🔢 mathematics

Counterexamples to the total-current Hohenberg-Kohn conjecture

This paper refutes the total-current Hohenberg-Kohn conjecture by presenting three counterexamples involving two spinless fermions in three dimensions, where distinct Hamiltonians with different magnetic fields share the same ground-state particle density and zero total current.

Original authors: Andre Laestadius, Mihály A. Csirik

Published 2026-10-21

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Andre Laestadius, Mihály A. Csirik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the microscopic world of atoms and molecules, scientists rely on a powerful shortcut to understand how matter behaves. Instead of tracking the frantic motion of every single electron in a system—a task that quickly becomes impossible as the number of particles grows—they focus on the electron density. This is simply a map showing where the electrons are most likely to be found at any given moment. A famous rule from the 1960s, known as the Hohenberg-Kohn theorem, established that for most systems, this density map is enough to uniquely determine the forces acting on the electrons. If you know where the electrons are, you can, in principle, work backward to find the electric fields and other forces that created that specific arrangement. This idea is the bedrock of modern computational chemistry, allowing researchers to simulate complex materials without solving the impossible equations for every individual particle.

However, when magnetic fields enter the picture, the rules become more complicated. A magnetic field does not just push electrons; it also makes them swirl, creating a physical flow or current. Scientists wondered if a combination of the electron density and this total current flow would be enough to uniquely identify the magnetic and electric forces at play, just as the density alone does for non-magnetic systems. This idea, known as the total-current Hohenberg-Kohn conjecture, suggested that if two different systems produced the exact same map of electron locations and the exact same pattern of electron flow, they must be governed by the same underlying forces, up to a simple shift in perspective. It was a hopeful assumption that promised to extend the power of density-based shortcuts into the realm of magnetism.

A team of researchers has now shown that this hope is unfounded for systems containing two electrons. By constructing three specific, mathematically rigorous examples, they demonstrated that it is possible to create two completely different magnetic environments that produce identical electron densities and identical zero-flow currents. In these scenarios, the electrons sit in the exact same spots and do not move in a way that creates a net flow, yet the magnetic fields holding them there are fundamentally different and cannot be transformed into one another. This discovery proves that for two-electron systems, knowing the density and the total current is not enough to uniquely determine the magnetic forces. The researchers did not merely suggest this might happen; they provided concrete, proven counterexamples where the logic of the conjecture breaks down, forcing a reevaluation of how magnetic systems are modeled in theoretical chemistry.

The researchers built their argument on a clever observation about how quantum systems behave when time is reversed. In the quantum world, if you take a stable state of an electron system and flip the direction of time, the electrons end up in the exact same locations, but their flow direction reverses. If a system happens to have a state where the electrons are not flowing at all—where the total current is zero—then reversing time changes nothing about the flow, because zero reversed is still zero. However, reversing time also flips the direction of the magnetic field. This means that a system with a magnetic field pointing one way and a system with the same field pointing the opposite way can both produce a state with zero current and the same electron density. If the researchers could prove that this zero-current state was the most stable, or ground, state for the system, they would have two different magnetic setups that look identical from the perspective of density and current.

To make this work, the team had to solve a difficult puzzle: finding a magnetic field where the electrons naturally settle into a state with zero net flow, even though the field itself is strong and swirling. In their first example, they designed a smooth, non-uniform magnetic field and showed that two specific electron orbits could cancel each other out perfectly. One electron orbit carried a flow in one direction, while the other carried an equal flow in the opposite direction. When combined, these flows canceled point by point, leaving a total current of zero, while the magnetic field remained active and non-zero. They proved mathematically that this specific arrangement was indeed the most stable state the system could achieve. Because the state was unique and stable, the system with the field pointing one way and the system with the field pointing the other way were both valid, stable ground states with identical densities and zero currents, yet they were governed by different magnetic forces.

The researchers then extended this finding to more realistic conditions. In a second example, they showed that this phenomenon occurs even when the forces holding the electrons are smooth and bounded, fitting within the standard mathematical categories used by physicists. They constructed a scenario where the existence of such a state was guaranteed by the continuous nature of the mathematical functions involved, ensuring that the result was not a fluke of a specific, weird setup but a robust feature of the theory. Finally, they tackled the most challenging case: a system where the two electrons repel each other, as they naturally do, while sitting in a magnetic field generated by a theoretical object called a Dirac monopole. This magnetic field is unique because it radiates outward from a single point like a magnetic version of an electric charge, a configuration that cannot be described by a single smooth magnetic field everywhere in space. Even with this complex interaction and the repulsion between electrons, they proved that a unique, stable ground state exists with zero current.

The implications of these findings are significant for the way scientists model magnetic materials. The total-current Hohenberg-Kohn conjecture was a proposed foundation for a method called current-density functional theory, which aims to simplify the calculation of magnetic properties. The new results show that this foundation is shaky for systems with more than one electron. If two different magnetic setups can produce the same density and current, then a computer program trying to reverse-engineer the magnetic forces from that data would face a fundamental ambiguity. It would not be able to tell which of the two different magnetic fields was the correct one. This does not mean the method is useless, but it does mean that the simple rule of "one density and current equals one set of forces" is false. The researchers have demonstrated that the relationship between the observable behavior of electrons and the invisible forces acting on them is more complex than previously thought, revealing a hidden layer of ambiguity in the quantum world that must be accounted for in future theories.

Summary of "Counterexamples to the total-current Hohenberg–Kohn conjecture"

Problem Statement
The Hohenberg–Kohn (HK) theorem establishes that the ground-state particle density uniquely determines the scalar external potential (up to an additive constant) for non-degenerate systems. In the presence of magnetic fields, a natural extension known as the total-current HK conjecture posits that the ground-state particle density ( $\rho$ ) and the total physical current ( $\mathbf{j}$ ) jointly determine the scalar potential ( $v$ ) and the vector potential ( $\mathbf{A}$ ) up to an additive constant and a gauge transformation, respectively. While this uniqueness holds for a single particle ( $N=1$ ) under suitable regularity conditions, its validity for interacting or multi-particle systems ( $N \geq 2$ ) has been an open question. This paper addresses whether the pair $(\rho, \mathbf{j})$ uniquely determines the Hamiltonian for two spinless fermions in three spatial dimensions.

Methodology
The authors construct explicit counterexamples by exploiting the properties of time-reversal symmetry and complex conjugation. The core logical mechanism is established in Lemma 2.2: if a Hamiltonian $H_{v, \mathbf{A}}$ possesses a non-degenerate ground state $\Psi$ with an identically zero total current ( $\mathbf{j} = 0$ ) in a non-zero magnetic field ( $\nabla \times \mathbf{A} \neq 0$ ), then the complex conjugate state $\bar{\Psi}$ is the ground state of the Hamiltonian $H_{v, -\mathbf{A}}$ .

Both Hamiltonians share the same scalar potential $v$ and interaction.
Both share the same density $\rho$ (since $|\Psi|^2 = |\bar{\Psi}|^2$ ).
Both share the same total current $\mathbf{j}=0$ (since the current of $\bar{\Psi}$ in field $-\mathbf{A}$ is $-\mathbf{j}_{\Psi} - \rho(-\mathbf{A}) = -0 - \rho(-\mathbf{A}) = \rho\mathbf{A}$ ? No, the lemma states $\mathbf{j}_{-\mathbf{A}} = -\mathbf{j}_{\mathbf{A}}$ . If $\mathbf{j}_{\mathbf{A}}=0$ , then $\mathbf{j}_{-\mathbf{A}}=0$ ).
However, the vector potentials $\mathbf{A}$ and $-\mathbf{A}$ are gauge-inequivalent if $\nabla \times \mathbf{A} \neq 0$ .

The challenge lies in constructing such a state $\Psi$ that is the unique, non-degenerate ground state of the system. The authors achieve this through three distinct constructions:

Explicit Noninteracting Example: Using cylindrical coordinates, they design a smooth, bounded vector potential where two occupied orbitals (in angular momentum sectors $m=0$ and $m=-1$ ) carry paramagnetic and diamagnetic currents that cancel pointwise.
Bounded Potential Noninteracting Example: They extend the construction to potentials within standard functional spaces ( $L^{3/2} + L^\infty$ for scalars, $L^3 + L^\infty$ for vectors), utilizing the intermediate value theorem to fix a normalization parameter ensuring the two orbitals are the lowest energy levels.
Interacting Example with Monopole Fields: They introduce Coulomb repulsion ( $\kappa=1$ ) and an isotropic harmonic trap, coupled with a Dirac monopole magnetic field (charge $g = \pm 1/2$ ). They employ a trial-state estimate to prove that the non-degenerate interacting ground state is rotationally invariant, forcing the current to vanish via the continuity equation. Crucially, the authors demonstrate that this correlated ground state is not a single Slater determinant, distinguishing it from the noninteracting case.

Key Results
The paper presents three specific counterexamples demonstrating the failure of the total-current HK conjecture for $N=2$ :

Theorem 3.1 (Noninteracting):
- (i) An explicit construction with a Gaussian density and smooth, bounded potentials where the total current vanishes despite a non-zero magnetic field.
- (ii) A construction using potentials in the standard sum spaces $V$ and $\mathcal{A}$ , proving the failure persists even under standard regularity assumptions.
Theorem 4.1 (Interacting): A counterexample involving Coulomb repulsion and Dirac monopole fields. The ground state is shown to be isolated and nondegenerate, rotationally invariant, and possessing zero total current. The theorem explicitly states that this ground state is not a single Slater determinant. The magnetic fluxes for the two gauge-inequivalent configurations differ ( $+2\pi$ vs $-2\pi$ ).

In all cases, the ground states are non-degenerate, ruling out degeneracy as the cause of the failure. The results show that distinct, gauge-inequivalent Hamiltonians can yield identical ground-state density and total current pairs.

Significance and Claims
The authors claim that their work definitively refutes the total-current HK conjecture for two spinless fermions. The significance of these findings lies in:

Limitations of Current-Density Functional Theory (CDFT): The results imply that for $N \geq 2$ , the ground-state pair $(\rho, \mathbf{j})$ does not uniquely determine the external potentials. Consequently, a Kohn–Sham formulation based on matching $(\rho, \mathbf{j})$ cannot infer the uniqueness of its effective potentials from these variables alone.
Distinction from $N=1$ : The work highlights a fundamental difference between single-particle and multi-particle systems in magnetic fields, showing that the uniqueness theorem valid for $N=1$ does not extend to $N=2$ .
Theoretical Rigor: The paper provides rigorous proofs involving operator theory, spectral analysis, and the use of monopole connections (requiring patch-wise vector potentials) to handle topological constraints that prevent globally smooth vector potentials.

The authors explicitly state that they do not address the recovery of ground-state densities via variational principles (the HK variational problem) but focus strictly on the uniqueness question. They also note that while their interacting example uses monopole connections (which are not globally smooth vector potentials), the noninteracting examples serve as counterexamples even within the standard class of smooth, bounded potentials.

More like this