pith. sign in

arxiv: 2604.12549 · v1 · submitted 2026-04-14 · 🪐 quant-ph

Utility of NISQ devices: optimizing experimental parameters for the fabrication of Au atomic junction using gate-based quantum computers

Takumi Kanezashi , Daisuke Tsukayama , Jun-ichi Shirakashi , Tetsuo Shibuya , Hiroshi Imai This is my paper

Pith reviewed 2026-05-10 15:17 UTC · model grok-4.3

classification 🪐 quant-ph
keywords NISQ devicesquantum optimizationatomic junctionsfeedback-controlled electromigrationparameter optimizationcombinatorial optimizationAu atomic junctionsnanofabrication
0
0 comments X

The pith

A NISQ device optimizes experimental parameters for fabricating gold atomic junctions more accurately than a quantum annealer on large problems.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests whether gate-based quantum computers can autonomously tune the many parameters needed for feedback-controlled electromigration to create precise gold atomic junctions. The authors recast the tuning task as a combinatorial optimization problem and solve it on a real NISQ device, then compare the quality of the solutions against results from a D-Wave quantum annealer. They report that the NISQ solutions show lower residual energies and remain higher quality as the number of parameters grows. A sympathetic reader would care because successful automation removes a major barrier to reliable, scalable production of atomic-scale devices without constant human adjustment.

Core claim

By mapping the multi-parameter optimization of feedback-controlled electromigration to a combinatorial problem, the NISQ device produces approximate solutions with lower residual energies and higher quality than a previously used D-Wave quantum annealer, especially for large-scale instances. This establishes that gate-based NISQ hardware can be applied directly to experimental parameter search in nanoscale fabrication.

What carries the argument

The reduction of feedback-controlled electromigration parameter tuning to a combinatorial optimization problem solved approximately on a gate-based NISQ device.

If this is right

  • Fabrication of Au atomic junctions can proceed with reduced reliance on manual parameter adjustment.
  • NISQ hardware offers a practical advantage over annealing hardware for this class of experimental optimization when the problem size increases.
  • Lower residual energies translate into tighter control over atomic migration rates during the electromigration process.
  • The same mapping technique can be reused for other materials or junction geometries once the formulation is established.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Closed-loop systems that feed NISQ-optimized parameters back into live experiments could shorten iteration cycles in nanoscale device development.
  • The observed NISQ advantage may change if future annealers improve or if better error mitigation becomes available on gate-based platforms.
  • This application points to a broader near-term role for NISQ devices in guiding physical fabrication steps rather than purely numerical benchmarks.

Load-bearing premise

The experimental parameter search for feedback-controlled electromigration can be accurately expressed as a combinatorial optimization problem that present-day NISQ devices can solve without large mapping or noise errors.

What would settle it

Running the NISQ-chosen parameters in actual fabrication experiments and finding that the resulting gold junctions show higher failure rates, poorer conductance quantization, or lower stability than junctions made with parameters from classical solvers or the annealer.

Figures

Figures reproduced from arXiv: 2604.12549 by Daisuke Tsukayama, Hiroshi Imai, Jun-ichi Shirakashi, Takumi Kanezashi, Tetsuo Shibuya.

Figure 1
Figure 1. Figure 1: FIG. 1. Overview of the autonomous system for optimizing and scheduling FCE experimental parameters using a gate [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Experimental parameter ( [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Residual energy as a function of the number of or [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Average [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
read the original abstract

Feedback-controlled electromigration (FCE) enables precise regulation of atomic migration by carefully optimizing multiple experimental parameters. However, manually fine-tuning these parameters poses significant challenges. This study investigated the feasibility of autonomously fabricating Au atomic junctions through gate-based quantum computing using a noisy intermediate-scale quantum (NISQ) device, which effectively approximates solutions to combinatorial optimization problems. We compared the computational accuracy of the NISQ device against a previously reported D-Wave quantum annealer. The results indicate that the NISQ device achieved lower residual energies and produced higher-quality approximate solutions for large-scale problems than the quantum annealing system.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript investigates the use of gate-based NISQ devices to autonomously optimize multiple experimental parameters for feedback-controlled electromigration (FCE) in fabricating Au atomic junctions. It formulates the task as a combinatorial optimization problem and compares the NISQ device's approximate solutions (via lower residual energies and higher-quality results on large instances) against those from a previously reported D-Wave quantum annealer.

Significance. If the comparison holds after proper validation, the result would indicate a concrete experimental application where NISQ hardware can outperform quantum annealing for parameter tuning in nanotechnology, supporting broader claims of NISQ utility beyond toy problems. The work attempts to connect quantum optimization directly to a physical fabrication process.

major comments (2)
  1. [Abstract] Abstract: The headline claim that the NISQ device 'achieved lower residual energies and produced higher-quality approximate solutions for large-scale problems' is unsupported because the abstract (and visible text) supplies no description of the QUBO/Ising encoding of the FCE parameters, the quantum algorithm (QAOA/VQE/etc.), variable count, embedding details, residual-energy normalization relative to the true minimum, or error-mitigation protocol. This directly undermines assessment of the NISQ vs. D-Wave comparison.
  2. [Abstract] Abstract/Results: The central performance advantage is load-bearing on the assumption that the same combinatorial objective is solved with identical ground-state energy scale on both platforms; without showing the mapping fidelity, chain-break handling, or post-processing, the reported lower residual energies could be an artifact of differing normalizations or noise models.
minor comments (1)
  1. [Abstract] The phrase 'large-scale problems' is undefined; adding the number of experimental parameters, qubits, or problem instances would clarify the scope.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight the need for greater clarity in the abstract and explicit validation of the NISQ-D-Wave comparison. We address each point below and will revise the manuscript to incorporate the suggested clarifications.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The headline claim that the NISQ device 'achieved lower residual energies and produced higher-quality approximate solutions for large-scale problems' is unsupported because the abstract (and visible text) supplies no description of the QUBO/Ising encoding of the FCE parameters, the quantum algorithm (QAOA/VQE/etc.), variable count, embedding details, residual-energy normalization relative to the true minimum, or error-mitigation protocol. This directly undermines assessment of the NISQ vs. D-Wave comparison.

    Authors: We agree that the abstract is too concise and should be expanded to support the headline claims without requiring the reader to consult the full text. The manuscript details the QUBO encoding of the FCE parameters, employs QAOA on the NISQ hardware, reports problem sizes with up to 20 variables for the large instances, normalizes residual energy as (E_approx - E_exact)/E_exact using classically computed minima, and applies standard readout-error mitigation. We will revise the abstract to include a brief statement on the algorithm, problem scale, and normalization procedure. revision: yes

  2. Referee: [Abstract] Abstract/Results: The central performance advantage is load-bearing on the assumption that the same combinatorial objective is solved with identical ground-state energy scale on both platforms; without showing the mapping fidelity, chain-break handling, or post-processing, the reported lower residual energies could be an artifact of differing normalizations or noise models.

    Authors: The identical QUBO formulation is solved on both platforms, with the ground-state energy obtained from an exact classical solver and used for normalization in both cases. D-Wave results incorporate the chain-break handling and post-processing described in the referenced prior work. We will add an explicit paragraph in the Results section (and a short note in the abstract) confirming the shared objective function, identical normalization, and post-processing steps to remove any ambiguity in the comparison. revision: yes

Circularity Check

0 steps flagged

No circularity: external benchmark comparison to prior D-Wave study

full rationale

The paper formulates FCE parameter optimization as a combinatorial problem and reports NISQ results against an external, previously reported D-Wave benchmark. No self-definitional equations, fitted inputs renamed as predictions, or load-bearing self-citations appear in the abstract or described derivation. The comparison relies on independent platform outputs rather than internal renormalization or ansatz smuggling. This is the standard case of a self-contained empirical comparison.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides insufficient detail to enumerate free parameters, axioms, or invented entities. The optimization is presumed to involve mapping experimental parameters to a quantum-solvable form, but no specifics are given.

pith-pipeline@v0.9.0 · 5416 in / 1043 out tokens · 53393 ms · 2026-05-10T15:17:52.357761+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages · 1 internal anchor

  1. [1]

    Huntington and A

    H. Huntington and A. Grone, Journal of Physics and Chemistry of Solids20, 76 (1961)

    work page 1961

  2. [2]

    Black, IEEE Trans

    J. Black, IEEE Trans. Electron Devices16, 338 (1969)

    work page 1969

  3. [3]

    Ho and T

    P. Ho and T. Kwok, Reports on Progress in Physics52, 301 (1989)

    work page 1989

  4. [4]

    Etzion, I

    M. Etzion, I. Blech, and Y. Komem, Journal of Applied Physics46, 1455 (1975)

    work page 1975

  5. [5]

    Strachan, D

    D. Strachan, D. Smith, D. Johnston, T.-H. Park, M. J. Therien, D. Bonnell, and A. Johnson, Applied Physics Letters86, 043109 (2005)

    work page 2005

  6. [6]

    J.CampbellandR.Knobel,AppliedPhysicsLetters102, 023105 (2013)

    work page 2013

  7. [7]

    Kanamaru, M

    Y. Kanamaru, M. Ando, and J.-i. Shirakashi, Journal of Vacuum Science & Technology B33, 02B106 (2015)

    work page 2015

  8. [8]

    Bolotin, F

    K. Bolotin, F. Kuemmeth, A. Pasupathy, and D. Ralph, Applied Physics Letters84, 3154 (2004)

    work page 2004

  9. [9]

    Arzubiaga, F

    L. Arzubiaga, F. Golmar, R. Llopis, F. Casanova, and L. E. Hueso, AIP Advances4, 117126 (2014)

    work page 2014

  10. [10]

    Rattalino, V

    I. Rattalino, V. Cauda, P. Motto, T. Limongi, G. Das, L. Razzari, F. Parenti, E. Di Fabrizio, A. Mucci, L. Sch- enetti,et al., RSC Advances2, 10985 (2012)

    work page 2012

  11. [11]

    M. L. Perrin, E. Burzurí, and H. S. van der Zant, Chem- ical Society Reviews44, 902 (2015)

    work page 2015

  12. [12]

    Johnson, A

    S. Johnson, A. Sundararajan, D. Hunley, and D. Stra- chan, Nanotechnology21, 125204 (2010)

    work page 2010

  13. [13]

    Q. Wang, R. Liu, D. Xiang, M. Sun, Z. Zhao, L. Sun, T. Mei, P. Wu, H. Liu, X. Guo,et al., ACS Nano10, 9695 (2016)

    work page 2016

  14. [14]

    Schirm, M

    C. Schirm, M. Matt, F. Pauly, J. C. Cuevas, P. Nielaba, and E. Scheer, Nature Nanotechnology8, 645 (2013)

    work page 2013

  15. [15]

    M. Ito, M. Yagi, M. Shimada, and J.-i. Shirakashi, AIP Advances8, 105005 (2018)

    work page 2018

  16. [16]

    Nakamura, Y

    Y. Nakamura, Y. A. Pashkin, and J. Tsai, Nature398, 786 (1999)

    work page 1999

  17. [17]

    Y. A. Pashkin, T. Yamamoto, O. Astafiev, Y. Nakamura, D. Averin, and J. Tsai, Nature421, 823 (2003)

    work page 2003

  18. [18]

    Itami, Y

    S.-i. Itami, Y. Tomoda, R.-t. Yasutake, and J.-i. Shi- rakashi, Journal of Nanoscience and Nanotechnology10, 7464 (2010)

    work page 2010

  19. [19]

    Iwata, T

    Y. Iwata, T. Sakurai, and J. Shirakashi, AIP Advances 10, 065301 (2020)

    work page 2020

  20. [20]

    Sakai, Y

    S. Sakai, Y. Hirata, M. Ito, and J.-i. Shirakashi, Scientific Reports9, 16211 (2019)

    work page 2019

  21. [21]

    Yoneda, M

    Y. Yoneda, M. Shimada, A. Yoshida, and J.-i. Shirakashi, Applied Physics Express16, 057001 (2023)

    work page 2023

  22. [22]

    Preskill, Quantum2, 79 (2018)

    J. Preskill, Quantum2, 79 (2018)

    work page 2018

  23. [23]

    Bravyi, D

    S. Bravyi, D. Gosset, and R. König, Science362, 308 (2018)

    work page 2018

  24. [24]

    Bravyi, D

    S. Bravyi, D. Gosset, R. König, and M. Tomamichel, Na- ture Physics16, 1040 (2020)

    work page 2020

  25. [25]

    Y. Kim, A. Eddins, S. Anand, K. X. Wei, E. Van Den Berg, S. Rosenblatt, H. Nayfeh, Y. Wu, M. Zale- tel, K. Temme,et al., Nature618, 500 (2023)

    work page 2023

  26. [26]

    Cerezo, A

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio,et al., Nature Reviews Physics3, 625 (2021). 8

    work page 2021

  27. [27]

    Peruzzo, J

    A. Peruzzo, J. McClean, P. Shadbolt, M.-H. Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O’brien, Nature Communications5, 4213 (2014)

    work page 2014

  28. [28]

    J. R. McClean, J. Romero, R. Babbush, and A. Aspuru- Guzik, New Journal of Physics18, 023023 (2016)

    work page 2016

  29. [29]

    A Quantum Approximate Optimization Algorithm

    E. Farhi, J. Goldstone, and S. Gutmann, arXiv preprint arXiv:1411.4028 (2014)

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  30. [30]

    Quantum Algorithms for Fixed Qubit Architectures

    E. Farhi, J. Goldstone, S. Gutmann, and H. Neven, arXiv preprint arXiv:1703.06199 (2017)

    work page Pith review arXiv 2017

  31. [31]

    T. Miki, R. Okita, M. Shimada, D. Tsukayama, and J.-i. Shirakashi, AIP Advances12, 035247 (2022)

    work page 2022

  32. [32]

    Nannicini, Physical Review E99, 013304 (2019)

    G. Nannicini, Physical Review E99, 013304 (2019)

    work page 2019

  33. [33]

    IBM Quantum,https://quantum-computing.ibm.com/

    work page

  34. [34]

    Umeno and K

    A. Umeno and K. Hirakawa, Applied Physics Letters94, 162103 (2009)

    work page 2009

  35. [35]

    Umeno and K

    A. Umeno and K. Hirakawa, Physica E: Low-dimensional Systems and Nanostructures42, 2826 (2010)

    work page 2010

  36. [36]

    K. M. Nakanishi, K. Fujii, and S. Todo, Physical Review Research2, 043158 (2020)

    work page 2020

  37. [37]

    Choi, Quantum Information Processing7, 193 (2008)

    V. Choi, Quantum Information Processing7, 193 (2008)

    work page 2008

  38. [38]

    Choi, Quantum Information Processing10, 343 (2011)

    V. Choi, Quantum Information Processing10, 343 (2011)

    work page 2011

  39. [39]

    com/en/stable/

    D-Wave Ocean SDK,https://docs.ocean.dwavesys. com/en/stable/

    work page

  40. [40]

    dwave-neal,https://github.com/dwavesystems/ dwave-neal

    work page

  41. [41]

    P. D. Nation, H. Kang, N. Sundaresan, and J. M. Gam- betta, PRX Quantum2, 040326 (2021)

    work page 2021

  42. [42]

    P. K. Barkoutsos, G. Nannicini, A. Robert, I. Tavernelli, and S. Woerner, Quantum4, 256 (2020)

    work page 2020

  43. [43]

    Y. Chai, L. Funcke, T. Hartung, K. Jansen, S. Kühn, P. Stornati, and T. Stollenwerk, Physical Review Applied 20, 064025 (2023)

    work page 2023

  44. [44]

    A. Glos, A. Krawiec, and Z. Zimborás, npj Quantum In- formation8, 39 (2022)

    work page 2022

  45. [45]

    Sciorilli, L

    M. Sciorilli, L. Borges, T. L. Patti, D. García-Martín, G. Camilo, A. Anandkumar, and L. Aolita, Nature Com- munications16, 476 (2025)

    work page 2025