Pith. sign in

REVIEW 4 major objections 4 minor 59 references

The Ridge Integration Method and its Application to Molecular Sieving, Demonstrated for Gas Purification via Graphdiyne Membranes

T0 review · 4 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a single membrane plane can serve as the ridge of a transition, yielding molecular-dynamics-level gas permeation rates from about one hundred potential-energy evaluations.

desk verdict The core method is a real step forward for low-cost membrane permeation screening, but the 'within 50%' claim is an average, not a pointwise guarantee. read the letter →

arxiv 2502.06654 v1 pith:TOA2PIPZ submitted 2025-02-10 physics.chem-ph physics.app-phphysics.comp-ph

Christian W. Binder , Johannes K. Krondorfer , Andreas W. Hauser This is my paper · ORCID
classification physics.chem-phphysics.app-phphysics.comp-ph
keywords ridgeintegrationmolecularsievinggraphdiynegasseparationpermeationratetransitionstatetheorypartitionfunctionl1-quadrature
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 is trying to establish a cheap route to accurate gas permeation rates through two-dimensional membranes: instead of running molecular dynamics or trusting a harmonic transition-state approximation, it counts Boltzmann-weighted crossings of a single dividing surface, the membrane plane, which it calls the ridge. For methane, nitrogen, and carbon dioxide passing through a graphdiyne pore, the method is claimed to match molecular-dynamics benchmark rates over a 500 K temperature range within about 50% (for CO2 above 200 K) while consuming roughly one hundred single-point potential-energy evaluations. Because the cost is independent of barrier height and rare-event statistics, this would make density-functional-theory-level permeance and selectivity predictions feasible for membrane materials. The paper also uses the method to argue that low-cost force fields are unreliable for such predictions and to give DFT-based selectivities for natural-gas purification via graphdiyne.

What carries the argument

The central object is the ridge, a hypersurface $\tilde{R}(x) = \hat{n}_R \cdot x - b = 0$ that divides the reaction volume into reactant and product sides; for a planar membrane it is the membrane plane $t_3 = 0$ in roto-translational coordinates. The carrying identity is the crossing-volume rate formula, which for a linear molecule reduces to $P_{\mathrm{trans}}(\delta t) = \frac{\delta t}{Z_{\mathrm{TR}}} \int_\Omega dt_1\,dt_2\,dr_1\,dr_2\, e^{-\beta V(t_1,t_2,t_3=0,r_1,r_2)} \det(S_{\mathrm{TRV}}) (2\pi\beta)^{-1/2}$, with $Z_{\mathrm{TR}}$ the roto-translational partition sum; the pore-transition rate is this ratio. The method's practical engine is the numerical evaluation of these low-dimensional integrals using $\ell^1$-quadrature, whose abscissas are seeded by a low-level potential and whose weights are fixed by a simplex procedure, so that roughly one hundred high-level single-point evaluations suffice.

What would settle it

Launch a large set of Boltzmann-weighted classical trajectories at the ridge (the membrane plane) with positive perpendicular velocity and measure the fraction that reaches the opposite adsorption basin without recrossing. The paper's CO2-at-100 K numbers imply that this fraction is about one third; any pore or temperature for which the fraction falls clearly below one would falsify the perfect-dividing-surface postulate on which the ridge-integration rate rests.

Watch

Extended reading notes

Core claim

The central claim is that pore permeation can be treated as a classical transition across a ridge, and that in molecular sieving the ridge is essentially the geometrical plane of the membrane. In this picture the transition probability per unit time is a ratio of two finite-dimensional integrals: the ridge integral, which samples the Boltzmann factor over molecular translations and rotations with the center of mass in the membrane plane, and the roto-translational partition sum over the accessible volume. The paper evaluates these integrals numerically with Monte Carlo importance sampling and with $\ell^1$-quadrature, drawing abscissas from a low-level potential and correcting with a small high-level basis set, so that the high-level potential is needed at only about 100 points. On a shared GFN-FF potential energy surface, the predicted transition-event counts for CH4 and N2 track the MD benchmark within about a factor of 1.5 across 100-600 K, and for CO2 above 200 K, while unmodified Eyring theory deviates by roughly an order of magnitude; the paper therefore claims to combine MD-level accuracy with TS-theory-level cost, enabling DFT-based rate predictions.

Load-bearing premise

Every crossing of the membrane plane with positive perpendicular velocity is counted as a completed transition; when a molecule can linger inside the pore instead of completing the crossing, as for CO2 at 100 K, the method counts roughly three times as many transitions as the molecular-dynamics benchmark.

Editorial extensions

If this is right

  • If the central claim is correct, permeance and selectivity screening of two-dimensional membranes can be run at density-functional-theory accuracy, because each pore needs only about one hundred high-level single-point evaluations instead of millions.
  • Eyring theory and its ad hoc entropy corrections become dispensable for sieving problems: for the graphdiyne test cases they deviate from the molecular-dynamics benchmark by about an order of magnitude or more.
  • Low-cost force fields are inadequate as the sole energy predictor for pore propagation; their predicted rates and selectivities can be off by one to two orders of magnitude relative to density-functional results.
  • The ridge formulation extends naturally to barrierless permeation, where no rate-determining transition state exists, provided the ridge remains a valid dividing surface.
  • For natural-gas purification, the DFT-level predictions indicate that graphdiyne is highly CO2-permeable and CH4-retentive, with selectivities and flow rates that make it a candidate membrane material for CO2 removal.

Reading between the lines

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

  • Going beyond the paper, the same ~100-evaluation budget would make DFT-level screening of many pore geometries and functionalizations practical, because each membrane's ridge integral and partition sum are independent single-point calculations.
  • Because the low-temperature CO2 failure is traced to the no-recrossing postulate, a short ridge-initiated trajectory correction could extend the method into that regime and would be testable against the same benchmark.
  • The division into a low-level sampling measure and a high-level correction suggests machine-learned potentials could act as the low-level measure, with l1-quadrature weights providing a built-in convergence check.
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

4 major / 4 minor

Summary. The paper proposes a 'ridge integration' method for computing molecular permeation rates through two-dimensional membranes from a single dividing hypersurface (the membrane plane), without running molecular dynamics. The method is derived from a classical phase-space crossing-volume expression, expressed in translational-rotational-vibrational coordinates, and evaluated numerically via Monte Carlo importance sampling and l1-quadrature. It is applied to CH4, N2, and CO2 permeation through a graphdiyne pore, with rates compared against GFN-FF molecular dynamics benchmarks at 100-600 K; the authors also use the method with DFT energy predictors to estimate realistic gas selectivities.

Significance. If the accuracy claims held as stated, this would be a substantial methodological advance: the paper reports MD-level permeation rates with about 100 single-point PES evaluations and no fitted parameters, which would make DFT-level rate calculations for membrane sieving practical. The validation design is a genuine strength: the ridge results and the MD benchmark use the same potential energy surface, removing energy-model mismatch as a confounding factor, and the authors provide preliminary code in a public repository. However, the headline 'within 50%' accuracy claim is not supported pointwise by the paper's own tables, and the central prefactor in the working equation is not derived consistently with the earlier momentum integrals. These issues are fixable, but they currently prevent the paper from establishing its central quantitative claim.

major comments (4)
  1. [Section V and Tables I-II] The claim that the ridge integration method 'reproduces the benchmark result obtained from molecular dynamics simulations over a temperature range of 500 K within 50%' is not supported by the reported data. For CO2 at 100 K, ridge integration gives 283.72 events/ns while the MD benchmarks are 145.06 (Table I, frozen), 107.04 (Table II, frozen), and 91.85 (Table II, unconstrained), i.e. discrepancies of factors 2.0 to 3.1; for CH4 at 100 K the ridge result 0.47 is a factor of 2.1 below the frozen MD value 1.00. The paper itself attributes the CO2 deviation to 'multiple counting of incomplete transitions' in Section III B 1. The statement in Section V should be restricted to the regime where the comparison is actually pointwise accurate, or should be replaced by the averaged accuracy of Table III with explicit per-case deviations.
  2. [Equations (9), (14), (18) and Appendix A] The momentum integrals are written inconsistently. Equations (9) and (14) use Boltzmann factors of the form exp(-beta p_i^2), while Equation (18) and Appendix A, Eq. (A3), use exp(-beta p^2/(2m)) and yield the prefactor 1/sqrt(2*pi*beta). In mass-weighted coordinates, the classical Boltzmann factor is exp(-beta p^2/2) for kinetic energy p^2/2; the printed form in Eqs. (9) and (14) changes the value of the momentum integral and therefore the prefactor of the working rate expression. Because the numerical rates in Section III inherit this prefactor through Eq. (18), the derivation must be reconciled and the convention stated explicitly.
  3. [Tables I and II] The MD benchmark values are not internally consistent between tables. For CO2 at 100 K the frozen-pore benchmark is 145.06 events/ns in Table I but 107.04 in Table II; for N2 at 500 K the frozen-pore benchmark is 32.45 in Table I but 35.02 in Table II. Since the MD numbers define the reference for all accuracy statements, these discrepancies must be resolved and their source (simulation length, counting protocol, or a typographical error) reported. The discrepancy directly affects the calculation of the 'relative accuracy' values in Table III.
  4. [Section III C, Table III] The relative accuracy of 1.5 for the ridge method is defined as the averaged relative deviation from the unconstrained MD benchmark over all temperatures and molecules. Because the 100 K CO2 and CH4 points deviate by factors of 2 to 3, an averaged factor of 1.5 does not justify the unqualified 'within 50%' phrasing used elsewhere. Please report the full distribution of per-case deviations, e.g. median and maximum, and state the averaging domain explicitly whenever a single accuracy number is quoted.
minor comments (4)
  1. [Section III C] The cross-references to the numerical integration methods are swapped: the text refers to 'Monte Carlo importance sampling (see Section II E 2)' and 'l1-quadrature (Section II E 1)', but in Section II E subsection 1 is Monte Carlo integration and subsection 2 is l1-quadrature.
  2. [Section II D] There is a typo, 'Futhermore', and the notation det(STRV) in Eq. (18) is printed with inconsistent subscript spacing compared with the matrix STRV defined in Eq. (15).
  3. [Equation (9)] The index ranges in Eq. (9) appear inconsistent: the position integrals run over n-1 variables labelled i=0,...,n-1 while the momentum factor contains p_n and a product over i=0,...,n. Please correct the index conventions so the dimensions of the position and momentum spaces match those used in Appendix A.
  4. [Figure 6] The caption lists the panels as '(a) ridge integral of N2 (Equation 18)' and '(b) partition sum N2 (Equation 19)', but the figure panels are labeled (a) and (b) with the same descriptions; please make the caption self-explanatory about which panel is the ridge integral and which is the partition sum, and ensure the in-text reference to 'Figure 6b' is consistent.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ridge rate expression is derived from a classical phase-space integral and validated against an external MD benchmark, with no rate fitted to the target data.

full rationale

The central rate expression (Eqs. 18 and 20) is obtained by integrating the Boltzmann-weighted flux through the membrane plane t3=0 and dividing by the full partition sum; this is a direct transcription of the crossing-volume integral in Eq. 9, and no parameter is fitted to the MD benchmark. The MD simulations serve only as an external reference, and the paper reports the comparison as validation. The 'ridge' concept is attributed to Ionova and Carter, and the paper explicitly states that its own definition is approximative and less strict, so the dividing-surface choice is a stated modeling assumption rather than a result imported by self-citation. The low-level/high-level quadrature in Section II E is an importance-sampling variance-reduction scheme, not a fit of the final rates to the benchmark. Self-citations appear only as background context (e.g., prior nanoporous-graphene studies) and are not load-bearing for the derivation. The paper itself flags the CO2 100 K failure (Section III B 1) as an overcounting of incomplete transitions caused by the confining adsorption minimum; this is a physical limitation of the no-recrossing postulate, not a circular step. The 'within 50%' claim in Section V is an averaged statement (Table III gives relative accuracy 1.5) and is not pointwise valid for CO2 at 100 K, but that is a correctness/presentation concern outside the scope of circularity analysis.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The method rests on standard classical statistical mechanics plus five domain assumptions listed above, and on two protocol parameters (sphere radius, update interval). It introduces no new physical entities and fits no free parameters to the benchmark data.

free parameters (2)
  • Confining sphere radius = 3 Å (N2, CH4), 4 Å (CO2)
    Chosen to fully accommodate the adsorption minima and to limit counted transition events to the pore center. It defines the integration domain for both the ridge method and the MD benchmark, and therefore affects the numerical values of the rates.
  • MD position update interval = 200 fs
    The benchmark counts a transition each time the stored z-coordinate of the center of mass changes sign across the pore plane. The interval affects the count for fast or recrossing trajectories, especially CO2 at 100 K, where the paper reports interval-dependent counts.
assumptions (5)
  • domain assumption No-recrossing dividing surface: every crossing of the membrane plane with positive normal velocity completes a transition
    Introduced in Section II A (the 'ridge' definition) and Section II D (t3=0). This is the Eyring-type postulate the method inherits; it fails for barrierless CO2 at low T, where incomplete transitions are counted.
  • domain assumption Potential separates into vibrational and roto-translational parts, and vibrational degrees do not change during propagation
    Appendix B, Eq. B1. Justified by intramolecular forces being much stronger than van der Waals forces; enables cancellation of vibrational partition sums and neglect of nu-dependence in Jacobians.
  • domain assumption Pore is static (frozen)
    Section II C: the pore appears as an external potential and its motion is neglected. The paper later tests flexible-pore MD and finds similar counts, supporting the assumption.
  • domain assumption Low-level potential approximates the high-level potential well enough for importance sampling and l1-quadrature
    Section II E: the quadrature measure is built from the low-level potential; accuracy relies on e^-beta(V_hl - V_ll) being smooth. The paper shows force fields deviate by two orders of magnitude from DFT, which limits this assumption.
  • domain assumption Gas molecules are non-interacting and pore clogging is negligible for selectivity predictions
    Section III E: justified by computing average occupation numbers of adsorption minima approximately 0.01 at 1 atm and 300 K.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Ridge Integration Method and its Application to Molecular Sieving, Demonstrated for Gas Purification via Graphdiyne Membranes." pith.science (2026). https://pith.science/paper/TOA2PIPZ

@misc{pith2026250206654,
  author       = {Pith},
  title        = {Pith review of: The Ridge Integration Method and its Application to Molecular Sieving, Demonstrated for Gas Purification via Graphdiyne Membranes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TOA2PIPZ}},
  note         = {Machine review of arXiv:2502.06654}
}
read the original abstract

Eyring theory provides a convenient approximation to the rate of a chemical reaction, as it uses only local information evaluated near extremal points of a given potential energy surface. However, in cases of pronounced anharmonicity and particularly low-lying vibrational frequencies, deviations from the correct reaction rate can become substantial. Molecular Dynamics simulations, on the other hand, are very costly at higher levels of theory, and of limited use since molecular reactions are `rare' events and hence statistically less accessible. In this article, we present an alternative description for problems of gas separation and storage via two-dimensional materials such as porous graphene or flat metal-organic frameworks. Taking geometric advantage of the typical problem setting, our method is based on a statistical analysis of molecular trajectories near the so-called `ridge', a hypersurface which divides the reaction volume into a reactant and a product side. It allows for more realistic predictions of permeabilities and selectivities, e.g. derived from density functional theory, but without the considerable costs of a full molecular dynamics simulation on the corresponding Born-Oppenheimer potential energy surface. We test our method on the example of methane separation from nitrogen and carbon dioxide via a graphdiyne membrane.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

59 extracted references · 48 canonical work pages

  1. [1]

    (24) In this case, the weights {wi} in Equation 22 are chosen to be 1 N , where {xi} values are drawn from a uniform distribution over the entire integration domain

    Monte Carlo Integration For integrands exhibiting low variance we identify Monte Carlo integration as a suitable method since the error of an integral calculated via the latter is directly linked to the integrands variance [36], which can be estimated via σ2(f ) = 1 N − 1 NX i=1 (f (xi) − ⟨f ⟩)2 . (24) In this case, the weights {wi} in Equation 22 are cho...

    work page

  2. [2]

    polynomials, are integrated ex- actly

    Integration via l1-quadrature Another popular integration routine is Gauss quadrature, where abscissas and weights are chosen such that a certain set of basis functions, e.g. polynomials, are integrated ex- actly. The convergence of the numerical integration is thereby done by increasing the set of basis functions. A suitable choice for polynomial basis f...

    work page

  3. [3]

    Chakraborty, I.-H

    G. Chakraborty, I.-H. Park, R. Medishetty, and J. J. Vittal, Two-dimensional metal-organic framework materials: Synthesis, structures, properties and applications, Chemical Reviews 121, 3751 (2021), pMID: 33630582, https://doi.org/10.1021/acs.chemrev.0c01049

    work page doi:10.1021/acs.chemrev.0c01049 2021

  4. [4]

    A direct comparison of the MD results and the prediction methods is given in Figure 4a

    Frozen pore We start with the propagation of CH 4. A direct comparison of the MD results and the prediction methods is given in Figure 4a. Uncertainties for the simulations, depicted by er- rorbars for the MD data points, have been determined statistically. The remaining relative uncertainty of the observed transition events lies at around 10 %. Besides t...

    work page

  5. [5]

    Unconstrained pore Since an unconstrained pore represents a more realistic case, we want to compare this com- putationally more challenging scenario to the results obtained for the frozen pore. The model system is identical to the structure shown in Figure 3, but now only the bond-saturating H 20 atoms at the benzene rings are kept at fixed positions duri...

    work page

  6. [6]

    H. Wang, M. Wang, X. Liang, J. Yuan, H. Yang, S. Wang, Y. Ren, H. Wu, F. Pan, and Z. Jiang, Organic molecular sieve membranes for chemical separations, Chem. Soc. Rev. 50, 5468 (2021)

    work page 2021

  7. [7]

    X. Gao, H. Liu, D. Wang, and J. Zhang, Graphdiyne: synthesis, properties, and applications, Chem. Soc. Rev. 48, 908 (2019)

    work page 2019

  8. [8]

    Hauser, N

    A. Hauser, N. Mardirossian, J. Panetier, M. Head-Gordon, A. Bell, and P. Schwerdtfeger, Functionalized graphene as a gatekeeper for chiral molecules: An alternative concept for chiral separation, Ang. Chem. Int. Ed. 53, 9957 (2014)

    work page 2014

Show all 59 references
  1. [9]

    Hauser and P

    A. Hauser and P. Schwerdtfeger, Nanoporous graphene sheets for gas separation, inMechanical and Chemical Properties, Handbook of Graphene Science, Vol. 4, edited by M. Aliofkhazraei, N. Ali, W. Milne, C. Ozkan, S. Mitura, and J. Gervasoni (Taylor and Francis, 2016) Chap. 25, p...

    2016

  2. [10]

    L. Ding, Y. Wei, L. Li, T. Zhang, H. Wang, J. Xue, L.-X. Ding, S. Wang, J. Caro, and Y. Gogotsi, Mxene molecular sieving membranes for highly efficient gas separation, Nature Communications 9, 155 (2018)

    2018

  3. [11]

    Hauser and P

    A. Hauser and P. Schwerdtfeger, Methane-selective nanoporous graphene membranes for gas purification, Phys. Chem. Chem. Phys. 14, 13292 (2012)

    2012

  4. [12]

    A. W. Hauser and P. Schwerdtfeger, Methane-selective nanoporous graphene membranes for gas purification, Physical Chemistry Chemical Physics 14, 13292 (2012)

    2012

  5. [13]

    Goodrow, A

    A. Goodrow, A. T. Bell, and M. Head-Gordon, Transition state-finding strate- gies for use with the growing string method, J. Chem. Phys. 130, 244108 (2009), https://doi.org/10.1063/1.3156312

    2009 doi

  6. [14]

    P. G. Bolhuis and C. Dellago, Practical and conceptual path sampling issues, The European Physical Journal Special Topics 224, 2409 (2015). 35

    2015

  7. [15]

    Henkelman and H

    G. Henkelman and H. J´ onsson, Improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points, J. Chem. Phys. 113, 9978 (2000), https://doi.org/10.1063/1.1323224

    2000 doi

  8. [16]

    Henkelman, B

    G. Henkelman, B. P. Uberuaga, and H. J´ onsson, A climbing image nudged elastic band method for finding saddle points and minimum energy paths, J. Chem. Phys. 113, 9901 (2000)

    2000

  9. [17]

    J´ onsson, G

    H. J´ onsson, G. Mills, and K. W. Jacobsen, Nudged elastic band method for finding mini- mum energy paths of transitions, in Classical and Quantum Dynamics in Condensed Phase Simulations (1998) pp. 385–404

    1998

  10. [18]

    In order to determine this number, we store the z-position of the center of mass of the respective molecule every 200 fs

    and via Eyring theory (Equation 5), followed by a comparison to the MD simulation benchmark. In order to determine this number, we store the z-position of the center of mass of the respective molecule every 200 fs. Each time the difference between that stored position and the ...

  11. [19]

    A. Behn, P. M. Zimmerman, A. T. Bell, and M. Head-Gordon, Efficient exploration of reaction paths via a freezing string method, J. Chem. Phys. 135, 224108 (2011), https://doi.org/10.1063/1.3664901

    2011 doi

  12. [20]

    P. M. Zimmerman, Growing string method with interpolation and optimization in internal coordinates: Method and examples, J. Chem. Phys. 138, 184102 (2013), https://doi.org/10.1063/1.4804162

    2013 doi

  13. [21]

    A. A. Peterson, Acceleration of saddle-point searches with machine learning, J. Chem. Phys. 145, 074106 (2016), https://doi.org/10.1063/1.4960708

    2016 doi

  14. [22]

    Koistinen, F

    O.-P. Koistinen, F. B. Dagbjartsd´ ottir, V. ´Asgeirsson, A. Vehtari, and H. J´ onsson, Nudged elastic band calculations accelerated with gaussian process regression, J. Chem. Phys. 147, 152720 (2017), https://doi.org/10.1063/1.4986787

    2017 doi

  15. [23]

    Meyer, K

    R. Meyer, K. Schmuck, and A. Hauser, Machine Learning in Computational Chemistry: An Evaluation of Method Performance for Nudged Elastic Band Calculations, Journal of Chemical Theory and Computation 15, 6513 (2019)

    2019

  16. [24]

    Eyring, The activated complex in chemical reactions, The Journal of Chemical Physics 3, 107 (1935)

    H. Eyring, The activated complex in chemical reactions, The Journal of Chemical Physics 3, 107 (1935)

    1935

  17. [25]

    I. V. Ionova and E. A. Carter, Ridge method for finding saddle points on potential energy surfaces, The Journal of Chemical Physics 98, 6377 (1993), https://doi.org/10.1063/1.465100

    1993 doi

  18. [26]

    R. A. Miron and K. A. Fichthorn, The step and slide method for finding saddle points on multidimensional potential surfaces, The Journal of Chemical Physics 115, 8742 (2001), 36 https://doi.org/10.1063/1.1412285

    2001 doi

  19. [27]

    Chawla, H

    M. Chawla, H. Saulat, M. Masood Khan, M. Mahmood Khan, S. Rafiq, L. Cheng, T. Iqbal, M. I. Rasheed, M. Z. Farooq, M. Saeed, N. M. Ahmad, M. B. Khan Ni- azi, S. Saqib, F. Jamil, A. Mukhtar, and N. Muhammad, Membranes for co2 /ch4 and co2/n2 gas separation, Chemical Engineering ...

    2020 doi

  20. [28]

    Chang, Y

    M. Chang, Y. Zhao, Q. Yang, and D. Liu, Microporous metal–organic frameworks with hy- drophilic and hydrophobic pores for efficient separation of ch4/n2 mixture, ACS Omega 4, 14511 (2019), https://doi.org/10.1021/acsomega.9b01740

    2019 doi

  21. [29]

    S. U. Nandanwar, D. R. Corbin, and M. B. Shiflett, A review of porous adsorbents for the separation of nitrogen from natural gas, Industrial & Engineering Chemistry Research 59, 13355 (2020), https://doi.org/10.1021/acs.iecr.0c02730

    2020 doi

  22. [30]

    X. Hao, Z. Li, H. Hu, X. Liu, and Y. Huang, Separation of ch4/n2 of low concentrations from coal bed gas by sodium-modified clinoptilolite, Frontiers in Chemistry 6, 633 (2018)

    2018

  23. [31]

    Carreon, Molecular sieve membranes for n2/ch4 separation, Journal of Materials Research 33, 1 (2017)

    M. Carreon, Molecular sieve membranes for n2/ch4 separation, Journal of Materials Research 33, 1 (2017)

    2017

  24. [32]

    M. G. Evans and M. Polanyi, Some applications of the transition state method to the calcu- lation of reaction velocities, especially in solution, Trans. Faraday Soc. 31, 875 (1935)

    1935

  25. [33]

    Isaacson, Anharmonic effects on the transition state theory rate constant, The Journal of Chemical Physics 117, 8778 (2002)

    A. Isaacson, Anharmonic effects on the transition state theory rate constant, The Journal of Chemical Physics 117, 8778 (2002)

    2002

  26. [34]

    Jensen, Introduction to computational chemistry(John wiley & sons, 2017)

    F. Jensen, Introduction to computational chemistry(John wiley & sons, 2017)

    2017

  27. [35]

    Pt´ aˇ cek, F.ˇSoukal, and T

    P. Pt´ aˇ cek, F.ˇSoukal, and T. Opravil, Introduction to the transition state theory, Introducing the Effective Mass of Activated Complex and the Discussion on the Wave Function of this Instanton , 27 (2018)

    2018

  28. [36]

    Grimme, Supramolecular binding thermodynamics by dispersion-corrected density func- tional theory, Chemistry - A European Journal 18, 9955 (2012)

    S. Grimme, Supramolecular binding thermodynamics by dispersion-corrected density func- tional theory, Chemistry - A European Journal 18, 9955 (2012)

    2012

  29. [37]

    Pracht and S

    P. Pracht and S. Grimme, Calculation of absolute molecular entropies and heat capacities made simple, Chem. Sci. 12, 6551 (2021)

    2021

  30. [38]

    R. F. Ribeiro, A. V. Marenich, C. J. Cramer, and D. G. Truhlar, Use of solution-phase vibra- tional frequencies in continuum models for the free energy of solvation, The Journal of Physical Chemistry B 115, 14556 (2011), pMID: 21875126, https://doi.org/10.1021/jp205508z. 37

    2011 doi

  31. [39]

    L. H. Sprowl, C. T. Campbell, and L. ´Arnad´ ottir, Hindered translator and hindered rotor models for adsorbates: Partition functions and entropies, The Journal of Physical Chemistry C 120, 9719 (2016), https://doi.org/10.1021/acs.jpcc.5b11616

    2016 doi

  32. [40]

    J. W. Ochterski, Vibrational analysis in gaussian (1999), https://gaussian.com/wp- content/uploads/dl/vib.pdf

    1999

  33. [41]

    Weinzierl, Introduction to monte carlo methods, arXiv: High Energy Physics - Phenomenol- ogy 10.48550/arXiv.hep-ph/0006269 (2000)

    S. Weinzierl, Introduction to monte carlo methods, arXiv: High Energy Physics - Phenomenol- ogy 10.48550/arXiv.hep-ph/0006269 (2000)

    2000 doi arXiv

  34. [42]

    J. D. Jakeman and A. Narayan, Generation and application of multivariate polynomial quadra- ture rules, Computer Methods in Applied Mechanics and Engineering 338, 134 (2018)

    2018

  35. [43]

    Glaubitz, Stable high-order cubature formulas for experimental data, Journal of Computa- tional Physics 447, 110693 (2021)

    J. Glaubitz, Stable high-order cubature formulas for experimental data, Journal of Computa- tional Physics 447, 110693 (2021)

    2021

  36. [44]

    Spicher and S

    S. Spicher and S. Grimme, Robust atomistic modeling of materials, organometallic, and biochemical systems, Angewandte Chemie International Edition 59, 15665 (2020), https://onlinelibrary.wiley.com/doi/pdf/10.1002/anie.202004239

    2020 doi

  37. [45]

    A. H. Larsen, J. J. Mortensen, J. Blomqvist, I. E. Castelli, R. Christensen, M. Du lak, J. Friis, M. N. Groves, B. Hammer, C. Hargus, E. D. Hermes, P. C. Jennings, P. B. Jensen, J. Kermode, J. R. Kitchin, E. L. Kolsbjerg, J. Kubal, K. Kaasbjerg, S. Lysgaard, J. B. Maronsson, T...

    2017

  38. [46]

    Grimme, C

    S. Grimme, C. Bannwarth, and P. Shushkov, A robust and accurate tight-binding quantum chemical method for structures, vibrational frequencies, and noncovalent interactions of large molecular systems parametrized for all spd-block elements (z= 1–86), Journal of chemical theory ...

    2017

  39. [47]

    A Head-Gordon damping function with α = 4 and ω0 = 100 cm−1 has been employed for the continuous interpolation between rotational and vibrational contributions

  40. [48]

    Bannwarth, S

    C. Bannwarth, S. Ehlert, and S. Grimme, Gfn2-xtb—an accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions, Journal of Chemical Theory and Computation 15, 1652 (2019),...

    2019 doi

  41. [49]

    Bannwarth, E

    C. Bannwarth, E. Caldeweyher, S. Ehlert, A. Hansen, P. Pracht, J. Seib- ert, S. Spicher, and S. Grimme, Extended tight-binding quantum chem- istry methods, WIREs Computational Molecular Science 11, e1493 (2021), https://wires.onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1493

    2021 doi

  42. [50]

    S. Grimme, Semiempirical gga-type density functional constructed with a long- range dispersion correction, Journal of Computational Chemistry 27, 1787 (2006), https://onlinelibrary.wiley.com/doi/pdf/10.1002/jcc.20495

    2006 doi

  43. [51]

    T. H. Dunning, Gaussian basis sets for use in correlated molecular calculations. i. the atoms boron through neon and hydrogen, The Journal of Chemical Physics 90, 1007 (1989), https://doi.org/10.1063/1.456153

    1989 doi

  44. [52]

    Lin, G.-D

    Y.-S. Lin, G.-D. Li, S.-P. Mao, and J.-D. Chai, Long-range corrected hybrid density functionals with improved dispersion corrections, Journal of Chemical Theory and Computation 9, 263 (2013)

    2013

  45. [53]

    Shao et al., Advances in molecular quantum chemistry contained in the Q-Chem 4 program package, Mol

    Y. Shao et al., Advances in molecular quantum chemistry contained in the Q-Chem 4 program package, Mol. Phys. 113, 184 (2015)

    2015

  46. [54]

    Avogadro Chemistry, Avogadro

  47. [55]

    Faramawy, T

    S. Faramawy, T. Zaki, and A.-E. Sakr, Natural gas origin, composition, and processing: A review, Journal of Natural Gas Science and Engineering 34, 34 (2016)

    2016

  48. [56]

    S. M. Fruehwirth, R. Meyer, and A. W. Hauser, Chiral Separation via Molecular Siev- ing: A Computational Screening of Suitable Functionalizations for Nanoporous Graphene, ChemPhysChem 19, 2331 (2018)

    2018

  49. [57]

    Schrier, Carbon dioxide separation with a two-dimensional polymer membrane, ACS applied materials & interfaces 4, 3745 (2012)

    J. Schrier, Carbon dioxide separation with a two-dimensional polymer membrane, ACS applied materials & interfaces 4, 3745 (2012)

    2012

  50. [58]

    L. Zhao, P. Sang, S. Guo, X. Liu, J. Li, H. Zhu, and W. Guo, Promising monolayer membranes for co2/n2/ch4 separation: Graphdiynes modified respectively with hydrogen, fluorine, and oxygen atoms, Applied Surface Science 405, 455 (2017)

    2017

  51. [59]

    F. J. Keil, R. Krishna, and M.-O. Coppens, Modeling of diffusion in zeolites, Reviews in Chemical Engineering 16, 71 (2000). 39

    2000

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.