source electrode, eager to tunnel through the quantum dot to the drain. In the simplest case, it can only do so if there is an empty state at the same energy level on the other side. But what if the electron could give up a little bit of its energy along the way? Suppose the artificial atom has an excited state with energy $\Delta$ above its ground state. A tunneling electron, whose energy is provided by the bias voltage $V$, can tunnel across, give a "kick" of energy $\Delta$ to the dot, and continue on its way. It's like striking a tiny, quantized bell.

This process, however, can only happen if the electron has enough energy to give away in the first place. This means its initial energy, which is at most $e|V|$, must be greater than or equal to the excitation energy $\Delta$. So, as we slowly increase the bias voltage across the SET, nothing happens... until suddenly, at $e|V| = \Delta$, a new channel for current opens up. The differential conductance, $dI/dV$, will show a sharp step. As we increase the voltage further, we see another step every time $e|V|$ matches the energy of another excited state of our artificial atom. The conductance-voltage graph becomes a staircase, where the location of each step precisely maps out the internal energy spectrum of the quantum dot.

This principle is not limited to quantum dots. With a Scanning Tunneling Microscope (STM), we can position a sharp tip just angstroms away from a single molecule on a surface. By applying a voltage, we can make electrons tunnel through the molecule. If we tune the voltage just right, these tunneling electrons can inelastically give up their energy to excite a specific vibration of the molecule—say, stretching a particular chemical bond. This can be used not just to measure the vibrational energy, but to induce a chemical reaction. We can break a bond, or change the molecule's shape, one molecule at a time. This is the ultimate form of chemical control, all made possible by the simple act of a single electron giving up a quantum of its energy.

So far, we have talked about the electron's charge and energy. But the electron has another, deeply quantum property: spin. It acts like a tiny bar magnet, an internal compass that can point "up" or "down". Can our single-electron tools sense this as well? The answer is a resounding yes, and it opens the door to the fields of spintronics and quantum computation.

If we place our artificial atom in a magnetic field $B$, the two spin states (up and down) will split by an amount of energy called the Zeeman energy, $\Delta_E = g \mu_B B$, where $g$ is a number characteristic of the electron in its specific material environment. Using inelastic tunneling spectroscopy, we can once again find this energy. We sweep the bias voltage $V$, and when $e|V|$ becomes equal to the Zeeman energy, a new tunneling channel opens up: an electron tunnels and flips the spin of the electron trapped on the dot. This creates symmetric steps in the conductance at $V = \pm \Delta_E / e$, from which we can directly measure the Zeeman splitting with incredible precision, even determining the effective $g$-factor for a single electron in a nanostructure.

The connection to spin becomes even more profound in a double quantum dot system. Imagine two quantum dots side-by-side, acting as a single, more complex artificial molecule. An electron can hop from the source onto the first dot, then to the second dot, and finally to the drain. Let's watch the crucial step: the transition from a state with one electron on each dot, denoted $(1,1)$, to a state with two electrons on the second dot, $(0,2)$.

The two electrons in the $(1,1)$ configuration can have their spins aligned (a "triplet" state) or anti-aligned (a "singlet" state). Now, consider the $(0,2)$ state. According to the Pauli exclusion principle, two electrons in the same orbital state must have their spins anti-aligned; they must form a singlet. The transition from a $(1,1)$ singlet to a $(0,2)$ singlet is allowed. But what if the two electrons in the $(1,1)$ state happen to form a triplet? Since single-electron tunneling conserves spin, the transition $(1,1)_\text{Triplet} \to (0,2)_\text{Singlet}$ is strictly forbidden! The electron is stuck. The current stops. This phenomenon is known as ​​Pauli spin blockade​​. The device acts as a perfect spin filter, allowing singlets to pass and blocking triplets. This blockade provides a powerful method for reading out the spin state of electrons, a critical requirement for building a quantum computer where information is stored in the spins of individual electrons.

The world of quantum physics is full of strange and wonderful phenomena. What happens when single-electron effects, governed by charging energy $E_C$, collide with another titan of quantum mechanics: superconductivity? In a superconductor, electrons form "Cooper pairs" with charge $2e$ and are bound together by an energy $\Delta$, the superconducting gap.

Consider a superconducting single-electron transistor (SSET), where the island and leads are all superconducting. This sets up a fascinating competition. The charging energy $E_C$ wants to add electrons to the island one at a time. But superconductivity resists this! An island with an even number of electrons can have them all neatly paired up. But an island with an odd number of electrons must have at least one lone, unpaired electron—a "quasiparticle"—which costs an energy of at least $\Delta$ to create. This is known as the parity effect.

So, who wins? If the charging energy is dominant ($E_C > \Delta$), the system behaves much like a normal SET, with current flowing via single-electron tunneling. But if the superconducting gap is dominant ($\Delta > E_C$), it is energetically forbidden to have an odd number of electrons on the island. The ground state of the island will always have an even number of electrons. In this case, charge cannot be transported by single electrons. Instead, it is transported by the coherent tunneling of whole Cooper pairs! This is a second-order quantum process, a subtle dance where two electrons tunnel in a correlated way, preserving the even-parity of the island. The stunning consequence is that the conductance oscillations as a function of gate voltage become $2e$-periodic, not $1e$-periodic. The device counts in twos! This beautiful effect is a direct signature of the interplay between single-electron physics and superconductivity, though it is fragile—a single stray quasiparticle "poisoning" the island can break the spell and restore the familiar $1e$ periodicity.

Finally, single-electron tunneling allows us to probe states of matter that defy our conventional understanding of electrons. In most metals, electrons behave as individual particles (albeit quantum ones). But in a one-dimensional wire, strong repulsive interactions can cause them to lose their identity entirely. They organize into collective, sound-like waves of charge and spin. This bizarre state of matter is called a Tomonaga-Luttinger liquid. What happens when we tunnel an electron into such a world? The result is remarkable. The current does not simply switch on; it follows a strange power-law relationship with voltage, $I \propto V^{\alpha+1}$. The exponent $\alpha$ is not a simple integer but is directly related to the strength of the electron-electron interactions in the wire. The tunneling electron acts as a probe, and the current it creates carries back a signature of the exotic collective physics of the world it entered.

From building a quantum standard of current to performing chemistry one molecule at a time, from reading the spin of a single electron to revealing the secret rules of exotic new materials, the applications of single-electron tunneling are as profound as they are diverse. A principle of elegant simplicity—that you can't chop an electron in half—gives rise to a tool of unparalleled power and subtlety, revealing time and again the inherent beauty and unity of the quantum universe.