Onsager-Casimir Relations

In the physical world, processes are rarely isolated. A temperature gradient can generate an electric voltage, and an electric current can transport heat. These ​​coupled flows​​ are central to understanding systems away from thermodynamic equilibrium. For a long time, the relationships governing these cross-phenomena were treated as independent empirical facts, creating a knowledge gap and requiring extensive, separate measurements. This article addresses this by exploring the profound symmetry that secretly governs them: the ​​Onsager-Casimir relations​​.

The following chapters will guide you through this fundamental principle. In "Principles and Mechanisms," we will uncover the origins of this symmetry, tracing it back to the elegant concept of microscopic time-reversal, and see how it is modified by external magnetic fields. Following that, "Applications and Interdisciplinary Connections" will demonstrate the remarkable predictive power of these relations, showing how they forge unexpected connections between thermoelectric effects in solids, complex behaviors in soft matter, and even the physics of neutron stars.

Imagine you are in a kitchen. You turn on a stove burner under a pot of water. The most obvious thing that happens is that heat flows from the burner into the water. But other things happen too, don't they? The water starts to circulate, creating currents. Steam is produced, which can build up pressure. The world is full of these interconnected processes. When you push on one part of a system, other, seemingly unrelated parts respond. In physics, we call these ​​coupled flows​​.

A classic example is in a simple piece of metal. If you create a voltage difference across it, an electric current flows. That’s Ohm's law. If you create a temperature difference, heat flows. That's the law of heat conduction. But what happens if you do both? Or what if doing one causes the other? In many materials, applying a temperature gradient can generate a voltage (the ​​Seebeck effect​​), and driving an electric current can cause the material to heat up or cool down (the ​​Peltier effect​​). It's a two-way street. Heat flow and electric current are coupled.

When things are not too wild—that is, when the system is close to a state of calm equilibrium—these couplings are often beautifully simple and linear. A flow (like electric current, $J_e$, or heat current, $J_q$) is proportional to the "forces" driving it (like a voltage gradient, $X_e$, or a temperature gradient, $X_q$). We can write this down like a recipe:

$J_e = L_{ee} X_e + L_{eq} X_q$

$J_q = L_{qe} X_e + L_{qq} X_q$

The coefficients, the $L_{ij}$s, are the "transport coefficients." They tell us how a given force produces a given flow. $L_{ee}$ is just related to the familiar electrical conductivity. $L_{qq}$ is related to thermal conductivity. But the interesting parts are the ​​cross-coefficients​​: $L_{eq}$ tells you how much electric current you get from a temperature difference (the Seebeck effect), and $L_{qe}$ tells you how much heat is carried along by an electric current (the Peltier effect).

For centuries, physicists and engineers measured these coefficients for different materials. They were just numbers, properties of the material you had to look up in a book. It seemed that $L_{eq}$ and $L_{qe}$ were two completely independent facts about the material. You would need two separate, difficult experiments to measure them. But would you?

In 1931, a Norwegian-American chemist named Lars Onsager dropped a bombshell. He showed, using an argument of breathtaking elegance, that these cross-coefficients are not independent at all. In fact, they must be equal.

$L_{eq} = L_{qe}$

This is one of the most profound and, frankly, useful statements in all of non-equilibrium physics. It's called an ​​Onsager reciprocal relation​​. Think about what it means. The coefficient that describes how a temperature gradient creates a voltage is exactly the same as the coefficient that describes how a voltage gradient creates a heat flow. The two-way street is perfectly symmetric! This immediately cuts the number of experiments you need to do in half. If you measure the Seebeck effect, you automatically know the Peltier effect, and vice versa. It’s like finding a cheat code for thermodynamics.

This symmetry isn't just limited to heat and electricity. It applies to any pair of coupled irreversible processes near equilibrium. For instance, in a chemical reaction network, the influence of reaction A's driving force (its "affinity") on the rate of reaction B is the same as the influence of reaction B's affinity on the rate of reaction A. It's a universal law.

So, where does this miraculous symmetry come from? It does not come from the Second Law of Thermodynamics. The Second Law tells us that entropy must increase (or stay the same), which translates to a condition on the symmetric part of the $L$ matrix—ensuring that overall, processes are dissipative. The symmetry of the cross-terms, however, comes from a much deeper, spookier place: ​​microscopic reversibility​​.

Imagine you have a movie of a box full of gas molecules, bouncing off each other and the walls. Now, play the movie in reverse. Does it look strange? No, not really. Each collision, viewed in reverse, is a perfectly valid collision. The underlying laws of motion—whether it's Newton's laws for billiard balls or Schrödinger's equation for atoms—are time-reversal symmetric. They don't have a preferred direction for the arrow of time.

Onsager's genius was to connect this microscopic time-invariance to the macroscopic transport coefficients. His argument, simplified, goes something like this: In any system at equilibrium, there are constant, tiny, random fluctuations. A few more molecules might bunch up here, or a little hot spot might appear there for a split second. Because the system is in equilibrium, these fluctuations, on average, die away and return to equilibrium. Now consider the correlation between a fluctuation of one type (say, in temperature) and a later fluctuation of another type (say, in electric potential). The principle of microscopic reversibility implies that the correlation between a temperature fluctuation now and a voltage fluctuation a short time $\tau$ later is the same as the correlation between a voltage fluctuation now and a temperature fluctuation a time $\tau$ later. When you work through the mathematics, this deep symmetry of fluctuations near equilibrium is what forces the macroscopic coefficients $L_{ij}$ to be symmetric.

This beautiful story seems perfect. But we have to be careful. What if we introduce something into our system that does have a sense of time's arrow? The classic example is a ​​magnetic field​​.

Imagine a single electron moving through a magnetic field. It follows a curved path due to the Lorentz force. Now, let's play the movie backwards. The electron retraces its path, but its velocity is reversed at every point. With a reversed velocity, the Lorentz force ($q \mathbf{v} \times \mathbf{B}$) also flips its direction, and the electron would curve the wrong way to stay on the original path! To make the reversed movie look right, you have to do something else: you must also reverse the direction of the magnetic field vector, $\mathbf{B} \to -\mathbf{B}$.

Quantities like position, density, and energy are ​​time-even​​; they look the same in the reversed movie. Quantities like velocity, momentum, and magnetic field are ​​time-odd​​; they flip their sign.

So, what does this do to our reciprocal relations? Onsager and, later, Hendrik Casimir figured this out. The simple symmetry is modified. The new rule, known as the ​​Onsager-Casimir reciprocal relations​​, states that the coefficients are related, but you must flip the sign of the magnetic field in the comparison:

$L_{ij}(\mathbf{B}) = L_{ji}(-\mathbf{B})$

This is for the simple case where the quantities being transported (like charge and heat) are both time-even. Let's look at the consequences for the coefficients. The diagonal coefficients, like electrical resistance, must obey $L_{ii}(\mathbf{B}) = L_{ii}(-\mathbf{B})$. This means the resistance of a material should be an even function of the magnetic field—its value should be the same for a "north pole up" field as a "north pole down" field. This is a concrete, testable prediction!

The off-diagonal terms are even more interesting. For something like the Hall effect, where an electric field in one direction ($X_x$) produces a current in a perpendicular direction ($J_y$), the relation is $L_{yx}(\mathbf{B}) = L_{xy}(-\mathbf{B})$. Experiments and theory show that the Hall coefficient ($L_{yx}$) reverses its sign when the magnetic field is reversed. This is perfectly consistent with the Onsager-Casimir relations!. A non-zero Hall effect is, in a way, a direct macroscopic manifestation of broken time-reversal symmetry by the magnetic field.

We can now state the full, glorious rule that includes all possibilities. What if the quantities themselves are time-odd? Spin, for example, is like a tiny magnetic moment—it's time-odd. The full Onsager-Casimir relation includes a "parity factor," $\epsilon_i$, for each quantity, which is $+1$ if it's time-even (like charge) and $-1$ if it's time-odd (like spin).

$L_{ij}(\mathbf{B}) = \epsilon_i \epsilon_j L_{ji}(-\mathbf{B})$

Let's see what this means for different couplings:

• ​​Charge-Charge (cc):​​ $\epsilon_c = +1$. So $L^{cc}(\mathbf{B}) = (+1)(+1) L^{cc}(-\mathbf{B})$, just as we saw.
• ​​Spin-Spin (ss):​​ $\epsilon_s = -1$. So $L^{ss}(\mathbf{B}) = (-1)(-1) L^{ss}(-\mathbf{B}) = L^{ss}(-\mathbf{B})$. The spin conductivity tensor has the same symmetry as the charge conductivity tensor.
• ​​Charge-Spin (cs):​​ This is the cool one! This describes how a voltage can create a spin current (the spin Hall effect) or how a spin accumulation can create a charge current (the inverse spin Hall effect). We have $\epsilon_c=+1$ and $\epsilon_s=-1$. The relation becomes: $L^{cs}(\mathbf{B}) = (+1)(-1) L^{sc}(-\mathbf{B}) = - L^{sc}(-\mathbf{B})$

The coupling between charge and spin has an intrinsic minus sign in its reciprocity! This sign is not a quirk; it is a direct consequence of the fundamental time-reversal properties of charge and spin.

You might think this is all just elegant theory, but these principles are a workhorse in modern condensed matter physics and spintronics. Consider a high-tech "nonlocal spin valve" device where you inject spin-polarized electrons at one point and detect them at another. We can define a resistance $R_A$ for this process. Now, we swap the injector and detector and measure a new resistance, $R_B$.

Are $R_A$ and $R_B$ the same? Our first instinct, thinking about simple circuits, might be "yes." But the device contains ferromagnets with magnetization $\mathbf{M}$ (which is time-odd, just like $\mathbf{B}$). The Onsager-Casimir relations give us the definitive answer:

$R_A(\mathbf{B}, \mathbf{M}) = R_B(-\mathbf{B}, -\mathbf{M})$

So, in general, for a fixed magnetic field and magnetization, $R_A \neq R_B$. This "non-reciprocal" behavior isn't a failure of the theory or a sign of some messy nonlinear effect. It is a direct, predicted signature of broken time-reversal symmetry! The only way to restore the equality is to reverse all the time-odd quantities. Modern experiments use precisely this signature to probe the intricate interplay of charge and spin in new materials.

From the simple observation of coupled flows to the esoteric world of spintronics, the Onsager-Casimir relations provide a golden thread. They are a stunning example of how a deep symmetry in the microscopic world—the indifference of physical law to the direction of time's arrow—imposes powerful and practical constraints on the macroscopic world we observe and engineer. It is a beautiful piece of physics.

After a journey through the fundamental principles of near-equilibrium thermodynamics, you might be left with a feeling of abstract satisfaction. We have a powerful theoretical tool, born from the subtle idea of microscopic time-reversal symmetry. But what is it for? What good is knowing that the matrix of kinetic coefficients possesses a certain symmetry? The answer, it turns out, is wonderfully far-reaching. The Onsager-Casimir relations are not just a piece of theoretical housekeeping; they are a master key, unlocking hidden connections across vast and disparate fields of science. They reveal a secret harmony in the world of irreversible processes, a kind of "if you tell me this, I can tell you that" which would otherwise seem magical. Let’s now explore this magic, and see how this one principle weaves its way through solids, fluids, and even the hearts of dead stars.

Let's begin with something familiar: a piece of metal. We know that if we apply a voltage, a current flows. If we put this metal in a magnetic field, something more interesting happens. The electrons, as they flow, are deflected sideways by the magnetic force, piling up on one side of the conductor. This creates a transverse voltage, an effect every student of physics knows as the Hall effect. The Onsager-Casimir relations are deeply embedded here. They demand a specific symmetry between the components of the resistivity tensor, $\rho_{xy}(B) = -\rho_{yx}(B)$. This isn't just a convenient property; it's a fundamental constraint. Indeed, microscopic theories like the simple Drude model must, and do, obey this rule. It is this very relationship that allows us to find the famous result for the Hall coefficient, $R_H = -1/(ne)$, a workhorse of condensed matter physics that lets us count the charge carriers in a material and determine their sign.

But the story gets richer when we stir heat into the mix. Consider two peculiar, yet very real, phenomena known as thermomagnetic effects.

• First, imagine sending a longitudinal flow of heat down a metallic bar that sits in a magnetic field. Amazingly, a transverse electric voltage appears. This is the ​​Nernst effect​​. It's as if the heat flow, guided by the magnetic field, somehow "steers" the electrons sideways.

• Now, consider the reverse experiment. Forget the heat flow, and instead drive an electric current longitudinally down the same bar. The reciprocal magic happens: a transverse temperature gradient develops. Heat is shuttled to one side. This is the ​​Ettingshausen effect​​.

On the surface, these two effects seem like distinct curiosities. Why should they have anything to do with one another? One involves a thermal gradient creating a voltage; the other involves a current creating a thermal gradient. This is where Onsager steps in and reveals the hidden script. The theory predicts an unequivocal, quantitative relationship between the Nernst coefficient $N$ and the Ettingshausen coefficient $P$. The relationship is startlingly direct: $N T = \kappa_{xx} P$, where $T$ is the temperature and $\kappa_{xx}$ is the thermal conductivity. This isn't a vague analogy; it's a rigid equation. Measuring one effect allows you to predict the outcome of the other. This remarkable prediction, confirmed by experiment, was one of the first great triumphs of the theory, transforming it from an elegant speculation into a powerhouse of physics. It assures us that even when we are building phenomenological models for new materials, any proposed transport coefficients must obey these symmetries to be physically plausible.

The unity of physics lies in the fact that its fundamental principles don't care whether they are acting on electrons in a copper wire or proteins in a cell. The Onsager-Casimir relations are no exception. Their domain extends far beyond the rigid lattice of a crystal into the squishy, complex world of soft matter and biophysics.

Imagine a dilute solution of charged macromolecules, like DNA or proteins, subjected to various forces. In one experiment, we apply an electric field and watch the molecules drift. If we also apply a magnetic field, we might observe a small sideways motion—a kind of ionic Hall effect. In a second, completely different experiment, we turn off the electric field and instead physically drag the molecules through the solvent, and we again observe a transverse drift. Is there a connection between the sideways drift caused by an electric push and the one caused by a mechanical pull? Common sense offers no guide, but the Onsager relations do. They establish a precise link between the response coefficients of these two seemingly unrelated processes, showing that nature's accounting is impeccably consistent.

The stage gets even more exotic when we enter the realm of liquid crystals, the substances that make our computer and television screens work. These materials consist of rod-like molecules that possess orientational order. Here, the irreversible processes include not only the flow of heat and charge, but also the "flow" of orientation—the collective rotation of the molecules. This leads to truly mind-bending cross-effects.

• In one effect, dubbed the ​​thermomechanical torque​​, applying a temperature gradient across a liquid crystal can create a torque that tries to reorient the molecules. You can literally twist the microscopic structure of the material just by heating it unevenly.

• The reciprocal phenomenon, predicted by Onsager's relations, is that forcing the molecules to rotate must generate a flow of heat! This "rotational heat flux" is profoundly non-intuitive, yet it is a necessary consequence of the underlying symmetry. The relations provide a direct, simple formula connecting the two coefficients: $C = -\alpha/T$.

In special "chiral" fluids, composed of molecules that have a handedness (like a screw), similar couplings can exist between an electric field and the fluid's rotation. Applying an electric field can induce a bulk torque on the fluid (the electrorotational effect), while physically rotating the fluid can induce an electric polarization current (the gyroelectric effect). Once again, the Onsager-Casimir relations connect the coefficients, $\alpha$ and $\beta$, of these two phenomena with stunning simplicity: $\alpha = -\beta$. One effect is the perfect anti-mirror of the other, a relationship that would be utterly mysterious without the principle of microscopic reversibility.

To truly appreciate the universality of these relations, we must push them to the frontiers of physics—to the ultra-cold world of quantum mechanics and the ultra-dense realm of astrophysics.

Consider a Bose-Einstein Condensate (BEC), a state of matter where millions of atoms behave as a single quantum entity. If we stir this quantum fluid, we can create a "quantum vortex," a tiny, stable whirlpool where the circulation is quantized. What happens when this vortex interacts with heat? The theory of irreversible processes predicts a "thermal Magnus force": a temperature gradient will push the vortex sideways, perpendicular to the gradient. Now for the reciprocal prediction: if you turn off the temperature gradient and instead drag the vortex through the BEC with an external force, the vortex itself must act as a heat pump, carrying a heat current perpendicular to its motion. The coefficient describing this transported heat is directly related to the coefficient of the thermal Magnus force. This is a profound link between a macroscopic concept (thermodynamics) and a purely quantum phenomenon (a quantized vortex).

From the impossibly cold, let's journey to the impossibly dense: the core of a neutron star. Here, matter exists as a bizarre soup of neutrons, protons, electrons, and possibly even free quarks in a "color-superconducting" state, all under the influence of crushing gravity and colossal magnetic fields. Physicists trying to model these objects need to understand how things like baryon number and lepton number are transported. This is a problem of coupled flows, a perfect playground for Onsager's theory. Even in this extreme environment, the transport coefficients that relate the flow of baryons to a gradient in lepton chemical potential must be the transpose of the coefficients relating the flow of leptons to a gradient in baryon chemical potential (with the appropriate reversal of the magnetic field). Moreover, the theory shows that this symmetry is robust. Even if you impose physical constraints, like trapping the leptons so their net current is zero, the resulting effective transport coefficients for the remaining particles still obey the reciprocity principle. This "old" principle of thermodynamics remains an indispensable tool for understanding some of the most exotic objects in the universe.

Throughout our examples, you may have noticed the term "Onsager-Casimir." The addition of Casimir's name is crucial. It reminds us that we must be careful about how our variables behave under time reversal. A magnetic field $\mathbf{B}$ or an angular velocity $\mathbf{\Omega}$ are "odd" under time reversal; running the movie backwards makes them point in the opposite direction. Most fluxes, like electric current or heat flow, are also odd. However, it is possible to construct fluxes and forces that are even under time reversal.

The full beauty of the Onsager-Casimir relation is that it accounts for this. The relation states $L_{ik} = \eta_i \eta_k L_{ki}$, where $\eta_i$ is the time-reversal signature ($+1$ for even, $-1$ for odd) of the flux $J_i$ (and any external odd fields like $\mathbf{B}$ are reversed in $L_{ki}$). If we couple two normal (odd) fluxes, $\eta_i \eta_k = (-1)(-1) = +1$, and we get the simple Onsager relation $L_{ik} = L_{ki}$. But what if we couple an odd flux to an even one? Then $\eta_i \eta_k = (-1)(+1) = -1$, and the reciprocity picks up a minus sign: $L_{ik} = -L_{ki}$. This is precisely the origin of the negative sign in the chiral fluid relation $\alpha = -\beta$, where an odd flux (polarization current) was coupled to a force conjugate to an even flux (torque).

This principle is a universal grammar for the physics of change. It doesn't predict the magnitude of any single transport coefficient—that depends on the messy details of microscopic interactions. But it does dictate the relationships between them. It reveals a hidden, symmetrical logic that governs the world's constant, irreversible slide toward equilibrium. From a computer chip to a liquid crystal display, from a swimming bacterium to the core of a star, this single, elegant principle of symmetry is always at work, conducting an unseen harmony in the universe.