Resonating Valence Bond (RVB) State

In the realm of condensed matter physics, the collective behavior of strongly interacting electrons gives rise to some of the most fascinating and challenging phenomena, from exotic magnetism to high-temperature superconductivity. Classical descriptions often fail in this quantum world, where simple ordered patterns are replaced by dynamic, entangled states of matter. One of the most powerful and influential ideas proposed to navigate this complex landscape is the Resonating Valence Bond (RVB) state. Initially conceived by Linus Pauling and later revitalized by P.W. Anderson, the RVB concept provides a framework for understanding systems where quantum fluctuations prevent conventional magnetic ordering, creating a novel liquid-like phase of matter. This article explores the theoretical foundations and profound implications of the RVB state. The first chapter, "Principles and Mechanisms," will deconstruct the state from its building blocks—spin singlets—to explain how resonance, frustration, and entanglement lead to the emergence of a Quantum Spin Liquid. Following this, the "Applications and Interdisciplinary Connections" chapter will examine how this abstract concept provides a compelling explanation for high-temperature superconductivity and forges deep connections to modern frontiers like topological order and quantum computation.

Imagine a world populated by tiny quantum magnets—the spins of electrons. In a certain class of materials called antiferromagnets, these spins have a simple social rule: neighbors must always point in opposite directions. On a one-dimensional chain, this is easy to achieve: up, down, up, down, and so on. But what happens in the richer, more complex world of two or three dimensions? What happens when the very layout of the atoms, the very fabric of their society, makes it impossible for everyone to satisfy this simple rule? This is where our story begins, in a strange and beautiful quantum realm where spins give up on simple static order and instead form a dynamic, fluctuating liquid.

Let's first think about the most perfect expression of the antiferromagnetic rule for just two spins. How can two neighbors be "as antiparallel as possible"? Quantum mechanics gives us a beautiful answer: they can form a ​​spin singlet​​. A singlet is a unique two-spin state, written as $\ket{s_{ij}} = \frac{1}{\sqrt{2}}(\ket{\uparrow_i \downarrow_j} - \ket{\downarrow_i \uparrow_j})$. This isn't just spin '$i$' up and spin '$j$' down; it's a true quantum superposition. You cannot say which spin is up and which is down. They are inseparably entangled, locked in a symmetric embrace with zero total spin. If you measure one to be up, the other is guaranteed to be down, and vice versa. This singlet pair, also called a ​​valence bond​​, is the fundamental building block of our story.

Now, let's try to build a state for an entire lattice of spins using only these perfect pairs. We could, for instance, imagine covering the whole lattice with a static, non-overlapping pattern of these valence bonds. This configuration is called a ​​valence bond solid (VBS)​​. It’s like tiling a floor with dominoes, where each domino represents a singlet. However, as we will see, this static, frozen picture is often not what nature chooses.

Let's test this idea in a simple "quantum laboratory"—a tiny square made of four spins at the corners, labeled 1, 2, 3, and 4. We can imagine two simple ways to tile this square with our singlet dominoes. We could pair them horizontally, creating the state $\ket{\psi_H} = \ket{s_{12}} \otimes \ket{s_{34}}$, or vertically, with $\ket{\psi_V} = \ket{s_{23}} \otimes \ket{s_{41}}$.

Which configuration is better? Let's consider the energy. For the standard antiferromagnetic ​​Heisenberg Hamiltonian​​, $H = J \sum_{\langle i, j \rangle} \mathbf{S}_i \cdot \mathbf{S}_j$ (where $J > 0$), a single valence bond on a link has an energy of $-3J/4$. So for a static covering like $\ket{\psi_H}$, the two bonds in the covering contribute an energy of $2 \times (-3J/4) = -3J/2$. The other two links on the square, (2,3) and (4,1), connect spins from different singlets, and their average interaction energy is zero. The total energy is therefore $-3J/2$. With 4 bonds on the square, the energy per bond is $-3J/8$. For comparison, a simple classical ​​Néel state​​ (a checkerboard of alternating up and down spins) has an energy of $-J/4$ per bond. Our static pair configuration is therefore energetically favorable compared to the classical one. However, this static configuration is still not the true ground state of the quantum model on the square lattice. This leads to a deep puzzle: why is this simple, intuitive picture of static pairs incomplete? This is where the magic of quantum mechanics re-enters the stage.

This is the central idea of the ​​Resonating Valence Bond (RVB)​​ state. The system exists in a quantum superposition of all possible singlet coverings. For our little square, the simplest RVB state is a combination of the two patterns: $\ket{\psi_{RVB}} \propto (\ket{\psi_H} + \sigma \ket{\psi_V})$, where $\sigma$ is some phase factor.

When the Hamiltonian acts on, say, the horizontal state $\ket{\psi_H}$, it can break the singlets on bonds (1,2) and (3,4) and re-form them into the vertical configuration $\ket{\psi_V}$. This means the Hamiltonian has non-zero off-diagonal matrix elements between the different covering patterns. In the language of physics, the patterns "resonate" with each other. This resonance is a purely quantum effect, and like all resonances in physics, it leads to a dramatic consequence: it lowers the energy. The true ground state is a dynamic, fluctuating soup where the singlet bonds are not static but are constantly rearranging. This energy gained from fluctuation is called ​​resonance energy​​, and it is the key to the stability of the RVB state.

This superposition is not a random jumble. It's a highly choreographed quantum dance with strict rules. For example, on a bipartite lattice like our square (where sites can be divided into two sublattices, A and B, such that neighbors are always on different sublattices), the relative sign $\sigma$ in our superposition is not arbitrary. A fundamental property of antiferromagnets on such lattices, known as the ​​Marshall-Peierls sign rule​​, dictates that the sign must be negative, so $\sigma = -1$. This hidden order within the superposition ensures that the wavefunction has the correct structure to minimize the antiferromagnetic exchange energy.

Another fascinating subtlety is that the different valence bond coverings are not mathematically orthogonal to each other. The overlap between the horizontal state $\ket{\psi_H}$ and the vertical state $\ket{\psi_V}$ on our square is non-zero. This overcompleteness of the valence bond basis makes calculations famously difficult, but it's an essential feature of the physics. Nature hasn't provided a conveniently neat and tidy basis for us; the underlying reality is a web of interconnected, non-orthogonal possibilities.

This web results in a state of profound entanglement. Even in the simple 4-site RVB state, the entanglement (measured by a quantity called ​​concurrence​​) between two adjacent spins is significant. This tells us that even though the horizontal state $\ket{\psi_H}$ has no direct bond between spins 2 and 3, the resonance with the vertical state $\ket{\psi_V}$ establishes a powerful quantum link between them. Every spin is entangled with multiple partners, not just one. The bonds are not fixed but are shared and delocalized across the whole system.

So, if RVB is such a great idea, why isn't everything an RVB state? The answer is that on simple, "unfrustrated" lattices like the square lattice, quantum fluctuations are not strong enough to completely melt the classical Néel order. The true ground state is still a long-range ordered antiferromagnet, albeit with a reduced moment due to quantum zero-point motion.

The RVB picture truly comes into its own in the realm of ​​geometric frustration​​. Imagine trying to apply the antiferromagnetic rule—all neighbors must be antiparallel—on a triangular lattice. Pick a single triangle of spins. If spin 1 is up and spin 2 is down, what should spin 3 do? It's a neighbor to both. It cannot be antiparallel to both spin 1 and spin 2 simultaneously. The system is "frustrated"; it cannot satisfy all the interaction constraints at the same time.

In such geometrically frustrated systems, simple magnetic order is severely weakened or completely destroyed. The classical ground state might be a complex, non-collinear arrangement (like a 120° ordering of spins on the triangle), but the strong quantum fluctuations present in spin-1/2 systems can tip the balance entirely. Here, the idea of a fluctuating, disordered liquid of singlets is no longer just a clever variational ansatz; it becomes a leading candidate for the true ground state of the system. Frustration cracks open the door for a new phase of matter to emerge.

What have we described? A phase of matter that is a symmetric, dynamic superposition of singlet pairs, possessing no static magnetic order. This is the definition of a ​​Quantum Spin Liquid (QSL)​​. Let's look at its bizarre properties.

• ​​No Local Order:​​ A magnet is characterized by a local order parameter—the average magnetization $\langle \mathbf{S}_i \rangle$, which is non-zero. In a QSL, the state is a total spin singlet, meaning it's invariant under spin rotations. A consequence of this symmetry is that the local magnetization at every site is identically zero: $\langle \mathbf{S}_i \rangle = \mathbf{0}$. Similarly, unlike a VBS which has a static dimer pattern that breaks lattice symmetry, a QSL is a fluid that respects all symmetries of the lattice. It's a state defined not by what it is, but by what it is not.

• ​​Short-Range Correlations:​​ In an ordered magnet, spin correlations persist over infinite distances. In a short-range RVB state, which describes a gapped QSL, the correlation between two spins $\langle \mathbf{S}_i \cdot \mathbf{S}_j \rangle$ dies off exponentially with the distance between them. In other, more exotic "long-range" or "critical" RVB states, the correlations can decay much slower, as a power-law, without ever establishing true long-range order.

• ​​Topological Order:​​ Perhaps the most profound property of many QSLs is that their ground state is ​​topologically ordered​​. This means the system has a ground state degeneracy that depends on the global topology of the lattice—for instance, whether it's on a sphere or a torus (a donut). On a torus, a simple $Z_2$ spin liquid has four distinct, degenerate ground states. These states are locally identical and cannot be distinguished by any local measurement. They only differ in a global, topological property, like the parity of singlets crossing the hole of the donut. This is a new kind of order, beyond the standard paradigm of symmetry breaking, and it's a direct consequence of long-range entanglement.

• ​​Fractionalization:​​ If you try to create an excitation in a normal magnet, you create a magnon, a spin-wave that carries spin-1. In a QSL, the elementary excitation is something far stranger. When you break a singlet bond (spin-0), you create two un-paired spins. In the RVB liquid, these two halves can separate and wander off independently. These emergent, deconfined quasiparticles are called ​​spinons​​, and each carries spin-1/2. An integer spin excitation has fractionalized into two half-integer excitations! It's as if you could break a bar magnet and get an isolated North pole and an isolated South pole that can move around freely.

This journey into the world of quantum spins, frustration, and fractionalization might seem like a theoretical fantasy. But it has a deep and direct connection to one of the greatest mysteries in modern physics: high-temperature superconductivity.

In the materials that exhibit this phenomenon, electrons are strongly interacting. A huge on-site Coulomb repulsion prevents two electrons from ever occupying the same atomic site. This is the ​​single-occupancy constraint​​. How can we describe a state of mobile electrons that also respect this rule?

This is where P.W. Anderson had his revolutionary insight. He proposed that the ground state of these materials could be an RVB state of electrons. But how do we build it? We can start with a well-understood wavefunction for paired electrons, the Bardeen-Cooper-Schrieffer (BCS) state of a conventional superconductor, which unfortunately contains many configurations with two electrons on the same site. Then, we apply a mathematical tool called the ​​Gutzwiller projector​​.

This projector, $\mathcal{P}_G = \prod_i (1 - n_{i\uparrow} n_{i\downarrow})$, acts like a quantum sieve. For each site $i$, the operator $n_{i\uparrow} n_{i\downarrow}$ checks if the site is doubly occupied. If it is not, the operator $(1 - n_{i\uparrow}n_{i\downarrow})$ multiplies the state by $1$, leaving it alone. If the site is doubly occupied, it multiplies the state by $1-1=0$, annihilating that configuration entirely. The projector simply filters the initial BCS state, throwing away all components that violate the single-occupancy constraint.

Mathematically, this Gutzwiller projector becomes perfectly efficient in the limit of infinite repulsion, where it is equivalent to applying a correlator factor $\exp(-g \sum_i n_{i\uparrow} n_{i\downarrow})$ with $g \to \infty$. What survives this projection is a valid state of strongly correlated electrons—a mobile liquid of singlets. The hope, which continues to drive research today, is that the coherent motion of these charged, resonating singlet pairs is the very mechanism that allows for superconductivity at unprecedentedly high temperatures. The RVB state, born from the abstract puzzles of quantum magnetism, may hold the key to a technological revolution.

Now that we have grappled with the fundamental machinery of the Resonating Valence Bond (RVB) state, let us embark on a journey to see what it is for. It is one thing to build an abstract object in the mind, but its true worth is revealed only when we see what it can explain, what new worlds it opens up, and what new questions it forces us to ask. The RVB concept is not merely a curiosity; it is a veritable theoretical laboratory, a playground where some of the deepest ideas in modern physics have been conceived and tested. It has profoundly reshaped our understanding of magnetism, given us our most compelling narrative for high-temperature superconductivity, and forged unexpected bridges to the furthest frontiers of quantum information and topology.

Before the notion of RVB, the world of magnetism seemed relatively tidy. At low temperatures, the rambunctious spins in a material were expected to settle down, "freezing" into a regular, repeating pattern—a kind of crystal of magnetic moments. This preference for order is strong. But what if the stage on which the spins perform their dance is designed to frustrate their every move? Consider a triangular lattice, where each spin has six nearest neighbors. If a spin wants to be anti-aligned with its neighbors, as an antiferromagnetic interaction $J$ demands, it immediately runs into a paradox. If spin A points up, and its neighbor B points down, what is their common neighbor C to do? It cannot satisfy both A and B. This geometric frustration breaks the spell of simple order.

This is the natural habitat of the RVB state. Instead of picking a direction and ordering, the spins can relinquish their individual identities and form a collective quantum fluid of tentative partnerships—the singlet bonds. But how good is this idea? We can try to test it. When we analyze a variational RVB wavefunction on a square lattice, we find its energy is an improvement over the classical Néel state. However, it is not the true ground state, which for the unfrustrated square lattice is a long-range ordered antiferromagnet (though with a quantum-mechanically reduced magnetic moment). This is not a failure of the RVB idea, but a profound lesson. It teaches us that the quantum world is subtle; our simple approximations, such as assuming different bond configurations are completely independent, can miss the very quantum coherence that gives the RVB state its advantage. In a similar vein, if we formulate a more general variational state on the triangular lattice and then apply these strong approximations, we find that the final energy is completely independent of the details we tried to optimize. These calculations are invaluable, as they reveal the critical importance of the subtle quantum overlaps between different singlet configurations—this is the "resonance" that gives the state its name and its stability.

The RVB state doesn't exist in a vacuum; it is a contender in a grand competition of quantum phases. On the triangular lattice, it competes directly with a more complex, but still ordered, $120^\circ$ magnetic structure. At half-filling (one electron per site), the ordered state typically wins. However, if we introduce a small number of mobile holes by "doping" the system, the tables can turn. The holes find it much easier to move through the fluctuating, liquid-like background of the RVB state than through the rigid magnetic order. This kinetic energy advantage can be enough to tip the balance, causing the system to dissolve its magnetic order and melt into an RVB spin liquid above a certain critical doping level. This reveals the RVB state not as an exotic outlier, but as a key player in the rich phase diagrams of strongly correlated materials.

So, this RVB state is a liquid of singlet bonds. Can we describe the motion of this liquid? The full quantum mechanics of all the spins is forbiddingly complex. But physics often progresses by finding a simpler, effective description that captures the essential action. Here, the essential action is the "resonance," the reconfiguration of the singlets. On a square lattice, the simplest "move" that a liquid of nearest-neighbor singlets can make involves four spins on a square plaquette. Imagine a plaquette with two parallel vertical dimers. The spins can rearrange themselves into two horizontal dimers.

This simple "flip" is the elementary motion of the dimer liquid. Astonishingly, one can show that the entire complex dynamics of the underlying Heisenberg spin model, when projected into this RVB subspace, boils down to a simple-looking Hamiltonian for the dimers themselves. This is the ​​Quantum Dimer Model (QDM)​​. It has a kinetic term, with amplitude $t$, that corresponds to this flipping process on a plaquette, and a potential term, with strength $v$, that gives a different energy to a plaquette depending on whether it is "flippable" or not. A complex story of interacting spins becomes a simpler, yet still fantastically rich, story of dancing dimers. This model is a beautiful example of emergence, where the low-energy collective behavior is described by degrees of freedom—the dimers—that are different from the microscopic constituents.

For decades, magnetism and superconductivity were seen as antithetical. Superconductivity, as described by the Bardeen-Cooper-Schrieffer (BCS) theory, involves electrons forming pairs (Cooper pairs) that move without resistance. Magnetism, on the other hand, tends to break these pairs. The discovery in 1986 of superconductivity in copper-oxide materials (cuprates) at unprecedentedly high temperatures shattered this paradigm. These materials are, in their undoped state, antiferromagnetic Mott insulators—materials that should be metals but are forced into insulating behavior by immensely strong electron-electron repulsion. How could superconductivity possibly emerge from such a state?

It was here that the RVB concept had its most triumphant moment. P. W. Anderson proposed a revolutionary idea: perhaps the "glue" holding the Cooper pairs together was not the lattice vibrations of conventional BCS theory, but the very same magnetic interactions that cause antiferromagnetism. The parent state is a Mott insulator, which can be thought of as a realization of the half-filled Hubbard model in the limit of strong on-site repulsion $U$. In this limit, the system is better described by the $t-J$ model, which captures the essential physics: electrons can hop between neighboring sites (the $t$ term) but are forbidden from ever occupying the same site, and their spins interact antiferromagnetically (the $J$ term).

The undoped parent state is an RVB spin liquid, a sea of pre-formed, charge-neutral spin singlets. Now, what happens when you dope this system by removing a few electrons? You create mobile "holes." The core of the RVB theory of superconductivity is that the elementary particles in this strange new world are not electrons, but separated spin and charge carriers: neutral, spin-1/2 ​​spinons​​ and spinless, charge-$e$ ​​holons​​. The spinons are the constituents of the singlets, already paired up in the parent RVB state. The holons are the mobile holes. If these bosonic holons undergo Bose-Einstein condensation—a collective quantum phenomenon where they all fall into a single coherent state—something magical happens. Their coherence "unveils" the pre-existing pairing of the spinons, converting the neutral singlet pairs into charged, mobile electron pairs. The system becomes a superconductor.

This beautiful narrative can be given a concrete mathematical form. We start with a standard BCS wavefunction, which describes a gas of paired electrons, but which completely ignores the strong repulsion. Then, we apply a "Gutzwiller projector," an operator whose sole job is to go through the wavefunction and mercilessly eliminate any configuration where two electrons are on the same site. The result, a Gutzwiller-projected BCS state, is the variational embodiment of the RVB superconductor. Furthermore, because the pairing is mediated by short-range antiferromagnetic correlations that are repulsive at zero distance, the pairs must arrange themselves in a clever way to avoid each other. On the square lattice, this naturally leads to a pairing state with $d_{x^2-y^2}$ symmetry, where the pair wavefunction changes sign between the horizontal and vertical directions—one of the key experimental signatures of the cuprate superconductors.

The RVB idea extends far beyond the stage of frustrated magnets and superconductors, providing crucial links to both real-world experiments and the most abstract frontiers of modern theory.

How would an experimentalist "see" a spin liquid? An ordered magnet has spins arranged in a periodic pattern, which acts like a crystal lattice for neutrons. A beam of neutrons will scatter off this pattern and produce sharp, bright spots—magnetic Bragg peaks—at specific angles. A liquid, by contrast, has no long-range order. Pouring water on the floor doesn't create a crystal. Similarly, the liquid-like correlations in an RVB state do not produce sharp peaks. Instead, neutron scattering experiments would reveal broad, diffuse patterns of intensity, a direct fingerprint of the short-range, fluctuating singlet correlations.

Perhaps the most profound connection is to the field of topology. An RVB spin liquid is one of the simplest examples of a system with ​​topological order​​. This is a new kind of order that is not described by any local pattern of spins, but by the global, topological properties of the quantum wavefunction itself. This long-range entanglement can be quantified by a remarkable quantity called the ​​topological entanglement entropy​​, denoted $\gamma$. It can be extracted from the entanglement of a region with its surroundings through an ingenious subtraction scheme that cancels out all the non-universal, boundary-dependent parts. For the simplest RVB states, which realize a so-called $\mathbb{Z}_2$ topological order, this universal number is found to be $\gamma = \ln(2)$. This value is not arbitrary; it is a direct reflection of the exotic "anyon" particles that live in this state—the spinons and their dual partners, the visons. This connects the RVB state directly to the quest for building a fault-tolerant quantum computer, whose qubits could be encoded in the topological properties of such anyons.

Finally, the abstract idea of an RVB state has found concrete expression in the language of modern computational physics. Tensor networks, such as ​​Projected Entangled-Pair States (PEPS)​​, provide a natural framework for representing the entanglement structure of strongly correlated states. An RVB state can be elegantly constructed as a PEPS, where the local tensors are built to enforce the singlet-pairing structure. This powerful representation allows for high-precision numerical calculations of properties like the correlation length, connecting the spectrum of the system's transfer matrix directly to how quickly correlations decay.

From an intuitive picture of frustrated spins, the RVB concept has blossomed into a unifying theme, weaving together magnetism, superconductivity, and topology. It stands as a powerful testament to how a simple, physically motivated idea can illuminate a vast and complex landscape, forever changing the way we think about the collective quantum world.