Pith. sign in

REVIEW 2 major objections 6 minor 53 references

Modified Marrone-Treanor model: parameterization and benchmarking for five-species air

T0 review · 2 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The Modified Marrone-Treanor two-temperature model, fitted entirely to ab initio trajectory data, reproduces direct molecular simulation of five-species air in post-shock heat baths at low computational cost.

desk verdict The most complete MMT parameter set for five-species air to date, with genuinely good DMS benchmarks and an honest Park comparison; just remember the validation is same-source, which the authors openly admit. read the letter →

arxiv 2506.20521 v2 pith:OG6XRMLT submitted 2025-06-25 physics.chem-ph physics.flu-dyn

Erik Torres , Thomas Gross , Graham V. Candler , Thomas E. Schwartzentruber This is my paper
classification physics.chem-phphysics.flu-dyn PACS 82.20.Kh47.40.Ki
keywords ModifiedMarrone-Treanormodeltwo-temperaturenonequilibriumchemistryfive-speciesairquasiclassicaltrajectorycalculationsdirectmolecularsimulationpotentialenergysurfacesvibrationalrelaxationZeldovichreactions
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

This paper completes a two-temperature chemistry model for five-species air (N2, O2, NO, N, O) by converting quasiclassical-trajectory and direct-molecular-simulation data from ab initio potential energy surfaces into analytical rate parameters, vibrational relaxation times, and energy-coupling corrections that any computational fluid dynamics code can evaluate. The central demonstration is that the resulting Modified Marrone-Treanor (MMT) model reproduces direct molecular simulation benchmark solutions with high accuracy in zero-dimensional heat baths that mimic strong post-shock nonequilibrium, at a computational cost comparable to the standard Park model. If the claim holds, hypersonic flow solvers gain a low-cost route to first-principles-based chemistry predictions that would otherwise require far more expensive molecular simulation. The paper also reports that the two models disagree sharply: MMT converts N2 into N noticeably more slowly below 10000 K and produces several times more nitric oxide at all temperatures, a difference traced to QCT-derived Zeldovich rate coefficients that exceed Park's values.

What carries the argument

The central object is the MMT vibrational nonequilibrium factor, $Z(T,T_v) = \tilde{Q}_v(T)\,\tilde{Q}_v(T_p)/[\tilde{Q}_v(T_v)\,\tilde{Q}_v(-V)]$, a ratio of approximate vibrational partition functions $\tilde{Q}_v$ evaluated at the trans-rotational temperature, the vibrational temperature, and two pseudotemperatures $T_p$ and $V$ that couple the local thermal state to the dissociation temperature and to reaction-specific parameters. This factor multiplies a modified Arrhenius rate $k_{\rm Arr}(T)$ and a non-Boltzmann factor $f_k^{\rm NB}=0.5$, and its companion formula for the average vibrational energy removed per dissociation, $\langle\varepsilon_{v,a}\rangle^{\rm MMT-NB} = f_\varepsilon^{\rm NB} k_B(\Theta_v/(e^{\Theta_v/T_p}-1) - \Theta_D/(e^{\Theta_D/T_p}-1))$, fixes the vibrational energy-chemistry coupling consistently with the rate expression. A second mechanism is the DMS-derived double-exponential fit for pair-wise vibrational relaxation times, $\tau_{v,a,s} = (p_0/p)[\exp(A^{\rm low}T^{-1/3}+B^{\rm low}) + \exp(A^{\rm high}T^{-1/3}+B^{\rm high})]$, which replaces Millikan-White correlations in the regime where they are known to fail. Detailed balance ties recombination to the same physics through $k_{\rm rec} = k_{\rm diss}^{\rm MMT-NB}(T,T)/K_c^{\rm eq}(T)$, and composition-dependent non-Boltzmann factors allow the model to recover thermal rates as the gas approaches equilibrium.

What would settle it

Shock-tube measurements of species time-histories in dissociating air at roughly 8000-12000 K would settle the claim: in particular, ultraviolet laser-absorption measurements of NO mole fraction in shock-heated N2/O2 mixtures test the signature high Zeldovich rates, since a measured peak NO level near the several-times-lower Park-model value rather than the MMT value would falsify the QCT-derived exchange rate coefficients, and the same data would test the predicted slower N2-to-N conversion below 10000 K.

Watch

Extended reading notes

Core claim

The paper claims that the Modified Marrone-Treanor two-temperature model, parameterized end-to-end from ab initio potential energy surfaces, reproduces the full five-species air chemistry of direct molecular simulation (DMS) with high accuracy in isothermal and adiabatic heat baths spanning 8000-20000 K, covering the vibrational excitation phase, the quasi-steady-state dissociation plateau, and the eventual approach to thermochemical equilibrium. The load-bearing structure is the MMT dissociation rate coefficient $k_{\rm diss}^{\rm MMT-NB}(T,T_v) = k_{\rm Arr}(T)\,Z(T,T_v)\,f_k^{\rm NB}$, whose vibrational nonequilibrium factor $Z$ is a ratio of approximate vibrational partition functions at pseudotemperatures that encode preferential dissociation from high vibrational levels, together with an analytical formula for the average vibrational energy removed per reaction and non-Boltzmann correction factors calibrated to trajectory data (0.5 for the rate, 0.85 for energy removal). Vibrational relaxation times are refitted from the same DMS data in a form that departs from Millikan-White behavior at high temperature, and detailed balance through the equilibrium constant makes three-body recombination drive the mixture correctly toward equilibrium. Benchmark comparisons deliberately exclude collision pairs for which no potential energy surface exists, notably the N2-NO and O2-NO pairs, whose rate parameters are reused from similar reactions only in the full-model version. In direct comparison with the Park TTv model, the MMT model predicts significantly slower conversion of N2 into N below 10000 K and significantly more NO production at all temperatures, which the authors trace to their Zeldovich rate coefficients being several times larger than Park's.

Load-bearing premise

The entire parameter set and the benchmark solutions it is validated against come from the same ab initio potential energy surfaces and the same quasiclassical-trajectory and direct-molecular-simulation approximations, and the paper deliberately replaces experimental validation with agreement against those simulations, so any shared error in the surfaces or the approximations is inherited by the model.

Editorial extensions

If this is right

  • Hypersonic CFD codes can evaluate five-species air chemistry analytically at roughly the cost of the Park model while carrying the content of ab initio QCT and DMS trajectory data.
  • Predicted shock-layer composition shifts relative to Park-model results, with N2-to-N conversion slowed below 10000 K and peak NO mole fractions several times higher, affecting radiative-heating and surface-catalysis estimates.
  • The DMS-derived relaxation-time tables replace Millikan-White correlations and their high-temperature correction for air species pairings in the range where those legacy correlations deviate most.
  • Because recombination is imposed through detailed balance, the model exhibits quasi-steady-state dissociation rates and a proper approach to thermochemical equilibrium, suiting it to both short-time post-shock and longer-time relaxation calculations.

Reading between the lines

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

  • The validation loop is internal: the MMT parameters and the DMS benchmark solutions derive from the same potential energy surfaces, so their close agreement measures self-consistency rather than fidelity to real air, and the decisive test will come from high-temperature shock-tube measurements of the same species histories.
  • If the high Zeldovich rates survive experimental scrutiny, aerothermal designs built on the Park model may have systematically underpredicted nitric oxide production and its radiative signature in hypersonic shock layers.
  • A testable extension suggested by the parameter structure: running MMT and DMS at matched heat-bath conditions while varying the initial vibrational-to-translational temperature ratio would check whether quasi-steady-state dissociation rates scale linearly with the thermal rate at any $T_v/T$, as the constant factors 0.5 and 0.85 imply.
  • The multi-electronic-surface enhancement factor for oxygen dissociation (16/3) is deliberately left as a limiting case, so the five O2-dissociation pre-exponentials are the first parameters new nonadiabatic trajectory studies would update.
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

2 major / 6 minor

Summary. This paper provides a complete parameterization of the Modified Marrone-Treanor (MMT) two-temperature dissociation model for five-species air (N2, O2, NO, N, O): modified Arrhenius parameters for 21 reactions, MMT pseudo-temperature parameters (U and T*), per-reaction and recommended uniform non-Boltzmann correction factors, and pair-wise vibrational relaxation times, all derived from quasiclassical trajectory (QCT) and direct molecular simulation (DMS) data on the Minnesota ab initio potential energy surfaces. The central claim is that a CFD implementation of MMT reproduces DMS reference solutions with high accuracy in zero-dimensional isothermal (8,000-20,000 K) and adiabatic heat baths representative of strong post-shock nonequilibrium. The paper further examines constant versus composition-dependent non-Boltzmann factors as the mixture approaches equilibrium, quantifies the influence of the speculative O2 multi-electronic-surface enhancement factor η = 16/3, and compares against Park's two-temperature model, finding slower N2-to-N conversion below 10,000 K and substantially higher peak NO mole fractions, attributed to significantly higher QCT-derived Zeldovich rate coefficients.

Significance. If the underlying PESs are faithful, this is a substantial practical contribution: the first complete, PES-derived MMT parameter set for five-species air, ready for CFD implementation at a computational cost comparable to Park's model. The strengths deserve explicit credit: the parameter tables (Tables 3-5) are complete and internally consistent; the benchmarking protocol is unusually careful (collision channels absent from DMS are excluded, thermodynamics are made PES-consistent via Table S2, missing relaxation pairs are set to infinity); the model demonstrably exhibits quasi-steady-state dissociation, detailed balance, and proper approach to equilibrium (Figs. 7-9); the sensitivity to the two main modeling choices (variable non-Boltzmann factors and the η factor) is quantified in Secs. IV-V; and the comparison with Park's model yields falsifiable predictions (slower N2-to-N conversion below 10,000 K, higher peak NO) that experiments could test. The authors disclose the missing-PES gaps and the surrogate-parameter strategy explicitly.

major comments (2)
  1. [Sec. III; Secs. II.D.1-II.D.2 and II.F, Tables 2 and 5 (notes [a])] The benchmark comparisons in Sec. III establish that the MMT model can emulate DMS solutions obtained on the same Minnesota PES set, and the Introduction states this explicitly: “This approach replaces traditional model validation, because nowhere in the process do we rely on experimental data.” The framing is honest and internally consistent, but the paper's broader claims — that the parameter tables enable “efficient CFD simulations of nonequilibrium air chemistry in hypersonic flows” and are proposed “for realistic air simulations” (Sec. II.D.2) — rest on two assumptions that deserve quantitative treatment rather than acknowledgment alone. First, any systematic bias in the PESs (e.g., high-energy repulsive regions, neglected electronically nonadiabatic transitions) appears on both sides of the comparison, so the close agreement in Figs. 7 and 8 does not test absolute accuracy; a concrete experimental anchor would materially strengthen the paper. In particular, Sec. VI reports that the QCT-derived Zeldovich rates are 2-6 times higher than Park's (Fig. 13), a strong, testable claim against existing shock-tube Zeldovich data; the authors should make that comparison or explicitly justify why Park's empirically calibrated rates are not a meaningful reference. Second, the surrogate assignments for reactions 3, 8, 11, and 12 (Table 2) and for the relaxation pairs N2-NO, O2-NO, NO-N2, and NO-O2 (Table 5) are excluded from the benchmarks by design, leaving the full model's predictions for those channels entirely unvalidated; an inexpensive sensitivity run that isolates the surrogate reactions (full-model solutions with and without reactions 3, 8, 11, and 12) would quantify their impact. I recommend scoping the conclusions accordingly, e.g., “consistent with the Minnesota PES-based DMS data” rather than “accurate for real air.”
  2. [Sec. II.D.3, Table 4] The recommendation of uniform non-Boltzmann factors f_k_NB = 0.5 and f_eps_NB = 0.85 is justified by the claim that per-reaction values “cluster within fairly narrow bands” and “mostly range between 0.4 and 0.5” for the rate factor. Table 4, however, lists f_k_NB = 0.285 for reaction 6 (O2 + N2), 0.360 for reaction 1 (N2 + N2), and 0.555 for reaction 15 (NO + O); the uniform value is 75 percent larger than the fitted value for reaction 6, and reactions 1 and 6 are among the most important dissociation channels in air. Since all Sec. III benchmarks use the uniform values, the claimed quantitative agreement in Figs. 7 and 8 is meaningful only if the heat-bath solutions are insensitive to these factor variations, which is not demonstrated. I request a sensitivity study that reruns the isothermal and adiabatic benchmark cases with the per-reaction values from Table 4 instead of the uniform pair. If the resulting profiles are nearly identical, the paper should state this explicitly and the per-reaction column can be presented as a diagnostic rather than a recommended input; if the profiles differ appreciably, the uniform-factor recommendation and the benchmarking claim that relies on it need to be revisited. The statistical-uncertainty argument in Sec. II.D.3 motivates pooling, but the magnitude of the effect on the benchmark profiles is the missing piece.
minor comments (6)
  1. [Sec. II.A, page 8] The sentence listing reactions evaluated with the MMT functional form says “reactions 1-12 and 14-15 in Table 2,” but reactions 11 and 12 re-use reaction 13's parameters, which Tables 2 and 4 treat as thermal and non-preferential; the correct list is reactions 1-10 and 14-15, and an implementer following the current text would mis-assign the rate treatment.
  2. [Nomenclature] The symbol n appears to be used for both number density and the modified Arrhenius temperature exponent, which is confusing in a manuscript whose tables list n as the exponent; consider renaming the number density.
  3. [Fig. 10 caption and Sec. VI] Several typographical artifacts remain (“the tree MMT paramet er sets of Table 7” in the Fig. 10 caption, “dash-totted lines” in Sec. VI); a careful proofreading pass is needed.
  4. [Sec. II.D.2 and Sec. V] The η = 16/3 enhancement is applied to the O2-dissociation rate coefficients, and because the recombination coefficients are obtained from detailed balance at (T, T), the same factor implicitly enhances the recombination direction; in the full model, which uses NASA-Lewis thermodynamics whose electronic partition functions already include excited O2 states, please clarify whether the enhanced recombination rate is intended and discuss any double-counting concern.
  5. [Table 5, note [a]] The implementation instruction to set A_low = A_high = 0 and B_low = B_high to a “very large” constant leaves the actual value undefined; for reproducibility, specify the numerical value or the criterion used.
  6. [Sec. IV] The variable non-Boltzmann factor formula is taken from Eq. (31) of Ref. [1], which is a preprint; since Sec. IV recommends variable factors for production use, consider restating the formula in this paper so that the recommendation is self-contained.

Circularity Check

3 steps flagged · score 5.0 of 10

Partial circularity via same-source validation: the non-Boltzmann factors, relaxation-time fits, and Arrhenius rates are calibrated to the same Minnesota-PES DMS/QCT data that generates the benchmark heat baths, so the reported 'high accuracy' against DMS is partly a restatement of the fit; independent content remains in the Zeldovich chemistry, adiabatic coupling, and Park-vs-MMT comparison.

  1. fitted input called prediction [Sec. I, fifth paragraph; benchmark design in Sec. III]
    "In this work we use DMS solutions in two complementary ways. First, to propose and calibrate the non-Boltzmann correction factors in the MMT model ... Second, to produce benchmark test cases against which CFD solutions with the MMT model will be compared. Since the DMS benchmarks and MMT parameters are both derived from the same set of ab initio PESs, close agreement between solutions obtained with the two methods constitutes the main metric of success. This approach replaces traditional model validation, because nowhere in the process do we rely on experimental data."

    The benchmark targets are not external to the fitting data: the same DMS solutions on the same Minnesota PESs supply the non-Boltzmann factors, the vibrational relaxation times, and (via QCT on identical PESs) every Arrhenius rate in Table 3. The abstract's headline claim that the model 'reproduces direct molecular simulation benchmark solutions with high accuracy' is therefore a consistency check between a reduced-order closure and its own generation source, which the paper itself acknowledges by calling it a replacement for validation by experiment. Because the 0D heat-bath integrations couple all fitted ingredients and the paper also compares against the independent Park model, the circularity is genuine but partial.

  2. fitted input called prediction [Sec. II.C and Sec. II.D.3, Eqs. (11) and (19), Table 4]
    "Eq. (19) contains another non-Boltzmann correction factor f_eps^NB, which was introduced to account for the reduction in vibrational energy removed during the quasi-steady-state (QSS) dissociation phase. As will be shown next, this simple correction factor can be calibrated to reproduce <eps_v>_diss observed in DMS calculations for N2, O2 and NO dissociation [27] with a range of collision partners. ... These values were derived from an analysis of the combined QCT and DMS data in Refs. [36] and [27] and are specific to each individual reaction."

    The two non-Boltzmann factors are, by construction, the measured DMS QSS deviations from thermal equilibrium: the rate factor is the near-constant ratio by which DMS QSS dissociation rates fall below thermal QCT rates, and the energy factor scales Eq. (19) onto the DMS QSS per-reaction energy removal. Substituting the calibrated constants into Eqs. (11) and (19) makes the model reproduce the DMS QSS dissociation rate and vibrational energy loss essentially by definition. The close QSS-phase agreement in the Sec. III heat baths is therefore a restatement of that calibration rather than an independent prediction of the coupled system.

1 more flagged steps
  1. self definitional [Sec. II.F, Eq. (22), Table 5 notes [a]-[b]; applied in Sec. III benchmarks]
    "The analytical expression chosen to curve-fit the DMS-derived pair-wise vibrational relaxation times (repeated from Eq. (19) in Ref. [1]) has the form [Eq. (22)]. ... Thus, in the model comparisons against DMS in Sec. III we will assume that the respective relaxation times tend toward infinity (see footnote [a] in Table 5), mimicking the behavior of the DMS reference solutions."

    The CFD relaxation model is built from curve fits of the very DMS data (Ref. [28]) from which the DMS heat-bath reference solutions are generated, and for species pairs with missing PES coverage the paper instructs the CFD model to set tau -> infinity expressly to 'mimic the behavior of the DMS reference solutions.' The vibrational-temperature time-histories in the benchmarks are thus imposed to agree with DMS by construction for those channels, rather than predicted. What remains genuinely unforced is the chemistry-relaxation coupling within the resolved channels.

full rationale

Main finding: partial circularity via same-source calibration and benchmarking, explicitly acknowledged in Sec. I ('This approach replaces traditional model validation, because nowhere in the process do we rely on experimental data'). The abstract's headline accuracy claim against DMS is weakened because the fitted inputs (non-Boltzmann factors, relaxation-time fits, Arrhenius rates on the same PESs) and the benchmark targets (DMS heat baths) share identical provenance in the Minnesota ab initio PES set and the same-group QCT/DMS machinery. Specific reductions: (i) the QSS dissociation rate and per-reaction vibrational energy removal are built from the DMS QSS ratios, making the QSS-phase agreement a restatement of that fit; (ii) relaxation channels with missing PESs are set to mimic the DMS behavior (tau -> infinity, Table 5 note [a]). Substantial independent, non-circular content remains: Zeldovich-driven NO production against the external Park model (Sec. VI), adiabatic T-Tv coupling, approach-to-equilibrium behavior (Sec. IV), and the visible discrepancies the paper itself reports (NO and O mole fractions at 1-50 us in Fig. 8a; N2 time lag; rotational nonequilibrium at 15,000-20,000 K). The load-bearing premise that DMS is a faithful stand-in for real air rests on externally generated PESs (Truhlar group, Refs. [5-15]) and an established method, so this is not a self-citation-chain or uniqueness-import circularity. The paper also flags its own missing support and ad hoc assumptions explicitly (Sec. II.D.2: 'no ab initio results in the literature to fully support either assumption'; Sec. VII: 'remaining ad hoc assumptions'), which lessens the charge of hidden circularity while confirming that the benchmark demonstrates internal consistency rather than external validation. The score of 5 reflects a central claim that partially reduces to its own fitting source while retaining genuine independent content.

Assumptions & free parameters 6 free parameters · 9 assumptions · 0 invented entities

The central claim rests on a large set of fitted parameters (rate coefficients, MMT U and T*, non-Boltzmann factors, relaxation times) and on assumptions that the ab initio PESs and DMS method are faithful, that missing collision pairs can be replaced by surrogates, and that a two-temperature collapse with combined vibronic relaxation is adequate. No new physical entities are introduced.

free parameters (6)
  • Arrhenius pre-exponential A and temperature exponent n for 21 reactions = A ranging 1.15e+11 to 2.9e+14 (units cm3/mol-s-K^n), n ranging -1.66 to 0.77
    Least-squares fits to QCT thermal rate coefficients on ab initio PESs (Table 3); these set the thermal rates.
  • MMT pseudo-temperature parameters U and T* = U from 0.30 to 0.72, T* from about -3.0e6 K to 3.9e6 K
    Fitted to reproduce QCT vibrational energy change per dissociation at thermal equilibrium (Table 4, Eq. 16).
  • Common non-Boltzmann rate factor f_k_NB = 0.5
    Collapsed from reaction-specific values 0.285 to 0.555 to reproduce DMS QSS dissociation rates; applied to all MMT reactions (Table 4, Sec. II.D.3).
  • Common non-Boltzmann energy factor f_eps_NB = 0.85
    Collapsed from reaction-specific values 0.80 to 0.90 to reproduce DMS QSS vibrational energy removal; applied to Eq. (19) for all MMT reactions.
  • Vibrational relaxation fit parameters a_low, b_low, a_high, b_high = a_low ranges 221.0 to -24.76 K^(1/3), b_low ranges -24.83 to -15.37, etc.
    Curve fits to DMS-derived pair relaxation times (Table 5, Eq. 22); these control Tv relaxation in CFD.
  • O2 multi-electronic-surface enhancement factor eta_O2 = 16/3
    Not fitted here; taken from degeneracy counting of O2 excited states. Presented as an optional limiting case with no direct ab initio confirmation, strongly affecting O2 consumption.
assumptions (9)
  • domain assumption Ab initio PESs from the Minnesota group accurately describe high-temperature air collisions.
    All QCT and DMS results rely on these surfaces (Sec. I, Refs. 5-14); no experimental validation of the surfaces is provided.
  • domain assumption QCT with classical trajectories on PESs yields accurate rate coefficients and vibrational energy changes.
    QCT is the basis for all Arrhenius and MMT fits (Sec. I).
  • domain assumption DMS is a sufficiently accurate stand-in for physical reality in heat baths.
    The paper explicitly replaces traditional experimental validation with DMS consistency (Sec. I), assuming the PES-based particle method is faithful.
  • domain assumption Two-temperature model: trans-rotational mode at T, vibronic mode at Tv, instantaneous rotational equilibrium.
    Governing equations in Sec. II.A and the benchmarks rely on this collapse; DMS itself shows rotational nonequilibrium early on.
  • ad hoc to paper Missing PES collision pairs can be represented by surrogate reaction and relaxation parameters.
    Reactions 3, 8, 11 and 12 reuse parameters from other pairs; N2-NO and O2-NO relaxation times are set to surrogate values (Secs. II.D.2 and II.F).
  • ad hoc to paper Common non-Boltzmann factors are transferable across reactions and temperatures.
    Reaction-specific factors vary from 0.285 to 0.555, yet the paper adopts uniform 0.5 and 0.85, citing statistical uncertainty (Sec. II.D.3).
  • domain assumption Landau-Teller form with a two-slope Arrhenius fit describes vibrational relaxation.
    Eqs. (20) to (22) assume this functional form over the whole temperature range.
  • domain assumption Detailed balance with equilibrium constants from PES-derived partition functions determines recombination rates.
    Recombination rates are not computed directly; they follow from k_rec = k_diss / K_eq (Sec. II.A), relying on PES-consistent thermodynamics (Secs. S1 and S2).
  • domain assumption Vibrational and electronic modes relax together at a common timescale.
    The model treats combined vibrational-electronic energy at Tv (Eq. 5 and Sec. S3.C); the text acknowledges this is an approximation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Modified Marrone-Treanor model: parameterization and benchmarking for five-species air." pith.science (2026). https://pith.science/paper/OG6XRMLT

@misc{pith2026250620521,
  author       = {Pith},
  title        = {Pith review of: Modified Marrone-Treanor model: parameterization and benchmarking for five-species air},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OG6XRMLT}},
  note         = {Machine review of arXiv:2506.20521}
}
read the original abstract

We present updated parameters for five-species air (N2, O2, NO, N and O) reactions to be used with the Modified Marrone-Treanor two-temperature model. The vibrational relaxation and chemical reaction rates are derived from quasiclassical trajectory calculations and direct molecular simulations using ab initio potential energy surfaces. The resulting model enables efficient computational fluid dynamics simulations of nonequilibrium air chemistry in hypersonic flows. We show that the model reproduces direct molecular simulation benchmark solutions with high accuracy in zero-dimensional heat baths representative of strong nonequilibrium post-shock conditions. The model's analytical expressions for dissociation rate coefficient and vibrational energy change per reaction ensure that the correct amount of energy is transferred between the vibrational and trans-rotational modes. Detailed balance is imposed for three-body recombination reactions and our simulations exhibit quasi-steady-state dissociation rates and proper approach to thermochemical equilibrium. In direct comparison with the Park TTv model, the Modified Marrone-Treanor model predicts significantly slower conversion of N2 into N below 10000 K and significantly more NO production at all temperatures. This is likely due to its significantly higher Zeldovich reaction rates compared to Park.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

53 extracted references · 49 canonical work pages

  1. [1]

    Modified Marrone-Treanor dissociation model: formulation and benchmarking for diatom/atom mixtures

    Chaudhry, R. S., Torres, E., Schwartzentruber, T. E., and Ca ndler, G., “Modified Marrone-Treanor dissociation model: formulation and benchmarking for diatom/atom mixtu res,” , 2025. doi:10.48550/arXiv.2506.06236, URL https://arxiv.org/abs/2506.06236, submitted for review, preprint available on arxiv.org

    work page Pith review arXiv doi:10.48550/arxiv.2506.06236 2025

  2. [2]

    Park, C., Nonequilibrium hypersonic aerothermodynamics, Wiley, New York, 1990

    work page 1990

  3. [3]

    Review of chemical-kinetic problems of futur e NASA missions. I - Earth entries,

    Park, C., “Review of chemical-kinetic problems of futur e NASA missions. I - Earth entries,” Journal of Thermophysics and Heat Transfer, Vol. 7, No. 3, 1993, pp. 385–398. doi:10.2514/3.431

    work page doi:10.2514/3.431 1993

  4. [4]

    Review of c hemical-kinetic problems of future NASA missions. II - Mars entries,

    Park, C., Howe, J., Jaffe, R., and Candler, G., “Review of c hemical-kinetic problems of future NASA missions. II - Mars entries,” Journal of Thermophysics ans Heat Transfer , Vol. 8, No. 1, 1994, pp. 9–23

    work page 1994

  5. [5]

    Global ab initio ground-state potential energy surface of N4,

    Paukku, Y ., Yang, K., Varga, Z., and Truhlar, D., “Global ab initio ground-state potential energy surface of N4,” The Journal of Chemical Physics, Vol. 139, 2013, p. 044309. doi:10.1063/1.4811653

    work page doi:10.1063/1.4811653 2013

  6. [6]

    An improved potential energy surface and multi-temperature quasiclas sical trajectory calculations of N2 + N2 dissociation reactions,

    Bender, J., Valentini, P., Nompelis, I., Paukku, Y ., Varga, Z., Truhlar, D., Schwartzentruber, T., and Candler, G.,“An improved potential energy surface and multi-temperature quasiclas sical trajectory calculations of N2 + N2 dissociation reactions,” The Journal of Chemical Physics , Vol. 143, 2015, p. 054304. doi:10.1063/1.4927571

    work page doi:10.1063/1.4927571 2015

  7. [7]

    Global triplet potential energy surfaces for theN2 (/u1D44B1Σ) +O(3/u1D443) → NO(/u1D44B2Π) + N(4/u1D446) reaction,

    Lin, W., Varga, Z., Song, G., Paukku, Y ., and Truhlar, D., “Global triplet potential energy surfaces for theN2 (/u1D44B1Σ) +O(3/u1D443) → NO(/u1D44B2Π) + N(4/u1D446) reaction,” The Journal of Chemical Physics , Vol. 144, 2016, p. 024309. doi:10.1063/1.4938241

    work page doi:10.1063/1.4938241 2016

  8. [8]

    Potential energy surface of triplet N2O2,

    Varga, Z., Meana-Pañeda, R., Song, G., Paukku, Y ., and Tr uhlar, D., “Potential energy surface of triplet N2O2,” The Journal of chemical physics, Vol. 144, No. 2, 2016, p. 024310. doi:10.1063/1.4939008

    work page doi:10.1063/1.4939008 2016

Show all 53 references
  1. [9]

    Potential energy surfaces of quintet and sing let O4,

    Paukku, Y ., Yang, K., Varga, Z., Song, G., Bender, J., and Truhlar, D., “Potential energy surfaces of quintet and sing let O4,” The Journal of Chemical Physics , Vol. 147, 2017, p. 034301. doi:10.1063/1.4993624

    2017 doi

  2. [10]

    Potential ener gy surfaces for O + O2 collisions,

    Varga, Z., Paukku, Y ., and Truhlar, D., “Potential ener gy surfaces for O + O2 collisions,” The Journal of Chemical Physics , Vol. 147, 2017, p. 154312. doi:10.1063/1.4997169

    2017 doi

  3. [11]

    Potential ener gy surface of triplet O4,

    Paukku, Y ., Varga, Z., and Truhlar, D., “Potential ener gy surface of triplet O4,” The Journal of Chemical Physics , Vol. 148, 2018, p. 124314. doi:10.1063/1.5017489

    2018 doi

  4. [12]

    Many-body p ermutationally invariant polynomial neural network poten tial energy surface for N4,

    Li, J., Varga, Z., Truhlar, D., and Guo, H., “Many-body p ermutationally invariant polynomial neural network poten tial energy surface for N4,” Journal of Chemical Theory and Computation , Vol. 16, No. 8, 2020, pp. 4822–4832. doi:10.1021/acs.jctc . 0c00430

    2020 doi

  5. [13]

    Potential energy surfaces for high-energy N + O2 collisions,

    Varga, Z., Liu, Y ., Li, J., Paukku, Y ., Guo, H., and Truhl ar, D., “Potential energy surfaces for high-energy N + O2 collisions,” The Journal of Chemical Physics , Vol. 154, No. 8, 2021, p. 084304. doi:10.1063/5.0039771

    2021 doi

  6. [14]

    Potential energy surface f or high-energy N + N2 collisions,

    Varga, Z., and Truhlar, D., “Potential energy surface f or high-energy N + N2 collisions,” Phys. Chem. Chem. Phys. , Vol. 23, 2021, pp. 26273–26284. doi:10.1039/D1CP04373K. 73

    2021 doi

  7. [15]

    POTLIB: An Online Library of Potential Energ y Surfaces,

    Varga, Z., Meana-Pañeda, R., Song, G., Paukku, Y ., and T ruhlar, D. G., “POTLIB: An Online Library of Potential Energ y Surfaces,” , 2021. URL comp.chem.umn.edu/potlib/

    2021

  8. [16]

    Reactive scattering c ross sections III: Quasiclassical and semiclassical metho ds,

    Truhlar, D., and Muckerman, J., “Reactive scattering c ross sections III: Quasiclassical and semiclassical metho ds,” Atom- Molecule Collision Theory , Springer, 1979, pp. 505–566

    1979

  9. [17]

    Quasiclassical Trajectory Analysis of Nitroge n for High- Temperature Chemical Kinetics,

    Chaudhry, R., Bender, J., Schwartzentruber, T., and Ca ndler, G., “Quasiclassical Trajectory Analysis of Nitroge n for High- Temperature Chemical Kinetics,” Journal of Thermophysics and Heat Transfer , Vol. 32, No. 4, 2018, pp. 833–845. doi: 10.2514/1.T5484

    2018 doi

  10. [18]

    Bird, G., Molecular Gas Dynamics and the Direct Simulation of Gas Flow s, Oxford University Press, 1994

    1994

  11. [19]

    Monte Carlo direct simulation of rotationa l relaxation of nitrogen through high total temperature sho ck waves using classical trajectory calculations,

    Koura, K., “Monte Carlo direct simulation of rotationa l relaxation of nitrogen through high total temperature sho ck waves using classical trajectory calculations,” Physics of Fluids, Vol. 10, 1998, p. 2689. doi:10.1063/1.869782

    1998 doi

  12. [20]

    Oxygen tra nsport Properties estimation by CT-DSMC simulations,

    Bruno, D., Frezzotti, A., and Ghiroldi, G., “Oxygen tra nsport Properties estimation by CT-DSMC simulations,” Physics of Fluids, Vol. 27, No. 5, 2015, p. 057101

    2015

  13. [21]

    Direct molecular simulation of n itrogen dissociation based on an ab initio potential energy surface,

    Valentini, P., Schwartzentruber, T., Bender, J., Nomp elis, I., and Candler, G., “Direct molecular simulation of n itrogen dissociation based on an ab initio potential energy surface,” Physics of Fluids, Vol. 27, 2015, p. 086102. doi:10.1063/1.4929394

    2015 doi

  14. [22]

    Dynamics of nitrogen dissociation from direct molecular simulation,

    Valentini, P., Schwartzentruber, T., Bender, J., and C andler, G., “Dynamics of nitrogen dissociation from direct molecular simulation,” Physical Review Fluids, Vol. 1, 2016, p. 043402. doi:10.1103/PhysRevFluids.1.04 3402

    2016 doi

  15. [23]

    First principles simulation of reacting hypersonic fl ow over a blunt wedge,

    Grover, M. S., Verhoff, A. M., Valentini, P., and Bisek, N . J., “First principles simulation of reacting hypersonic fl ow over a blunt wedge,” Physics of Fluids, Vol. 35, No. 8, 2023, p. 086106. doi:10.1063/5.0161570

    2023 doi

  16. [24]

    Vibrational Energy Transfer and Collision-Induced Dissociation in O + O2 Collisions,

    Grover, M., Schwartzentruber, T., Varga, Z., and Truhlar, D., “Vibrational Energy Transfer and Collision-Induced Dissociation in O + O2 Collisions,” Journal of Thermophysics and Heat Transfer, Vol. 33, No. 3, 2019, pp. 797–807. doi:10.2514/1.T5551

    2019 doi

  17. [25]

    Dir ect molecular simulation of internal energy relaxation and dissociation in oxygen,

    Grover, M., Torres, E., and Schwartzentruber, T., “Dir ect molecular simulation of internal energy relaxation and dissociation in oxygen,” Physics of Fluids, Vol. 31, 2019, p. 076107. doi:10.1063/1.5108666

    2019 doi

  18. [26]

    Direct molecul ar simulation of oxygen dissociation across normal shocks,

    Torres, E., and Schwartzentruber, T., “Direct molecul ar simulation of oxygen dissociation across normal shocks, ” Theoretical and Computational Fluid Dynamics , Vol. 36, No. 1, 2022, pp. 41–80. doi:10.1007/s00162-021-0 0596-6

    2022 doi

  19. [27]

    High-Temperature Nonequilibrium Air Chemistry from F irst Principles,

    Torres, E., Geistfeld, E. C., and Schwartzentruber, T. E., “High-Temperature Nonequilibrium Air Chemistry from F irst Principles,” Journal of Thermophysics and Heat Transfer , Vol. 38, No. 2, 2024, pp. 260–291. doi:10.2514/1.T6863

    2024 doi

  20. [28]

    Characteris tic vibrational and rotational relaxation times for air spe cies from first- principles calculations,

    Torres, E., and Schwartzentruber, T. E., “Characteris tic vibrational and rotational relaxation times for air spe cies from first- principles calculations,” Journal of Thermophysics and Heat Transfer, Vol. 39, No. 2, 2025, pp. 223–249. doi:10.2514/1.T7042

    2025 doi

  21. [29]

    Toward a b initio molecular simulation of reacting air: Mach 15 air flo w over a blunt wedge,

    Valentini, P., Grover, M. S., and Bisek, N. J., “Toward a b initio molecular simulation of reacting air: Mach 15 air flo w over a blunt wedge,” Journal of Fluid Mechanics , Vol. 997, 2024, p. A27. doi:10.1017/jfm.2024.772. 74

    2024 doi

  22. [30]

    Direct Molecular Simulation of Nonequilibrium Dilute Gase s,

    Schwartzentruber, T., Grover, M., and Valentini, P., “ Direct Molecular Simulation of Nonequilibrium Dilute Gase s,” Journal of Thermophysics and Heat Transfer , Vol. 32, No. 4, 2018, pp. 892–903. doi:10.2514/1.T5188

    2018 doi

  23. [31]

    Nikitin, E., Theory of elementary atomic and molecular processes in gase s, Clarendon Press, 1974

    1974

  24. [32]

    Coefficients for C alculating Thermodynamic and Transport Properties of Indi vidual Species,

    McBride, B., Gordon, S., and Reno, M., “Coefficients for C alculating Thermodynamic and Transport Properties of Indi vidual Species,” Tech. rep., NASA Technical Memorandum 4513, 1993

    1993

  25. [33]

    Chemical Relaxation wit h Preferential Dissociation from Excited Vibrational Leve ls,

    Marrone, P., and Treanor, C., “Chemical Relaxation wit h Preferential Dissociation from Excited Vibrational Leve ls,” The Physics of Fluids, Vol. 6, 1963, p. 1215. doi:10.1063/1.1706888

    1963 doi

  26. [34]

    doi:10.2514/6.1988-458

    Park, C., Two-temperature interpretation of dissociation rate data for N2 and O2, 1988. doi:10.2514/6.1988-458

    1988 doi

  27. [35]

    Rate parameters for coupled vibration-dissociation in a generalized SSH ap proxi- mation,

    Sharma, S. P., Huo, W. M., and Park, C., “Rate parameters for coupled vibration-dissociation in a generalized SSH ap proxi- mation,” Journal of Thermophysics and Heat Transfer , Vol. 6, No. 1, 1992, pp. 9–21. doi:10.2514/3.312

    1992 doi

  28. [36]

    Modeling and Analysis of Chemical Kinet ics for Hypersonic Flows in Air,

    Chaudhry, R., “Modeling and Analysis of Chemical Kinet ics for Hypersonic Flows in Air,” , 2018. URL https://hdl.handle.net/11299/201709, University of Minnesota - Doctoral Thesis

    2018

  29. [37]

    Consistent kine tic–continuum dissociation model I. Kinetic formulation,

    Singh, N., and Schwartzentruber, T., “Consistent kine tic–continuum dissociation model I. Kinetic formulation, ” The Journal of Chemical Physics, Vol. 152, No. 22, 2020, p. 224302. doi:10.1063/1.5142752

    2020 doi

  30. [38]

    Consistent kine tic-continuum dissociation model. II. Continuum formulat ion and verification,

    Singh, N., and Schwartzentruber, T., “Consistent kine tic-continuum dissociation model. II. Continuum formulat ion and verification,” The Journal of Chemical Physics , Vol. 152, No. 22, 2020, p. 224303. doi:10.1063/1.5142754

    2020 doi

  31. [39]

    Nonequilibrium internal energy distributions during dissociation,

    Singh, N., and Schwartzentruber, T., “Nonequilibrium internal energy distributions during dissociation,” Proceedings of the National Academy of Sciences , Vol. 115, No. 1, 2018, pp. 47–52. doi:10.1073/pnas.171384 0115

    2018 doi

  32. [40]

    Systematics of Vibra tional Relaxation,

    Millikan, R. C., and White, D. R., “Systematics of Vibra tional Relaxation,” The Journal of Chemical Physics , Vol. 39, No. 12, 1963, pp. 3209–3213. doi:10.1063/1.1734182

    1963 doi

  33. [41]

    Investigation of oxygen dissociation and vibrational relaxation at tempe ratures 4000-10800 K,

    Ibraguimova, L. B., Sergievskaya, A. L., Levashov, V . Y ., Shatalov, O. P., Tunik, Y . V ., and Zabelinskii, I. E., “Investigation of oxygen dissociation and vibrational relaxation at tempe ratures 4000-10800 K,” The Journal of Chemical Physics , Vol. 139, No. 3, 2013, p. 034...

    2013 doi

  34. [42]

    Vibrati onal relaxation time measurements in shock-heated oxygen and air from 2000 K to 9000 K using ultraviolet laser absorption,

    Streicher, J. W., Krish, A., and Hanson, R. K., “Vibrati onal relaxation time measurements in shock-heated oxygen and air from 2000 K to 9000 K using ultraviolet laser absorption,”Physics of Fluids, Vol. 32, No. 8, 2020, p. 086101. doi:10.1063/5.0015890

    doi

  35. [43]

    Coupled vibr ation-dissociation time-histories and rate measurements in shock- heated, nondilute O2 and O2 − Ar mixtures from 6000 to 14000 K,

    Streicher, J., Krish, A., and Hanson, R., “Coupled vibr ation-dissociation time-histories and rate measurements in shock- heated, nondilute O2 and O2 − Ar mixtures from 6000 to 14000 K,” Physics of Fluids , Vol. 33, No. 5, 2021, p. 056107. doi:10.1063/5.0048059

    2021 doi

  36. [45]

    High-te mperature vibrational relaxation and decomposition of sho ck-heated nitric oxide: II. Nitrogen dilution from 1900 to 8200 K,

    Streicher, J. W., Krish, A., and Hanson, R. K., “High-te mperature vibrational relaxation and decomposition of sho ck-heated nitric oxide: II. Nitrogen dilution from 1900 to 8200 K,” Physics of Fluids , Vol. 34, No. 11, 2022, p. 116123. doi: 10.1063/5.0122787

    1900 doi

  37. [46]

    Nonequilibri um Dissociation and Recombination Models for Hypersonic Fl ows,

    Singh, N., and Schwartzentruber, T. E., “Nonequilibri um Dissociation and Recombination Models for Hypersonic Fl ows,” AIAA Journal, Vol. 60, No. 5, 2022, pp. 2810–2825. doi:10.2514/1.J06115 4

    2022 doi

  38. [47]

    doi:10.2514/6.1984- 1730, AIAA 84-1730

    Park, C., Problems of rate chemistry in the flight regimes of aeroassis ted orbital transfer vehicles, 1984. doi:10.2514/6.1984- 1730, AIAA 84-1730

    1984 doi

  39. [48]

    Diabatic Potential Energy Surfaces and Semiclassical Multi-State D ynamics for Fourteen Coupled 3A ’ States of O3,

    Varga, Z., Shu, Y ., Ning, J., and Truhlar, D., “Diabatic Potential Energy Surfaces and Semiclassical Multi-State D ynamics for Fourteen Coupled 3A ’ States of O3,” Electronic Structure, Vol. 4, No. 4, 2022, p. 047002. doi:10.1088/2516-1075/ac9 4ac

    2022 doi

  40. [49]

    Diabatic States of Molecules,

    Shu, Y ., Varga, Z., Kanchanakungwankul, S., Zhang, L., and Truhlar, D. G., “Diabatic States of Molecules,” The Journal of Physical Chemistry A , Vol. 126, No. 7, 2022, pp. 992–1018. doi:10.1021/acs.jpca .1c10583

    2022 doi

  41. [50]

    Quasi-classical trajectory analysis of three-body coll ision induced recombination in neutral nitrogen and oxygen,

    Geistfeld, E. C., Torres, E., and Schwartzentruber, T. , “Quasi-classical trajectory analysis of three-body coll ision induced recombination in neutral nitrogen and oxygen,” The Journal of Chemical Physics , Vol. 159, No. 15, 2023, p. 154111. doi: 10.1063/5.0163942

    2023 doi

  42. [51]

    Objective molecular dynamics investigation of d issociation and recombination kinetics in high-temperature nitrogen,

    Pahlani, G., Torres, E., Schwartzentruber, T., and Jam es, R. D., “Objective molecular dynamics investigation of d issociation and recombination kinetics in high-temperature nitrogen, ” Physics of Fluids , Vol. 35, No. 6, 2023, p. 067111. doi:10.1063/5. 0150492

    2023 doi

  43. [52]

    State-to-state study of non-equili brium recombination of oxygen and nitrogen molecules,

    Macdonald, R. L., “State-to-state study of non-equili brium recombination of oxygen and nitrogen molecules,” The Journal of Chemical Physics, Vol. 160, No. 13, 2024, p. 134307. doi:10.1063/5.0195238

    2024 doi

  44. [53]

    Computer Program for Calcu lation of Complex Chemical Equilibrium Compositions and Applications II. Users Manual and Program Description,

    McBride, B., and Gordon S., “Computer Program for Calcu lation of Complex Chemical Equilibrium Compositions and Applications II. Users Manual and Program Description,” NA SA Reference Publication 1311, NASA Lewis Research Center, 1996

    1996

  45. [54]

    Velocity-Slip and Temperature-Jump Effects in Near- Continuum Hypersonic Flows,

    Bhide, P. M., Nompelis, I., Schwartzentruber, T., and C andler, G., “Velocity-Slip and Temperature-Jump Effects in Near- Continuum Hypersonic Flows,” AIAA Journal, Vol. 59, No. 10, 2021, pp. 3815–3830. doi:10.2514/1.J0598 90. 76

    2021 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.