Fault-Tolerant Quantum Computations of Vibrational Wave Functions

Marco Majland, Rasmus Berg Jensen, Patrick Ettenhuber, Irfansha Shaik, N. T. Zinner, Ove Christiansen

TL;DR

This paper presents a comprehensive framework for simulating molecular vibrations on fault-tolerant quantum computers using qubitization. By combining high-order tensor decomposition to simplify the Hamiltonian with graph-coloring algorithms for circuit parallelization, the authors demonstrate a path toward simulating complex molecules with over 100 vibrational modes while maintaining chemical accuracy.

What Problem Does This Paper Address?

While quantum computing research has heavily favored the "electronic structure problem" (calculating electron energies), the "nuclear motion problem"—how atoms vibrate within a molecule—is equally vital for understanding chemical reactivity, thermodynamics, and spectroscopy.

Classical computers struggle with vibrational wave functions because the complexity scales exponentially with the number of atoms. Previous quantum approaches often relied on "Trotterization," which introduces approximation errors that require many gates to resolve. This paper tackles the challenge of fault-tolerant resource estimation: how can we encode a vibrational Hamiltonian efficiently enough to run on a future error-corrected quantum computer without the gate count exploding?

Methods

The authors introduce several key innovations to streamline the simulation:

  1. Qubitization & LCU: Instead of Trotterization, they use qubitization to encode the Hamiltonian as a Linear Combination of Unitaries (LCU). This approach generally requires fewer Toffoli gates (the "expensive" gates in fault-tolerant computing).
  2. Tensor Decomposition: The vibrational Hamiltonian (specifically the Potential Energy Surface or PES) often contains a massive number of interaction terms. The authors use Canonical Polyadic (CP) and Tucker decompositions to find low-rank approximations of these terms. This reduces the number of operators that need to be "loaded" onto the quantum computer.
  3. Graph Coloring for Parallelization: To reduce circuit depth, the authors group commuting operators so they can be executed in parallel. They map this grouping task to a classical graph coloring problem, comparing naive, greedy, and optimal planning algorithms.
  4. Coordinate Systems: They explore both standard rectilinear (normal) coordinates and curvilinear (polyspherical) coordinates. The latter is more complex to implement but far better for describing large-amplitude motions like molecular torsions.

Datasets

The algorithms were benchmarked on a variety of molecular systems:

  • Small Molecules: Formaldehyde ($H_2CO$) and Hydrogen Peroxide ($H_2O_2$) were used to test coordinate systems and the accuracy of tensor decompositions.
  • Large Scale Benchmarks: A series of Polycyclic Aromatic Hydrocarbons (PAHs), ranging from PAH1 (benzene, 30 modes) to PAH8 (156 modes).
  • Potential Energy Surfaces (PES): Generated using high-level electronic structure theory (CCSD(T)) for small molecules and semi-empirical methods (GFN2-xTB) for the larger PAH chains.

Key Insights and Ideas

  • Triangular vs. Diagonal Representations: The authors compared different ways to represent one-mode operators. Surprisingly, they found that the triangular representation usually requires fewer gates than the diagonal one. While diagonalizing reduces the number of terms, the "cost of admission"—the gates needed to perform the basis transformation—outweighs the savings for the relatively small basis sets used in vibrational problems.
  • Controllable Error: Tensor decomposition error is not systematic; it behaves stochastically. By setting a single threshold ($\epsilon_{LR}$), researchers can effectively control the trade-off between gate count reduction and chemical accuracy (defined here as an error of $< 1 \text{ cm}^{-1}$).
  • The Power of Greedy Grouping: For large molecules, finding the perfect way to parallelize the circuit (optimal graph coloring) is computationally impossible. However, the authors found that a weighted greedy algorithm provides results very close to the theoretical optimum and scales efficiently to hundreds of modes.

Key Results

  • Massive Term Reduction: Tensor decomposition reduced the number of terms in the Hamiltonian by one to two orders of magnitude for larger molecules.
  • PAH8 Simulation: For the largest molecule tested (PAH8, 156 vibrational modes), the estimated cost was approximately $2.01 \times 10^{11}$ Toffoli gates and ~10,000 qubits.
  • Parallelization Speedup: The grouping algorithms reduced the circuit depth for large PAH chains by roughly two orders of magnitude compared to a naive sequential encoding.
  • Chemical Accuracy: The authors successfully demonstrated that "chemically accurate" simulations are possible even with significant low-rank approximations of the Hamiltonian.

Limitations and Future Work

  • Runtime Realities: Even with these optimizations, $10^{11}$ Toffoli gates would take several days to execute on proposed fault-tolerant hardware (assuming 40ms per gate). Further reductions are likely needed for practical utility.
  • Basis Set Optimization: The study used a fixed number of "modals" (basis functions) per mode. Optimizing the basis set size for each specific vibration could further prune the Hamiltonian.
  • Beyond Energy: This work focuses on calculating ground-state energies. Future work could extend these qubitization techniques to calculate transition intensities and time-resolved dynamics.

Reference: Majland, M., et al. (2025). Fault-tolerant quantum computations of vibrational wave functions. arXiv:2508.16253v1 [quant-ph].

Summary updated

Abstract

Quantum computation of vibrational properties of molecules is a promising platform to obtain computational advantages for computational chemistry. However, fault-tolerant quantum computations of vibrational properties remain a relatively unexplored field in quantum computing. In this work, we present different algorithms for efficient encodings of vibrational Hamiltonians using qubitization. Specifically, we investigate different encoding representations, high-order tensor decomposition to obtain low rank approximations for the vibrational Hamiltonian, rectilinear, and polyspherical coordinate systems, parallelization, and grouping algorithms. To investigate the performance of the different methods, we perform benchmark computations for both small and large molecules with more than one hundred vibrational modes with two- and three-mode coupled Hamiltonian. We present the first resource estimates for qubitization using multimode vibrational Hamiltonians with 1 cm<sup>-1</sup> spectral resolution for molecules with up to one hundred vibrational modes. Additionally, we show that tensor decomposition and operator parallelization effectively reduce the T gate depth of the quantum algorithm.
BibTeX
@article{majland2025fault,
  author = {Marco Majland and Rasmus Berg Jensen and Patrick Ettenhuber and Irfansha Shaik and N. T. Zinner and Ove Christiansen},
  title = {{Fault-Tolerant Quantum Computations of Vibrational Wave Functions}},
  journal = {Journal of Chemical Theory and Computation},
  year = {2025},
  doi = {10.1021/acs.jctc.5c01419},
}

Loading related content...

Figure