⚛️ lattice

(2+1)D quantum electrodynamics at finite density on a quantum computer

This paper presents a proof-of-principle simulation of (2+1)D quantum electrodynamics with two fermion flavors at finite density on a quantum computer, utilizing a gauge-invariant ansatz and classical optimization to identify phase transitions on IBM hardware while discussing future scalability.

Original authors: Emil Otis Rosanowski, Arianna Crippa, Lena Funcke, Paulo Vitor Itaborai, Karl Jansen, Simran Singh

Published 2026-07-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Emil Otis Rosanowski, Arianna Crippa, Lena Funcke, Paulo Vitor Itaborai, Karl Jansen, Simran Singh

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

Imagine the universe as a giant, complex video game. In this game, the rules that govern how particles interact are written in a language called "gauge theory." One of the most important rules is Quantum Electrodynamics (QED), which describes how light and matter dance together.

For decades, scientists have tried to simulate these dances on supercomputers. But when the game gets crowded (a state called "finite density," like a packed concert hall), the math becomes so messy that even the world's fastest computers crash. This is where quantum computers come in. Instead of trying to solve the messy math on a regular computer, they try to become the quantum system, mimicking the particles directly.

This paper is a "proof of concept" report. The authors successfully taught a real quantum computer to simulate a specific, simplified version of this dance: QED in 2+1 dimensions (two spatial dimensions plus time) with two types of particles (flavors) packed into a tiny, crowded space.

Here is a breakdown of what they did, using everyday analogies:

1. The Setup: A Tiny, Crowded Grid

Imagine a chessboard that is only 2 squares by 2 squares. This is the "lattice" where the simulation happens.

The Players: There are two types of fermions (particles like electrons) on this board. Let's call them "Red" and "Blue."
The Rules: The particles can't just float anywhere; they are tethered by invisible strings (gauge links) that connect them. These strings must follow a strict rule called Gauss's Law. Think of Gauss's Law like a strict bouncer at a club: "For every person who enters, someone else must leave, or the total number of people in the room must stay balanced." If the bouncer isn't satisfied, the simulation is invalid.
The Challenge: The researchers wanted to see what happens when they push more "Red" and "Blue" particles into this tiny room by increasing the "chemical potential" (basically, the pressure to add more people).

2. The Method: The Variational Quantum Eigensolver (VQE)

Since current quantum computers are noisy (like a radio with static), they can't run complex calculations perfectly. So, the authors used a strategy called VQE.

The Analogy: Imagine you are trying to find the lowest point in a foggy valley (the ground state energy). You can't see the whole valley. Instead, you send out a drone (the quantum circuit) with a specific shape. You tweak the drone's wings (parameters) to see if it flies lower.
The Loop: The quantum computer runs the drone, measures how low it went, and sends the data back to a classical computer. The classical computer acts as the pilot, adjusting the drone's wings to make it fly lower next time. They repeat this until they find the absolute lowest point.

3. The Discovery: The "Level Crossing"

The main goal was to see how the number of Red vs. Blue particles changes as they squeeze more pressure into the system.

The Result: They found that as they increased the pressure, the system didn't just slowly change. Instead, it suddenly flipped.
The Analogy: Imagine a seesaw. At first, it's balanced. Then, you add a little weight, and it stays balanced. But then, suddenly, snap—the seesaw flips completely to the other side.
What they saw: The simulation showed "level crossings." This means the system suddenly switched from having more Red particles to having more Blue particles (or vice versa) at specific pressure points. They identified these "tipping points" on their tiny 2x2 grid.

4. The Hardware Reality Check

The authors didn't just run this on a simulator; they ran it on a real quantum computer made by IBM (the "ibm marrakesh" device).

The Noise Problem: Real quantum computers are like trying to build a house of cards in a hurricane. The "noise" (static) can knock the cards over.
What Worked: Surprisingly, the computer was very good at counting the particles (the "number operator"). Even with the noise, the count was accurate.
What Struggled: The computer had a harder time measuring the total energy of the system accurately because the energy calculation is much more sensitive to the "static."
The Outcome: Despite the noise, they could still see the "tipping points" (the level crossings). They estimated the pressure where the flip happens, and it was close to what their perfect simulations predicted.

5. Why This Matters (According to the Paper)

The authors are careful to say this is just the first step.

The Scale: They only used a 2x2 grid. Real physics happens on much larger grids.
The Future: They showed that their method works on a tiny scale. They believe that with better software (called "adaptive ansätze"), they could eventually scale this up to larger grids.
The Goal: This isn't about building a new battery or curing a disease today. It's about proving that we can teach quantum computers to handle the complex, crowded rules of particle physics that classical computers cannot solve.

In summary: The team successfully taught a noisy quantum computer to play a tiny, 2x2 version of a complex particle physics game. They proved that even with the computer's "static," it could spot the exact moment the game rules force the particles to suddenly rearrange themselves. This is a small but crucial victory in the race to simulate the universe's most crowded and chaotic environments.

Problem Statement

Simulating lattice gauge theories, particularly Quantum Chromodynamics (QCD), at finite density remains a significant challenge for classical computational methods due to the sign problem and the complexity of real-time dynamics. While lower-dimensional models like the (1+1)D Schwinger model have been successfully simulated on quantum hardware, extending these studies to (2+1)D Quantum Electrodynamics (QED3) with multiple fermion flavors at finite density has not yet been demonstrated on quantum devices. QED3 is physically relevant as an effective description for quantum spin liquids and cuprate superconductors, yet its phase structure at finite density is not fully understood analytically.

Methodology

The authors employ the Variational Quantum Eigensolver (VQE) to simulate a (2+1)D QED3 system with two fermion flavors on a small lattice.

Model Formulation: The study utilizes a lattice Hamiltonian with staggered fermions and compact gauge links. The total Hamiltonian includes electric field, magnetic plaquette, mass, kinetic hopping, and chemical potential terms. The electric field operators are truncated to a finite range ( $l=1$ ).
Gauge Invariance: To enforce Gauss's law, the authors use a gauge-invariant ansatz. For the specific $2 \times 2$ lattice with open boundary conditions (OBC), Gauss's law allows three of the four gauge links to be expressed in terms of the remaining dynamical link. This reduces the gauge-field Hilbert space significantly, requiring the simulation of only one dynamical gauge link while retaining the full 2D connectivity of the fermionic sector.
Quantum Circuit Ansatz: The variational ansatz consists of three components:
1. Fermion Encoding: Uses Jordan-Wigner mapping to convert fermionic degrees of freedom (two flavors per site) into qubits.
2. Gauge Encoding: Employs a Gray code encoding for the truncated electric field states on two qubits.
3. Entanglement: Includes controlled $U^{xy}$ gates to couple the matter and gauge sectors while preserving gauge invariance, along with inter-flavor mixing layers and phase correction ( $R_Z$ ) gates.
Optimization: The circuit contains 24 variational parameters. Classical simulations are performed using the COBYLA optimizer to find optimal parameters, which are then deployed on quantum hardware. The authors note that more advanced optimizers like NFT were deemed unsuitable due to the specific structure of their gate generators.
Hardware Execution: Inference runs were conducted on the IBM ibm_marrakesh device. The authors applied Zero-Noise Extrapolation (ZNE) using gate folding to mitigate hardware noise.

Key Contributions

First Proof-of-Principle: This work presents the first demonstration of QED3 with multiple fermion flavors at finite density on quantum hardware.
Protocol Development: The authors generalize a previous QED3 construction to include two fermion flavors and finite density, explicitly encoding Gauss's law into the circuit structure to reduce qubit requirements.
Phase Structure Analysis: The paper derives a formula for the critical chemical potentials where density-induced level crossings occur in (2+1)D, analogous to previous (1+1)D studies.
Noise Analysis: The authors provide a quantitative analysis of operator sensitivity, demonstrating that the number operator is significantly more robust against hardware noise than the Hamiltonian due to differences in their operator norms (condition numbers).

Results

Classical Simulations: Noiseless classical simulations on a $2 \times 2$ lattice with gauge coupling $g=5$ successfully identified density-induced level crossings. The particle number difference between the two flavors ( $\langle \hat{N}_1 - \hat{N}_2 \rangle$ ) exhibited step-like behavior, indicating transitions between different particle number sectors.
Intermediate Region: A small intermediate phase was observed near the transition region. The authors found that this region becomes narrower and harder to resolve as the coupling strength $g$ increases, while the entanglement entropy between the gauge link and fermion system is highest in this intermediate region.
Hardware Inference: When deployed on the IBM quantum processor, the simulation successfully captured the general trend of the level crossings. The critical point was estimated at $\mu \approx 13 \pm 0.7$ $μ \approx 13 \pm 0.7$ .
- The number operator results ( $\hat{N}_1, \hat{N}_2$ ) remained close to expected values despite noise.
- The energy expectation values showed larger deviations from the exact diagonalization (ED) results, consistent with the higher condition number of the Hamiltonian.
- Due to circuit depth limitations, the hardware could not resolve the narrow intermediate region, resulting in an averaged transition point rather than two distinct critical points.

Significance and Claims

The paper claims to be a "first step" toward the real-time evolution of finite-density systems in (2+1)D. The authors emphasize that while current hardware limitations (noise and circuit depth) prevent the resolution of fine-grained phase structures on larger lattices, the study validates the feasibility of the proposed gauge-invariant VQE protocol.

The authors explicitly state that no analytical or numerical results exist for the level-crossing points in (2+1)D QED for the studied parameter regime, making the identification of these transitions on quantum hardware a novel contribution. They conclude that while the current $2 \times 2$ system is small, the method is scalable in principle using advanced adaptive ansätze (such as SC-ADAPT-VQE) for larger systems, though this remains a subject for future investigation. The work serves as a benchmark for future studies aiming to explore the phase diagram of QED3 and related condensed matter systems on quantum computers.