primaryNitrogenLibrary: Lin2013a CH3NO2 isomerization entries (#127–#129) applied outside their pressure range

Abierto
  • ¿Algo no está bien?
#753 0 comentarios 0 reacciones 0 asignados Ver en GitHub

Nadie ha tomado este issue todavía.

Evaluación

Dificultad
5/5
Tiempo estimado
Más de una semana
Aptitud para principiantes
12/100
Tipo de issue
Error
Claridad
Bastante claro
Estado de actividad
Activo
Stack tecnológico
python
Área
databases

Línea de trabajo

The entries are in input/kinetics/libraries/primaryNitrogenLibrary/reactions.py, indices 127 to 129, and the validation comparison is against Rasmussen et al. 2008 Table 3. Reading the paper's pressure dependence and the RMG rate-form documentation is needed first. Done would mean a pressure-dependent representation for the CH3NO2/CH3ONO network that reproduces the source paper's branching, with the isomerization flux counted once.

Escrito por el modelo de indexación a partir del texto del issue.

Descripción

Summary

Three primaryNitrogenLibrary entries from Zhu, Raghunath & Lin, J. Phys. Chem. A 117 (2013) 7308 ([Lin2013a]) are stored in forms valid only at low pressure, but RMG applies them at all pressures. Above roughly 1 atm the CH3NO2 → CH3ONO channel then dominates nitromethane decomposition. The source paper says the opposite: CH3 + NO2 accounts for more than 90% of the products above 200 Torr.

In a CH4/O2/NOx mechanism generated with default ('auto') libraries, this makes the predicted CH3NO2 25–400× too low compared with high-pressure flow-reactor data.

Affected entries (input/kinetics/libraries/primaryNitrogenLibrary/reactions.py)

index label stored as issue
127 CH3NO2 <=> CH3ONO ThirdBody(arrheniusLow = 7.05e54 T^-10.94 exp(-64344 cal/mol / RT)) This is exactly the paper's low-pressure limit for the roaming isomerization (p. 7311: k0 = 1.17e31 T^-10.94 exp(-32379/T) cm3 molecule-1 s-1; 1.17e31 × N_A = 7.05e54; 32379 K = 64344 cal/mol). As a third-body rate with no fall-off or k∞ cap, it grows linearly with [M] at every pressure. The paper gives no k∞ or fall-off parameters for this channel.
128 CH3NO2 <=> CH3O + NO pressure-independent Arrhenius, elementary_high_p = False The longDesc already says "valid for 10e-5 to 760 torr". It is the chemically activated CH3NO2 → CH3ONO* → CH3O + NO fit, but it is applied at all pressures.
129 CH3NO2 <=> CH2O + HNO same as 128 Same: the chemically activated CH3ONO* → CH2O + HNO fit, valid ≤ 760 Torr, applied at all pressures.

Entries 126 (CH3NO2 <=> CH3 + NO2, k∞) and 130 (CH3ONO <=> CH3O + NO, k∞) look fine.

127, 128 and 129 all describe the same roaming-isomerization flux (to CH3ONO, and then to its two product channels). When all three are loaded together, that flux is represented up to three times.

Evidence 1: comparison with the source paper

At 1000 K and 1 atm, entry 127 × [M] gives about 1.1e3 s⁻¹. The paper's Fig. 2 shows about 10–20 s⁻¹ for the isomerization (read by eye) against kd760 ≈ 180 s⁻¹ for fission. Fig. 3 shows the isomerization branching ratio is ≤ 0.05 at 760 Torr above about 600 K.

At 800 K, entry 127 gives 4.9, 49 and 487 s⁻¹ at 1, 10 and 100 bar. That exceeds the C–N fission rate in the same mechanism by 9×, 24× and 108×.

Evidence 2: effect on a generated mechanism compared with experiment

Setup:

  • An RMG model for CH4/O2/NO/NO2 with reactionLibraries = 'auto' (which loads primaryNitrogenLibrary and NOx2018).
  • Reactors spanning 700–1200 K and 1.07–100 bar; move-to-core tolerance 0.001.
  • Simulated at the exact conditions of Rasmussen, Rasmussen & Glarborg, Combust. Flame 154 (2008) 529, Table 3: isothermal PFR, τ = C/T, N2 bath, measured inlet NO/NO2 split.

Peak CH3NO2 in ppm. Measured values were read from the paper's Figs. 3–4 (±5 ppm).

exp P (bar) measured 'auto' model same model without entry 127
DN (φ≈1.1) 100 167 0.4 163
EN (φ≈1.15) 50 109 0.8 82
FN (φ≈1.05) 20 27 0.5 14
GN (φ≈0.04) 100 115 0.9 136
HN (φ≈0.04) 50 107 2.1 98
IN (φ≈0.05) 20 63 2.5 53

Rate-of-production analysis shows entry 127 consumes about half of the CH3NO2 formed. The CH3ONO it makes falls apart at once to CH3O + NO, so CH3NO2 acts as a chain-propagating shortcut instead of a chain-terminating reservoir.

The full 'auto' model still gets the onset temperature roughly right. That is a compensating error: removing entry 127 alone fixes CH3NO2 but delays the onset by 25–50 K at 20–50 bar. This suggests the true isomerization rate is small but not zero, and that a pressure-dependent representation, rather than plain removal, is the right long-term fix.

Suggested fix: Chebyshev k(T,P) for the CH3NO2 / CH3ONO network

Replace entries 127–129 with Chebyshev kinetics covering the full range RMG uses: about 300–3000 K and 1e-5 Torr to ≥ 100 bar.

The data could come from a master-equation calculation (e.g. Arkane) on the paper's PES:

  • energies in Fig. 1 (CH3 + NO2 59.8, RTS1 ≈ 57–59, TS2 69.4, CH3ONO ≈ 2, CH3O + NO 42.9 kcal/mol);
  • ⟨ΔE⟩down = 400 cm⁻¹ for Ar, as in the paper.

A single Chebyshev fit gives each channel the correct fall-off. That means:

  • the isomerization dominates at very low pressure;
  • CH3 + NO2 dominates above about 200 Torr;
  • the isomerization flux is counted once, not three times.

Ideally, entry 126 (CH3NO2 <=> CH3 + NO2, currently k∞ only) would come from the same calculation, so all channels of the network are consistent.

An interim option, if a new calculation is not practical soon: restrict 127–129 to their valid range with PLOG at ≤ 760 Torr, built from the paper's figures. Anything above 760 Torr would be an extrapolation.

Environment

  • RMG-database fc7bb138f. The entries are unchanged on current main (6391cc4f7).
  • RMG-Py 5c33e07e0.
  • The entries came in with the 2018 nitrogen-library refactor (f5062c2cd).

I can share the RMG input files, the analysis scripts and the full comparison on request.

Lenguaje dominante
Python
Estrellas
131
Forks
163
Sin PR fusionados en 30 d

Preparar el entorno

Este proyecto no incluye contenedor de desarrollo, Dockerfile ni guía de contribución, así que la configuración corre por tu cuenta: empieza por su README y consulta nuestra guía para la primera contribución para los pasos generales.

Primeros pasos

  1. Lee el issue completo y luego la guía de contribución del proyecto.
  2. Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
  3. Haz un fork del repositorio y trabaja en una rama.
  4. Abre un pull request que haga referencia al número del issue.

Más de ReactionMechanismGenerator/RMG-database

Todos los issues de ReactionMechanismGenerator/RMG-database

Issues similares

Más issues de Python

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.