Pith. sign in

REVIEW 1 major objections 1 minor 13 references

Decoding Transient X-ray Absorption Spectra of Acetylacetone With Multireference Algebraic Diagrammatic Construction Theory

T0 review · 1 major / 1 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The transient carbon K-edge spectrum of acetylacetone is governed by evolving proton-sharing geometries on the $S_2$ and $S_1$ surfaces, with the long-time triplet features assigned to central-carbon 1s excitations into low-lying triplet…

desk verdict Solid singlet-side story with a genuinely new method combination and a clean marker for proton-sharing geometries, but the long-time triplet assignment rests on a thermalized T1 ensemble that the dynamics do not actually produce. read the letter →

arxiv 2608.09747 v1 pith:M4HBO2WV submitted 2026-08-10 physics.chem-ph

Bennett W. Clark , Donna H. Odhiambo , Haden Dickerson , Alexander Yu. Sokolov This is my paper · ORCID
classification physics.chem-ph
keywords transientX-rayabsorptionspectroscopyacetylacetonemultireferencealgebraicdiagrammaticconstructioncore-valenceseparationexcited-stateintramolecularhydrogentransfersurface-hoppingdynamicscarbonK-edgespectrumtripletstate
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper reports trajectory-averaged simulations of the transient carbon K-edge X-ray absorption spectrum of acetylacetone, combining multireference algebraic diagrammatic construction (MR-ADC) for core-excited states with surface-hopping dynamics that generate nonequilibrium molecular geometries. It argues that the 20-200 fs spectrum is shaped by continuously evolving distributions of structures on the $S_2$ and $S_1$ surfaces, not by a handful of stationary geometries. The weak absorption at 279.5-281.5 eV is traced to transient, nearly symmetric proton-sharing configurations, while the 284-286 eV envelope reflects C1s excitations modulated by proton transfer, bond alternation, and ring opening. At 7-10 ps, the principal peaks are assigned to central-carbon 1s excitations into low-lying triplet π orbitals, with the second peak's intensity recovered only when the extended semi-internal MR-ADC(2)-SX approximation is used.

What carries the argument

The load-bearing machinery is the ensemble-averaged MR-ADC spectrum: multireference algebraic diagrammatic construction (MR-ADC) with the core-valence separation approximation, applied to geometries sampled from fewest-switches surface-hopping trajectories driven by multistate complete active space second-order perturbation theory. The specific protocol MR-ADC(2)/CASSCF(10e,8o)/cc-pwCVDZ was selected by benchmarking the ground-state carbon K-edge spectrum. The central observable is the correlation between the O-H/H-O bond-length deviation (how symmetric the proton-sharing configuration is) and the computed intensity at 279.5-281.5 eV, together with natural transition orbital analysis (which represents each excitation by its dominant occupied-virtual orbital pair) showing that the intensity tracks the spatial overlap of the C3 1s core orbital with the accepting π-type orbital.

What would settle it

A computational experiment that propagates the $S_2$/$S_1$ dynamics while biasing the enolic proton to remain localized on one oxygen should eliminate the 279.5-281.5 eV band; if the band persists in such an ensemble, the paper's assignment to proton-sharing geometries would be falsified.

Watch

Extended reading notes

Core claim

The central discovery is that the transient X-ray absorption of acetylacetone is governed by the evolving nuclear geometry distribution on the singlet $S_2$ and $S_1$ states rather than by electronic-state populations or single-point spectra alone. At 20-100 fs, absorption between 279.5 and 281.5 eV is enhanced specifically for nearly symmetric, $C_{2v}$-like proton-sharing structures formed during excited-state intramolecular hydrogen transfer; the enhancement arises because the accepting π-type orbital localizes on the central carbon C3, increasing its overlap with the C3 1s core orbital. The 284-286 eV band is a single broad, geometry-dependent manifold of C1s transitions that appears doublet-like only because of overlap with the ground-state bleach near 285 eV. In the long-lived triplet spectrum, the features near 281.4 and 283.8 eV are assigned to C3 and C4 1s excitations into low-lying triplet π orbitals, and the weaker intensity of the second feature under MR-ADC(2) is corrected by the MR-ADC(2)-SX approximation.

Load-bearing premise

The long-time triplet assignment rests on assuming that thermally sampled geometries around the relaxed $T_1$ minimum, propagated for 50 fs, faithfully represent the actual triplet molecules formed by intersystem crossing, even though the singlet trajectories carried no spin-orbit coupling and therefore never generated the triplet population dynamically.

Editorial extensions

If this is right

  • The 279.5-281.5 eV absorption can serve as a real-time marker of the short-lived proton-sharing geometry reached during excited-state intramolecular hydrogen transfer.
  • The apparent doublet at 284-286 eV is not two distinct excited-state bands; it is one broad excited-state envelope split by the ground-state bleach, so transient difference spectra should be interpreted with the bleach included.
  • Single-point calculations at stationary geometries are insufficient for assigning TR-XAS of acetylacetone; vibrationally averaged ensembles change peak positions, widths, and intensities.
  • The long-time 7-10 ps spectrum is dominated by central-carbon (C3 and C4) 1s excitations into low-lying triplet π orbitals of $T_1$.
  • For open-shell triplet core spectra, MR-ADC(2) underestimates the 283.8 eV intensity, and the semi-internal extended variant MR-ADC(2)-SX is needed to reproduce the experimental peak ratio.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the proton-sharing marker should generalize to other intramolecularly hydrogen-bonded enols; substituting the bridging proton with deuterium would be a direct test, since the zero-point spread along the O-H-O coordinate should change the 279.5-281.5 eV intensity.
  • Editorial inference: because the singlet trajectories excluded spin-orbit coupling, the simulated $T_1$ ensemble may not capture the vibrational distribution at the moment of intersystem crossing; including spin-orbit couplings in the dynamics would test whether the 7-10 ps spectrum survives a realistic crossing distribution.
  • Editorial inference: the bleach-overlap explanation predicts that the apparent 284-286 eV doublet should shift as the ground-state population recovers, so time-resolved analysis after separating the bleach contribution could expose the underlying state-specific envelope.
  • Editorial inference: the same ensemble-averaged MR-ADC pipeline could be extended to other ultrafast carbon K-edge experiments where multiconfigurational excited states and nonadiabatic motion coexist, provided the active space captures the relevant valence and core orbitals.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

1 major / 1 minor

Summary. The manuscript reports time-resolved simulations of the carbon K-edge X-ray absorption spectrum of acetylacetone (AcAc), combining XMS-CASPT2 fewest-switches surface-hopping dynamics with core-valence-separated MR-ADC(2) calculations of C1s excitation spectra. The central claims are: (i) the early-time (20–200 fs) transient spectrum is governed by continuously evolving distributions of nuclear geometries on the S2 and S1 surfaces rather than by electronic-state populations alone; (ii) the weak 279.5–281.5 eV absorption (peak a) is a marker of transient, nearly symmetric proton-sharing (C2v-like) configurations; (iii) the 284–286 eV doublet arises from a broad geometry-dependent manifold of C1s excitations whose apparent splitting is largely caused by overlap with the ground-state bleach; and (iv) the long-time (7–10 ps) spectrum is dominated by triplet T1 features near 281.4 and 283.8 eV assigned to C3 and C4 1s excitations into low-lying triplet pi orbitals. The paper also introduces and benchmarks a semi-internal extended variant, MR-ADC(2)-SX, which improves the description of the triplet-state spectrum.

Significance. If the assignments hold, the paper provides a mechanistic connection between time-resolved carbon K-edge spectra and specific nuclear coordinates (proton transfer, bond alternation, ring opening, triplet formation), which is a valuable step beyond stationary-geometry interpretations. The work has clear strengths: it uses a multireference core-excitation method across nonequilibrium ensembles; it demonstrates the necessity of nuclear sampling for these spectra; it provides a state-resolved analysis; and it offers a falsifiable structural correlation (Figure 10) that is not purely an artifact of the fitted energy shifts. The early-time geometric assignments and the bleach-splitting argument are supported by internal consistency and by comparison with experiment. However, the long-time triplet assignment rests on an independently constructed T1 ensemble that is not dynamically connected to the S1 population, and the overall comparison with experiment relies on several fitted parameters (energy shifts, Lorentzian widths, intensity scaling). These issues weaken but do not invalidate the central mechanistic picture.

major comments (1)
  1. [Section 3, Figures 14 and 15] The comparison with experiment uses multiple fitted parameters: global energy shifts (−3.5 eV for single-point, −3.25 eV for vibrationally averaged, and a different −2.0 eV for the MR-ADC(2)-SX spectrum in Figure 15), Lorentzian broadening widths (0.4 and 0.2 eV), and a uniform intensity scaling factor. These are fitted to the same experimental TR-XAS spectra being interpreted, which adds a circularity burden to statements such as 'good agreement' and 'markedly improved description.' The central geometric correlation (Figure 10) and the bleach-splitting argument are less affected because they rely on relative intensities and energy trends, but the absolute peak-energy assignments and the quantitative claims of agreement should be couched more carefully. In particular, the use of a different energy shift for the MR-ADC(2)-SX triplet spectrum means that part of the observed improvement in Figure 15 is absorbed by the extra shift parameter; the text should state this explicitly and discuss the sensitivity of the assignment to the shift choice.
minor comments (1)
  1. [Figure 11 caption] Typo: 'non-equillibrium' should be 'non-equilibrium.'

Circularity Check

0 steps flagged · score 0.0 of 10

The paper's central transient-XAS assignments are self-contained: fitted energy shifts and intensity scaling only align overall position and amplitude, while the key state-resolved and geometry-dependent assignments follow from trajectory ensembles, difference-spectrum analysis, and NTO characterization rather than from the fitted constants.

full rationale

The derivation chain is not circular. The authors benchmark MR-ADC(2) against the experimental ground-state XAS spectrum and then apply rigid energy shifts (-3.5 eV for single-point and -3.25 eV for vibrationally averaged spectra) together with a uniform intensity scale factor when comparing difference spectra. These parameters align overall peak positions and amplitudes but do not determine relative peak shapes, state-resolved intensities, or geometry correlations, so the central assignments are not forced by them. The assignment of the 279.5-281.5 eV feature to nearly symmetric proton-sharing geometries is derived from trajectory sampling and NTO analysis (Figures 10 and 11), not from the fitted shift. The interpretation of the 284-286 eV doublet as a single excited-state envelope split by overlap with the ground-state bleach is obtained by computing the excited-state envelope and then subtracting the calculated ground-state spectrum; although the bleach position is aligned by the energy shift, the conclusion that the doublet arises from bleach overlap is a genuine calculational result. The long-time T1 assignment is based on an independently constructed 300 K Wigner ensemble around the CASPT2 T1 minimum propagated for 50 fs; the paper explicitly states that spin-orbit coupling was not included, so this is a stated modeling assumption rather than a prediction derived from the experimental spectrum, and it does not reduce to the target data by construction. The MR-ADC(2), MR-ADC(2)-SX, and MR-ADC(2)-X variants are systematic electronic-structure approximations, not parameters fitted to the transient spectra. Self-citations to MR-ADC methodology appear throughout, but they are method references rather than load-bearing uniqueness claims or ansatz justifications. The lack of a dynamically generated triplet ensemble is a limitation and a correctness risk, but it is not circularity under the standards applied here. Overall, no step in the paper's claimed derivation is equivalent to its own input by definition or by construction.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claims rest on a small number of fitted spectral parameters, a dynamics method whose accuracy is assumed from prior benchmarks, and a triplet ensemble that is not dynamically generated by intersystem crossing. No new physical entities are introduced.

free parameters (4)
  • Energy shift for single-point MR-ADC(2) spectra = -3.5 eV
    Applied to align the calculated ground-state bleach maximum with the experimental TR-XAS spectrum.
  • Energy shift for vibrationally averaged MR-ADC(2) spectra = -3.25 eV
    Applied to align the vibrationally averaged ground-state spectrum with experiment.
  • Lorentzian broadening width = eta = 0.4 eV single-point, 0.2 eV averaged
    Chosen by hand to represent phenomenological broadening in the computed spectra.
  • Uniform intensity scaling factor = unspecified
    Applied when comparing calculated difference spectra with experimental signals; the value is not stated.
assumptions (5)
  • domain assumption XMS-CASPT2/SA-CASSCF(10e,8o)/cc-pVDZ surface-hopping trajectories faithfully sample the nonadiabatic relaxation of acetylacetone.
    The entire transient ensemble and all spectral averages depend on this assumption; the dynamics are compared with prior simulations but not with a fully converged quantum dynamics benchmark.
  • ad hoc to paper A single global energy shift corrects core-excitation energies for all electronic states and geometries.
    The shift is fitted to the ground-state bleach and then applied to S1, S2, and T1 spectra; state-dependent errors in core-excitation energies would break this assumption.
  • ad hoc to paper A 300 K Wigner ensemble propagated for 50 fs on the T1 surface represents the experimental 7 to 10 ps triplet population.
    Spin-orbit coupling was not included in the dynamics, so the triplet nuclear distribution is not formed by the actual intersystem crossing process; the paper acknowledges this construction in Section 3.
  • domain assumption The CVS-MR-ADC(2) framework accurately describes carbon K-edge transitions in acetylacetone.
    This is supported by the ground-state benchmark and prior MR-ADC X-ray spectroscopy studies, but the accuracy for non-equilibrium excited-state geometries is not independently verified.
  • domain assumption The CASSCF(10e,8o) active space provides a balanced description of both the dynamics and the core-excited states.
    The active space is adopted from prior AcAc studies and selected after benchmarking against the ground-state XAS spectrum, but the same active space is then used for all excited-state and non-equilibrium calculations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Decoding Transient X-ray Absorption Spectra of Acetylacetone With Multireference Algebraic Diagrammatic Construction Theory." pith.science (2026). https://pith.science/paper/M4HBO2WV

@misc{pith2026260809747,
  author       = {Pith},
  title        = {Pith review of: Decoding Transient X-ray Absorption Spectra of Acetylacetone With Multireference Algebraic Diagrammatic Construction Theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M4HBO2WV}},
  note         = {Machine review of arXiv:2608.09747}
}
read the original abstract

Time-resolved X-ray absorption spectroscopy (TR-XAS) offers an element- and site-specific probe of coupled electronic and nuclear dynamics, but its interpretation requires methods that can treat multiconfigurational excited states across nonequilibrium nuclear ensembles. Here, we report time-resolved simulations of the acetylacetone (AcAc) transient X-ray absorption spectra by combining surface-hopping dynamics from complete active space second-order perturbation theory with multireference algebraic diagrammatic construction (MR-ADC). The simulated 20-200 fs spectra show a good agreement with experimental measurements and reveal that the TR-XAS response is governed by continuously evolving distributions of molecular geometries on the singlet potential energy surfaces (S2 and S1). In particular, absorption at 279.5-281.5 eV is enhanced for transient, nearly symmetric proton-sharing configurations, whereas the 284-286 eV profile reflects geometry-dependent C1s excitations modulated by proton transfer, bond alternation, and ring opening. For the long-time T_1 spectrum (7-10 ps), the principal features near 281.4 and 283.8 eV are assigned to central C 1s excitations into low-lying triplet pi-orbitals. Together, our simulations provide a more complete mechanistic picture of AcAc photorelaxation by directly linking ultrafast carbon K-edge signals to proton transfer, skeletal reorganization, internal conversion, and triplet formation.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

  1. [1]

    Decoding Transient X-ray Absorption Spectra of Acetylacetone With Multireference Algebraic Diagrammatic Construction Theory

    Decoding Transient X-ray Absorption Spectra of Acetylacetone With Multireference Algebraic Diagram- matic Construction Theory Bennett W. Clark,a Donna H. Odhiambo,a Haden Dickerson,a and Alexander Yu. Sokolov∗a Time-resolved X-ray absorption spectroscopy (TR-XAS) offers an element- and site-specific probe of coupled electronic and nuclear dynamics, but it...

    work page Pith review arXiv 2026

  2. [2]

    Natural transition orbital analysis assigns peak 1 pre- dominantly to a C3 1s excitation into a singly occupiedπ-type or- bital delocalized across the central conjugated framework

    and yields a markedly improved description of the energies and relative intensities of the first two experimental peaks (Fig- ure 15). Natural transition orbital analysis assigns peak 1 pre- dominantly to a C3 1s excitation into a singly occupiedπ-type or- bital delocalized across the central conjugated framework. Peak 2 arises primarily from a C4 1s exci...

    work page 2004

  3. [63]

    Scutelnic, D

    12 V. Scutelnic, D. S. Tikhonov, S. Marggi Poullain, E. A. Haugen, A. Attar, L. Barreau, K. Chang, A. P. Fidler, J. D. Gaynor, Y.-C. Lin, J. Yang, E. G. Champenois, X. Wang, X. Shen, D. Luo, F. Ji, S. Weathersby , Y. V. Vishnevskiy , M. Schnell, T. J. A. Wolf, M.-F. Lin and S. R. Leone,J. Phys. Chem. Lett., 2025, 16, 5068–5075. 13 H. Nakanishi, H. Morita ...

    work page 2025

  4. [158]

    Faber, E

    22 R. Faber, E. F. Kjønstad, H. Koch and S. Coriani,J. Chem. Phys., 2019,151, 144107. 23 A. F ¨orster and L. Visscher,J. Chem. Theory Comput., 2022,18, 6779–6793. 24 B. Helmich-Paris,Int. J. Quantum Chem., 2021,121, e26559. 25 K. Andersson, P.- ˚A. Malmqvist and B. O. Roos,J. Chem. Phys., 1992,96, 1218–1226. 26 C. Angeli, R. Cimiraglia and J.-P. P. Malrie...

    work page 2019

  5. [302]

    Hentschel, R

    3 M. Hentschel, R. Kienberger, C. Spielmann, G. A. Rei- der, N. Milosevic, T. Brabec, P. Corkum, U. Heinzmann, M. Drescher and F. Krausz,Nature, 2001,414, 509–513. 4 S.-C. Lin, C.-C. Chang, S.-Y. Chiu, H.-T. Pai, T.-Y. Liao, C.-S. Hsu, W.-H. Chiang, M.-K. Tsai and H. M. Chen,Nat. Commun., 2020,11,

    work page 2001

  6. [1061]

    40 C. E. V. de Moura and A. Y. Sokolov,Phys. Chem. Chem. Phys., 2022,24, 8041–8046. 41 C. E. V. de Moura and A. Y. Sokolov,Phys. Chem. Chem. Phys., 2022,24, 4769–4784. 42 I. M. Mazin and A. Y. Sokolov,J. Chem. Theory Comput., 2023, 19, 4991–5006. 43 C. E. V. De Moura and A. Y. Sokolov,J. Phys. Chem. A, 2024, 128, 5816–5831. 44 N. P. Gaba, C. E. V. de Mour...

    work page 2022

  7. [1242]

    55 B. O. Roos, P. R. Taylor and P. E. M. Sigbahn,Chem. Phys., 1980,48, 157–173. 56 P. E. M. Siegbahn, J. Alml ¨of, A. Heiberg and B. O. Roos,J. Chem. Phys., 1981,74, 2384–2396. 57 H.-J. Werner and W. Meyer,J. Chem. Phys., 1981,74, 5794–

    work page 1980

  8. [2008]

    Schirmer, inMany-Body Methods for Atoms, Molecules and Clusters, ed

    49 J. Schirmer, inMany-Body Methods for Atoms, Molecules and Clusters, ed. J. Schirmer, Springer International Publishing, Cham, 2018, pp. 195–204. 50 A. Trofimov, I. Krivdina, J. Weller and J. Schirmer,Chem. Phys., 2006,329, 1–10. 51 J. Oddershede, P. Jørgensen and D. L. Yeager,Comput. Phys. Rep., 1984,2, 33–92. 52 J. Oddershede, inMethods in Computation...

    work page 2018

Show all 13 references
  1. [2025]

    64 S. Mai, P. Marquetand and L. Gonz ´alez,WIREs Comput. Mol. Sci., 2018,8, e1370. 65 J. E. Runeson,J. Chem. Phys., 2025,163, 154105. 66 G. Granucci and M. Persico,J. Chem. Phys., 2007,126, 134114. 67 T. Shiozaki, W. Gy ˝orffy , P. Celani and H.-J. Werner,J. Chem. Phys., 2011,...

    2018

  2. [3525]

    5 K. Miao, S. Chen and J. Zhou,Sustain. Mater . T echnol., 2025, 45, e01608. 6 J. Kim, K. H. Kim, K. Y. Oang, J. H. Lee, K. Hong, H. Cho, N. Huse, R. W. Schoenlein, T. K. Kim and H. Ihee,Chem. Com- mun., 2016,52, 3734–3749. 7 E. A. Haugen, D. Hait, V. Scutelnic, T. Xue, M. Hea...

    2025

  3. [5053]

    59 L. S. Cederbaum,Phys. Rev. A, 1987,35, 622–631. 60 J. C. Tully ,J. Chem. Phys., 1990,93, 1061–1071. 61 S. Mukherjee, R. S. Mattos, J. M. Toldo, H. Lischka and M. Barbatti,J. Chem. Phys., 2024,160, 154306. 62 A. Jain and A. Sindhu,ACS Omega, 2022,7, 45810–45824. 63 S. Mai, B...

    1987

  4. [5801]

    Werner and P

    58 H. Werner and P. J. Knowles,J. Chem. Phys., 1985,82,

    1985

  5. [9143]

    Liu and D

    69 W. Liu and D. Peng,The Journal of Chemical Physics, 2009, 131, 031104. 70 T. Shiozaki,WIREs Computational Molecular Science, 2018,8, e1331. + P V S O B M / B N F < Z F B S > < W P M > 1–14|13 71 S. Mai, D. Avagliano, M. Heindl, P. Marquetand, M. F. S. J. Menger, M. Oppel, F...

    2009

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.