The reviewed record of science sign in
Pith

arxiv: 2607.06380 · v1 · pith:TRQFTRUF · submitted 2026-07-07 · physics.chem-ph

Reaction Pathway Detection using Machine-Learned Energy Potentials -- Decomposition of Energized CF₃CHOO

Cangtao Yin , Markus Meuwly This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-07-08 07:39 UTCglm-5.2pith:TRQFTRUFrecord.jsonopen to challenge →

classification physics.chem-ph
keywords reactiondecompositionpathwaysapproacheschoodynamicsenergizedenergy
0
0 comments X

The pith

Machine-learned dynamics reveal HFC-23 forms from Criegee decomposition

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

This paper uses a machine-learned potential energy surface to run 10,000 molecular dynamics trajectories of the energized Criegee intermediate CF₃CHOO, a species formed during atmospheric ozonolysis of hydrofluoroolefins. The central claim is that the decomposition of this intermediate proceeds through multiple competing, non-statistical fragmentation pathways that static reaction-path calculations cannot predict. The simulations yield a 14% probability of forming HFC-23 (CHF₃), a potent greenhouse gas, which qualitatively agrees with the experimental yield of roughly 8%. The authors argue that excess internal energy redirects reaction outcomes away from minimum-energy pathways, and that the energy landscape alone is not a useful proxy for reaction kinetics because dynamical bottlenecks, mode-specific energy flow, and momentum distributions govern which products form.

Core claim

The decomposition of energized CF₃CHOO is governed by non-statistical dynamics: product branching is controlled not by barrier heights or thermodynamic stability but by how excess energy is distributed among internal degrees of freedom and how it flows on sub-picosecond timescales. The intermediate I3, though short-lived, branches into at least five distinct product channels, and the lowest-energy pathway is not the most traveled. A machine-learned PES trained on MP2 data reproduces reference energies to within ~0.1 kcal/mol across a 150 kcal/mol range and enables enough trajectories to map the full branching tree, including a 14% HFC-23 yield consistent with experiment.

What carries the argument

A PhysNet neural network trained on 41,252 MP2/aug-cc-pVTZ energy and force evaluations provides a global potential energy surface. Ten thousand classical MD trajectories are launched from thermally sampled initial structures of CF₃CHOO with 32 kcal/mol of internal energy (mimicking post-ozonolysis conditions), each propagated for up to 1 ns. Trajectory analysis identifies which transition states and intermediates are crossed, and stretched-exponential fits to survival probabilities quantify non-RRKM behavior through the Kohlrausch exponent β.

If this is right

  • Atmospheric models that assume RRKM statistics or follow only minimum-energy paths may systematically misestimate HFC-23 yields from hydrofluoroolefin ozonolysis, with consequences for greenhouse gas inventories.
  • The approach generalizes to other energized Criegee intermediates and short-lived atmospheric species where explicit ab initio dynamics is computationally infeasible.
  • The finding that barrier heights do not predict branching ratios challenges the use of static reaction diagrams as kinetic proxies for highly energized intermediates.
  • The 1 ns simulation window corresponds to tropospheric collision frequencies, making the gas-phase results directly relevant to the first atmospheric collision cycle.

Where Pith is reading between the lines

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

  • If the 32 kcal/mol internal energy assignment is off by even 4 kcal/mol, the fraction of trajectories crossing the first barrier changes from ~30% to <10%, which would propagate nonlinearly into the HFC-23 yield. A full sensitivity surface over energy would clarify whether the 14% result is robust or an artifact of a single energy choice.
  • The paper notes that CASPT2 lowers the TS1 barrier by ~2 kcal/mol, which would increase the TS1 crossing rate. Since all downstream yields are conditioned on crossing TS1, the HFC-23 yield could be systematically underestimated by the MP2-based PES.
  • The ~70% of trajectories that remain in I1 at 1 ns represent a reservoir whose fate on longer timescales or under collisional conditions is unaddressed; these could contribute additional product channels if collisional stabilization does not intervene first.

Load-bearing premise

The internal energy assigned to CF₃CHOO (32 kcal/mol, roughly half the ~70 kcal/mol released in the parent ozonolysis) is an approximation without rigorous justification, and the results are highly sensitive to it: lowering the energy to 28 kcal/mol reduces the TS1 crossing fraction from ~30% to under 10%, which would drastically change all downstream product yields.

What would settle it

If a full sensitivity analysis over internal energies from 28–35 kcal/mol shows that the HFC-23 yield varies from below 5% to above 20%, the quantitative agreement with experiment would be revealed as coincidental rather than predictive.

Figures

Figures reproduced from arXiv: 2607.06380 by Cangtao Yin, Markus Meuwly.

Figure 1
Figure 1. Figure 1: Reaction diagram showing all the reaction pathways studied in this work. Panel A: [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Upper panel shows a representative trajectory for the first step in the unimolecular [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Normalized probability distributions P(rCC, rOH) while sampling I3 well for each of the direct (left) and indirect (right) channels. For the indirect channel, the C–C stretch (with some admixture of the O–H separation) is the main progression coordinate, whereas the indirect pathway proceeds along a combination of the O–H and C–C bonds. The numbers in the contour lines indicate the normalized intensity of … view at source ↗
Figure 4
Figure 4. Figure 4: Tree diagram of reaction pathways showing the percentage of trajectories (relative [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The survival fractions N(τ )/N(0) for intermediates I2 (Panel A), I3 (Panel B), I4 (Panel C), and I5 (Panel D) corresponding to the reaction channels as shown in [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
read the original abstract

Characterization of the decomposition products of energized Criegee intermediates is essential for assessing their impact on the chemical evolution of the atmosphere. Here, a generic and microscopically resolved approach is used to determine the molecular fragmentation pathways and products for CF$_3$CHOO. They include, among others, direct formation of CO$_2$ + CHF$_3$ (HFC-23), HF + CO$_2$ + CF$_2$, and fragmentation routes that are not evident from static reaction path calculations alone. The computed probability for formation of HFC-23 of 14 \% qualitatively agrees with a value of $(7.9^{+0.4}_{-0.2})$ \% from recent measurements, given the differences in the two approaches. Non-statistical dynamics is found for almost all decomposition pathways and the simulations show that excess energy can redirect reaction outcomes away from minimum-energy pathways. The results highlight the power of machine-learned PESs to elucidate multi-step reaction mechanisms of atmospherically relevant intermediates beyond traditional Master equation/electronic structure approaches to provide molecular-level understanding of the role of dynamics.

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 / 7 minor

Summary. This manuscript studies the unimolecular decomposition of energized CF₃CHOO (a fluorinated Criegee intermediate) using molecular dynamics on a machine-learned potential energy surface (PhysNet) trained on MP2/aug-cc-pVTZ reference data. The authors train four independent neural network models with active learning (20 iterations, final dataset ~41,000 structures), validate the PES extensively (R² > 0.999 for energies, trajectory-level agreement with MP2), and run 10,000 reactive MD trajectories at 32 kcal/mol internal energy. They identify multiple competing fragmentation pathways—including direct CO₂ + CHF₃ (HFC-23) formation and several HF-containing channels—and report a 14% HFC-23 yield, which they compare to an experimental value of ~8%. Non-statistical dynamics is analyzed via stretched-exponential survival fits, revealing compressed-exponential behavior (β > 1) for I₃ and near-RRKM behavior for I₄ and I₅.

Significance. The paper makes a valuable contribution by applying a well-validated ML-PES to a problem of atmospheric relevance (HFC-23 formation from Criegee intermediate decomposition). The PES construction is rigorous: four independent PhysNet models, active learning with explicit disagreement-based sampling, trajectory-level validation against MP2, and open data availability (https://github.com/MMunibas/cf3choo). The identification of non-statistical pathways not evident from IRC calculations, and the systematic survival-fraction analysis across intermediates, provide genuinely new mechanistic insight. The comparison to experimental HFC-23 yields, while acknowledged as approximate, contextualizes the results. The finding that excess energy redirects reaction outcomes away from minimum-energy pathways is a substantive dynamical result.

major comments (1)
  1. §4 (Methods), final paragraph: The paper states that 'additional simulations were run with somewhat lower (28.0 kcal/mol) and higher (35.0 kcal/mol) amounts of internal energy,' but only the 28 kcal/mol result is reported ('fewer than 10% of trajectories cross TS1'). The 35 kcal/mol outcome is never mentioned again. This is a load-bearing omission because the central quantitative claim (14% HFC-23 yield) depends directly on the fraction of trajectories crossing TS1 (29.6% at 32 kcal/mol). The paper's own sensitivity probe at 28 kcal/mol demonstrates extreme sensitivity: a 4 kcal/mol decrease reduces TS1 crossing from 29.6% to <10%. Without the upper-bound result, the robustness of the 14% yield cannot be assessed. If the 35 kcal/mol simulations yield a substantially higher HFC-23 probability, the 'qualitative agreement' with the 8% experimental value would need qualification. The authors
minor comments (7)
  1. Table S2: The product labels 'dioxi' and 'explosion' are unclear. Please define these categories or replace with descriptive names.
  2. §2.2: The sentence 'Note that the percentages of pathways leaving I3 do not sum to the 29% of trajectories arriving at I3 because only the major channels are shown here' could be clarified by stating the exact fraction accounted for by minor channels.
  3. Figure 1 caption: The panels A/B/C are labeled in the figure but the caption does not explicitly state which panel corresponds to which part of the mechanism. A brief description in the caption would help.
  4. §3 (Discussion): The comparison between the computed 14% and experimental 7.9% would benefit from a more quantitative discussion of how the unknown primary ozonide branching ratio and collisional stabilization affect the comparison. The current treatment is qualitative.
  5. Table 1: The entry for I3 → CO₂+CHF₃ reports τr = 0.08±0.00 ps. The uncertainty should be reported with at least one significant figure.
  6. §4 (Methods): 'dcomposition' should be 'decomposition'. Also, 'stratosphere' in the sentence about collision frequency should likely be 'troposphere' (the paper elsewhere correctly states troposphere).
  7. Figure S5: The caption mentions 'Panels A, B, and C' but the figure axes are not labeled with panel letters in the image provided. Please ensure panel labels are visible.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and the constructive evaluation. The referee raises one major point regarding the unreported 35 kcal/mol simulation results. We agree this is a legitimate omission and will address it in the revised manuscript.

read point-by-point responses

  1. Referee: The 35 kcal/mol simulation outcome is never mentioned after being stated in Methods. The 28 kcal/mol result shows extreme sensitivity of TS1 crossing (from 29.6% to <10%), so without the upper-bound result the robustness of the 14% HFC-23 yield cannot be assessed.

    Authors: The referee is correct that the 35 kcal/mol results are mentioned in the Methods section but never reported in the manuscript. This is an omission on our part, and we agree it should be addressed. We will include the 35 kcal/mol results in the revised manuscript. Specifically, we will report the TS1 crossing fraction and the HFC-23 yield at 35 kcal/mol, and we will add a brief discussion of how the product branching depends on internal energy across the three values (28, 32, and 35 kcal/mol). This will allow readers to assess the sensitivity of the central quantitative claim. We note that the 32 kcal/mol internal energy was chosen as physically motivated—approximately half of the ~70 kcal/mol released in the ozonolysis of HFO-1234ze(E)—and the 28 and 35 kcal/mol simulations serve as sensitivity probes around this value. Including the upper-bound result will make this sensitivity analysis complete. If the 35 kcal/mol simulations yield a substantially higher HFC-23 probability, we will qualify the 'qualitative agreement' statement accordingly, as the referee suggests. revision: yes

Circularity Check

0 steps flagged

No significant circularity found; derivation is self-contained against external benchmarks

full rationale

The paper's central claim — a 14% HFC-23 formation probability from CF₃CHOO decomposition — emerges from explicit MD simulations on a PhysNet PES trained against MP2/aVTZ reference data (an independent electronic structure method). The product yields are not fitted to experimental data; they arise from trajectory statistics (10000 trajectories). The comparison to the experimental yield of 7.9% (from Ref. 15, an independent group) serves as external validation. The PES is validated on structures outside the training set (Figure 2, test set R² > 0.9999), and the reaction energetics from the PES are cross-checked against MP2 single-point calculations (Figure 1, black vs. red values). Self-citations exist (Refs. 5–8, 17–18 by Yin; Ref. 36 PhysNet by Unke & Meuwly), but none are load-bearing for the central quantitative claim: the 14% yield is not defined in terms of any self-cited result, nor is it a fitted parameter renamed as a prediction. The 32 kcal/mol internal energy assignment is a physical approximation (half of ~70 kcal/mol from ozonolysis), not a circular definition. The sensitivity probe at 28 kcal/mol is reported and shows the expected decrease in TS1 crossing. The omission of 35 kcal/mol results is a completeness concern (correctness risk), not a circularity issue. The derivation chain (MP2 data → trained PES → MD trajectories → product yields → comparison with experiment) contains no step that reduces to its own inputs by construction.

Axiom & Free-Parameter Ledger

5 free parameters · 5 axioms · 0 invented entities

The paper introduces no new physical entities, particles, forces, or dimensions. All chemical species (intermediates I1-I5, transition states TS1-TS9, products) are standard molecular structures. The ML-PES is a computational tool, not a new physical entity. No free parameters are fitted to the target result (HFC-23 yield); the yield emerges from dynamics on a PES trained against external quantum chemistry data.

free parameters (5)
  • Internal energy of CF₃CHOO = 32.0 kcal/mol (16000 K)
    Chosen to represent approximately half the ~70 kcal/mol released from ozonolysis of HFO-1234ze(E). Directly determines trajectory outcomes. Sensitivity tested at 28 and 35 kcal/mol but not systematically.
  • Energy difference threshold for active learning = 1.0 kcal/mol
    Threshold for adding new geometries to the training set during the 20-iteration active learning loop. Affects final dataset composition but not a fitted parameter of the central claim.
  • Energy cutoff for dataset = 100 kcal/mol above reactant
    Structures above this energy removed from training set. Standard practice but affects PES coverage.
  • Simulation time = 1 ns
    Chosen to correspond to tropospheric collision frequency of 10^9/s. Affects observed product yields since 70% of trajectories remain unreacted.
  • Stretched-exponential parameters (τr, β) = Varied per channel (Table 1)
    Fitted to survival fractions for phenomenological characterization. Not used to derive the central claim (product yields).
axioms (5)
  • domain assumption MP2/aVTZ provides an adequate single-reference description of the CF₃CHOO PES for most stationary points
    Invoked in Methods and Discussion. T1 diagnostics (Table S3) support this for most species except TS1 (T1=0.040). CASPT2 calculations (Table S4) show TS1 barrier is ~2 kcal/mol lower, acknowledged but not corrected in the PES.
  • domain assumption Normal mode sampling at 300 K represents the vibrational ground state of I1
    Invoked in §2.2: 'Sampling from an equilibrium distribution at 300 K ensures that the initial structures are representative of the vibrational ground state region of I1.'
  • domain assumption Assigning ~32 kcal/mol to CF₃CHOO meaningfully approximates post-ozonolysis energy content
    Stated in Methods: 'The energy released following the reaction HFO-1234ze(E) + O₃ → CF₃CHOO + HCHO reactant is ~70 kcal/mol. Hence, assigning approximately half that value to CF₃CHOO is a meaningful, albeit approximate procedure.'
  • domain assumption 1 ns simulation time is meaningful for atmospheric relevance
    Invoked in Discussion: 'The time scale of the ML-MD simulations (1 ns) is meaningful, because it corresponds to typical collision frequencies of 10^9/s in the troposphere.'
  • standard math PhysNet neural network architecture can faithfully represent a global reactive PES for CF₃CHOO
    The PhysNet architecture (ref 36) is a published, validated framework. The paper provides extensive validation (Table S1, Figures S4, S5) supporting this assumption for the present system.

pith-pipeline@v1.1.0-glm · 19465 in / 3626 out tokens · 397453 ms · 2026-07-08T07:39:32.511386+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

300 extracted references · 300 canonical work pages

  1. [1]

    Algorithms for Macromolecular Dynamics and Constraint Dynamics , author =. Mol. Phys. , volume = 34, pages =

    work page

  2. [2]

    CHARMM: The Biomolecular Simulation Program , author =

    work page

  3. [3]

    Comparison of Simple Potential Functions for Simulating Liquid Water , author =

    work page

  4. [4]

    The structure of 2Zn pig insulin crystals at 1.5 \ resolution , author =. Philos. Trans. R. Soc. London, Ser. B , volume = 319, number = 1195, pages =

    work page

  5. [5]

    Update of the CHARMM All-Atom Additive Force Field for Lipids: Validation on Six Lipid Types , author =

    work page

  6. [6]

    The protein data bank: A computer-based archival file for macromolecular structures , author =

    work page

  7. [7]

    Biological Crystallography , publisher =

    The protein data bank , author =. Biological Crystallography , publisher =

    work page

  8. [8]

    A Computer Simulation Method for the Calculation of Equilibrium Constants for the Formation of Physical Clusters of Molecules: Application to Small Water Clusters , author =. J. Chem. Phys. , volume = 76, pages =

    work page

  9. [9]

    A Unified Formulation o the Constant Temperature Molecular-Dynamics Methods , author =. J. Chem. Phys , volume = 81, pages =

    work page

  10. [10]

    Canonical Dynamics: Equilibrium Phase-Space Distributions , author =. Phys. Rev. A , volume = 31, pages =

    work page

  11. [11]

    Molecular Dynamics Simulations at Constant Pressure and/or Temperature , author =. J. Chem. Phys. , volume = 72, pages =

    work page

  12. [12]

    Constant Pressure Molecular Dynamics for Molecular Systems , author =. Mol. Phys. , volume = 50, pages =

    work page

  13. [13]

    New Spherical-Cutoff Methods for Long-Range Forces in Macromolecular Simulation , author =. J. Comp. Chem. , volume = 15, pages =

    work page

  14. [14]

    Particle Mesh Ewald: An Nlog(N) Method for Ewald Sums in Large Systems , author =. J. Chem. Phys. , volume = 98, pages =

    work page

  15. [15]

    Geometric Numerical Integration Illustrated by the St

    Ernst Hairer and Christian Lubich and Gerhard Wanner , year = 2003, journal =. Geometric Numerical Integration Illustrated by the St

    work page 2003

  16. [16]

    Leveraging Symmetries of Static Atomic Multipole Electrostatics in Molecular Dynamics Simulations , author =

    work page

  17. [17]

    Atomic Multipoles: Electrostatic Potential Fit, Local Reference Axis Systems and Conformational Dependence , author =

    work page

  18. [18]

    Molecular dynamic simulations of OH-stretching overtone induced photodissociation of fluorosulfonic and chlorosulfonic acid , author =

    work page

  19. [19]

    Fluorosulfonic acid and chlorosulfonic acid: Possible candidates for OH-stretching overtone-induced photodissociation , author =

    work page

  20. [20]

    Overtone spectroscopy of sulfonic acid derivatives , author =

    work page

  21. [21]

    Photolysis of sulfuric acid vapor by visible solar radiation , author =

    work page

  22. [22]

    Spectroscopy of transition region species , author =

    work page

  23. [23]

    Theoretical studies of enzymic reactions: Dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme , author =

    work page

  24. [24]

    An empirical valence bond approach for comparing reactions in solutions and in enzymes , author =

    work page

  25. [25]

    RMDff: A smoothly transitioning, forcefield-based representation of kinetics for reactive molecular dynamics simulations , author =

    work page

  26. [26]

    , author =

    CHARMM: the biomolecular simulation program. , author =

    work page

  27. [27]

    , author =

    Competitive reaction pathways in vibrationally induced photodissociation of H _2 SO _4 . , author =

    work page

  28. [28]

    , author =

    HSO _3 Cl : a prototype molecule for studying OH-stretching overtone induced photodissociation. , author =

    work page

  29. [29]

    Laser excited N _2^+ in a 22-pole ion trap: Experimental studies of rotational relaxation processes , author =

    work page

  30. [30]

    Collision-induced rotational excitation in N 2+(2 g+, v= 0)--Ar: Comparison of computations and experiment , author =

    work page

  31. [31]

    unpublished , volume =

    Quantum and Quasiclassical Trajectory Studies of Rotational Relaxation in Ar-N _2^+ Collisions , author =. unpublished , volume =

    work page

  32. [32]

    Investigation of the ArN+ 2 ion by dissociative ionization of argon/nitrogen clusters , author =

    work page

  33. [33]

    Constructing multidimensional molecular potential energy surfaces from ab initio data , author =

    work page

  34. [34]

    Protein conformational dynamics probed by single-molecule electron transfer , author =

    work page

  35. [35]

    Recognition dynamics up to microseconds revealed from an RDC-derived ubiquitin ensemble in solution , author =

    work page

  36. [36]

    Proteins in Action: Femtosecond to Millisecond Structural Dynamics of a Photoactive Flavoprotein , author =

    work page

  37. [37]

    Science , volume = 254, number = 5030, pages =

    THE 1ST STEP IN VISION - FEMTOSECOND ISOMERIZATION OF RHODOPSIN , author =. Science , volume = 254, number = 5030, pages =. doi:10.1126/science.1925597 , issn =

    work page doi:10.1126/science.1925597

  38. [38]

    Science , volume = 266, number = 5184, pages =

    VIBRATIONALLY COHERENT PHOTOCHEMISTRY IN THE FEMTOSECOND PRIMARY EVENT OF VISION , author =. Science , volume = 266, number = 5184, pages =. doi:10.1126/science.7939680 , issn =

    work page doi:10.1126/science.7939680

  39. [39]

    Conical intersection dynamics of the primary photoisomerization event in vision , author =

    work page

  40. [40]

    Femtosecond pump-shaped-dump quantum control of retinal isomerization in bacteriorhodopsin , author =

    work page

  41. [41]

    An experimental study , author =

    Coherent control of the population transfer in complex solvated molecules at weak excitation. An experimental study , author =

    work page

  42. [42]

    Cell cycle-dependent degradation of a flagellar motor component requires a novel-type response regulator , author =

    work page

  43. [43]

    Aldridge and R

    P. Aldridge and R. Paul and P. Goymer and P. Rainey and U. Jenal , year = 2003, journal = MM, volume = 47, pages =. Role of the

    work page 2003

  44. [44]

    Cell cycle-dependent dynamic localization of a bacterial response regulator with a novel di-guanylate cyclase output domain , author =

    work page

  45. [45]

    G. J. Burton and G. B. Hecht and A. Newton , year = 1997, journal = JBact, volume = 179, pages =. Roles of the Histidine Protein Kinase

    work page 1997

  46. [46]

    Two-Component Signal Transduction , author =

    work page

  47. [47]

    , author =

    Metastable CO binding sites in the photoproduct of a novel cooperative dimeric hemoglobin. , author =

    work page

  48. [48]

    , author =

    Structural transitions upon ligand binding in a cooperative dimeric hemoglobin. , author =. Science , volume = 249, number = 4968, pages =

    work page

  49. [49]

    , author =

    Residue F4 plays a key role in modulating oxygen affinity and cooperativity in Scapharca dimeric hemoglobin. , author =

    work page

  50. [50]

    , author =

    Structural and thermodynamic aspects of cooperativity in the homodimeric hemoglobin from Scapharca inaequivalvis. , author =

    work page

  51. [51]

    Is cooperative oxygen binding by hemoglobin really understood? , author =

    work page

  52. [52]

    Protein Conformational Dynamics of Homodimeric Hemoglobin Revealed by Combined Time-Resolved Spectroscopic Probes , author =

    work page

  53. [53]

    , author =

    Cooperativity in Scapharca dimeric hemoglobin: simulation of binding intermediates and elucidation of the role of interfacial water. , author =

    work page

  54. [54]

    , author =

    Ligand migration and cavities within Scapharca Dimeric HbI: studies by time-resolved crystallo-graphy, Xe binding, and computational analysis. , author =

    work page

  55. [55]

    Ligand Migration and Binding in the Dimeric Hemoglobin of Scapharca inaequivalvis , author =

    work page

  56. [56]

    Cyclic di-guanosine-monophosphate comes of age: a novel secondary messenger involved in modulating cell surface structures in bacteria? , author =

    work page

  57. [57]

    Mechanisms of Cyclic-di-GMP Signaling in Bacteria , author =. Annu. Rev. Genet. , volume = 40, pages =

    work page

  58. [58]

    Allosteric Control of Cyclic di-GMP Signaling , author =

    work page

  59. [59]

    Ligand and interfacial dynamics in a homodimeric hemoglobin , author =

    work page

  60. [60]

    Studying reactive processes with classical dynamics: Rebinding dynamics in MbNO , author =

    work page

  61. [61]

    Atomistic Simulation of Adiabatic Reactive Processes Based on Multi-State Potential Energy Surfaces , author =

    work page

  62. [62]

    Dynamics of vibrational overtone excitations of H2SO4, H2SO4-H2O: Hydrogen-hopping and photodissociation processes , author =

    work page

  63. [63]

    Reactive force fields for proton transfer dynamics , author =. J. Comput. Chem , publisher =. doi:10.1002/jcc.20864 , issn =

    work page doi:10.1002/jcc.20864

  64. [64]

    Spectroscopy and dynamics of double proton transfer in formic acid dimer

    Mackeprang, Kasper and Xu, Zhen-Hao and Maroun, Zeina and Meuwly, Markus and Kjaergaard, Henrik G. Spectroscopy and dynamics of double proton transfer in formic acid dimer. Phys. Chem. Chem. Phys. doi:10.1039/C6CP03462D

    work page doi:10.1039/c6cp03462d

  65. [65]

    doi:10.1073/pnas.1306642110 , url =

    Proton transfer through the water gossamer , author =. doi:10.1073/pnas.1306642110 , url =

    work page doi:10.1073/pnas.1306642110

  66. [66]

    ab initio

    Simulation of proton transfer along ammonia wires: An “ab initio” and semiempirical density functional comparison of potentials and classical molecular dynamics , author=. 2002 , publisher=

    work page 2002

  67. [67]

    Mixed Grotthuss and Vehicle Transport Mechanism in Proton Conducting Polymers from Ab initio Molecular Dynamics Simulations , author =. Chem. Mater. , volume = 23, number = 6, pages =. doi:10.1021/cm102674u , url =

    work page doi:10.1021/cm102674u

  68. [68]

    doi:10.1126/science.aaf8425 , issn =

    Spectroscopic snapshots of the proton-transfer mechanism in water , author =. doi:10.1126/science.aaf8425 , issn =

    work page doi:10.1126/science.aaf8425

  69. [69]

    doi:10.1073/pnas.1420734111 , eprint =

    Site-specific vibrational spectral signatures of water molecules in the magic H3O+(H2O)20 and Cs+(H2O)20 clusters , author =. doi:10.1073/pnas.1420734111 , eprint =

    work page doi:10.1073/pnas.1420734111

  70. [70]

    doi:10.1021/acs.jpca.5b10649 , url =

    Diffuse Vibrational Signature of a Single Proton Embedded in the Oxalate Scaffold, HO2CCO2– , author =. doi:10.1021/acs.jpca.5b10649 , url =

    work page doi:10.1021/acs.jpca.5b10649

  71. [71]

    Kinetic isotope effect in malonaldehyde determined from path integral Monte Carlo simulations

    Huang, Jing and Buchowiecki, Marcin and Nagy, Tibor and Vanicek, Jiri and Meuwly, Markus. Kinetic isotope effect in malonaldehyde determined from path integral Monte Carlo simulations. Phys. Chem. Chem. Phys. doi:10.1039/C3CP53698J

    work page doi:10.1039/c3cp53698j

  72. [72]

    Proton Diffusion Mechanisms and Conductivity , author =

    Thermodynamics and Proton Transport in Nafion: II. Proton Diffusion Mechanisms and Conductivity , author =. J. Electrochem. Soc. , volume = 152, number = 3, pages =. doi:10.1149/1.1859814 , url =

    work page doi:10.1149/1.1859814

  73. [73]

    Quantum Modeling of Complex Molecular Systems

    Jean-Louis Rivail and Manuel Ruiz-Lopez and Xavier Assfeld. Quantum Modeling of Complex Molecular Systems

    work page

  74. [74]

    A generalized reactive force field for nonlinear hydrogen bonds: Hydrogen dynamics and transfer in malonaldehyde

    Yang, Yonggang and Meuwly, Markus. A generalized reactive force field for nonlinear hydrogen bonds: Hydrogen dynamics and transfer in malonaldehyde. J. Chem. Phys. doi:http://dx.doi.org/10.1063/1.3447701

    work page doi:10.1063/1.3447701

  75. [75]

    doi:10.1021/acs.jpca.5b01863 , issn =

    Infrared and Near-Infrared Spectroscopy of Acetylacetone and Hexafluoroacetylacetone , author =. doi:10.1021/acs.jpca.5b01863 , issn =

    work page doi:10.1021/acs.jpca.5b01863

  76. [76]

    On the relationship between infrared spectra of shared protons in different chemical environments: A comparison of protonated diglyme and protonated water dimer , author =

    work page

  77. [77]

    Application of Molecular Dynamics Simulations in Molecular Property Prediction. 1. Density and Heat of Vaporization , author =

    work page

  78. [78]

    and Zhang, Yuhong and Pastor, Richard W

    Feller, Scott E. and Zhang, Yuhong and Pastor, Richard W. and Brooks, Bernard R. Constant pressure molecular dynamics simulation: The Langevin piston method

    work page

  79. [79]

    Solvents and Solvent Effects in Organic Chemistry , author =

    work page

  80. [80]

    Particle mesh Ewald: An N.log(N) method for Ewald sums in large systems

    Darden, Tom and York, Darrin and Pedersen, Lee. Particle mesh Ewald: An N.log(N) method for Ewald sums in large systems

    work page

Showing first 80 references.