Error-corrected phase estimation averaged over variable grids on a trapped-ion quantum computer: hyperacuity spectra of a CO molecule adsorbed onto χ \chi -Fe 5 C 2 Preprint: APS/123-QED

Taichi Kosugi, Hirofumi Nishi, Keito Kasebayashi, Hiroki Takahashi, Yu-ichiro Matsushita · 2026

Read Error-corrected phase estimation averaged over variable grids on a trapped-ion quantum computer: hyperacuity spectra of a CO molecule adsorbed onto χ \chi -Fe 5 C 2 Preprint: APS/123-QED by Taichi Kosugi, Hirofumi Nishi, Keito Kasebayashi, Hiroki Takahashi, Yu-ichiro Matsushita (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-29674.txt.

Introduction

Taichi Kosugi and four colleagues test a sharper quantum phase estimation method on a trapped-ion machine. Their QAVG blends low-detail scans from shifted starting points to rebuild excitation spectra more finely than one scan allows. They run it end to end for carbon monoxide on an iron carbide surface, using plain circuits and error-shielded logic circuits. Even with noisy counts, rebuilt spectra beat base scan spacing by far, and smoothed cost maps cut false dips that stall tuning.

Catalog record

CREATORTaichi Kosugi, Hirofumi Nishi, Keito Kasebayashi, Hiroki Takahashi, Yu-ichiro Matsushita
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66394
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.29674

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Error-corrected phase estimation averaged over variable grids on a trapped-ion quantum computer: hyperacuity spectra of a CO molecule adsorbed onto χ \chi -Fe 5 C 2 Preprint: APS/123-QED Taichi Kosugi, Hirofumi Nishi, Keito Kasebayashi, Hiroki Takahashi, Yu-ichiro Matsushita Abstract Quantum phase estimation (QPE) is an underlying technology for extracting the excitation spectra of many-electron systems, yet its practical use on current hardware is hindered by low grid resolution and environmental noises. Here we propose QPE averaged over variable grids (QAVG), a vernier-type approach that combines low-resolution QPE with multiple origin shifts and physically motivated continuous parametrization to reconstruct the spectra accurately. We introduce this approach into an end-to-end workflow for the ab initio -based model system for a CO molecule adsorbed onto the χ \chi -Fe 5 C 2 surface. We perform experiments on Quantinuum H2-2 using both physical QPE circuits and logical QPE circuits encoded in the Steane code with offline bit-flip correction. We demonstrate that QAVG accurately reconstructs the spectra with deviations much smaller than the nominal QPE resolution, even when the noisy histograms are used. The cost landscapes averaged over the shifted grids substantially suppress the local minima arising from the spectral leakage, thereby stabilizing the optimization of trial parameters. These results indicate that QAVG provides a robust route to quantum simulations of correlated spectra toward the era of early-fault-tolerant quantum computers. I Introduction Quantum algorithms that are expected to outperform classical ones have been proposed in various scientific fields. Many of these algorithms are designed as combinations of a small number of elementary quantum subroutines that serve as foundational building blocks. One of the most prominent among them is quantum phase estimation (QPE) 50 ; 43 ; 9 ; 78 ; 30 ; 79 ; 5 ; 66 ; 67 , variants of which basically extract the eigenvalues of an implementable unitary operator (or an implementable exponentiated Hermitian operator) from an input many-qubit state overlapping with the corresponding eigenstates. QPE is particularly efficient for applications in quantum chemistry, where it is known to enable the acquisition of physical quantities such as one-particle Green’s functions (GFs) 33 and linear-response functions 34 ; 73 in second-quantized formalism more efficiently than classical algorithms. It is because such QPE sampling does not require diagonalization for the huge Hilbert subspaces where the excited states reside as long as the initial state is already prepared. These quantities provide essential information for understanding the static and dynamic properties particularly in systems where correlation effects play significant roles. To perform such sophisticated algorithms with high precision, it is essential to implement them in a fault-tolerant (FT) manner 50 based on logical states by using high-fidelity physical qubits and appropriate error-detecting or error-correcting codes. Given the recent development of quantum hardware 7 ; 3 ; 68 ; 12 ; 64 , much more attention is paid to practical applications of quantum computation based on logical qubits than ever. We find such works that used real quantum computers as follows. Yamamoto et al. 81 performed the Bayesian QPE for estimating the ground state energy of a hydrogen molecule based on the [[6,4,2]] error-detecting code, which is a member of the family of the [[k+2,k,2]] Iceberg codes 69 , using a trapped-ion quantum computer. Zhong et al. 17 performed the variational quantum eigensolver (VQE) experiments on a superconducting quantum computer to prepare the ground state of a hydrogen molecule, where they employed a hybrid protocol that integrates Pauli twirling and probabilistic error cancellation based on the [[4,2,2]] code 77 ; 63 . Nishi et al. 55 performed the probabilistic imaginary-time evolution (PITE) 41 ; 39 ; 38 ; 51 ; 37 ; 54 ; 80 ; 56 ; 14 ; 31 based on the [[6,4,2]] code using a trapped-ion quantum computer to prepare the ground and excited states of NV center models for promising materials as spin qubits. He et al. 25 carried out experiments of the quantum approximation optimization algorithm (QAOA) based on the Iceberg codes up to 20 logical qubits by using a trapped-ion quantum computer. Nishi et al. 53 performed QPE sampling for the excited states of FePO 4 based on the [[6,4,2]] code using a trapped-ion quantum computer to obtain the x-ray absorption spectroscopy (XAS) spectra. Yamamoto et al. 82 performed QPE for a hydrogen molecule based on the [[7,1,3]] error-correcting code, also known as the Steane code 71 , using a trapped-ion quantum computer to estimate the ground state energy by maximizing the likelihood. It is noted here that Mayer et al. 45 benchmarked the logical quantum Fourier transform (QFT), which will be undoubtedly one of the most important elementary technologies in the FT quantum computation (FTQC) era, for three logical qubits in the Steane code by using a trapped-ion quantum computer. They adopted recursive gate teleportation 76 ; 4 ; 75 ; 45 for implementing logical non-Clifford operations. Iron-based catalysts in Fischer–Tropsch synthesis (FTS) are known to play crucial roles in chemical industry. Such catalysts actually undergo the carburization process to form various iron carbides. Amongst them, \chi -Fe 5 C {}_{2}, also known as Hägg carbide, is particularly interesting since its surface with a high Miller index (510) has been experimentally detected 83 ; 65 ; 28 and the calculations predicted its high thermodynamic stability compared to other facets 59 . In addition, CO activation on the surfaces of \chi -Fe 5 C 2 is important for the initiation of the FTS. Realistic elucidation and description of the activation process for a CO molecule on transition metal surfaces 74 ; 27 from a viewpoint of electronic theory demands the incorporation of electronic correlation, as discussed later. We therefore adopt an end-to-end workflow in the present study for \chi -Fe 5 C 2 where a CO molecule is adsorbed onto the (510) surface, as depicted in Figs. 1 (a) and (b). The workflow starts from electronic-structure calculations based on the density functional theory (DFT), from which the maximally localized Wannier orbitals (MLWOs) 44 essential for the adsorption are constructed on a classical computer. The second-quantized model Hamiltonian for them is then transformed into qubit representation. For obtaining the one-particle GF of the model system, we propose QPE averaged over variable grids (QAVG), which we apply to physical and logical qubits on a trapped-ion quantum computer. This approach enables us to determine the physical quantities based on observed histograms for the QPE circuits characterized by shifts of energy origin smaller than the grid resolution. The philosophy of QAVG reminds us of the hyperacuity (see, e.g., Refs. 16 ; 26 ), that achieves rather fine spatial resolution compared to the distance between the centers of photoreceptors on the retina by making use of multiple neural information sources. Although that is just an analogy, the QAVG cost function in fact has much fewer local minima than that for a single origin shift and it thus facilitates the parameter optimization, as will be demonstrated later. Figure 1: (a) Unit cell used for DFT calculations in the present study. This periodic system mimics the non-periodic adsorption system, where a CO molecule is adsorbed onto the (510) surface of \chi -Fe 5 C 2 . This figure was drawn by using VESTA 46 . (b) Schematic energy

Read the complete text