One-to-one quantum simulation of the low-dimensional frustrated quantum magnet TmMgGaO$_4$ with 256 qubits

This paper demonstrates a one-to-one quantum simulation of the frustrated quantum magnet TmMgGaO$_4$ using a 256-qubit Rydberg-based simulator, achieving quantitative agreement with laboratory magnetisation data to validate the material's microscopic Hamiltonian and enabling the investigation of its antiferromagnetic phase transitions, quantum fluctuations, and non-equilibrium dynamics.

The Gist

Imagine you are trying to understand how a complex, chaotic crowd behaves during a panic. You could try to write down equations for every single person, but the math would get so huge that even the world's fastest supercomputers would crash. Or, you could try to build a tiny, perfect model of the crowd in a lab to watch what happens.

This paper is about doing exactly that, but with atoms instead of people, and quantum physics instead of panic.

Here is the story of how a team of scientists used a "quantum simulator" to solve a mystery about a special magnetic material called TmMgGaO4.

1. The Mystery: A Frustrated Magnet

The scientists were studying a crystal called TmMgGaO4. Inside this crystal, tiny magnetic particles (spins) are arranged in a triangle pattern.

• The Problem: Imagine three friends sitting in a triangle, and each wants to sit next to a friend who is facing the opposite direction. If Friend A faces North, Friend B must face South. But then Friend C is stuck: if they face North, they clash with B; if they face South, they clash with A.
• The Result: This is called "frustration." The particles can't decide on a single order, so they get stuck in a chaotic, quantum "dance" where they fluctuate wildly. This makes the material behave in exotic, unpredictable ways that are incredibly hard to calculate with normal computers.

2. The Tool: A "Quantum Playground"

Instead of trying to solve the math on a supercomputer, the team (from a company called Pasqal) built a Quantum Simulator.

• The Setup: They trapped 256 individual Rubidium atoms in a grid using invisible laser tweezers.
• The Trick: They excited these atoms to a high-energy state called a Rydberg state. In this state, the atoms act like giant magnets that can "feel" each other from far away, mimicking the interactions inside the real crystal.
• The Scale Shift: This is the coolest part. In the real crystal, the atoms are packed incredibly tight (about 3 Angstroms apart—trillions of times smaller than a hair). In the simulator, the atoms are spaced about 10 micrometers apart (visible to the naked eye).
  • Analogy: It's like taking a microscopic city and blowing it up to the size of a football stadium. You can now walk through the streets and see exactly what the "citizens" (atoms) are doing, which is impossible in the real, tiny crystal.

3. The Experiment: Two Worlds, One Answer

The team did a "twin study" to see if their simulator was accurate.

• World A (The Real Crystal): They took a real piece of TmMgGaO4 and put it in a massive, super-cold magnetic lab. They measured how it reacted to magnetic fields.
• World B (The Simulator): They programmed their 256-atom simulator to act exactly like the crystal. They ran the same magnetic "test" on the atoms.

The Result: The two worlds matched perfectly. The simulator predicted the exact same magnetic behavior as the real crystal. This proved that their quantum simulator is a faithful "digital twin" of the real material.

4. The Discovery: Seeing the Invisible

Because the simulator is so big and visible, they could see things the real crystal hides:

• The "Elbow" in the Curve: They found a specific point where the material changes its magnetic state (a phase transition). The simulator showed exactly how the atoms rearrange themselves into a specific pattern (1/3 of them flip direction) as the magnetic field changes.
• The "Snapshot" Power: In the real crystal, you can only see the average behavior of billions of atoms. In the simulator, they can take a "snapshot" of every single atom at once. They saw that even when the material looks disordered, there are tiny, hidden pockets of order forming and dissolving. This helped them prove that the chaos wasn't caused by "dirty" impurities in the crystal, but by pure quantum mechanics.

5. The Future: Time Travel (Sort of)

The most exciting part is that the simulator can run fast-forward.

• In the real crystal, the atoms move and react on a picosecond timescale (one trillionth of a second). It's too fast for any camera to catch.
• In the simulator, the same events happen on a microsecond timescale (one millionth of a second).
• Analogy: It's like watching a movie of a flower blooming. In real life, it takes days. In the simulator, they can watch it happen in seconds, frame by frame. They used this to watch how the atoms "thermalize" (settle down) after a sudden jolt, something that is currently impossible to calculate with classical supercomputers.

The Big Picture

This paper is a milestone because it moves quantum simulation from "theoretical cool" to "practical tool."

• Before: We used quantum simulators to show off cool physics concepts.
• Now: We are using them to measure real materials with perfect accuracy, acting as a bridge between the messy real world and the clean world of math.

It's like finally having a crystal ball that doesn't just guess the future, but actually simulates the physics of materials so accurately that we can design better batteries, superconductors, or quantum computers without ever having to build them first.

Technical Summary

1. Problem Statement

Low-dimensional quantum materials, particularly frustrated magnets, exhibit exotic properties driven by enhanced quantum fluctuations. Understanding their microscopic origins is central to condensed matter physics. However, studying these systems faces two major bottlenecks:

• Classical Numerical Limitations: Classical simulation methods (e.g., Monte Carlo, Tensor Networks) struggle with large-scale, highly entangled systems, especially in non-equilibrium regimes or when simulating real-time dynamics in two dimensions.
• Experimental Limitations: While solid-state experiments can measure macroscopic properties (like magnetization), they often lack the spatial resolution (single-site) and temporal resolution (picosecond scales) required to observe microscopic quantum dynamics and correlations directly.
• The Gap: Previous analogue quantum simulators have largely focused on demonstrating universal physical phenomena rather than performing quantitative, material-specific comparisons with real-world solid-state experiments.

The specific system of interest is TmMgGaO4, a layered frustrated rare-earth magnet with Tm$^{3+}$ ions forming a 2D triangular lattice. It is a candidate for studying 2D frustrated quantum magnetism, but its complex phase diagram and dynamics remain challenging to fully characterize.

2. Methodology

The authors employed a hybrid approach, combining macroscopic solid-state experiments with a large-scale analogue quantum simulator to perform a "one-to-one" simulation.

A. The Quantum Simulator (PASQAL)

• Platform: A neutral-atom quantum processing unit (QPU) based on Rydberg atoms (specifically $^{87}$Rb).
• Scale: Up to 256 qubits arranged in a programmable triangular lattice.
• Encoding: Qubits are encoded in the ground state $|g\rangle$ and a highly excited Rydberg state $|r\rangle$.
• Hamiltonian Mapping: The system simulates a Transverse Field Ising Model (TFIM) on a triangular lattice. The Rydberg Hamiltonian is mapped to the material's effective Hamiltonian:
  $$\hat{H}_{TMGO} = J_1 \sum_{\langle i,j \rangle} \hat{\sigma}^z_i \hat{\sigma}^z_j + J_2 \sum_{\langle\langle i,j \rangle\rangle} \hat{\sigma}^z_i \hat{\sigma}^z_j + \sum_i (\Delta_x \hat{\sigma}^x_i - \Delta_z \hat{\sigma}^z_i)$$
  • Rescaling: The lattice constant of the material ($\sim 3$ Å) is rescaled to $\sim 10$ µm in the simulator. The energy scale is rescaled by a factor $\alpha_{QPU} \approx 1.5 \times 10^{-5}$, allowing picosecond material dynamics to be observed on microsecond timescales in the simulator.
• Protocols:
  • Equilibrium: Quasi-adiabatic state preparation to reach ground states.
  • Non-Equilibrium: Sudden quenches of the longitudinal magnetic field to study post-quench dynamics.

B. The Solid-State Experiment

• Material: Single-crystal TmMgGaO4 grown via the optical floating-zone method.
• Measurements: Macroscopic AC and DC magnetic susceptibility ($\chi_{AC}^z$) and magnetization measurements performed at the National High Magnetic Field Laboratory (NHMFL).
• Conditions: Temperatures down to 20 mK and magnetic fields up to 18 T.

C. Numerical Benchmarks

• DMRG (Density Matrix Renormalization Group): Used to generate phase diagrams and benchmark ground states.
• TDVP (Time-Dependent Variational Principle) with MPS: Used for short-time dynamics simulations. The authors highlight that these simulations become computationally prohibitive (requiring weeks of runtime and massive memory) for large systems and long times, unlike the QPU.
• QMC (Quantum Monte Carlo): Used to simulate thermal equilibrium states and estimate effective temperatures for post-quench dynamics.

3. Key Contributions

1. First Quantitative One-to-One Simulation: Demonstrated a direct, quantitative agreement between a macroscopic solid-state material and a 256-qubit quantum simulator across a phase transition.
2. Validation of Microscopic Models: Validated the effective 2D microscopic Hamiltonian for TmMgGaO4 by showing that the simulator reproduces the material's magnetization curves and critical points.
3. Microscopic Insight into Quantum Fluctuations: Used site-resolved snapshots from the QPU to analyze quantum correlations, connecting them to integrated inelastic neutron scattering data. This confirmed that the observed paramagnetic features are driven by quantum fluctuations rather than quenched disorder.
4. Access to Non-Equilibrium Dynamics: Successfully simulated post-quench dynamics on picosecond-equivalent timescales, observing thermalization and entanglement growth that are inaccessible to classical methods.

4. Key Results

A. Equilibrium Phase Transition

• Magnetization Agreement: The magnetization curves ($M^z$) extracted from the QPU ground states matched the experimental AC susceptibility data at $T=20$ mK with excellent quantitative agreement.
• Critical Point Identification:
  • The quantum phase transition from a paramagnet to a 1/3-ordered antiferromagnetic phase was identified.
  • Experimental Critical Point: $\Delta_z^q / J_1 \approx 3.88 \pm 0.12$.
  • Simulator Critical Point: $\Delta_z^q / J_1 \approx 3.87^{+0.44}_{-0.36}$ (for $N=256$).
  • The agreement confirms the validity of the effective Hamiltonian and the simulator's ability to capture the thermodynamic limit behavior despite finite size.
• Structure Factor: The QPU measured the structure factor $S^{zz}_{QPU}(q_{1/3})$, showing the emergence of Bragg peaks corresponding to the 1/3 order, confirming long-range correlations.

B. Quantum Fluctuations and Disorder

• Fluctuation Analysis: The variance of magnetization $(\Delta M^z)^2$ measured on the QPU was compared to integrated inelastic neutron scattering data. The qualitative agreement suggests the intermediate paramagnetic regime is governed by quantum fluctuations (specifically the transverse field $\Delta_x$) rather than quenched disorder, refuting previous hypotheses about disorder-driven phases.
• Snapshot Analysis: Single-shot measurements revealed local 1/3-ordered patches superimposed on a polarized background within the critical region, providing a microscopic view of the "quantum paramagnet."

C. Non-Equilibrium Dynamics

• Post-Quench Evolution: The system was quenched from a paramagnetic product state. The QPU tracked the evolution of nearest-neighbor correlations $C^{zz}_1$ over time.
• Thermalization: The long-time dynamics of the closed quantum system were found to agree with equilibrium predictions from QMC at an effective temperature, providing evidence of eigenstate thermalization.
• Classical vs. Quantum Performance:
  • Classical MPS simulations (TDVP) converged only for short times ($tJ_1 \le 1$) before requiring bond dimensions ($D$) that exceeded available GPU memory (141 GB).
  • The QPU performed the same simulation in ~1 day, whereas the equivalent classical simulation would take ~2 weeks and still fail to converge for long times.

5. Significance

• Paradigm Shift in Simulation: This work moves quantum simulation beyond "proof-of-principle" demonstrations of universal models to predictive modeling of specific, complex materials. It establishes a workflow: (1) Model certification via macroscopic comparison, (2) Microscopic analysis via quantum snapshots, and (3) Prediction of inaccessible regimes.
• Bridging Scales: It successfully bridges the gap between macroscopic thermodynamic measurements and microscopic quantum dynamics, validating the "analogue" nature of the simulator.
• Overcoming Classical Barriers: The study provides concrete evidence that neutral-atom quantum simulators can solve problems (specifically 2D non-equilibrium dynamics of frustrated magnets) that are currently intractable for the most advanced classical supercomputers.
• Future Outlook: The methodology opens the door to studying other quantum magnets, including exotic clock phases and 3D systems, and validates the use of quantum simulators as a standard tool for condensed matter physics research.