Pith. sign in

REVIEW 3 major objections 4 minor 2 references

Recommended Second Virial Coefficients for Nitrogen and Oxygen

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Retuned ab initio pair potentials, constrained by reanalyzed density and refractometry data, produce recommended second virial coefficients for nitrogen and oxygen with experimental-quality uncertainty across 20–3000 K.

desk verdict A careful, useful retuning of B(T) for N2 and O2, with a real reproducibility catch in the nitrogen higher-virial inputs. read the letter →

arxiv 2607.24634 v1 pith:SIXNS7DI submitted 2026-07-27 physics.chem-ph

Robert Hellmann , Giovanni Garberoglio , Allan H. Harvey This is my paper · ORCID
classification physics.chem-ph
keywords secondvirialcoefficientnitrogenoxygenabinitiopairpotentialpath-integralMonteCarloreanalysisrefractometrythermophysicalproperties
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 establishes recommended values for the second virial coefficient $B(T)$—the leading measure of how much a gas's density departs from ideal-gas behavior—for nitrogen over 20–3000 K and oxygen over 20–2000 K, each accompanied by an expanded-uncertainty curve. The route combines previously published ab initio pair potentials with selected high-accuracy experiments: the nitrogen potential is tuned to reanalyzed density data whose uncertainty is sharply reduced by constraining the fourth through sixth virial coefficients to calculated values, while the oxygen potentials are tuned to new refractometry data. Quantum effects are included exactly with path-integral Monte Carlo, so the curves are valid over a temperature range far wider than any single experiment covers. If the recommendations hold, they give metrology, atmospheric science, and reference equations of state a firmer basis for the nonideal behavior of air's two main components.

What carries the argument

The load-bearing object is the two-body potential-energy surface (pair PES): an ab initio rigid-rotor interaction potential for N$_2$ and three spin-coupled surfaces (quintet, triplet, singlet) for O$_2$, each modified by one physical parameter so that the calculated $B(T)$ matches selected experiments. The $B(T)$ values are then computed with path-integral Monte Carlo (PIMC), which represents each quantum rotor pair as ring polymers and evaluates the Mayer-$f$ average exactly, including nuclear quantum effects; a quadratic Feynman–Hibbs (QFH) approximation supplies the higher virial coefficients $C$–$F$ used to constrain the nitrogen density isotherms and confirms the PIMC results at most temperatures. The recommended values are carried by the correlation of Eq. (15) with the parameters in Table 2, and the low-temperature uncertainties by Eq. (16).

What would settle it

Measure $B(T)$ for nitrogen between 130 and 340 K with a technique that does not rely on calculated higher virial coefficients, such as low-density speed-of-sound or dielectric-constant gas thermometry, and compare with the recommended curve; a deviation outside the claimed expanded uncertainty of 0.05 cm$^3$ mol$^{-1}$ over 150–500 K would show the higher-virial constraint is biased.

Watch

Extended reading notes

Core claim

The central claim is that slightly retuning previously published ab initio pair potentials—scaling the difference between the CCSDT(Q) and CCSD(T) levels by 0.472 for nitrogen, and adjusting the isotropic $C_8$ dispersion coefficient of each of the three spin-coupled oxygen surfaces—makes the $B(T)$ calculated from the potentials agree with selected high-accuracy measurements while retaining a temperature range far wider than any experiment. For nitrogen, the most influential measurements are values of $B$ obtained by reanalyzing density isotherms with the fourth through sixth virial coefficients fixed to semiclassically calculated values; for oxygen, the tuning target is a single refractometry data set. The paper reports the recommended $B(T)$ as smooth correlations (Eq. (15) with Table 2) for 20–3000 K (N$_2$) and 20–2000 K (O$_2$), with expanded ($k=2$) uncertainties of 0.05 cm$^3$ mol$^{-1}$ for nitrogen over 150–500 K and 0.1 cm$^3$ mol$^{-1}$ for oxygen over 300–400 K, rising linearly outside those ranges. Quantum effects are included exactly with path-integral Monte Carlo, and the semiclassical quadratic Feynman–Hibbs results agree closely except below about 50–60 K.

Load-bearing premise

The nitrogen recommendations stand on the accuracy of the calculated fourth-through-sixth virial coefficients—theory values from a preliminary pair potential and a nonpublic three-body surface—used to constrain the reanalysis of density measurements; if those values are biased, the extracted $B$ values and the tuned potential inherit the bias.

Editorial extensions

If this is right

  • The recommended $B(T)$ curves provide low-density boundary conditions for updating the nitrogen reference equation of state (2000) and the oxygen reference equation of state (1985), which the paper identifies as outdated.
  • Applications in flow metering, refractometry-based pressure standards, and humidity or length metrology can use air's main components with uncertainties close to the best single measurements over a temperature span no experiment covers.
  • Viscosity and thermal conductivity values from the original potentials remain valid, because the tuning shifts them by far less than their uncertainties.
  • The tuned potentials form the foundation for a revised second virial coefficient of dry air, once the N$_2$–O$_2$ cross coefficient and the small argon contributions are added.
  • The unexplained discrepancies with existing acoustic virial data for nitrogen show that new acoustic measurements are needed to test and refine these recommendations.

Reading between the lines

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

  • An independent, openly available three-body surface for nitrogen would let the $D,E,F$ constraint be tested directly; a shift in $D$ beyond the assumed 2000 cm$^9$ mol$^{-3}$ would move the extracted $B$ values by a comparable amount.
  • The paper's explanation of the refractometry offset implies that a future ab initio calculation of the second refractivity virial coefficient $B_R$ could adjudicate between the recommended $B(T)$ and the values derived from refractivity.
  • No independent low-temperature oxygen data exist; the rescaled-tuning-difference method predicts larger low-temperature uncertainty for oxygen than nitrogen, so new measurements below 300 K would be the decisive test.
  • The per-spin-surface $B$ shifts in oxygen stay nearly equal up to 373 K, suggesting that the one-parameter tuning result does not depend on which detailed part of the potential is adjusted; the same strategy should transfer to other spin-coupled molecular pairs.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper recommends values for the second virial coefficient B(T) and its expanded uncertainty for molecular nitrogen (20-3000 K) and oxygen (20-2000 K). The recommendations are obtained by tuning previously published ab initio pair potentials to selected high-accuracy experimental data, with full quantum treatment via path-integral Monte Carlo. For nitrogen, the authors reanalyze density isotherms with the third through sixth virial coefficients constrained to values calculated using a preliminary pair PES and a new nonadditive three-body PES, and then retune the pair PES (scaling factor 0.472) to the extracted B values. For oxygen, the three singlet/triplet/quintet PES are retuned by shifting the isotropic C8 dispersion coefficient so that B at 293.15 K matches the refractometry data of Egan. The paper provides smooth correlations for B(T) and U(B), tables of parameters, PIMC data, and Fortran code for the tuned potentials.

Significance. If the recommended values are reliable, they are directly useful for pressure metrology, buoyancy corrections, humidity metrology, and future reference equations of state for air. The paper has clear strengths: the PIMC treatment of quantum effects is the appropriate rigorous method, the QFH semiclassical calculations provide a useful internal cross-check, the uncertainty estimation is transparent, and the authors provide the tuned potential routines and PIMC data as supplementary material. The central concern, discussed below, is that for nitrogen the B values used for tuning were extracted from isotherms constrained by higher virial coefficients from a preliminary and partly unpublished theory, so the final recommendation may inherit a systematic bias that is not fully captured by the quoted uncertainty.

major comments (3)
  1. [Section 2.1 and Table 1] The B values used to retune the nitrogen pair potential are extracted from isotherms with D, E, and F constrained to values calculated with a preliminary pair PES (scaling 0.52) and an unpublished nonadditive three-body PES. The final pair PES is then tuned to those extracted B values (scaling 0.472). Consequently, any systematic error in the constrained higher virial coefficients propagates directly into the extracted B values and into the final recommended B(T), and the internal agreement shown in Fig. 1 cannot validate D, E, and F. The paper's 2000 cm9 mol-3 uncertainty estimate for D is a heuristic: it is not derived from a comparison of D computed with the preliminary and final pair potentials, and no uncertainty is assigned to E and F. Table 1 reinforces this concern: at 130 K, Ccalc-Cexp = -48 cm6 mol-2 exceeds the reported U(Cexp) = 35 cm6 mol-2, indicating that the constrained fits may still contain systematic bias in the strongly correlated pair (B, C). I recommend that the authors publish the three-body PES, recompute D, E, and F with the final pair potential as well as the preliminary one, and perform sensitivity tests in which D (and, at least approximately, E and F) are varied within physically plausible bounds; the resulting effect on B should be incorporated into U(B).
  2. [Section 4.1 and Figure 2] The acoustic second virial coefficient data of Ewing and Trusler and of Boyes deviate from the recommended correlation in a way that the authors do not explain quantitatively. Because these data provide an independent check that was not used in the tuning, the claim that U(B) = 0.05 cm3 mol-1 for nitrogen in the range 150-500 K is not supported by the full body of evidence. The authors should either reconcile the acoustic data with the correlation within a defensible uncertainty budget or quote a larger uncertainty that encompasses the discrepancy.
  3. [Section 5] The low-temperature uncertainty estimate for nitrogen is obtained by rescaling the difference between the original and retuned potentials so that it equals 0.05 cm3 mol-1 at 150 K; this is a calibration to the tuning data rather than an independent bound on the potential error. The high-temperature estimate, a linear growth to 0.3 cm3 mol-1 at 3000 K, is justified by a statement about molecular flexibility effects in H2, which is a different molecule with different rotational constants. The authors should support these estimates with actual calculations of the flexibility correction for N2 and O2, or with a direct estimate of the rigid-rotor truncation error.
minor comments (4)
  1. [Section 2.1] The procedure for assigning the plateau ranges in the virial fits is described as subjective; please provide the assigned minimum and maximum pressures for each isotherm in the supplementary material so that the analysis is reproducible.
  2. [Section 3.3] The statement that the PIMC calculations have an accuracy of 10 ppm would be more convincing if convergence data or a convergence plot were included in the supplementary material.
  3. [Section 10] Data availability via 'reasonable request' is not sufficient for a recommended reference dataset; the nonadditive three-body PES for N2 should be deposited in a public repository.
  4. [Eq. (13)] The notation in Eq. (13) is potentially confusing because B appears both as the second density virial coefficient and as a refractivity-related coefficient; please add a clarifying phrase or a different symbol for the latter.

Circularity Check

2 steps flagged · score 4.0 of 10

Nitrogen B(T) is tuned to isotherm reanalyses whose higher-virial corrections come from the same authors' unpublished PES family; the oxygen recommendation is a transparent fit to independent refractivity data.

  1. other [Section 2.1, paragraphs 2-5 (Tuning of Pair Potentials; nitrogen)]
    "We calculated the third to sixth virial coefficients, i.e., C, D, E, and F, at the temperatures at which Nowak et al. and McLinden and Lösch-Will performed their ρ(p) measurements, using the quadratic Feynman–Hibbs (QFH) modification of a version of the pair PES tuned to the data of Egan and Yang. Each considered isotherm was fitted as p(ρ) by a sixth-order virial expansion in which D, E, and F were constrained to the respective semiclassically calculated values, while B and C and initially also the zero-density limit of p/(ρRT) were fitted."

    The B(T) targets used to tune the nitrogen pair PES were not directly measured. They were extracted from density isotherms by fixing D, E, and F to values computed from a preliminary version of the same pair PES family (scaling 0.52) plus an unpublished three-body PES from the same group. The final potential is then adjusted to match these extracted B values, so the recommended B(T) is, by construction, close to B values that already encode the theoretical higher-virial corrections of the potential family being tuned. Any systematic error in D, E, or F from the preliminary surface is absorbed into the extracted B and propagates into the final recommendation. The agreement shown in Fig.

  2. self citation load bearing [Section 1, paragraph 8; Section 2.1; Section 10 (Data Availability)]
    "Perhaps most important, it is now possible to reanalyze isotherms from previous ρ(p) experiments using fourth and higher-order virial coefficients calculated from the pair potential and a new ab initio three-body PES developed also within the MQB-Pascal project by one of us (RH), which is available upon reasonable request. ... The nonadditive three-body PES for N2 used for calculating higher virial coefficients is available upon reasonable request to RH."

    The central improvement claimed for nitrogen—the 'greatly reduced uncertainty' in B from reanalyzed densities—rests on D, E, F computed from an ab initio three-body PES that is not published and is available only on request from the same author who tunes the pair PES. No external benchmark, independent calculation, or machine-checkable artifact is provided for this surface. The only validation offered is agreement with B and C values extracted using the same surface, and those extracted B values are then used as the tuning targets for the final pair PES. Thus the unpublished three-body PES is load-bearing for the final nitrogen recommendation, and its accuracy cannot be checked outside the present paper's fitted data.

full rationale

The paper is transparent that its recommendations are fits: previously published ab initio pair potentials are tuned to selected high-accuracy experimental B(T) values, and the PIMC B(T) from the tuned potentials is then correlated. For oxygen, the Egan refractivity data are independent of the potential-tuning loop, and the O2 recommendation is a clearly described calibration, not a circular derivation. For nitrogen, however, the tuning targets are not raw experimental B values. The B values in Table 1 were obtained by fitting isotherms with D, E, and F fixed to values computed from a preliminary version of the same pair PES family and an unpublished three-body PES from the same group; the final pair PES is then tuned to those very B values. This creates a partial loop: the final B(T) inherits any systematic error in D, E, F, and the internal agreement in Fig. 1 is a consistency check with the tuning target rather than an independent test. The paper's heuristic estimate of the D uncertainty (2000 cm9/mol) is propagated, but it does not cover a potential bias in the unpublished three-body PES, and the acoustic-virial deviations noted in Sec. 4.1 are not incorporated into U(B). Still, the central output is not masquerading as an independent first-principles prediction, and the underlying CCSDT(Q) pair PES and the PIMC treatment provide substantial independent content; a score of 4 reflects this partial, data-construction circularity rather than a fully forced self-citation chain.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central claim rests on two explicit tuning parameters (the nitrogen scaling factor and the oxygen C8 shift) and on several domain assumptions, most notably the accuracy of the unpublished three-body PES used to constrain the nitrogen density-data reanalysis. No new physical entities are introduced.

free parameters (2)
  • Nitrogen tuning scaling factor s = 0.472 (original 0.5, preliminary 0.52)
    Section 2.1: the difference between CCSDT(Q) and CCSD(T) levels is scaled by s to match reanalyzed experimental B values; this is the single tuning parameter.
  • Oxygen C8 dispersion coefficient shift = Adjusted so that each spin-surface B at 293.15 K increases by 0.27 cm3 mol-1
    Section 2.2: the isotropic part of C8 for the site-site interaction in each of the three PES is varied to match Egan's refractometry data; effectively one parameter per surface, noted by the authors as effectively a single tuning parameter.
assumptions (6)
  • domain assumption The rigid-rotor approximation is adequate for N2 and O2 in the considered temperature ranges; intramolecular flexibility contributions to B(T) are neglected or handled only in uncertainty estimates.
    Section 3.1 uses a rigid rotor Hamiltonian (Eq. 5) and fixed bond lengths. Section 5 notes flexibility effects grow with temperature and are handled as a growing uncertainty estimate, not computed explicitly.
  • domain assumption Nuclear exchange effects and ortho-para distinctions are negligible for the partition functions at temperatures above 10 K.
    Section 3.2 omits the intermolecular exchange term based on the D2 result of Ref. 33; the paper states this is a fortiori true for N2 and O2.
  • ad hoc to paper A single-parameter tuning (scaling of the CCSDT(Q)-CCSD(T) difference for N2; C8 dispersion coefficient shift for O2) is sufficient to correct the dominant uncertainty in the ab initio pair potentials.
    Introduced in Sections 2.1 and 2.2 to match selected experimental B(T). The validity of this reduction is assumed, though the temperature-insensitivity of the shift is partially checked for O2 at one additional temperature.
  • domain assumption The ab initio three-body PES for N2 (developed by one of the authors and available on request) is accurate enough to calculate D, E, F used to constrain the experimental isotherm fits.
    Section 2.1 states the higher virial coefficients were calculated using this PES; no independent verification is provided.
  • domain assumption The QFH semiclassical approximation is accurate for computing the higher virial coefficients D,E,F and is sufficiently accurate for B down to about 60 K (within 1% at 20 K), so it can be used in the reanalysis and in uncertainty estimates.
    Section 4.1 validates QFH against PIMC for B, but no such validation is given for D,E,F.
  • domain assumption For oxygen, the three spin surfaces (quintet, triplet, singlet) with statistical weights 5/9, 3/9, 1/9 provide a complete description below 2000 K.
    Section 2.2; the first excited electronic state a1Δg becomes significant above 2000 K, so calculations stop at 2000 K.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Recommended Second Virial Coefficients for Nitrogen and Oxygen." pith.science (2026). https://pith.science/paper/SIXNS7DI

@misc{pith2026260724634,
  author       = {Pith},
  title        = {Pith review of: Recommended Second Virial Coefficients for Nitrogen and Oxygen},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SIXNS7DI}},
  note         = {Machine review of arXiv:2607.24634}
}
abstract

We provide recommended values for the second virial coefficient, $B(T)$, and its uncertainty, for molecular nitrogen and oxygen. The temperature range covered is $20-3000$ K for nitrogen and $20-2000$ K for oxygen. The recommendations are based on tuning previously published state-of-the-art ab initio pair potentials so that the $B(T)$ calculated from the potentials match selected high-accuracy experimental data; for nitrogen the tuning utilizes values of $B$ derived from literature density data with greatly reduced uncertainty by analyzing the data with the aid of ab initio calculated higher virial coefficients. Quantum effects on $B$ are fully included with the path-integral Monte Carlo method. The resulting $B(T)$ have uncertainties similar to those of the best experimental data, but cover a much wider temperature range.

Figures

Figures reproduced from arXiv: 2607.24634 by the authors.

Figure 1
Figure 1. FIG. 1. Deviations of selected experimental data for the second virial coefficient [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Deviations of the available experimental data for the second acoustic virial coefficient [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Deviations of the experimental data of Weber [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

  1. [1]

    Equation for the determination of the density of moist air (1981/91),

    1R. S. Davis, “Equation for the determination of the density of moist air (1981/91),” Metrologia 29, 67–70 (1992). 2A. Picard, R. S. Davis, M. Gläser, and K. Fujii, “Revised formula for the density of moist air (CIPM-2007),” Metrologia45, 149–155 (2008). 3R. W. Hyland and A. Wexler, “Formulations for the thermodynamic properties of dry air from 173.15 K t...

    work page 1992

  2. [2026]

    Determination of fundamental properties of nitrogen from first principles. III. Temperature and frequency dependence of the molecular polarizability and magnetic susceptibility

    13G. Garberoglio, C. Gaiser, R. M. Gavioso, A. H. Harvey, R. Hellmann, B. Jeziorski, K. Meier, M. R. Moldover, L. Pitre, K. Szalewicz, and R. Underwood, “Ab initiocalculation of fluid prop- 22 erties for precision metrology,” J. Phys. Chem. Ref. Data52, 031502 (2023). 14A. H. Harvey and G. Garberoglio, “Avoiding interpolation errors for computed second vi...

    work page Pith review arXiv doi:10.48550/arxiv.2607.21261 2023

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.