arXiv:2610.09616
Sample-based quantum simulation of vibrational structure of polyyne chains
Sarah Mostame, Tanvi P. Gujarati, Alberto Baiardi, Abhijit Mitra, Dimitar Trenev, Sumathy Raman
Abstract
Predicting anharmonic vibrational spectra is a complex computational task, which becomes quickly infeasible as molecular size and the number of vibrational basis functions increase, leading to an exponential growth of the underlying Hilbert space. Here, we employ a sample-based quantum diagonalization (SQD) framework to compute anharmonic vibrational energies and spectra. The approach combines quantum state preparation and sampling with configuration recovery, diagonalization of the Hamiltonian projected in a subspace of basis configurations, and calculation of transition-dipole moments to obtain both vibrational energies and infrared spectra. We investigate two approaches for sampling, namely, vib-LUCJ and vib-SqDRIFT and apply the method to study the polyyne molecules C$_2$H$_2$ and C$_4$H$_2$, reaching representations requiring up to 104 qubits. Beyond energies, transition-dipole calculations reveal that infrared intensities and excited-state composition continue to evolve with basis size through anharmonic mixing and intensity redistribution, even when low-energy eigenvalues appear converged. These results show that sample-based quantum subspace methods can capture spectroscopic information in vibrational spaces far beyond the regime of full diagonalization. Moreover, our simulation identify multi-state recovery and scalable projected-space treatment as central challenges for further scaling sample-based approaches.
AI Summary
The paper asks whether sample-based quantum diagonalization (SQD) can recover anharmonic vibrational energies and infrared spectra without constructing the full vibrational product space. It adapts two sampling strategies—vibrational local unitary cluster Jastrow (vib-LUCJ), initialized from classical vibrational coupled-cluster amplitudes, and randomized time-evolution sampling (vib-SqDRIFT)—to unary-encoded modal Hamiltonians. Hardware samples undergo occupancy-constrained configuration recovery, followed by classical projected-Hamiltonian diagonalization and transition-dipole evaluation. Calculations treat acetylene, $\mathrm{C_2H_2}$, and diacetylene, $\mathrm{C_4H_2}$, using two-mode-truncated potential and dipole surfaces and 2–8 modals per mode. The largest representation uses 104 qubits for 13 modes, with physical dimension $8^{13}\approx5.5\times10^{11}$. At this size, vib-LUCJ retains 26,437 configurations and vib-SqDRIFT 16,039; their ground-state energies differ from VCC[3] references by 7.71 and 3.45 $\mathrm{cm}^{-1}$, respectively. Exact diagonalization at accessible truncations provides checks of energies and dominant infrared features. At larger modal bases, transition strengths and excited-state composition continue changing even when low-energy eigenvalues appear comparatively stable. The stated limitations include approximate, nonvariational VCC references, which prevent interpreting these deviations as absolute errors; unmatched sampling and post-processing costs between the two strategies; and sampling parameters that were not systematically optimized. Comparisons with controlled classical configuration-selection methods and systematically converged vibrational benchmarks remain future work. Multi-state configuration recovery and scalable projected-space calculations are identified as obstacles to further scaling. The calculations do not establish an advantage over classical selected-subspace methods.
Error found: The largest-space infrared spectra lack controlled excited-state and transition-strength benchmarks, so their accuracy is not established by the reported ground-state energy comparisons. The attribution of spectral changes to modal-basis enlargement and anharmonic mixing is not controlled against selected-subspace reconstruction errors, and some sampling and recovery resources change simultaneously. The eight-modal C4H2 vib-LUCJ calculation specifies 50 eigenstates in its table but 100 states in its spectral captions, without explaining whether additional states were computed afterward.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 19 pages, 9 figures