pith. sign in

arxiv: 2605.22196 · v1 · pith:B5VHAULSnew · submitted 2026-05-21 · ❄️ cond-mat.mtrl-sci · cs.ET

Toward the Rational Design of Molecular Field-Coupled Nanocomputing Candidates

Federico Ravera , Leonardo Medrano Sandonas , Andrea Vezzoli , Yuri Ardesi , Mariagrazia Graziano , Gianluca Piccinini , Gianaurelio Cuniberti This is my paper

Pith reviewed 2026-05-22 05:12 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci cs.ET PACS 85.35.-p73.63.-b
keywords Molecular Field-Coupled NanocomputingMolFCNelectrostatic logicconformational samplingcharge transcharacteristicsmolecular candidate validationdevice-level propagationultra-low-power computing
0
0 comments X

The pith

A framework called LUFFY shows that conformationally stable electrostatic charge responses in molecules enable reliable logic signal transfer in field-coupled nanocomputing.

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

The paper introduces LUFFY as a systematic way to screen and validate molecules for MolFCN, a beyond-CMOS approach where information moves through electrostatic fields instead of charge currents. It starts with 27 synthetically accessible molecules, samples their shapes, and computes how they respond to input voltages in both neutral and oxidized forms by tracking aggregated charge. These responses are averaged over energy to account for real-world flexibility and then tested for how well they support signal propagation between molecules. The work demonstrates that molecules meeting this electrostatic robustness criterion can maintain stable information flow at the device scale. If correct, this directly connects choices made at the molecule level to working circuit behavior in ultra-low-power logic systems.

Core claim

The central claim is that energy-averaged V_in-to-Aggregated-Charge Transcharacteristics extracted from neutral and oxidized states of sampled molecular conformations provide a transferable predictor of stable electrostatic information propagation in MolFCN architectures, as shown by linking molecular responses to device-level functionality across the tested set.

What carries the argument

The V_in-to-Aggregated-Charge Transcharacteristics (VACTs), which capture the field-induced change in a molecule's total charge as a function of input voltage and serve as the bridge from single-molecule electrostatics to multi-molecule signal propagation.

If this is right

  • Molecules selected through this process support stable information transfer at the device level.
  • Conformationally robust electrostatic responses become a necessary design criterion for functional MolFCN operation.
  • Molecular structure can be mapped directly onto circuit-level information flow using the extracted descriptors.
  • The approach supplies a scalable route for discovering additional candidates without ad-hoc trial and error.

Where Pith is reading between the lines

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

  • The same response descriptors could be used to rank molecules for other electrostatic computing schemes that rely on field-driven polarization rather than transport.
  • Extending the averaging procedure to include explicit solvent models would test whether the current neutral-plus-oxidized sampling already captures the dominant effects.
  • The resulting library of VACT curves could serve as training data for faster machine-learning screens of much larger molecular spaces.

Load-bearing premise

That voltage-to-charge response curves averaged over conformations in neutral and oxidized states alone are enough to guarantee reliable signal transfer without explicit treatment of many-body interactions or solvent screening.

What would settle it

Device simulations or measurements in which molecules ranked highly by LUFFY VACTs nevertheless show signal loss or unstable logic states when placed in a realistic clocked chain that includes environmental polarization effects.

Figures

Figures reproduced from arXiv: 2605.22196 by Andrea Vezzoli, Federico Ravera, Gianaurelio Cuniberti, Gianluca Piccinini, Leonardo Medrano Sandonas, Mariagrazia Graziano, Yuri Ardesi.

Figure 1
Figure 1. Figure 1: Molecular dataset presentation. Geometry of the best conformer for the proposed molecules. [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Electrostatic descriptors for oxidized conformers. [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Conformer count and µx trends for different anchor–spacer combinations. 3 [PITH_FULL_IMAGE:figures/full_fig_p016_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Oxidized-state µx trends as a function of anchoring group (AG). 4 [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Neutral molecules electrostatic analysis. [PITH_FULL_IMAGE:figures/full_fig_p018_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Neutral molecules: µx variations for selected PG–SP combinations [PITH_FULL_IMAGE:figures/full_fig_p019_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Neutral molecules: µx variation for Trz-linked derivatives. 6 [PITH_FULL_IMAGE:figures/full_fig_p019_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Molecular structures of phenyl-substituted derivatives. [PITH_FULL_IMAGE:figures/full_fig_p020_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Oxidized state properties of phenyl-substituted molecules. [PITH_FULL_IMAGE:figures/full_fig_p020_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Neutral state properties of phenyl-substituted molecules. [PITH_FULL_IMAGE:figures/full_fig_p021_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Semi-static vs static Vin–µ and Eck–µ analysis. 9 [PITH_FULL_IMAGE:figures/full_fig_p022_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Neutral molecules electrostatic response under [PITH_FULL_IMAGE:figures/full_fig_p023_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Semi-static analysis for molecule 6 under different clock fields. 14 [PITH_FULL_IMAGE:figures/full_fig_p027_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Semi-static analysis at Eck = 0 V/nm. 15 [PITH_FULL_IMAGE:figures/full_fig_p028_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Energy-averaged Vin–µ curve. 16 [PITH_FULL_IMAGE:figures/full_fig_p029_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Time evolution of dipole moment components in AIMD simulations. [PITH_FULL_IMAGE:figures/full_fig_p030_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Analysis of dipole moment variations for molecule 15. [PITH_FULL_IMAGE:figures/full_fig_p030_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Dipole moment dynamics for molecule 28. 18 [PITH_FULL_IMAGE:figures/full_fig_p031_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Clock signals and information propagation through a molecular wire. [PITH_FULL_IMAGE:figures/full_fig_p032_19.png] view at source ↗
read the original abstract

Molecular Field-Coupled Nanocomputing (MolFCN) is a promising beyond-CMOS paradigm in which information is propagated electrostatically rather than through charge transport, enabling ultra-low-power logic. Identifying molecules with stable logic states, efficient clock-field switching, and reliable information propagation, however, remains an open challenge. In this Letter, we introduce LUFFY (Layered Unified Framework for MolFCN systematic analYsis), a framework for the rational design and validation of molecular candidates for MolFCN architectures. Starting from 27 synthetically accessible molecules, we combine conformational sampling and electrostatic analysis in neutral and oxidized states to derive robust descriptors of molecular response. In particular, we extract the V${in}$-to-Aggregated-Charge Transcharacteristics (VACTs), capturing the field-induced charge response, and introduce energy-averaged models validated via ab initio molecular dynamics to account for conformational diversity. Finally, we use the resulting molecular responses to evaluate device-level propagation and demonstrate stable information transfer. These results directly link molecular structure to functional information flow, identifying conformationally robust electrostatic response as a key requirement for MolFCN operation. Overall, this work establishes a unified and transferable framework for the identification and validation of MolFCN molecular candidates, bridging molecular design and circuit-level functionality. By unifying previously fragmented approaches into a sustainable methodology, LUFFY enables rational and scalable molecular discovery and establishes a foundation for data-driven design strategies that accelerate the development of ultra-low-power information processing technologies.

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

1 major / 1 minor

Summary. The manuscript introduces LUFFY (Layered Unified Framework for MolFCN systematic analYsis), a workflow for rational design of molecular candidates for Molecular Field-Coupled Nanocomputing. From 27 synthetically accessible molecules, it performs conformational sampling and electrostatic analysis in neutral and oxidized states to extract V_in-to-Aggregated-Charge Transcharacteristics (VACTs), introduces energy-averaged models validated by ab initio molecular dynamics, and applies the resulting descriptors to evaluate device-level propagation, claiming demonstration of stable information transfer that links molecular structure to functional electrostatic information flow.

Significance. If the results hold, the work provides a valuable unified and transferable methodology that bridges molecular design with circuit-level functionality in the MolFCN paradigm. Strengths include the systematic starting set of 27 molecules, use of ab initio MD for conformational validation, and explicit connection of single-molecule electrostatic responses to propagation behavior, which could support scalable, data-driven discovery of ultra-low-power logic candidates.

major comments (1)
  1. [Abstract, final paragraph] Abstract, final paragraph: The central claim that energy-averaged VACTs from neutral and oxidized single-molecule states suffice to predict stable device-level information transfer rests on the untested assumption that intermolecular electrostatic coupling and solvent polarization effects can be neglected or absorbed into the averaged descriptors. No explicit many-body check or multi-molecule propagation simulation is described to confirm that the mean-field model remains accurate once these effects are restored, which is load-bearing for the device-level results.
minor comments (1)
  1. [Abstract] Abstract: The description of ab initio MD validation mentions conformational sampling but provides no quantitative details on the number of conformations sampled, error bars on the averaged VACTs, or comparison to experimental data; adding these would improve clarity without altering the framework.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive and detailed review of our manuscript on the LUFFY framework for MolFCN candidate design. We address the single major comment below and indicate the revisions planned for the next version.

read point-by-point responses

  1. Referee: [Abstract, final paragraph] Abstract, final paragraph: The central claim that energy-averaged VACTs from neutral and oxidized single-molecule states suffice to predict stable device-level information transfer rests on the untested assumption that intermolecular electrostatic coupling and solvent polarization effects can be neglected or absorbed into the averaged descriptors. No explicit many-body check or multi-molecule propagation simulation is described to confirm that the mean-field model remains accurate once these effects are restored, which is load-bearing for the device-level results.

    Authors: We appreciate the referee for pinpointing this foundational assumption. The LUFFY workflow derives VACT descriptors exclusively from single-molecule ab initio calculations in neutral and oxidized states, with conformational robustness introduced via energy averaging validated by ab initio molecular dynamics. These descriptors are subsequently employed to evaluate device-level propagation and demonstrate stable information transfer. We agree that the manuscript does not contain explicit many-body calculations or full multi-molecule propagation simulations that would directly test the restoration of intermolecular couplings and solvent polarization. While the systematic sampling and averaging steps are intended to capture essential response features in a transferable mean-field sense, the referee is correct that this leaves the accuracy of the approximation unverified at the many-body level. In the revised manuscript we will therefore add a new paragraph in the Discussion section that (i) explicitly states the mean-field scope and its limitations, (ii) discusses how solvent and intermolecular effects could modify the extracted VACTs, and (iii) outlines a computationally feasible route for future multi-molecule validation. This clarification will temper the device-level claims without altering the core single-molecule methodology or results. revision: partial

Circularity Check

0 steps flagged

No significant circularity in LUFFY framework derivation chain

full rationale

The paper describes a forward computational pipeline: ab initio MD conformational sampling on 27 molecules, extraction of V_in-to-Aggregated-Charge Transcharacteristics (VACTs) from neutral/oxidized states, energy-averaging of those responses, and subsequent use of the averaged descriptors to evaluate device-level electrostatic propagation. This workflow relies on external quantum-chemistry methods and does not reduce any claimed prediction or device-level result to a parameter fitted from the same target data by construction. No self-citations are invoked as load-bearing uniqueness theorems, no ansatzes are smuggled, and no renaming of known empirical patterns occurs. The derivation remains self-contained against external ab initio benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the premise that electrostatic charge response averaged over conformations is a sufficient proxy for logic-state stability and propagation fidelity; no explicit free parameters or new entities are named in the abstract, but the framework implicitly assumes that neutral/oxidized state calculations capture the relevant physics without solvent or many-body corrections.

axioms (1)
  • domain assumption Conformational sampling combined with single-molecule electrostatics in neutral and oxidized states is adequate to predict device-level information transfer.
    Invoked in the description of VACT extraction and energy-averaged models (abstract).

pith-pipeline@v0.9.0 · 5832 in / 1530 out tokens · 44624 ms · 2026-05-22T05:12:14.437348+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

56 extracted references · 56 canonical work pages

  1. [1]

    & Bhanja, S.Field-Coupled Nanocomputing: Paradigms, Progress, and Perspectives

    Anderson, N. & Bhanja, S.Field-Coupled Nanocomputing: Paradigms, Progress, and Perspectives. Lecture Notes in Computer Science (Springer Berlin Heidelberg, 2014)

    work page 2014

  2. [2]

    Listo, R.et al.Unfolding potential and challenges in molecular field-coupled nanocomputing.Nano Futur.9, 012501 (2025)

    work page 2025

  3. [3]

    S., Isaksen, B

    Lent, C. S., Isaksen, B. & Lieberman, M. Molecular quantum-dot cellular automata.J. Am. Chem. Soc.125, 1056–1063 (2003). 4.Lent, C. S., Tougaw, P. D., Porod, W. & Bernstein, G. H. Quantum cellular automata.Nanotechnology4, 49 (1993)

    work page 2003

  4. [4]

    & Lent, C

    Lu, Y ., Liu, M. & Lent, C. Molecular quantum-dot cellular automata: From molecular structure to circuit dynamics.J. applied physics102(2007)

    work page 2007

  5. [5]

    Pulimeno, A.et al.Understanding a bisferrocene molecular qca wire.Field-Coupled Nanocomputing: Paradigms, Progress, Perspectives307–338 (2014)

    work page 2014

  6. [6]

    In2023 IEEE Nanotechnology Materials and Devices Conference (NMDC), 212–213 (IEEE, 2023)

    Ravera, F.et al.A roadmap for molecular field-coupled nanocomputing actualization. In2023 IEEE Nanotechnology Materials and Devices Conference (NMDC), 212–213 (IEEE, 2023)

    work page 2023

  7. [7]

    In2025 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), vol

    Ferrero, E.et al.Technology and power-aware investigation of nanostructures for molecular field-coupled nanocomputing. In2025 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), vol. 1, 1–6 (IEEE, 2025)

    work page 2025

  8. [8]

    electric field driven switching of a silicon surface bound array of vertically oriented two-dot molecular quantum cellular automata.J

    Qi, H.et al.Molecular quantum cellular automata cells. electric field driven switching of a silicon surface bound array of vertically oriented two-dot molecular quantum cellular automata.J. Am. Chem. Soc.125, 15250–15259 (2003)

    work page 2003

  9. [9]

    & Cuniberti, G

    Santana-Bonilla, A., Medrano Sandonas, L., Gutierrez, R. & Cuniberti, G. Exploring the write-in process in molecular quantum cellular automata: a combined modeling and first-principle approach.J. Physics: Condens. Matter31, 405502 (2019)

    work page 2019

  10. [10]

    Electric-field inputs for molecular quantum-dot cellular automata circuits.IEEE Transactions on Nanotechnol

    Blair, E. Electric-field inputs for molecular quantum-dot cellular automata circuits.IEEE Transactions on Nanotechnol. 18, 453–460 (2019)

    work page 2019

  11. [11]

    & Lent, C

    Lu, Y . & Lent, C. S. Theoretical study of molecular quantum-dot cellular automata.J. Comput. Electron.4, 115–118 (2005). 13.Macrae, R. M. Mixed-valence realizations of quantum dot cellular automata.J. Phys. Chem. Solids177, 111303 (2023)

    work page 2005

  12. [12]

    & Blair, E

    Liza, N., Lu, Y . & Blair, E. P. Designing boron-cluster-centered zwitterionic y-shaped clocked qca molecules.Nanotech- nology33, 465201 (2022)

    work page 2022

  13. [13]

    & Halet, J.-F

    Groizard, T., Kahlal, S. & Halet, J.-F. Zwitterionic mixed-valence species for the design of neutral clocked molecular quantum-dot cellular automata.Inorg. Chem.59, 15772–15779 (2020)

    work page 2020

  14. [14]

    Arima, V .et al.Toward quantum-dot cellular automata units: thiolated-carbazole linked bisferrocenes.Nanoscale813–823 (2012)

    work page 2012

  15. [15]

    A., Dimitrov, V

    Walus, K., Schulhof, G., Jullien, G. A., Dimitrov, V . & Miller, A. Qcadesigner: A rapid design and simulation tool for quantum-dot cellular automata.IEEE Transactions on Nanotechnol.3, 26–31 (2004)

    work page 2004

  16. [16]

    & Graziano, M

    Ardesi, Y ., Beretta, G., Vacca, M., Piccinini, G. & Graziano, M. Impact of molecular electrostatics on field-coupled nanocomputing and quantum-dot cellular automata circuits.Electronics11, 276 (2022)

    work page 2022

  17. [17]

    & Graziano, M

    Ardesi, Y ., Wang, R., Turvani, G., Piccinini, G. & Graziano, M. Scerpa: A self-consistent algorithm for the evaluation of the information propagation in molecular field-coupled nanocomputing.IEEE Transactions on Comput. Des. Integr. Circuits Syst.39, 2749–2760 (2019)

    work page 2019

  18. [18]

    & Piccinini, G

    Ardesi, Y ., Turvani, G., Graziano, M. & Piccinini, G. Scerpa simulation of clocked molecular field-coupling nanocomputing. IEEE Transactions on very large scale integration (VLSI) systems29, 558–567 (2021)

    work page 2021

  19. [19]

    & Piccinini, G

    Ardesi, Y ., Pulimeno, A., Graziano, M., Riente, F. & Piccinini, G. Effectiveness of molecules for quantum cellular automata as computing devices.J. Low Power Electron. Appl.8, 24 (2018). 11/13

    work page 2018

  20. [20]

    & Graziano, M

    Ardesi, Y ., Mo, F., Vacca, M., Piccinini, G. & Graziano, M. Guesstimation of molecular ensemble electrostatics properties through scerpa-dft calculation: Molecular field-coupled nanocomputing as a case study.Adv. Theory Simulationse00812 (2025)

    work page 2025

  21. [21]

    E., Nuckolls, C., Hybertsen, M

    Venkataraman, L., Klare, J. E., Nuckolls, C., Hybertsen, M. S. & Steigerwald, M. L. Dependence of single-molecule junction conductance on molecular conformation.Nature442, 904–907 (2006)

    work page 2006

  22. [22]

    & Ratner, M

    Galperin, M., Nitzan, A. & Ratner, M. A. Molecular transport junctions: vibrational effects.J. Physics: Condens. Matter 19, 103201 (2007)

    work page 2007

  23. [23]

    C., Estroff, L

    Love, J. C., Estroff, L. A., Kriebel, J. K., Nuzzo, R. G. & Whitesides, G. M. Self-assembled monolayers of thiolates on metals as a form of nanotechnology.Chem. reviews105, 1103–1170 (2005). 26.Herrer, L.et al.Electrically transmissive alkyne-anchored monolayers on gold.Nanoscale11, 7976–7985 (2019)

    work page 2005

  24. [24]

    S.et al.A systematic study of methyl carbodithioate esters as effective gold contact groups for single-molecule electronics.Angewandte Chemie136, e202403577 (2024)

    Ward, J. S.et al.A systematic study of methyl carbodithioate esters as effective gold contact groups for single-molecule electronics.Angewandte Chemie136, e202403577 (2024). 28.Shi, T.et al.Recent advances in the syntheses of pyrroles.Green Synth. Catal.4, 20–34 (2023)

    work page 2024

  25. [25]

    Munir, R.et al.Synthesis and photovoltaic performance of carbazole (donor) based photosensitizers in dye-sensitized solar cells (dssc): A review.Top. Curr. Chem.383, 5 (2024)

    work page 2024

  26. [26]

    Yamada, R.et al.Single-molecule rectifiers based on voltage-dependent deformation of molecular orbitals in carbazole oligomers.Nanoscale10, 19818–19824 (2018)

    work page 2018

  27. [27]

    Mishra, P., Singh, N., Siddique, A., Joshi, S.et al.Multifunctional potential of triazole-based metal-organic frameworks: An overview.Chem. Inorg. Mater.100111 (2025)

    work page 2025

  28. [28]

    Buene, A. F. & Almenningen, D. M. Phenothiazine and phenoxazine sensitizers for dye-sensitized solar cells–an investigative review of two complete dye classes.J. Mater. Chem. C9, 11974–11994 (2021)

    work page 2021

  29. [29]

    & Jia, C.-Y

    Luo, J.-S., Wan, Z.-Q. & Jia, C.-Y . Recent advances in phenothiazine-based dyes for dye-sensitized solar cells.Chin. Chem. Lett.27, 1304–1318 (2016)

    work page 2016

  30. [30]

    & Grimme, S

    Bannwarth, C., Ehlert, S. & Grimme, S. Gfn2-xtb—an accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions.J. chemical theory computation15, 1652–1671 (2019)

    work page 2019

  31. [31]

    & Grimme, S

    Pracht, P., Bohle, F. & Grimme, S. Automated exploration of the low-energy chemical space with fast quantum chemical methods.Phys. Chem. Chem. Phys.22, 7169–7192 (2020)

    work page 2020

  32. [32]

    Pracht, P.et al.Crest—a program for the exploration of low-energy molecular chemical space.The J. Chem. Phys.160 (2024)

    work page 2024

  33. [33]

    Data11, 742 (2024)

    Medrano Sandonas, L.et al.Dataset for quantum-mechanical exploration of conformers and solvent effects in large drug-like molecules.Sci. Data11, 742 (2024)

    work page 2024

  34. [34]

    & Medrano Sandonas, L

    Hinostroza Caldas, A., Kokorin, A., Tkatchenko, A. & Medrano Sandonas, L. Assessing the performance of quantum- mechanical descriptors in physicochemical and biological property prediction.Digit. Discov.5, 803–818 (2026)

    work page 2026

  35. [35]

    Software update: The orca program system—version 5.0.Wiley Interdiscip

    Neese, F. Software update: The orca program system—version 5.0.Wiley Interdiscip. Rev. Comput. Mol. Sci.12, e1606 (2022)

    work page 2022

  36. [36]

    Yanai, T., Tew, D. P. & Handy, N. C. A new hybrid exchange–correlation functional using the coulomb-attenuating method (cam-b3lyp).Chem. physics letters393, 51–57 (2004)

    work page 2004

  37. [37]

    & Ahlrichs, R

    Weigend, F. & Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for h to rn: Design and assessment of accuracy.Phys. Chem. Chem. Phys.7, 3297 (2005)

    work page 2005

  38. [38]

    chemical physics150(2019)

    Caldeweyher, E.et al.A generally applicable atomic-charge dependent london dispersion correction.The J. chemical physics150(2019)

    work page 2019

  39. [39]

    & Graziano, M

    Ardesi, Y ., Gaeta, A., Beretta, G., Piccinini, G. & Graziano, M. Ab initio molecular dynamics simulations of field-coupled nanocomputing molecules.J. Integr. Circuits Syst.16, 1–8 (2021)

    work page 2021

  40. [40]

    T., Davies, D

    Butler, K. T., Davies, D. W., Cartwright, H. M., Isayev, O. & Walsh, A. Machine learning for molecular and materials science.Nature559, 547–555 (2018)

    work page 2018

  41. [41]

    Fromer, J. C. & Coley, C. W. An algorithmic framework for synthetic cost-aware decision making in molecular design. Nat. Comput. Sci.4, 440–450 (2024). 12/13

    work page 2024

  42. [42]

    freedom of design

    Medrano Sandonas, L.et al.“freedom of design” in chemical compound space: towards rational in silico design of molecules with targeted quantum-mechanical properties.Chem. Sci.14, 10702–10717 (2023)

    work page 2023

  43. [43]

    & Tkatchenko, A

    Hilfiker, M., Medrano Sandonas, L., Klähn, M., Engkvist, O. & Tkatchenko, A. Leveraging quantum mechanical properties to predict solvent effects on large drug-like molecules. In Clevert, D.-A., Wand, M., Malinovská, K., Schmidhuber, J. & Tetko, I. V . (eds.)AI in Drug Discovery, 47–57 (Springer Nature Switzerland, Cham, 2025)

    work page 2025

  44. [44]

    & Cuniberti, G

    Chen, L., Medrano Sandonas, L., Huang, S., Croy, A. & Cuniberti, G. Interpretable machine learning for quantum-informed property predictions in artificial sensing materials.arXiv preprint arXiv:2601.00503(2026)

    work page arXiv 2026

  45. [45]

    In2024 IEEE 24th International Conference on Nanotechnology (NANO), 424–429 (IEEE, 2024)

    Listo, R.et al.Unveiling charge dynamics in molecular field-coupled nanocomputing. In2024 IEEE 24th International Conference on Nanotechnology (NANO), 424–429 (IEEE, 2024). 50.Koch, N., Ueno, N. & Wee, A. T.The molecule-metal interface(Wiley Online Library, 2013)

    work page 2024

  46. [46]

    arXiv preprint arXiv:2604.00069(2026)

    Medrano Sandonas, L.et al.Perspective: Towards sustainable exploration of chemical spaces with machine learning. arXiv preprint arXiv:2604.00069(2026). Data availability The dataset generated in this work and scripts to use the LUFFY workflow are available in the LUFFY GitHub repository (https://github.com/federicoravera-sudo/LUFFY). Author contributions ...

    work page arXiv 2026

  47. [47]

    Geometry of the best conformer for the proposed molecules

    Molecular Dataset Figure 1: Molecular dataset presentation. Geometry of the best conformer for the proposed molecules. 2

    work page

  48. [48]

    Figure 3: Conformer count andµx trends for different anchor–spacer combinations

    Oxidized-State Electrostatic Properties Figure 2: Electrostatic descriptors for oxidized conformers. Figure 3: Conformer count andµx trends for different anchor–spacer combinations. 3 Figure 4: Oxidized-stateµ x trends as a function of anchoring group (AG). 4

    work page

  49. [49]

    5 Figure 6: Neutral molecules:µx variations for selected PG–SP combinations

    Neutral-State Electrostatic Properties Figure 5: Neutral molecules electrostatic analysis. 5 Figure 6: Neutral molecules:µx variations for selected PG–SP combinations. Figure 7: Neutral molecules:µx variation for Trz-linked derivatives. 6

    work page

  50. [50]

    Figure 9: Oxidized state properties of phenyl-substituted molecules

    Phenyl-Substituted Derivatives Figure 8: Molecular structures of phenyl-substituted derivatives. Figure 9: Oxidized state properties of phenyl-substituted molecules. 7 Figure 10: Neutral state properties of phenyl-substituted molecules. 8

    work page

  51. [51]

    9 Figure 12: Neutral molecules electrostatic response underEck

    Semi-Static vs Static Electrostatic Response Figure 11: Semi-static vs staticVin–µandE ck–µanalysis. 9 Figure 12: Neutral molecules electrostatic response underEck. 10

    work page

  52. [52]

    Specularity Coefficient To quantify molecular reflection symmetry, we define the specularity coefficientS. Given atomic coordinates R={⃗ ri}N i=1, the geometric center of mass is: ⃗ rCOM = 1 N NX i=1 ⃗ ri The radius of gyration is: Rg = 1 N NX i=1 ∥⃗ ri −⃗ rCOM∥ Coordinates are reflected along the Cartesian axes: ⃗ r(x) i = (−xi, y i, z i), ⃗ r(y) i = (xi...

    work page

  53. [53]

    Perturbation-Based Grouping Model In the aggregated-charge framework, the molecule is mapped onto a reduced three-point charge system, where atomic charges are partitioned into three groups. The resulting model is considered reliable only if it can reproduce the DFT reference values for the dipole moment components (µx,µ y,µ z) and for the output potentia...

    work page

  54. [54]

    Selecting only one conformer for SCERPA simulations would neglect conformational diversity

    Averaging VACTs As discussed in the main text, conformational analysis reveals that the same molecule may present markedly different dipole moments depending on the conformer, leading to different transcharacteristics. Selecting only one conformer for SCERPA simulations would neglect conformational diversity. To address this, we introduce the concept of a...

    work page

  55. [55]

    Figure 17: Analysis of dipole moment variations for molecule 15

    AIMD Validation Figure 16: Time evolution of dipole moment components in AIMD simulations. Figure 17: Analysis of dipole moment variations for molecule 15. 17 Figure 18: Dipole moment dynamics for molecule 28. 18

    work page

  56. [56]

    Information Propagation Test Figure 19: Clock signals and information propagation through a molecular wire. 19

    work page