Yang-Lee Zeros

The boiling of water, the magnetization of iron—these dramatic changes, known as phase transitions, are cornerstones of our physical world. Yet, a fundamental paradox lies at the heart of their theoretical description: the mathematical framework of statistical mechanics, when applied to any finite system, predicts only smooth, gradual changes. True sharpness seems impossible. How, then, does the real world manage these abrupt transformations? This question points to a profound gap in our understanding, a puzzle that was brilliantly solved by C. N. Yang and T. D. Lee in 1952.

Their solution was to venture beyond the realm of physical variables into the abstract landscape of the complex plane. This article explores their groundbreaking concept of Yang-Lee zeros. The first chapter, ​​Principles and Mechanisms​​, will uncover the core theory, explaining how invisible zeros of the partition function govern the existence of phase transitions and exploring elegant results like the Yang-Lee Circle Theorem. Subsequently, the chapter on ​​Applications and Interdisciplinary Connections​​ will reveal the astonishing universality of this idea, showcasing its power to unify concepts across statistical mechanics, quantum physics, and even the field theories describing fundamental forces.

Imagine trying to build a perfect, sharp mountain peak by piling up smooth, round stones. No matter how many stones you use, if the number is finite, the summit will always be a little bit rounded. You can make it steeper and steeper, but you'll never achieve a true mathematical point. This simple analogy captures a deep truth about the world of statistical mechanics: phase transitions, the sharp, dramatic changes we see when water boils or a magnet loses its power, are fundamentally phenomena of the infinite.

Let's think about a system in the grand canonical ensemble, like a gas of particles in a box of a fixed volume $V$ and temperature $T$. The system can exchange particles with a large reservoir, and its behavior is governed by a quantity called the ​​grand partition function​​, denoted by $\Xi$. This function is essentially a weighted sum over all possible numbers of particles, from zero up to the maximum that can fit. For a simple model like a lattice gas on $M$ sites, it takes the form of a polynomial:

Here, $z$ is a variable called the ​​fugacity​​, which is like a knob we turn to control the average number of particles in the box; it's related to the chemical potential $\mu$ by $z = \exp(\beta \mu)$, where $\beta = 1/(k_B T)$. The coefficients $Q_N$ are all positive numbers, representing the statistical weight of having exactly $N$ particles.

Now, all the thermodynamic properties we care about, like pressure, are derived from the logarithm of this function, $\ln \Xi$. A phase transition, in mathematical terms, is a point where this function is non-analytic—a point where its derivative is discontinuous or infinite, creating a "sharp corner" like our mountain peak. But look at our expression for $\Xi(z)$. For any real, physical value of fugacity ($z > 0$), every term in the sum is positive. The sum itself, therefore, can never be zero. Since $\Xi(z)$ is a smooth, positive, and non-zero function, its logarithm, $\ln \Xi(z)$, must also be perfectly smooth and analytic. There are no sharp corners. No phase transitions. This is a crucial conclusion: for any system of finite size, a true phase transition is impossible.

So where does the boiling of water come from? If finite systems can't have phase transitions, how does the real world, which we often model as a vast but finite collection of atoms, manage it? The answer, proposed in a groundbreaking insight by C. N. Yang and T. D. Lee in 1952, is that we must take two bold steps: allow the system size to become infinite (the ​​thermodynamic limit​​) and allow the fugacity $z$ to become a complex number.

While for real, positive $z$, the partition function $\Xi(z)$ can never be zero, it can be zero for certain values of $z$ in the complex plane. These complex roots are the celebrated ​​Yang-Lee zeros​​. They are the points where $\ln \Xi$ would blow up, its singularities. For any finite system, these zeros are like invisible mines floating in the complex sea, always kept at a safe distance from the real line where the physical world resides.

The magic happens in the thermodynamic limit ($M \to \infty$). As the system grows, the number of zeros also grows, and they begin to move. A phase transition occurs if, and only if, in this limit, the zeros migrate until they form a continuous line or curve that finally touches, or "pinches," the positive real axis at some point, say $z_c$. At that precise point, the singularity that was hidden in the complex plane crashes into the real world. The pressure, which is proportional to $\lim_{M\to\infty} \frac{1}{M}\ln \Xi$, suddenly becomes non-analytic. A phase transition is born.

This picture also gives us a beautiful way to understand fluctuations. The variance in the number of particles, a measure of how much the particle count fluctuates, is given by $\langle (\Delta N)^2 \rangle = z \frac{\partial \langle N \rangle}{\partial z}$. At a phase transition, like the boundary between liquid and gas, these fluctuations become enormous—the system can't decide which phase to be in. Mathematically, this corresponds to the second derivative of $\ln \Xi$ diverging, which is exactly what happens when a zero hits the real axis.

This might seem like a beautiful but rather abstract story. Can we actually find these zeros? For some simple but important models, the answer is a resounding yes. One of the most elegant results in all of statistical physics is the ​​Yang-Lee Circle Theorem​​. It states that for a whole class of models known as ferromagnetic Ising models (which describe magnets, but can also be mapped to lattice gases with attractive forces), all the Yang-Lee zeros lie exactly on a circle in the complex fugacity plane.

Let's consider the simplest case: a one-dimensional chain of magnetic spins. This model is famous for not having a phase transition at any non-zero temperature. And the Yang-Lee zeros tell us exactly why! Using a powerful technique called the transfer matrix method, one can calculate the location of the zeros precisely. It turns out they indeed lie on the unit circle $|z|=1$. However, they don't cover the entire circle. There is always a ​​gap​​ in the distribution of zeros around the point $z=1$, which corresponds to zero external magnetic field. Because of this gap, the zeros can never actually touch the real axis at the physically relevant point. No pinching, no phase transition. The theory works perfectly, predicting not only when transitions happen, but also when they don't.

What happens in systems that do have a phase transition, like the two-dimensional Ising model? At the critical temperature, $T_c$, the gap in the circle of zeros closes. The zeros become dense on the circle, right up to the point $z=1$. This boundary point, where the sea of zeros meets the land of real physics, is called the ​​Yang-Lee edge​​.

Near this edge, the density of zeros exhibits a fascinating universal behavior. If we parameterize the unit circle by an angle $\theta$, so $z=e^{i\theta}$, then the density of zeros near the edge ($\theta \to 0$) follows a power law:

The exponent $\sigma$ is called the ​​Yang-Lee edge singularity exponent​​. The amazing thing is that this exponent is connected to the standard critical exponents that are measured in experiments!

To see this, we take an audacious leap of faith, following an argument originally due to Michael Fisher. We know that at the critical temperature, the magnetization $M$ scales with a real magnetic field $H$ as $M \sim H^{1/\delta}$, where $\delta$ is a famous critical exponent. What if we assume this simple scaling law holds not just for real $H$, but can be analytically continued into the complex plane, for an imaginary magnetic field $H=iy$? And what if we further assume that the density of zeros $g(y)$ is proportional to the real part of this complex magnetization, $\text{Re}[M(T_c, H=iy)]$?

Performing this simple calculation, one finds a stunningly beautiful and simple result:

This is remarkable. The geometric distribution of abstract mathematical zeros in the complex plane is directly tied to a physical, measurable property of the critical point. For the 2D Ising model, it is known that $\delta = 15$, which would predict $\sigma = 1/15$. However, the problem is more subtle, as the mapping between variables is key. In fact, a different line of reasoning, connecting the problem to non-hermitian quantum mechanics, correctly predicts that for the 2D Ising model, the exponent is $\sigma = -1/2$. This negative value means the density of zeros actually diverges as it hits the real axis, a dramatic signature of criticality.

The story doesn't end with magnetic fields and fugacity. We can play the same game with temperature. The ​​canonical partition function​​, $Z_N(\beta)$, which describes a system with a fixed number of particles, can also be studied in the complex plane—this time, the plane of complex inverse temperature $\beta$. The zeros of $Z_N(\beta)$ are called ​​Fisher zeros​​.

Just like their Yang-Lee counterparts, Fisher zeros for a finite system lie off the real axis. In the thermodynamic limit, they can migrate and pinch the real $\beta$-axis to produce a temperature-driven phase transition. This reveals a grand, unified picture: phase transitions are governed by the analytic structure of partition functions, and the location of their zeros in the complex planes of their respective control variables (fugacity for Yang-Lee, temperature for Fisher) tells us everything.

This framework becomes extraordinarily powerful when combined with the theory of ​​finite-size scaling​​. By studying how the zeros move as we change the system size $L$, we can classify the nature of the phase transition.

• For a ​​first-order transition​​ (like boiling), where two phases coexist, the zero closest to the real axis moves very quickly. Its distance to the axis scales as $L^{-d}$, where $d$ is the spatial dimension of the system.
• For a ​​continuous (second-order) transition​​, the approach is much slower. The distance of the closest Fisher zero scales as $L^{-1/\nu}$, where $\nu$ is the critical exponent governing the divergence of the correlation length.

This provides a practical and powerful computational tool. To measure the critical exponent $\nu$ for a model, we can simulate it at various sizes $L$, find the Fisher zero closest to the real axis for each size, and see how its distance from the axis scales with $L$. The dance of the zeros, as the system size grows, plays out a universal tune, and by listening carefully, we can decode the fundamental exponents that govern the physics of critical phenomena. From a simple puzzle about sharp corners, we have journeyed into the complex plane and emerged with a profound and unified understanding of the cooperative behavior of matter.

We have journeyed into the abstract world of complex numbers to find the roots of a physical function, the partition function. We discovered that these mathematical "ghosts"—the zeros of Yang and Lee—are the hidden architects of phase transitions. But the story does not end with a pot of water boiling or a magnet losing its power. The true genius of this idea lies in its breathtaking universality. It is like discovering a strange key that not only opens one door but unlocks a whole series of doors into seemingly unrelated wings of the scientific mansion. Let us now take a walk through these rooms and witness the astonishing reach of Yang and Lee's insight.

Our first steps are in the familiar territory of statistical mechanics, the idea's birthplace. Here, the Yang-Lee framework provides a wonderfully precise lens for dissecting the behavior of interacting systems. For any finite system of spins, the partition function is a simple polynomial in the fugacity $z=e^{2\beta h}$, and its zeros can be found directly. More than just finding them, we can study how they behave. For instance, in a simple mean-field model, we can precisely calculate how the locations of the zeros "flow" across the complex plane as we change physical parameters like temperature or the coupling strength between spins. This gives us a dynamic picture of how the precursors to a phase transition evolve with the system's conditions.

Of course, real materials are not just amorphous blobs of interacting particles; they have structure. The geometry of a system's connections, its lattice, leaves a deep imprint on the distribution of its zeros. Moving beyond simple models, we can consider more intricate arrangements like a Husimi cactus—a lattice made of triangles sharing corners. For such a system, the zeros in the thermodynamic limit no longer appear as discrete points but coalesce into continuous arcs on the unit circle. The exact shape and extent of these arcs are determined by the lattice geometry and interaction strengths, providing a beautiful link between microscopic structure and macroscopic critical behavior.

The theory's power truly shines when we analyze more complex phase diagrams. Consider a material where spins can point not just up or down, but can also be in a non-magnetic "zero" state, as described by the Blume-Emery-Griffiths model. This model exhibits a rich tapestry of phase transitions, including lines of first-order transitions meeting second-order lines at special tricritical points. The Yang-Lee zeros provide a map to this complexity. By tuning a parameter in the model, one can observe a remarkable transformation: the zeros, once neatly arranged on the unit circle, can abruptly snap onto the negative real axis. This dramatic change in the geometry of the zero locus is not just a mathematical curiosity; it signals a fundamental change in the physical nature of the phase transition itself.

So far, we have spoken of thermal transitions driven by the chaos of heat. What happens when we cool a system to absolute zero? Temperature becomes irrelevant, but phase transitions can still occur, driven now by the weirdness of quantum mechanics itself. These are quantum phase transitions (QPTs), and the Yang-Lee concept makes the leap into this new realm with astonishing grace.

At zero temperature, the system settles into its ground state. The role of the partition function is now played by the ground state energy, and its non-analyticities as a function of some parameter—like a magnetic field—mark the QPTs. The Yang-Lee zeros become the complex "echoes" of these quantum critical points. For instance, in a quantum Ising model where a transverse magnetic field $\Gamma$ competes with the ferromagnetic ordering, we can find the quantum critical points by seeking where the energy landscape develops a flat inflection point. This condition pinpoints the locations of the critical fields in the complex plane, which are precisely the points where the zeros would accumulate in the thermodynamic limit.

This quantum extension is not limited to magnetism. In the world of ultracold atoms, physicists create exotic states of matter described by the Bose-Hubbard model. Here, the crucial knob is not a magnetic field but the chemical potential $\mu$, which controls the number of particles. The phase transition between a superfluid (where atoms move freely) and a Mott insulator (where they are locked in place) can be understood by studying the zeros of the grand partition function in the complex plane of the particle fugacity $z=e^{\beta\mu}$. Even for a tiny two-site system, the zeros trace a perfect circle, with a radius determined by the strength of the on-site particle interactions, elegantly mapping out the system's propensity to change its state.

The concept's journey into the quantum world leads to even more modern frontiers, such as quantum information theory. A key feature of a quantum state is its entanglement—the spooky connection between its parts. It turns out that the sensitivity of a subsystem's entanglement to changes in a system parameter can be quantified by a tool called the Entanglement Fisher Information. In a fascinating parallel to the original theory, one can study the zeros of this information-theoretic quantity in the complex plane. These "entanglement Fisher zeros" have been shown to act as signposts for quantum phase transitions, suggesting a profound and beautiful connection between criticality, information, and the very fabric of quantum entanglement.

The world of equilibrium, where things are static and settled, is an idealization. The real universe is a dynamic, fluctuating place. Remarkably, the Yang-Lee idea finds one of its most powerful applications in this chaotic realm of non-equilibrium physics.

Imagine tracking the total number of particles hopping across a junction in a microscopic conductor over a long time. This quantity fluctuates. The statistics of these fluctuations are encoded in a mathematical object called a moment generating function, which we can think of as a "dynamical partition function." The variable in this function is no longer related to a physical field, but is a purely mathematical "counting field" $\lambda$ that keeps track of the current. The zeros of this function in the complex $\lambda$-plane signal "dynamical phase transitions"—abrupt changes in the character of the system's fluctuations. Studying these zeros in models like the Symmetric Simple Exclusion Process (SSEP) reveals a hidden order in the seemingly random dance of particles.

This framework provides especially deep insights when applied to entropy production, the hallmark of irreversible, out-of-equilibrium processes. Consider a tiny quantum system, like a single two-level atom, coupled to two heat baths at different temperatures. It is constantly being kicked by the hot bath and cooled by the cold one, generating a flow of entropy. We can again define a generating function to count the exchanges of energy quanta with a bath. Its zeros—the dynamical Yang-Lee zeros—are not located arbitrarily. One zero is trivially at $z=1$, reflecting probability conservation. But the other, non-trivial zero is found at a location $z_0 = \exp((\beta_H - \beta_C)\epsilon)$ that is directly related to the temperature difference between the baths and the energy spacing of the system. This is a manifestation of the celebrated Fluctuation Theorems, linking the zeros to the fundamental thermodynamic forces driving the system away from equilibrium.

Our final stop is at the highest planes of theoretical physics, where the Yang-Lee concept forges a profound unity between the behavior of everyday materials and the fundamental laws of nature.

The power-law singularity observed in the density of zeros near a critical point—the Yang-Lee edge singularity—is not just a property of one model. It is a universal feature, shared by a vast class of systems. This universality points toward a deeper description. In one of the great triumphs of theoretical physics, it was shown that this singularity can be described exactly by a specific quantum field theory (QFT): a theory of a scalar field $\phi$ with a cubic self-interaction, $\phi^3$, and a purely imaginary coupling constant. This extraordinary link means we can use the powerful machinery of the renormalization group, developed for particle physics, to calculate the universal critical exponent $\sigma$ that governs the singularity. This bridges the worlds of condensed matter and high-energy physics in a spectacular way.

The idea of studying zeros of partition functions in a complex parameter plane extends naturally to the theories that describe the fundamental forces. In lattice gauge theory, the method used to simulate the strong nuclear force that binds quarks into protons and neutrons, the partition function depends on a coupling constant $\beta$. By allowing this coupling to be a complex variable, physicists can search for zeros. These zeros, often called Fisher zeros in this context, signal the presence of phase transitions, most notably the transition between a state of confined quarks and a deconfined quark-gluon plasma—the state of the universe moments after the Big Bang. Even a toy model on a single lattice plaquette reveals the existence of these zeros, offering a window into the phase structure of the universe's fundamental constituents.

From a simple magnet to the quantum vacuum, the message is clear. A beautifully simple mathematical question—"Where are the zeros?"—has become a unifying principle across physics. It reveals a hidden architecture underlying the collective behavior of matter and energy, a testament to the deep and often surprising connections that knit our physical world into a coherent whole.