Multiband Superconductivity

The foundational Bardeen-Cooper-Schrieffer (BCS) theory of superconductivity describes a beautifully simple scenario: a single sea of electrons condenses into one unified quantum state. While powerful, this picture assumes an idealized metal with a single electronic band. The reality in most materials is far more complex and interesting, featuring multiple distinct families of electrons. This complexity presents a significant challenge to the single-band model and opens the door to the richer physics of multiband superconductivity. This article explores this intricate world, addressing how superconductivity emerges when multiple electronic bands are at play. It will guide the reader through the core concepts that define this field, offering a comprehensive overview of its theoretical underpinnings and observable consequences. The journey begins by dissecting the fundamental rules of this multi-component system, exploring the principles and mechanisms governing the interaction between different superconducting condensates.

In the world of physics, we often start by envisioning the simplest possible scenario. For superconductivity, this ideal case is a perfect, uniform sea of electrons. When the temperature drops, a subtle attraction, perhaps mediated by the vibrations of the crystal lattice, binds electrons into pairs—Cooper pairs. These pairs all condense into a single, unified quantum state, opening up an energy gap, $\Delta$, below which no individual electron excitations can exist. It’s a beautiful, monolithic picture described by the Bardeen-Cooper-Schrieffer (BCS) theory. The reality within a crystal, however, is almost always more intricate and, as it turns out, far more interesting.

An electron moving through a crystal is not free. Its quantum mechanical wavefunction must respect the periodic arrangement of atoms, a constraint that structures the allowed electron energies into a series of "bands". In a metal, at least one of these bands is only partially filled, giving electrons the freedom to move and conduct electricity. The collection of the highest-energy electron states at zero temperature forms a surface in momentum space known as the ​​Fermi surface​​.

But what if a material has not just one, but two, three, or even more bands crossing this crucial Fermi energy? This is not an exotic exception; it is the rule for most real materials. Each of these bands can be thought of as a distinct family of electrons, each with its own character, its own Fermi surface. This is the stage for ​​multiband superconductivity​​.

When such a material decides to become a superconductor, it faces a choice. Does it form a single, uniform condensate from all the electron families? Or does each family form its own condensate, with its own character? The answer is the latter, and it leads to a wonderfully rich physics. We find that each band, say band $i$, can develop its own superconducting energy gap, $\Delta_i$. The material doesn't have a superconducting gap; it has a whole set of them.

How do we know this isn't just a theoretical fancy? We can look for the evidence. Imagine trying to probe the energy levels of such a material.

• ​​Tunneling Spectroscopy:​​ One of the most direct methods is to bring a sharp metallic tip extremely close to the superconductor's surface and measure the quantum-mechanical tunneling of electrons. The rate of tunneling reveals the density of available electronic states. In a simple single-gap superconductor, we find a desert at low energies, bounded by two sharp "coherence peaks" at energies corresponding to the gap, $\pm\Delta$. In a multiband superconductor, we see something more complex: multiple pairs of these peaks appear, one pair for each distinct energy gap present in the system. It's like finding evidence of several different species, each with its own characteristic energy. The classic example is magnesium diboride, $\mathrm{MgB}_2$, where experiments clearly resolve two gaps: a large one around $7 \text{ meV}$ on a set of two-dimensional $\sigma$ bands, and a smaller one around $2-3 \text{ meV}$ on three-dimensional $\pi$ bands.

• ​​Specific Heat:​​ The heat capacity of a material tells us how much energy it takes to raise its temperature. For a superconductor, this is related to the energy needed to break Cooper pairs and create excited "quasiparticles". Just below the critical temperature, $T_c$, a single-gap superconductor shows a characteristic sharp jump. In a two-band system, the story is subtler. The presence of a smaller gap means it's much easier to create quasiparticles at low temperatures than the larger gap would suggest. This leads to a tell-tale "shoulder" in the specific heat curve below the main transition—a signature that a second, low-energy channel for excitations has opened up.

• ​​Angle-Resolved Photoemission Spectroscopy (ARPES):​​ This remarkable technique is like a high-powered camera for electrons in a solid. It uses photons to knock electrons out of the material and measures their energy and momentum. By doing this, it can directly map out the electronic bands and, in the superconducting state, measure the size of the energy gap at every point on the Fermi surface. With ARPES, we can literally see that the gap is large on one sheet of the Fermi surface and small on another, providing irrefutable, pictorial evidence for the multiband nature of the state.

A fascinating question arises: if there are two or more distinct "superconducting families" in different bands, why do they all decide to become superconducting at the exact same temperature, $T_c$? Why don't we see one band condense at $T_{c1}$ and another at a lower temperature $T_{c2}$?

The answer lies in the fact that these bands are not isolated islands. Electrons, and more importantly, Cooper pairs, can scatter from one band to another. This ​​interband coupling​​ acts as a communication channel, locking the different condensates together into a single, cohesive whole.

We can describe this beautifully using the phenomenological Ginzburg-Landau theory. Here, we describe each condensate with its own order parameter, say $\psi_1$ and $\psi_2$. The free energy of the system includes terms for each band independently, but also a crucial coupling term, like $-\eta(\psi_1^* \psi_2 + \psi_2^* \psi_1)$, where $\eta$ is the coupling strength. This term acts like a spring connecting the two order parameters, forcing them to cooperate. Because of this coupling, the entire system undergoes a single, sharp phase transition at a well-defined critical temperature $T_c$. The resulting jump in the specific heat is a collective phenomenon, though its shape below $T_c$ still bears the marks of the multiple gaps involved. If this coupling $\eta$ were to vanish, we would indeed see two separate transitions. But in any real material, electrons can scatter between bands, ensuring that the superconducting transition is a unified, democratic event.

To get to the heart of how this works, we must look at the interactions that cause pairing in the first place. The onset of superconductivity can be understood as an instability. As we cool the metal, its "willingness" to form pairs increases. At $T_c$, this willingness, when combined with the attractive interaction, becomes infinite, and the paired state spontaneously emerges.

In a multiband system, this process is governed by a matrix equation. We can think of it intuitively like this: the final state, represented by a vector of gaps $(\Delta_1, \Delta_2, \dots)$, is determined by how the system responds to those very gaps, mediated by a matrix of interactions, $\hat{V}$. This matrix is the rulebook. Its diagonal elements, $V_{ii}$, describe the strength of the pairing attraction within a band, while its off-diagonal elements, $V_{ij}$, describe the strength of scattering of a pair from band $j$ to band $i$.

The system becomes superconducting when the following condition is met:

where $\hat{\chi}(T)$ is a diagonal matrix representing the intrinsic "susceptibility" of each band to pairing. But the most profound secret is that the structure of the final state—the relative sizes of the gaps—is encoded in the ​​eigenvectors​​ of the interaction matrix $\hat{V}$. The system chooses the eigenvector that corresponds to the strongest overall attraction (the largest eigenvalue), as this state is the most energetically favorable and yields the highest $T_c$.

This has remarkable consequences. For example, under certain simplifying conditions, one can show that the ratio of the two gaps at zero temperature is directly determined by a ratio of the interaction matrix elements. The microscopic interactions directly dictate the macroscopic properties of the superconducting state.

So far, we have tacitly assumed that the gaps, while different in magnitude, all have the same sign. We can call this an $s_{++}$ state, where the subscript 's' denotes the isotropic nature of the gap on each Fermi surface, and '++' denotes the common sign. But what if the interactions are more complicated? What if there is competition?

This is precisely the situation in many modern superconductors, like the iron-based families. The pairing mechanism is thought to arise from two competing forces: a general, phonon-mediated attraction, and a sharp, short-range Coulomb repulsion. The repulsion is a penalty for putting two electrons on the same atom.

How can a system that is repulsive still superconduct? It can perform a clever trick. Imagine two bands, or two "pockets" of the Fermi surface. If the gap on pocket 1 has a positive sign ($\Delta_1 > 0$) and the gap on pocket 2 has a negative sign ($\Delta_2 < 0$), the system has found a way to be "attractive on average" while avoiding the on-site repulsion. This is because the contribution to pairing from inter-pocket scattering depends on the product of the gaps. By making them have opposite signs, a repulsive interaction can actually aid superconductivity! This remarkable state is known as the ​​$s_{\pm}$ state​​.

The choice between an $s_{++}$ and an $s_{\pm}$ state becomes a contest. If the inter-pocket attraction is strong, the $s_{++}$ state wins. But if the on-site repulsion is sufficiently strong compared to the inter-pocket interaction, the system will favor the $s_{\pm}$ state to minimize its energy. The ground state of a superconductor is thus a delicate balance of competing interactions.

This idea of a sign-changing gap might seem like a theorist's abstraction. How could we ever tell if a gap is "positive" or "negative"? The phase of a quantum wavefunction isn't directly measurable. The answer, beautifully, comes from studying how the superconductor reacts to imperfections.

A famous result known as ​​Anderson's Theorem​​ states that non-magnetic impurities (like a zinc atom substituting a copper atom) have no effect on the critical temperature of a conventional s-wave superconductor. The impurities scatter electrons, but since the gap is the same everywhere, this scattering doesn't disrupt the pairing.

The situation is dramatically different in an $s_{\pm}$ superconductor. Imagine a Cooper pair living happily in band 1, enjoying a gap of $\Delta_1$. Suddenly, one of its electrons scatters off an impurity into band 2. In band 2, the world is inverted; it feels a gap of $\Delta_2 = -\Delta_1$. This violent change in the phase of the order parameter is extremely disruptive to the pair. This interband scattering process acts as a potent ​​pair-breaking mechanism​​, akin to what magnetic impurities do in a conventional superconductor.

The stunning consequence is that non-magnetic impurities will strongly suppress the critical temperature of an $s_{\pm}$ superconductor. This flagrant violation of Anderson's Theorem is a "smoking gun" signature for a sign-changing gap. The rate of $T_c$ suppression follows a universal formula derived by Abrikosov and Gor'kov, revealing a deep and unexpected unity in the physics of disorder. If the impurity scattering is strong enough, it can even lead to the emergence of a bizarre ​​gapless superconducting​​ state, where a finite superfluid density coexists with a continuous spectrum of low-energy quasiparticle excitations.

From the simple concept of multiple electron bands, a rich and complex world emerges. We find new states of matter, like the $s_{\pm}$ state, born from the competition between fundamental forces. We discover that their unique properties, like their fragility to impurities, provide the very clues we need to unravel their secrets. This is the inherent beauty of condensed matter physics: the most intricate and surprising collective phenomena often arise from the simple interplay of a few fundamental rules.

If the world of single-band superconductivity is a solo violin playing a beautiful, pure melody, then the realm of multiband superconductivity is a full symphony orchestra. Here, multiple sections—the brass, the woodwinds, the strings—each contribute their own voice. These are our different electronic bands, each forming its own "condensate" of Cooper pairs. Sometimes they play in perfect, harmonious unison; other times, they engage in a complex, contrapuntal dance. The resulting music is incredibly rich, leading to a spectacular array of new phenomena, solving old mysteries, and opening doors to states of matter that a single-band theorist could only dream of. Now that we have tuned our instruments by understanding the basic principles, let's sit back and listen to the performance.

How do we even know there's an orchestra playing, rather than just a single violin? We need to find a way to listen to the individual sections. Fortunately, the quantum world provides us with incredibly sensitive microphones.

One of the most direct methods is a technique called scanning tunneling spectroscopy (STS). Imagine bringing an atomically sharp metal tip exquisitely close to the surface of our superconductor. By applying a voltage $V$, we encourage electrons to "tunnel" across the vacuum gap. The ease with which they tunnel, measured as the differential conductance $G(V)$, gives us a direct map of the available electronic states in the superconductor at energy $eV$. For a single-band superconductor, we see one distinct energy gap. But for a two-band system, the spectrum is a beautiful superposition: we see two sets of gap features, one for each condensate! The total conductance is a weighted sum of the contributions from each band, each with its own characteristic gap $\Delta_i$. We can quite literally see the two gaps on our screen, providing a snapshot of the two coexisting condensates at play.

A more subtle, but equally profound, clue comes from the isotope effect. In the simple Bardeen–Cooper–Schrieffer (BCS) picture, the critical temperature $T_c$ depends on the mass $M$ of the vibrating ions that provide the pairing "glue." If you make the ions heavier, they vibrate more slowly, and $T_c$ goes down according to a precise law, $T_c \propto M^{-\alpha}$, with the isotope coefficient $\alpha$ predicted to be exactly $1/2$. For many superconductors, this works beautifully. But in magnesium diboride ($\mathrm{MgB_2}$), experiments found a boron isotope coefficient significantly smaller than $1/2$. Why? The multiband picture provides a perfect explanation. In $\mathrm{MgB_2}$, the pairing is dominated by high-frequency vibrations of the light boron atoms. However, this phonon "rhythm" is felt much more strongly by one set of electrons (the $\sigma$-band) than by another (the $\pi$-band). Since the final $T_c$ emerges from the cooperation of both condensates, the weaker coupling of the $\pi$-band effectively dilutes the influence of the boron mass on the overall transition temperature, reducing $\alpha$ below $1/2$. It's as if the string section of our orchestra is keeping perfect time with the conductor's baton, but the woodwinds are a bit less attentive, softening the overall rhythmic impact.

With multiple condensates, new dynamic possibilities emerge. They can not only move together, maintaining a fixed relative phase (this gives the familiar sound-like mode of a superfluid), but they can also oscillate against each other. This out-of-phase oscillation of the condensates' phase differences is a unique collective excitation called a ​​Leggett mode​​. It's a dance that simply cannot happen in a single-condensate system. Because the condensates are coupled (the different sections of the orchestra are listening to each other), this relative-phase dance isn't free; it costs energy. As a result, unlike the normal gapless "sound" mode, the Leggett mode has a finite energy gap, a characteristic frequency $\omega_L$ determined by the strength of the interband coupling $\Lambda$ and properties of the condensates themselves.

This is a beautiful theoretical prediction, but can we see this dance? It turns out we can, using Raman spectroscopy, a technique that involves shining light on the material and observing how it scatters. The light can kick the system into an excited state, and if the energy and, crucially, the symmetry of that excitation are right, we will see a corresponding signal. In a material like an iron-pnictide superconductor, the crystal structure has a particular symmetry (for example, the $D_{4h}$ point group). The possible Leggett modes, which correspond to different ways the relative phases of the multiple condensates can oscillate, also have specific symmetries. By working out the group theory, one can show that these modes can transform under symmetries like $A_{1g}$ or $B_{1g}$, which happen to be "Raman active". This means we can directly "see" the Leggett mode as a peak in the Raman spectrum, providing stunning experimental verification of this unique multiband phenomenon.

The multiband concept is not just an abstract idea; it is the key to understanding the behavior of entire classes of real-world materials.

• ​​The Poster Child: Magnesium Diboride ($\mathrm{MgB_2}$)​​. Discovered in 2001, the superconductivity in this simple ceramic at a surprisingly high $39\,\mathrm{K}$ was initially a puzzle. The solution was quintessentially multiband. $\mathrm{MgB_2}$ has two distinct electronic systems: strongly bonded, two-dimensional sheets of boron atoms giving rise to $\sigma$ bands, and a more three-dimensional network associated with $\pi$ bands. The key is that the light mass of boron allows for very high-frequency vibrations, and these vibrations couple very strongly to the electrons in the $\sigma$ bands, creating a large superconducting gap. The coupling to the $\pi$ bands is much weaker, creating a smaller second gap. The high $T_c$ is a direct result of this happy accident of nature: a combination of high-frequency phonons and strong coupling within a specific band, all made possible by the material's multiband structure.

• ​​The Unconventional Cousins: Iron-Based Superconductors​​. In the iron-based superconductors (FeBS), the story is even more fascinating. Here, simple calculations show that the electron-phonon "glue" is far too weak to explain the observed high critical temperatures. This forces us to look for an "unconventional" pairing mechanism, widely believed to be magnetic spin fluctuations. In this picture, the multiband structure is not just helpful—it's essential. The typical FeBS has hole-like Fermi surfaces at the center of the Brillouin zone and electron-like surfaces at the corners. The magnetic interaction provides an effective attraction for electrons on different types of surfaces, but a repulsion for electrons on the same type. The consequence is a "sign-changing" s-wave state, often called $s_{\pm}$, where the superconducting gap $\Delta$ is positive on the hole pockets and negative on the electron pockets. This delicate sign structure is a hallmark of unconventional pairing that is only possible in a multiband system.

• ​​Anisotropy's Roots: From Transport to Critical Fields​​. Why does a crystal conduct electricity better in one direction than another? Often, the answer lies in its multiband electronic structure. In the famous cuprate superconductor $\mathrm{YBa}_{2}\mathrm{Cu}_{3}\mathrm{O}_{7-x}$ (YBCO), the structure contains not only two-dimensional $\mathrm{CuO_2}$ planes but also quasi-one-dimensional $\mathrm{CuO}$ chains. By modeling this as a two-band system—one 2D band from the planes, one 1D band from the chains—we can naturally explain the observed transport anisotropy. The chains provide an extra, highly directional conduction path along the crystal's $b$-axis, making both the normal-state conductivity and the superconducting superfluid density larger in that direction. Similarly, the anisotropy of the upper critical field $H_{c2}$—the field at which superconductivity is destroyed—is a direct probe of the multiband nature. The overall anisotropy is a weighted average of the properties (like the Fermi velocities) of the constituent bands, allowing us to use a macroscopic measurement ($H_{c2}$) to deduce microscopic parameters of the individual condensates.

The most spectacular consequence of the multiband orchestra is when the sections play in a way that creates something entirely new—a form of music no one had ever heard before. This is what happens when the different bands have qualitatively different characteristics.

In a single-band world, superconductors are either "Type I" (which expel magnetic fields completely until a critical field $H_c$, where superconductivity abruptly vanishes) or "Type II" (which allow magnetic fields to penetrate in the form of quantized flux tubes called vortices, which repel each other). This behavior is governed by the ratio of two length scales: the magnetic penetration depth $\lambda$ and the coherence length $\xi$. But what if you have a two-band superconductor where one band is Type-I-like ($\xi_1 > \lambda$) and the other is Type-II-like ($\xi_2 \lambda$)?

The astonishing result is a new classification: ​​Type-1.5 superconductivity​​. In this state, the interaction between vortices is non-monotonic: they repel each other at short distances (due to the Type-II-like band) but attract each other at long distances (due to the Type-I-like band). This leads to a beautiful and bizarre thermodynamic state. Instead of forming a uniform lattice as in a Type-II superconductor, the vortices spontaneously condense into tightly packed clusters or "islands," separated by vast "seas" of field-free Meissner state. This remarkable "semi-Meissner" state, with its self-organized vortex clusters, can be directly imaged using techniques like scanning SQUID microscopy, providing a stunning visual confirmation of a qualitatively new state of matter born from the interplay of multiple condensates.

This interplay can even be harnessed. Imagine tuning a material by doping or pressure, such that we push the bottom of a new, "shallow" electronic band right to the Fermi level. This is called a Lifshitz transition. In this situation, virtual scattering of Cooper pairs from an existing superconducting band into this new band can dramatically alter the pairing strength. Under the right conditions, a "shape resonance" can occur when the small Fermi energy of the new band becomes comparable to the energy scale of the pairing glue. This can lead to a significant, non-monotonic enhancement of the critical temperature $T_c$. This opens up an exciting frontier: using the principles of multiband superconductivity to actively engineer and design materials with higher transition temperatures.

From explaining the properties of known materials to predicting and discovering entirely new states of matter, the concept of multiband superconductivity has proven to be one of the richest and most powerful ideas in modern condensed matter physics. It has shown us that when different quantum condensates are brought together, the resulting symphony is far greater than the sum of its parts. And the most exciting thing is that we feel the orchestra is still warming up; there is surely more beautiful music yet to be composed and discovered.