pith. sign in

arxiv: 2511.17767 · v2 · submitted 2025-11-21 · ⚛️ physics.chem-ph

Benchmarking Hartree-Fock and DFT for Molecular Hyperpolarizability: Implications for Evolutionary Design

Dominic Mashak , S. A. Alexander This is my paper

Pith reviewed 2026-05-17 19:54 UTC · model grok-4.3

classification ⚛️ physics.chem-ph
keywords molecular hyperpolarizabilitydensity functional theoryHartree-Fockevolutionary algorithmsnonlinear optical materialsbenchmarkingpush-pull chromophoresmolecular design
0
0 comments X

The pith

All thirty Hartree-Fock and DFT combinations produce identical pairwise rankings of molecular hyperpolarizability for the tested chromophores.

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

This paper benchmarks Hartree-Fock and five density functionals across six basis sets for calculating the first hyperpolarizability of five organic push-pull chromophores. It finds that every one of the thirty method combinations agrees perfectly on which molecules rank higher than others, even though absolute errors versus experiment remain moderate. The result matters because evolutionary algorithms for molecular design need fast fitness functions that correctly order candidates rather than report precise numbers. With this ordering preserved, any of the tested methods can guide searches for nonlinear optical materials without risk of selecting inferior structures due to method choice. The work therefore validates the use of inexpensive calculations like HF/3-21G as reliable ranking tools despite their 45 percent mean absolute percentage error.

Core claim

For the five chromophores examined, every combination of the five functionals and six basis sets yields exactly the same pairwise ordering of first hyperpolarizability values, thereby supporting their interchangeable use as evolutionary fitness functions even when absolute values deviate from experiment.

What carries the argument

The invariance of pairwise rankings of hyperpolarizability across all functional and basis-set combinations.

If this is right

  • Any of the thirty tested method combinations can serve as a fitness function for evolutionary optimization of hyperpolarizable molecules.
  • The fastest option, HF/3-21G, can be adopted without sacrificing ranking accuracy on similar systems.
  • Larger basis sets reduce absolute error but leave the molecular ordering unchanged.
  • Evolutionary searches can therefore minimize compute time while still identifying the best candidates according to these methods.

Where Pith is reading between the lines

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

  • If the ranking invariance extends to more diverse molecular families, evolutionary algorithms could screen millions of structures using only the cheapest calculations.
  • The same benchmarking logic could be applied to other molecular properties such as two-photon absorption cross sections.
  • A natural next test would apply the same thirty methods to a library of hundreds of chromophores to check whether ordering remains stable.

Load-bearing premise

The perfect ordering observed for these five specific molecules will continue to hold for the far larger and more diverse chemical spaces searched by evolutionary algorithms.

What would settle it

Any additional molecule where two different functional or basis-set combinations produce opposite orderings of hyperpolarizability values would disprove the claim of universal pairwise agreement.

Figures

Figures reproduced from arXiv: 2511.17767 by Dominic Mashak, S. A. Alexander.

Figure 1
Figure 1. Figure 1: compares error distributions across function￾als. HF exhibits the lowest median error (51.7% ± 23.5%). Hybrid functionals cluster tightly: PBE0 (55.2%), B3LYP (55.6%), CAM-B3LYP (54.1%), and M06-2X (54.0%), all within 1.5 percentage points [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (a) Mean wall time increases systematically with basis set size. (b) The Pareto frontier identifies optimal methods (red dots). Only HF/STO-3G and HF/3-21G achieve Pareto optimality. 3.4 Pareto-Optimal Methods Figure 2b maps all 30 methods in accuracy-speed space. Only HF/STO-3G and HF/3-21G achieve Pareto opti￾mality; no other method offers better accuracy without increased cost. HF/STO-3G achieves 60.5% … view at source ↗
Figure 4
Figure 4. Figure 4: provides a functional-basis heatmap of mean percentage errors [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: Calculated versus experimental hyperpolarizabilities for five representative functionals. Each panel shows one functional across all basis sets. Black dashed lines indicate regression fits; red solid lines show ideal y = x. All achieve R 2 > 0.99 with near-unity slopes. against systematic errors in diverse chemical space. The modest 3× cost increase is justified. However, this recommendation comes with cav… view at source ↗
read the original abstract

Evolutionary algorithms for molecular design require computationally efficient yet accurate fitness functions. We systematically benchmark Hartree-Fock and density functional theory for predicting molecular first hyperpolarizability ($\beta$), evaluating five functionals (HF, PBE0, B3LYP, CAM-B3LYP, M06-2X) across six basis sets against experimental data from five organic push-pull chromophores. For this dataset, HF/3-21G achieves 45.5% mean absolute percentage error with perfect pairwise ranking in 7.4 minutes per molecule. All 30 tested combinations of functional and basis sets maintain perfect pairwise agreement, validating their use as evolutionary fitness functions despite moderate absolute errors. Larger basis sets yield a lower percentage error compared to the experimental values than the difference with the functional. The preservation of pairwise rankings across all combinations of functionals and basis sets provides crucial guidance for evolutionary optimization of nonlinear optical materials.

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 / 2 minor

Summary. The manuscript benchmarks Hartree-Fock and five DFT functionals (PBE0, B3LYP, CAM-B3LYP, M06-2X) paired with six basis sets for computing the first hyperpolarizability β of five organic push-pull chromophores against experimental data. It reports that all 30 method combinations produce identical pairwise rankings of the five molecules, with HF/3-21G achieving 45.5% mean absolute percentage error in 7.4 minutes per molecule, and concludes that this ranking invariance validates the methods as fitness functions for evolutionary algorithms in nonlinear optical materials design despite moderate absolute errors.

Significance. The direct comparison to experimental data on five molecules supports the pairwise-ranking observation for this narrow class of push-pull chromophores. If the ranking preservation were shown to generalize, the work would offer useful practical guidance for selecting fast methods like HF/3-21G in evolutionary searches; however, the current evidence is limited to a small, structurally homogeneous set and does not yet establish broader utility.

major comments (2)
  1. Abstract and conclusions: the central claim that perfect pairwise agreement across all 30 combinations 'validates their use as evolutionary fitness functions' is load-bearing for the paper's implications but rests on an untested extrapolation; the five tested molecules share similar donor-π-acceptor motifs, and no additional validation on structurally diverse or EA-generated candidates is provided to support generalization to the chemical spaces searched by evolutionary algorithms.
  2. Results section (discussion of ranking): with a sample of only five molecules and no reported error bars, statistical tests, or cross-validation on a hold-out set, the observation of perfect agreement does not yet demonstrate robustness against the possibility of correlated errors within this molecular class.
minor comments (2)
  1. The statement that 'larger basis sets yield a lower percentage error compared to the experimental values than the difference with the functional' is unclear in phrasing; please reword for precision and support with explicit quantitative comparisons or a table.
  2. The abstract and main text would benefit from a brief table summarizing the five molecules (structures or key features) to help readers assess the structural diversity of the test set.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and constructive review. We have carefully considered the major comments and revised the manuscript to address the concerns regarding the scope of our claims and the limitations of the dataset. Our responses to each comment are provided below.

read point-by-point responses

  1. Referee: Abstract and conclusions: the central claim that perfect pairwise agreement across all 30 combinations 'validates their use as evolutionary fitness functions' is load-bearing for the paper's implications but rests on an untested extrapolation; the five tested molecules share similar donor-π-acceptor motifs, and no additional validation on structurally diverse or EA-generated candidates is provided to support generalization to the chemical spaces searched by evolutionary algorithms.

    Authors: We agree that the five chromophores are structurally homogeneous, all featuring donor-π-acceptor motifs, and that our study does not include validation on more diverse structures or on molecules generated by evolutionary algorithms. The perfect ranking preservation is an observation specific to this set. To address this, we have revised the abstract and conclusions to remove the implication of broad validation and instead state that the ranking invariance for this class of chromophores suggests potential utility as fitness functions, with the need for further testing on diverse candidates noted as future work. revision: yes

  2. Referee: Results section (discussion of ranking): with a sample of only five molecules and no reported error bars, statistical tests, or cross-validation on a hold-out set, the observation of perfect agreement does not yet demonstrate robustness against the possibility of correlated errors within this molecular class.

    Authors: The referee correctly points out the limitations of a small sample size without statistical analysis. The perfect agreement across methods is an empirical result for these five molecules, but it does not include error bars, statistical tests, or cross-validation. We have updated the results section to explicitly discuss these limitations and to present the finding as an observation that holds for the current dataset rather than a demonstration of general robustness. revision: yes

Circularity Check

0 steps flagged

No circularity: benchmarking rests on external experimental comparisons

full rationale

The paper computes hyperpolarizabilities for five push-pull chromophores using 30 HF/DFT functional-basis combinations and compares results to external experimental data. It reports observed perfect pairwise ranking preservation and moderate MAPE values but performs no derivation, fitting of parameters to the target ranking, or self-referential equations. The claim that these methods are suitable as EA fitness functions follows directly from the external-data agreement on this set; no step reduces a 'prediction' or 'result' to quantities defined inside the paper by construction. No self-citation chains, uniqueness theorems, or ansatzes are invoked as load-bearing premises. This is a standard empirical benchmarking study whose central observations are falsifiable against the cited experiments.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper relies on standard quantum-chemical approximations without introducing new free parameters, axioms beyond domain conventions, or invented entities.

axioms (1)
  • domain assumption Hartree-Fock and the chosen DFT functionals provide sufficiently accurate electron densities for hyperpolarizability ranking in these push-pull systems.
    Invoked implicitly when treating computed values as comparable to experiment for ranking purposes.

pith-pipeline@v0.9.0 · 5456 in / 1315 out tokens · 72545 ms · 2026-05-17T19:54:01.429917+00:00 · methodology

discussion (0)

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

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. CVT Archives and Chemical Embedding Measures for Multi-Objective Quality Diversity in Molecular Design

    physics.comp-ph 2026-04 unverdicted novelty 5.0

    CVT archives with learned chemical embeddings improve median global hypervolume and multi-objective quality diversity in NLO molecular design compared to grid-based archives.

  2. Multi-Objective Evolutionary Design of Molecules with Enhanced Nonlinear Optical Properties

    physics.comp-ph 2026-02 conditional novelty 4.0

    Evolutionary algorithms can discover molecules with improved nonlinear optical properties by simultaneously optimizing hyperpolarizability ratio, HOMO-LUMO gap, polarizability, and energy per atom.

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages · cited by 2 Pith papers

  1. [1]

    D. R. Kanis, M. A. Ratner, and T. J. Marks, De- sign and construction of molecular assemblies with large second-order optical nonlinearities. Quantum chemical aspects.Chemical Reviews, vol. 94, no. 1, pp. 195–242, 1994

    work page 1994

  2. [2]

    A., Finding Molecules with Large Hyperpolarizabilities.MATCH Com- mun

    Mashak, D., & Alexander, S. A., Finding Molecules with Large Hyperpolarizabilities.MATCH Com- mun. Math. Comput. Chem., vol. 94, no. 3, pp. 633– 644, 2025

    work page 2025

  3. [3]

    A., Finding Molecules with Specific Properties: Simulated Annealing vs

    Mashak, D., & Alexander, S. A., Finding Molecules with Specific Properties: Simulated Annealing vs. Evolution.Companion Proceedings of the Ge- netic and Evolutionary Computation Conference (GECCO ’25). ACM, 2025

    work page 2025

  4. [4]

    Burke, Perspective on density functional theory

    K. Burke, Perspective on density functional theory. J. Chem. Phys., vol. 136, no. 15, p. 150901, 2012

    work page 2012

  5. [5]

    R. G. Parr and W. Yang,Density-Functional The- ory of Atoms and Molecules. Oxford University Press, New York, 1989

    work page 1989

  6. [6]

    Champagne, P

    B. Champagne, P. Beaujean, M. de Wergifosse, M. H. Cardenuto, V. Li´ egeois, and F. Castet, Quan- tum chemical methods for predicting and interpret- ing second-order nonlinear optical properties: From small to extendedπ-conjugated molecules. InFron- tiers of Quantum Chemistry, pp. 117–138, Springer, 2018

    work page 2018

  7. [7]

    Plaquet, M

    A. Plaquet, M. Guillaume, B. Champagne, F. Castet, L. Ducasse, J.-L. Pozzo, and V. Rodriguez, In silico optimization of merocyanine-spiropyran compounds as second-order nonlinear optical molec- ular switches.Phys. Chem. Chem. Phys., vol. 10, no. 41, pp. 6223–6232, 2008

    work page 2008

  8. [8]

    S. R. Marder, Organic nonlinear optical materials: where we have been and where we are going.Chem. Commun., pp. 131–134, 2006

    work page 2006

  9. [9]

    L. R. Dalton, P. A. Sullivan, and D. H. Bale, Elec- tric field poled organic electro-optic materials: state of the art and future prospects.Chem. Rev., vol. 110, no. 1, pp. 25–55, 2010

    work page 2010

  10. [10]

    H. A. Kurtz, J. J. P. Stewart, and K. M. Dieter, Calculation of the nonlinear optical properties of molecules.J. Comput. Chem., vol. 11, no. 1, pp. 82–87, 1990

    work page 1990

  11. [11]

    Champagne and B

    B. Champagne and B. Kirtman, Evaluation of alter- native sum-over-states expressions for the first hy- perpolarizability of push-pullπ-conjugated systems. J. Chem. Phys., vol. 125, p. 024101, 2006. 5

    work page 2006

  12. [12]

    Adamo and V

    C. Adamo and V. Barone, Toward reliable density functional methods without adjustable parameters: The PBE0 model.J. Chem. Phys., vol. 110, pp. 6158–6170, 1999

    work page 1999

  13. [13]

    Ernzerhof and G

    M. Ernzerhof and G. E. Scuseria, Assessment of the Perdew–Burke–Ernzerhof exchange-correlation functional.J. Chem. Phys., vol. 110, pp. 5029–5036, 1999

    work page 1999

  14. [14]

    A. D. Becke, Density-functional thermochemistry. III. The role of exact exchange.J. Chem. Phys., vol. 98, pp. 5648–5652, 1993

    work page 1993

  15. [15]

    P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and M. J. Frisch, Ab initio calculation of vibrational ab- sorption and circular dichroism spectra using den- sity functional force fields.J. Phys. Chem., vol. 98, pp. 11623–11627, 1994

    work page 1994

  16. [16]

    Yanai, D

    T. Yanai, D. P. Tew, and N. C. Handy, A new hybrid exchange–correlation functional using the Coulomb- attenuating method (CAM-B3LYP).Chem. Phys. Lett., vol. 393, pp. 51–57, 2004

    work page 2004

  17. [17]

    Zhao and D

    Y. Zhao and D. G. Truhlar, The M06 suite of density functionals for main group thermochem- istry, thermochemical kinetics, noncovalent inter- actions, excited states, and transition elements. Theor. Chem. Acc., vol. 120, pp. 215–241, 2008

    work page 2008

  18. [18]

    Q. Sun, T. C. Berkelbach, N. S. Blunt, et al., PySCF: the Python-based simulations of chemistry framework.Wiley Interdiscip. Rev.: Comput. Mol. Sci., vol. 8, p. e1340, 2018

    work page 2018

  19. [19]

    J. L. Oudar and D. S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state dipole moment.J. Chem. Phys., vol. 66, pp. 2664–2668, 1977

    work page 1977

  20. [20]

    Blanchard-Desce, V

    M. Blanchard-Desce, V. Alain, P. V. Bedworth, et al., Large quadratic hyperpolarizabilities with donor–acceptor polyenes exhibiting optimum bond length alternation: correlation between structure and hyperpolarizability.Chem. Eur. J., vol. 3, pp. 1091–1104, 1997

    work page 1997

  21. [21]

    Castet, V

    F. Castet, V. Rodriguez, J.-L. Pozzo, L. Ducasse, A. Plaquet, and B. Champagne, Design and char- acterization of molecular nonlinear optical switches. Acc. Chem. Res., vol. 46, pp. 2656–2665, 2013

    work page 2013

  22. [22]

    Goedecker, Linear scaling electronic structure methods.Rev

    S. Goedecker, Linear scaling electronic structure methods.Rev. Mod. Phys., vol. 71, pp. 1085–1123, 1999

    work page 1999

  23. [23]

    G. E. Scuseria, Linear scaling density functional cal- culations with Gaussian orbitals.J. Phys. Chem. A, vol. 103, pp. 4782–4790, 1999

    work page 1999

  24. [24]

    Jensen, Atomic orbital basis sets.WIREs Com- put

    F. Jensen, Atomic orbital basis sets.WIREs Com- put. Mol. Sci., vol. 3, pp. 273–295, 2013

    work page 2013

  25. [25]

    D. E. Goldberg,Genetic Algorithms in Search, Op- timization, and Machine Learning. Addison-Wesley, Reading, MA, 1989

    work page 1989

  26. [26]

    K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, A fast and elitist multiobjective genetic algorithm: NSGA-II.IEEE Trans. Evol. Comput., vol. 6, pp. 182–197, 2002

    work page 2002

  27. [27]

    J. L. Br´ edas, D. Beljonne, V. Coropceanu, and J. Cornil, Charge-transfer and energy-transfer pro- cesses inπ-conjugated oligomers and polymers: a molecular picture.Chem. Rev., vol. 104, pp. 4971– 5004, 2004

    work page 2004

  28. [28]

    M. G. Kuzyk, Using fundamental principles to un- derstand and optimize nonlinear-optical materials. J. Mater. Chem., vol. 19, pp. 7444–7465, 2009

    work page 2009

  29. [29]

    W. J. Hehre, L. Radom, P. v. R. Schleyer, and J. A. Pople,Ab Initio Molecular Orbital Theory. John Wiley & Sons, New York, 1986

    work page 1986

  30. [30]

    T. H. Dunning Jr., Gaussian basis sets for use in cor- related molecular calculations. I. The atoms boron through neon and hydrogen.J. Chem. Phys., vol. 90, pp. 1007–1023, 1989

    work page 1989

  31. [31]

    Kirtman, S

    B. Kirtman, S. Bonness, A. Ramirez-Solis, B. Champagne, H. Matsumoto, and H. Sekino, Cal- culation of electric dipole (hyper)polarizabilities by long-range-correction scheme in density functional theory: A systematic assessment for polydiacety- lene and polybutatriene oligomers.J. Chem. Phys., vol. 128, p. 114108, 2008

    work page 2008

  32. [32]

    B¨ ack, D

    T. B¨ ack, D. B. Fogel, and Z. Michalewicz (Eds.), Handbook of Evolutionary Computation. IOP Pub- lishing Ltd. and Oxford University Press, 1997

    work page 1997

  33. [33]

    P. A. Sullivan and L. R. Dalton, Theory-inspired development of organic electro-optic materials.Acc. Chem. Res., vol. 43, pp. 10–18, 2010

    work page 2010

  34. [34]

    Y. V. Pereverzev, K. N. Gunnerson, O. V. Prezhdo, P. A. Sullivan, Y. Liao, B. C. Olbricht, A. J. P. Ake- laitis, A. K.-Y. Jen, and L. R. Dalton, Guest-host cooperativity in organic materials greatly enhances the nonlinear optical response.J. Phys. Chem. C, vol. 112, pp. 4355–4363, 2008

    work page 2008

  35. [35]

    Dreuw, J

    A. Dreuw, J. L. Weisman, and M. Head-Gordon, Long-range charge-transfer excited states in time- dependent density functional theory require non- local exchange.J. Chem. Phys., vol. 119, pp. 2943– 2946, 2003. 6

    work page 2003

  36. [36]

    R. Baer, E. Livshits, and U. Salzner, Tuned range- separated hybrids in density functional theory. Annu. Rev. Phys. Chem., vol. 61, pp. 85–109, 2010

    work page 2010

  37. [37]

    Kirtman and J

    B. Kirtman and J. M. Luis, On the contribution of mixed terms in response function treatment of vi- brational nonlinear optical properties.Int. J. Quan- tum Chem., vol. 111, pp. 839–847, 2011

    work page 2011

  38. [38]

    D. R. Bowler and T. Miyazaki,O(N) methods in electronic structure calculations.Rep. Prog. Phys., vol. 75, p. 036503, 2012. 7 Appendix This appendix presents the complete calculated hyperpolarizability values, wall times, and percentage errors for all 30 method combinations across all five test molecules. Tables are organized by molecule, with experimental...

    work page 2012