Quantum simulations of Green's functions for small superfluid systems

This paper presents and validates an end-to-end hybrid quantum-classical strategy for computing Green's functions in small superfluid systems by combining variational techniques for ground states with quantum subspace expansion for excited states, demonstrating high accuracy across normal-to-superfluid transitions and for odd-particle systems.

The Gist

The Big Picture: Predicting the Future of Tiny Systems

Imagine you are trying to predict the weather. In the world of quantum physics, scientists study "many-body systems"—groups of tiny particles (like atoms or electrons) that interact with each other. To understand how these systems behave, they use a mathematical tool called a Green's function.

Think of the Green's function as a "shadow" or a "fingerprint" of the system. If you know this fingerprint perfectly, you can predict almost anything about the system: its energy, how it reacts to changes, and even what happens if you add or remove a single particle.

The problem? Calculating this fingerprint for complex systems is incredibly hard. It's like trying to solve a massive jigsaw puzzle where the pieces keep changing shape. Traditional supercomputers struggle with this, especially when the system involves "superfluidity" (a state where particles flow without friction, like a dance floor where everyone moves in perfect sync).

The Solution: A Hybrid Team-Up

The authors of this paper propose a new strategy that uses a team-up between a classical computer and a quantum computer.

• The Classical Computer (The Manager): It handles the heavy planning, optimization, and organizing.
• The Quantum Computer (The Specialist): It handles the specific, tricky parts of the puzzle that are too hard for normal computers.

They call this a "hybrid quantum-classical" approach.

How the Strategy Works (The Three Steps)

The paper outlines a three-step recipe to build this "fingerprint":

1. Finding the "Home Base" (The Ground State)
First, the team needs to find the most stable, calm state of the system (the "ground state"). Imagine a crowded room where everyone is trying to find the most comfortable spot to stand.

• They use a technique called VQE (Variational Quantum Eigensolver).
• Think of this as a "trial and error" game. The quantum computer tries different arrangements of particles (like trying different dance formations). The classical computer checks the score and tells the quantum computer, "Try this move instead," until they find the perfect, most stable formation.
• The paper tested different "dance moves" (mathematical guesses) to see which one found the best formation fastest.

2. Exploring the "Neighbors" (Adding or Removing a Particle)
Once they have the perfect "Home Base" (with $N$ particles), they need to know what happens if they add one person ($N+1$) or take one away ($N-1$).

• In the past, calculating this was like trying to rebuild the whole puzzle from scratch.
• Here, they use a method called QSE (Quantum Subspace Expansion).
• The Analogy: Imagine you have a perfect photo of a group of friends. Instead of taking a new photo of the whole group with a new person, you use a special filter (the QSE) to mathematically "simulate" what the photo would look like if you added or removed a friend, based on the original photo. This is much faster and requires less computing power.

3. Assembling the Final Picture (The Green's Function)
Finally, they combine the "Home Base" info with the "Neighbor" info.

• They plug these pieces into a formula (the Lehmann representation) to construct the Green's function.
• This final result tells them the energy levels and behavior of the system, effectively creating the "fingerprint" they wanted.

What They Tested

To see if this works, they didn't use a real, messy nuclear reactor. Instead, they used a mathematical model called the "Richardson model" (or pairing model).

• The Analogy: Think of this as a "flight simulator." Before flying a real plane, pilots practice in a simulator that mimics the physics of flight but is controlled and predictable.
• This model is famous in physics because it creates strong "superfluid" effects (like the synchronized dance mentioned earlier). It's the perfect test bed to see if their new algorithm can handle complex, synchronized movements.

The Results: Did It Work?

The team ran their strategy on a computer that simulates a quantum computer (since real quantum computers are still noisy and error-prone).

• Accuracy: The results were very close to the "perfect" answer (which they calculated using a traditional supercomputer for comparison).
• The "Odd" Systems: A surprising bonus was that their method worked well for systems with an odd number of particles (where one particle is left without a partner), which are usually much harder to calculate.
• The Best "Dance Move": They tested several different ways to set up the initial quantum computer. They found that a specific method called ADAPT-VQE (which builds the solution step-by-step, adding one piece at a time) was the most efficient and accurate, especially when the particles were strongly interacting.

The Bottom Line

The paper demonstrates a proof of concept. It shows that by combining a classical computer's planning skills with a quantum computer's ability to handle complex quantum states, we can accurately predict the behavior of small superfluid systems.

They didn't build a new nuclear reactor or cure a disease. Instead, they built a better calculator for a specific type of physics problem. They proved that this hybrid team-up can solve a difficult puzzle that is currently too hard for standard computers, paving the way for future, more complex simulations of atomic nuclei.

Technical Summary

1. Problem Statement

Many-body Green's functions are a cornerstone for describing quantum $N$-body systems in nuclear physics, condensed matter, and quantum chemistry. They provide access to ground-state properties, excitation spectra, and particle-removal/addition energies. However, calculating these functions for interacting systems is computationally expensive on classical supercomputers, particularly for heavy nuclei or high-precision applications involving uncertainty quantification.

The paper addresses the challenge of computing one-body Green's functions for small superfluid systems (specifically those governed by pairing interactions) using hybrid quantum-classical algorithms. The goal is to leverage near-term quantum devices (simulated here) to overcome the scaling limitations of classical methods while maintaining accuracy across phase transitions (normal-to-superfluid).

2. Methodology

The authors propose an end-to-end hybrid quantum-classical strategy based on the Lehmann spectral representation of the Green's function. The method avoids direct time-evolution (which is noise-prone) and instead constructs the Green's function from the system's eigenstates and energies.

The workflow consists of four main steps:

1. Variational Ground State Preparation (QPU + CPU):

   • The $N$-body ground state $|\tilde{\Psi}^N_0\rangle$ is approximated using the Variational Quantum Eigensolver (VQE).
   • The paper tests three distinct ansatzes:
     • PAV-BCS: Particle-number projected BCS state (Variation After Projection).
     • VAP-BCS: Variation After Projection (optimizing parameters after projection).
     • ADAPT-VQE: Adaptive Derivative-Assembled Pseudo-Trotter ansatz. Three variants were tested:
       • ADAPT-St: Standard local gradient selection.
       • ADAPT-Min: Global energy minimization criterion.
       • ADAPT-Fix: Fixed operator ordering.
   • Encoding: The fermionic pairing Hamiltonian is mapped to qubits. For even systems, a pair-to-qubit encoding (reducing qubit count by half) is used for the ground state, which is then converted to particle-to-qubit encoding (Jordan-Wigner) to handle the $N \pm 1$ sectors required for the Green's function.

2. Quantum Subspace Expansion (QSE) for Neighbors (QPU + CPU):

   • To access the $(N \pm 1)$-body systems (odd neighbors), the method constructs a reduced subspace using operators $\{A^\pm_\alpha\}$ applied to the approximate ground state.
   • The study uses single-particle creation/annihilation operators ($a^\dagger_i, a_i$) to generate the basis states $|\phi^{N\pm1}_\alpha\rangle$.
   • Hamiltonian and Overlap Matrices: The QPU measures the expectation values required to build the Hamiltonian matrix ($H^{N\pm1}_{\beta\alpha}$) and the overlap matrix ($O^{N\pm1}_{\beta\alpha}$).
   • Diagonalization: A classical computer solves the generalized eigenvalue problem to obtain approximate eigenstates $|\tilde{\Psi}^{N\pm1}_k\rangle$ and eigenenergies $\tilde{E}^{N\pm1}_k$. This step effectively performs configuration mixing to recover excited states and correct the ground state of odd systems.

3. Green's Function Construction:

   • Using the Lehmann representation, the one-body Green's function $G_{ij}(\omega)$ is reconstructed using the computed spectroscopic amplitudes (matrix elements of creation/annihilation operators between $N$ and $N\pm1$ states) and energy differences.
   • Spectroscopic amplitudes are derived by combining the QPU-measured overlaps with the classical eigenvector coefficients from the QSE step.

4. Measurement Optimization:

   • The authors exploit symmetries in the pairing model (time-reversal and seniority) to reduce the number of required Pauli term measurements from $O(D^2)$ to $O(D)$, where $D$ is the number of doubly-degenerate levels.
   • Pauli grouping techniques are used to minimize circuit evaluations.

3. Key Contributions

• First Nuclear-Specific Green's Function Simulation: This is the first reported quantum simulation of Green's functions specifically tailored to nuclear many-body systems (using the Richardson pairing model).
• Hybrid Strategy Validation: The paper validates a specific workflow where the QPU handles state preparation and expectation value measurement, while the CPU handles optimization and subspace diagonalization (QSE).
• Comparison of Ansatzes: A rigorous benchmarking of symmetry-breaking (BCS) vs. symmetry-conserving (ADAPT-VQE) approaches.
• Odd-System Description: Demonstration that the QSE approach, starting from an even-system ground state, can accurately describe the ground states of neighboring odd systems ($N \pm 1$), a task often difficult for standard mean-field theories.

4. Results

The method was benchmarked on the Richardson pairing model with $N=8$ particles and $D=8$ levels (half-filling) across various interaction strengths ($g$).

• Ground State Accuracy:
  • ADAPT-VQE (specifically ADAPT-Min) achieved the highest fidelity (>99.9%) and lowest energy errors across all coupling strengths, including the superfluid phase transition.
  • VAP-BCS performed well, outperforming PAV-BCS, particularly in the superfluid regime.
  • PAV-BCS struggled at weak coupling (collapsing to the Hartree-Fock limit) but improved as coupling increased.
• Green's Function Accuracy:
  • The reconstructed Green's functions were in excellent agreement with exact Full Configuration Interaction (FCI) results.
  • Moments Analysis: The relative errors on the first four spectral moments were generally below a few percent.
    • ADAPT-Min yielded the lowest errors (often <0.5% for strong coupling).
    • VAP-BCS was comparable to ADAPT-Min in many cases.
    • PAV-BCS showed larger deviations, especially at weak coupling, though the Green's function remained qualitatively correct due to error cancellation in energy differences.
• Odd Systems:
  • The QSE method successfully reproduced the "odd-even staggering" in correlation energies.
  • Calculations for odd systems derived from the even-system ansatz were consistent whether approached from $N-1$ or $N+1$, confirming the method's robustness.

5. Significance and Future Outlook

• Proof of Principle: The study demonstrates that hybrid quantum-classical algorithms can accurately compute Green's functions for correlated superfluid systems, a critical step toward simulating realistic nuclear reactions and structure.
• Error Mitigation: The Lehmann-based approach offers inherent error mitigation. Since the final Green's function depends on energy differences and relative spectroscopic amplitudes, systematic errors in the absolute ground state energy may cancel out.
• Scalability: While the current results are for small systems, the authors identify that the primary scalability bottlenecks are the number of measurements (scaling with system size and operator pool complexity) and the size of the QSE subspace.
• Future Directions: The authors suggest that for realistic nuclear Hamiltonians, the operator pool in the QSE step may need to be expanded to include collective excitations (e.g., 2-particle-1-hole) to capture spectral fragmentation. The approach is also applicable to two-body Green's functions and other strongly correlated models like the Fermi-Hubbard model.

In conclusion, the paper establishes a viable pathway for using near-term quantum computers to solve complex many-body problems in nuclear physics, offering a promising alternative to classical methods for systems where strong correlations and phase transitions are present.