Source-linked AI summary
U(1) Wilson lattice gauge theories in digital quantum simulators
Christine Muschik, Markus Heyl, Esteban Martinez, Thomas Monz, Philipp Schindler, Berit Vogell, Marcello Dalmonte, Philipp Hauke, Rainer Blatt, Peter Zoller
TL;DR
Real-time lattice-gauge dynamics are difficult to calculate classically, motivating quantum simulation of the Schwinger model. The paper analyzes a gauge-invariant encoding that integrates out gauge fields into long-range interactions and shows how trapped-ion digital simulators can implement it efficiently. The protocol supports studies of particle production and observables such as entanglement entropy, while continuum-limit extrapolation remains system-size limited.
Problem
Classical methods cannot access lattice-gauge-theory nonequilibrium dynamics generally because real-time simulations suffer from the sign problem.
Method
The protocol analytically integrates out gauge fields, maps the model to a spin Hamiltonian with long-range interactions, and implements it digitally with trapped-ion-compatible gates.
Results
The encoded protocol preserves exact gauge invariance and realizes the Schwinger-model dynamics with N−2 time steps for the long-range interaction.
Takeaways & Limitations
The scheme enables quantum simulations of particle production and observables including entanglement entropy and vacuum persistence amplitudes.
Takeaways & Limitations
Reliable continuum-limit extrapolation requires larger systems than the available numerics can support as finite-size effects strengthen at smaller lattice spacing.
Abstract
from arXiv · showhide
Lattice gauge theories describe fundamental phenomena in nature, but calculating their real-time dynamics on classical computers is notoriously difficult. In a recent publication [Nature 534, 516 (2016)], we proposed and experimentally demonstrated a digital quantum simulation of the paradigmatic Schwinger model, a U(1)-Wilson lattice gauge theory describing the interplay between fermionic matter and gauge bosons. Here, we provide a detailed theoretical analysis of the performance and the potential of this protocol. Our strategy is based on analytically integrating out the gauge bosons, which preserves exact gauge invariance but results in complicated long-range interactions between the matter fields. Trapped-ion platforms are naturally suited to implementing these interactions, allowing for an efficient quantum simulation of the model, with a number of gate operations that scales only polynomially with system size. Employing numerical simulations, we illustrate that relevant phenomena can be observed in larger experimental systems, using as an example the production of particle--antiparticle pairs after a quantum quench. We investigate theoretically the robustness of the scheme towards generic error sources, and show that near-future experiments can reach regimes where finite-size effects are insignificant. We also discuss the challenges in quantum simulating the continuum limit of the theory. Using our scheme, fundamental phenomena of lattice gauge theories can be probed using a broad set of experimentally accessible observables, including the entanglement entropy and the vacuum persistence amplitude.
1. Introduction
The paper develops and analyzes an efficient trapped-ion protocol for digitally simulating real-time U(1) lattice gauge dynamics. It addresses gauge invariance, implementation of long-range interactions, experimental imperfections, scalability, and observables of nonequilibrium phenomena.
- Scope: The paper provides a detailed account of the experimental scheme and examines imperfections, scalability, and future trapped-ion implementations.It also situates the work alongside semiclassical and tensor-network approaches with restricted regimes of applicability.
- Motivation: Real-time lattice-gauge dynamics are difficult for classical methods because quantum Monte Carlo is obstructed by the sign problem.This limits access to nonequilibrium phenomena such as particle–antiparticle production.
- Motivation: The Schwinger model is a paradigmatic U(1) gauge theory exhibiting confinement, chiral symmetry breaking, and particle–antiparticle pair creation.Its reduced dimensionality makes it suitable for quantum simulation with moderate resources.
- Approach: The scheme combines digital quantum simulation with encoding techniques to represent the lattice model as a spin system.Fermionic matter occupies lattice sites, while gauge fields reside on links between sites.
- Approach: Analytically integrating out gauge fields preserves gauge invariance but produces anisotropic long-range interactions between spins.Trapped-ion platforms provide arbitrary-pair entangling gates and high-fidelity local operations suited to these interactions.
2. Digital quantum simulation of the Schwinger model
The Schwinger model is mapped exactly onto a pure spin Hamiltonian whose gauge fields are replaced by asymmetric long-range interactions. A digital protocol then realizes this Hamiltonian efficiently using local rotations and a single type of two-body interaction.
- Mapping: The encoding eliminates gauge fields exactly, yielding a pure spin Hamiltonian with long-range interactions while retaining the model’s gauge constraints.The electric fields are determined by the spin configuration through Gauss’s law.
- Encoding: Matter fields are represented by staggered one-component fermions, with occupied even and unoccupied odd sites denoting electrons and positrons.The lattice unit cell contains two sites, and vacuum configurations are defined by the complementary occupations.
- Hamiltonian: The encoded Hamiltonian separates into long-range spin–spin, nearest-neighbour hopping, and local mass or field terms.The long-range term represents Coulomb interactions, while hopping creates and annihilates particle–antiparticle pairs.
- Interaction structure: The long-range coupling is asymmetric: each spin couples constantly to spins on one side, while coupling to the other side decreases linearly with distance.The reverse-direction encoding is physically equivalent, but the implementation of this interaction is the main realization challenge.
3. Dynamics of particle production
The scheme enables quantum-simulation studies of vacuum decay, particle–antiparticle production, and entanglement dynamics. Pair production oscillates through competing creation and recombination, while electric-field energy suppresses pair separation and entanglement growth.
- Particle production: Increasing electric-field energy J raises the cost of separating pairs, strengthening recombination and reducing particle density.Figure 5 compares J/w = 0 and J/w = 1 at fixed m/w = 1; larger J/w produces smaller absolute ν(t) and larger oscillations.
- Observables: Particle–antiparticle production can be studied through vacuum persistence amplitudes and entanglement entropy, though entanglement measurements require more resources.Vacuum persistence requires local addressability, whereas entanglement entropy generally requires reconstructing density matrices through methods such as quantum state tomography.
- Particle production: The unstable vacuum initially produces particles rapidly, then undergoes recombination, yielding oscillatory particle density with a slowly decaying envelope.At late times, production and recombination balance into a steady state.
- Vacuum decay: In the continuum, vacuum decay rate λ(t) tracks particle density ν(t); on the lattice, this one-to-one relation breaks but their similarity remains visible numerically.The vacuum persistence amplitude measures deviation from the initial state and quantifies decay of the unstable vacuum.
- Entanglement dynamics: For J = 0, half-chain entanglement grows linearly because particles spread ballistically, but finite size cuts growth off at wt ∝ N/2.For nonzero J, the initial growth follows the free case only up to t_J = J^-1, after which entanglement production slows substantially.
- Entanglement dynamics: Larger electric-field energy reduces entanglement by making long-distance particle–antiparticle separation energetically unfavorable.The resulting dynamics supports fewer pairs shared between the two halves of the system.
4. Imperfections of the scheme and implementation in trapped ions
The trapped-ion implementation realizes the long-range spin model through digital, stroboscopic evolution, while addressing discretization and experimental errors. Numerical analyses indicate that relevant dynamics remain well resolved and robust against the dominant imperfections, although decoherence and minimum gate-window durations constrain accuracy.
- Implementation in trapped ions: The protocol implements the Schwinger-model spin dynamics digitally using trapped-ion spin-spin interactions with infinite-range coupling.Single-qubit operations and Mølmer-Sørensen-mediated interactions provide the required experimental controls; individual spins can be decoupled when needed.
- Discretization errors: The stroboscopic scheme approximates the target Hamiltonian through experimentally realizable Hamiltonians, introducing a controllable Trotter error.Increasing the number of time steps can improve the decomposition accuracy, but trapped-ion gate durations impose a lower bound on the step size.
- Discretization errors: The minimum step size is constrained by spin-motion decoupling requirements and by decoherence, limiting the accuracy achievable in the trapped-ion implementation.The basic time windows must satisfy ω_trap∆t_min ≫ 1, while decoherence creates additional practical restrictions.
- Discretization errors: As the Trotter step decreases, particle-density and half-chain-entropy dynamics converge rapidly toward the ideal evolution.For N = 10, the simulations compare T = 0.75/w, 1.5/w, and 3/w; wt = 5 corresponds to 16 ms for J0 = 4 kHz.
- Experimental errors: The dominant experimental imperfections are fluctuating Mølmer-Sørensen coupling strengths and collective dephasing, modeled through ensemble-averaged evolutions.The simulations draw coupling and dephasing fluctuations randomly and evaluate particle density and the rate function for ten ions.
- Experimental errors: These imperfections produce only minor corrections, while imperfect local operations are negligible compared with multi-qubit gate errors.Finite local-operation fidelities above Flocal = 0.99 are considered attainable, and hiding-level effects lead only to minor corrections.
- Error detection techniques: Postselection detects dominant hiding/unhiding errors and preserves the desired unitary evolution apart from a controllable residual error.Undetected failures require both hiding and unhiding pulses to fail on one ion, whereas detected runs are discarded, reducing data-acquisition rates.
5. Continuum limit
The continuum limit can in principle be reached by successive thermodynamic-limit extrapolations at decreasing lattice spacing, but smaller lattice constants amplify finite-size demands and state-preparation challenges.
- Continuum extrapolation: The continuum limit requires taking N →∞ before a →0, implemented by extrapolating at fixed lattice spacing and then decreasing a.The lattice and thermodynamic limits must be ordered correctly to reproduce the continuum theory.
- Initial-state preparation: A valid continuum limit also requires an initial state with correct long-wavelength properties rather than the spatially modulated bare vacuum.The paper therefore considers an alternative initial state and an adiabatic transformation whose gap does not close.
- Continuum extrapolation: 1/N corrections are not always sufficient for reliable fits, so 1/N 2 terms are also needed for available system sizes.Both correction orders are required to obtain good fits across the data points.
- Finite-size effects: Halving the lattice spacing to m/w = 0.5 strengthens finite-size effects, especially at larger times.For mt ≲3, extrapolation to N →∞ remains possible, but larger systems are needed for a fully reliable limit.
- Outlook: The main long-term challenges are preparing the initial state and reaching larger system sizes as the lattice spacing decreases.The authors find the continuum limit possible in principle, but the required system sizes exceed the utilized numerics in the more demanding case.
6. Conclusions and outlook
The protocol provides an exact-gauge-invariant, resource-efficient digital simulation of the Schwinger model and supports observables beyond conventional experiments. Future work concerns extending these capabilities to broader geometries and gauge theories.
- Conclusions: The encoded protocol preserves exact gauge invariance while simulating 2N −1 degrees of freedom with only N physical qubits.The encoding represents N matter fields and N −1 gauge fields using a spin chain.
- Conclusions: The complex long-range Hamiltonian requires only addressable single-qubit manipulations and one type of two-qubit gate, with gate count scaling linearly with system size.This supports efficient scaling to larger chain lengths on digital quantum simulators.
- Outlook: The simulations can access vacuum persistence amplitudes and entanglement entropy, observables described as inaccessible to conventional experiments.These observables connect the platform to questions about nonequilibrium entanglement and vacuum dynamics.
- Outlook: Extending the scheme to two spatial dimensions, ladder geometries, and non-Abelian gauge theories remains an important direction.The paper presents these extensions as long-term goals for controlled quantum simulations.
Appendix A. Entanglement in the encoded Schwinger model
In the charge-conserving encoded Schwinger model, entanglement across a spatial cut can be computed from the reduced spin system because the gauge fields are fixed by spin configurations. This correspondence fails when charge conservation is violated.
- Entanglement correspondence: The encoded model’s adjacent-block entanglement equals the original model’s half-chain entropy within a fixed charge sector.The gauge transformation and Jordan–Wigner transformation do not alter correlations between the left and right halves.
- Gauge constraints: Gauss’ law determines electric fields stepwise from spin configurations, allowing physical states to be represented as spin and gauge-field components on each side of a cut.Figure A1 illustrates the bipartition and the corresponding factorized left/right description.
- Entanglement correspondence: The gauge field at the cut carries no additional entanglement information once the spin state on either side is known.It can be inferred by measuring spins on either the left or right side separately.
- Limitation: The reduced-spin entanglement no longer represents the full-model entanglement when charge conservation is violated.Spin flips can create or annihilate a single charge; such states can be detected through nonzero total magnetisation and filtered by postselection.