betaFeedback

WorksPhysics and Astronomy

Collisional lineshapes of an N2_2-perturbed electronic transition in O2_2 from first principles of quantum theory validated by accurate experimental spectra

Abstract

We demonstrate the feasibility of predicting the lineshape of an electronic molecular transition collisionally perturbed by another molecule from first principles of quantum mechanics. We consider an N2_2-perturbed O2_2 AA-band transition, which is relevant to spectroscopic observations in Earth's atmosphere. This approach simultaneously incorporates the electronic structure, quantum scattering and lineshape effects required to accurately predict pressure- and temperature-dependence of observed lineshapes. The ab initio\textit{ab initio} lineshapes reproduce measurements to better than 0.6%. Theory also provides new physical constraints on beyond-Voigt lineshape parameters in fits of high-precision data, yielding fit residuals at the 0.08% level. This work motivates advances in molecular electronic structure calculations and demonstrates the potential to generate comprehensive spectroscopic reference data from first principles.

Introduction

The oxygen (O2_2) A-band comprises rovibronic transitions connecting vibrational ground states of the X3Σg−^3\Sigma_g^- and b1Σg+^1\Sigma_g^+ electronic terms of O2_2 (Figure 1 (a)). It plays a vital role in studies of the Earth’s atmosphere, where the high concentration of O2_2 and long absorption path lengths allow optical saturation of the weak spin-forbidden A-band lines [1]. This band spans an otherwise highly transparent near-infrared window around 760 nm, making it an important tool in atmospheric science with numerous calibration-free retrieval applications [1, 2, 3, 4, 5, 6, 7, 8, 9]. However, due to the prevalence of molecular collisions in the atmosphere, the accuracy of O2_2 retrievals is limited by uncertainty in the available collisional lineshape parameters. Additionally, accurate extraction of atmospheric properties from O2_2 spectroscopic measurements requires using beyond-Voigt lineshape profiles [1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25].

Figure 1. (a) The O2 A-band comprises rovibronic transitions between the lowest indicated vibrational vi=vf=0v_i=v_f=0 levels of the electronic terms X3Σg−X^{3}\Sigma_g^{-} (green curve) and b1Σg+b^{1}\Sigma_g^{+} (red curve) of O2. (b) (left blue box): P1P1P_1P_1 A-band transition (arrow) connecting the fine structure-resolved ∣X3Σg−,Na=1,j=1⟩|X^{3}\Sigma_g^{-},N_a=1,j=1\rangle and ∣b1Σg+,Na=0⟩|b^{1}\Sigma_g^{+},N_a=0\rangle rotational states of O2. (right orange box): Rotational states ∣X1Σg+,Nb⟩|X^{1}\Sigma_g^{+},N_b\rangle of the collision partner N2. (c) The angle-averaged potential energy surfaces (PESs) for the X (green curve) and b (red dashed curve) terms of O2. (d) Partial-wave contributions to the generalized cross-section Re⁡σλ(q)\operatorname{Re}\sigma_{\lambda}^{(q)} at high collisional kinetic energy for the N2-perturbed target transition (black solid line). Also shown is the analogous result for the H2-He case (gray dashed line).

Historically, collisional lineshape parameters for the O2_2 A-band have been investigated primarily through experiment [10, 11, 12, 13, 14, 15, 17, 18, 19, 20, 23, 24, 25, 26, 27, 28, 29, 30], with only a few semi-empirical [15, 16] or classical molecular dynamics calculations [31, 32]. The only fully quantum-mechanical treatment was given by Grimminkin et al. [33] for the case of pressure broadening and shift of the A-band lines due to collisions with He(1S0^1S_0) atoms. To date, no theoretical studies have investigated collisional lineshape effects in the O2_2 A-band perturbed by major atmospheric constituents entirely from first principles of quantum mechanics, nor have they provided any rigorous ab initio calculations of beyond-Voigt lineshape parameters for such systems.

In this Letter, we report a fully ab initio calculation of lineshape parameters for the P1P1 O2_2 A-band transition perturbed by collisions with 14^{14}N2_2 (Figure 1 (b)). Henceforth, all computed and measured spectra correspond to this absorption line and molecular system. We present a treatment of a molecular electronic transition collisionally perturbed by another molecule from first principles which simultaneously accounts for the electronic structure, scattering dynamics and beyond-Voigt lineshape effects. To validate the theory, we perform accurate measurements of the spectra across a broad pressure range relevant to atmospheric conditions. We show that the ab initio and experimental lineshapes agree to better than 0.6%, although there are discrepancies in the individual parameters. We discuss how limitations of the electronic structure calculations of the O2_2-N2_2 interaction potential preclude accurate prediction of the collisional shift parameter. We also estimate ab initio parameter bias by comparison to experiment, and we describe how higher-order ab initio lineshape parameters can be used to constrain fits of experimental spectra.

Theoretical methodology— To investigate the N2_2-induced collisional perturbations of the O2_2 A-band, the O2_2-N2_2 quantum scattering problem was solved for the ground (η=X\eta= X) and excited (η=b\eta= b) electronic states of the absorbing O2_2. To this end, two ab initio potential energy surfaces (PESs) VηV_\eta were calculated for 16^{16}O2(η)−14_2(\eta)-{}^{14}N2(X 1Σg+)_2(X\,^1\Sigma_g^+) interactions using explicitly correlated unrestricted coupled-cluster theory [34]. The present work reports a new VbV_b, while using the VXV_X from Ref. [35].

In principle, VXV_X and the associated O2_2-N2_2 scattering dynamics depend on the total electronic spin of O2_2. However, since the collisional kinetic energies relevant for atmospheric conditions (≈100\approx100 cm−1^{-1} to 300 cm−1^{-1}) are much larger than fine-structure corrections to rotational energies of O2_2 (≈2\approx2 cm−1^{-1}, see Figure 1 (b)), the spin dependence of the O2_2-N2_2 Hamiltonian may be neglected in quantum scattering calculations. The collisional problem was therefore formulated in terms of a spin-independent Hamiltonian (O2 and N2 labeled as a and b, respectively)

Hη=−ℏ22μR∂2∂R2+ℏ2l22μR2+Ha(η)+Hb+Vη(R,r^a,r^b),(1)H_{\eta}=-\frac{\hbar^{2}}{2\mu R}\frac{\partial^{2}}{\partial R^{2}}+\frac{\hbar^{2}l^{2}}{2\mu R^{2}}+\mathcal{H}_{a(\eta)}+\mathcal{H}_{b}+V_{\eta}(R,\hat{r}_{a},\hat{r}_{b}), \tag*{(1)}

where RR is the distance between nuclear mass centers of the colliding molecules, μ\mu is the reduced mass, ll is the relative orbital angular momentum and Hi\mathcal{H}_{i} is the rotational Hamiltonian of the ii-th isolated monomer whose orientation relative to the axis taken along RR is given by a unit vector r^i\hat{r}_{i}. The dependence of VηV_{\eta} on the monomer orientations leads to rotationally inelastic collisions. The angle-averaged PESs for both η\eta terms of O2 are shown in Figure 1(c).

Exact coupled-channels quantum scattering calculations were then performed for each η\eta using the BIGOS code [36, 37], and appropriate scattering boundary conditions were imposed [38] to obtain spin-independent SS-matrix elements. The non-zero total electronic spin of O2 was accounted for by transforming the SS-matrices to the representation that includes spin explicitly [35, 39] and is thus capable of correctly labeling the internal monomer states involved in the A-band transitions (Figure 1(b)).

The high conceptual and computational complexity of the scattering problem [40] presents a major challenge, illustrated schematically in Figure 2(a), due to the accuracy requirements of atmospheric lineshape modeling. First, the large reduced mass of the O2-N2 complex and the relevant collisional kinetic energies require an expansive partial-wave basis substantially larger than for simpler molecule-atom cases like H2-He (Figure 1(d)). Second, the O2-N2 scattering dynamics are dictated by two highly anisotropic PESs coupling distant rotational states NaN_a and NbN_b of the monomers which, when combined with the small rotational constants of both molecules, demands an immense basis set of internal states (up to 50 {Na,Nb}\{N_a,N_b\} pairs). The partial-wave and internal state requirements result in a large number of scattering channels (≈4000\approx4000 per parity) that must be included in the ab initioab\ initio SS-matrix calculations.

Figure 2. (a) Complexity of the quantum scattering problem. (b) Generalized cross-sections σλ(q)\sigma_{\lambda}^{(q)} as a function of collisional kinetic energy EkinE_{\rm kin} and rotational state NbN_b of the perturbing N2_2 (only four examples of NbN_b are shown, for clarity). The gray curve is the Maxwell distribution (in arbitrary units) at temperature T=296.8T=296.8 K. (c) Ab initio collisional lineshape parameters (Table I) as a function of temperature, with T=296.8T=296.8 K indicated by a vertical dashed line. The theoretical approach provides accurate lineshape parameters over a substantially wider temperature range than has been reported experimentally (gray vertical strip) [10, 11, 12, 13, 14, 15, 17, 18, 19, 20, 23, 24, 25, 26, 27, 28, 29, 30].

The transformed SS-matrices were used to compute the complex-valued generalized cross-sections σλ(q)\sigma_{\lambda}^{(q)} [40, 41, 42] necessary to calculate the collisionally perturbed lineshape parameters. In contrast to the more familiar differential scattering cross-sections, the generalized cross-sections depend on the rank of a 2q2^{q}-pole transition moment and on the index λ=0,1\lambda=0,1 associated with transport phenomena (e.g., Dicke narrowing), complicating the dependence on SS-matrix elements. For the A-band transitions driven by the O2 magnetic dipole moment, q=1q=1.

Figure 2(b) shows the ab initioab\ initio cross-sections σλ(q)\sigma_{\lambda}^{(q)} for the N2-perturbed P1P1P_1P_1 O2 A-band transition. It is clearly seen that the collisional effects are dominated by the real parts, Re⁡σλ(q)\operatorname{Re}\sigma_{\lambda}^{(q)}, which are nearly two orders of magnitude larger than the imaginary parts, Im⁡σλ(q)\operatorname{Im}\sigma_{\lambda}^{(q)}. The calculations cover more than 99.9% of the statistical distributions for the relative O2-N2 speeds and rotational states of the perturbing N2 at atmospherically relevant temperatures (≈300\approx300 K) in order to meet the requirements of atmospheric applications. This translates to a broad range of collisional kinetic energies, EkinE_{\rm kin}, and rotational states, NbN_b, with exceptionally high upper limits (Ekin≈1700 cm−1E_{\rm kin}\approx1700\ {\rm cm}^{-1}, Nb≈26N_b\approx26), see Figure 2(a). We note that it is possible to cover large values of EkinE_{\rm kin} and NbN_b using the extrapolated σλ(q)\sigma_{\lambda}^{(q)} without loss of accuracy, see End Matter.

The cross-sections relate the microscopic intermolecular interactions to the macroscopic transport-relaxation phenomena in the absorbing gas, allowing the computation of collisional parameters of the modeled lineshape profiles. Here, we generate the ab initioab\ initio collisional lineshape parameters of the modified Hartmann–Tran (mHT) profile [43] for the N2-perturbed O2 target line. Besides the standard Voigt collisional parameters (speed-averaged pressure broadening, Γ0\Gamma_0, and shift, Δ0\Delta_0), the mHT model accounts for the leading beyond-Voigt effects: speed dependence of broadening and shift (Γ2\Gamma_2 and Δ2\Delta_2, respectively), and velocity-changing collisions quantified by the complex-valued Dicke parameter, νopt\nu_{\mathrm{opt}} [43]. The mHT lineshape parameters were determined by averaging σλ(q)(Ekin;Nb)\sigma_{\lambda}^{(q)}(E_{\mathrm{kin}};N_b) over EkinE_{\mathrm{kin}} and NbN_b. In particular, λ=0\lambda=0 cross-sections were used to compute the speed-independent pressure broadening and shift [41, 44, 45]

Γ0+iΔ0=12πcpkBTvˉr∑NbρNb∫0∞dx xe−xσ0(q)(xkBT;Nb),(2)\Gamma_0+i\Delta_0=\frac{1}{2\pi c}\frac{p}{k_{\mathrm{B}}T}\bar{v}_r\sum_{N_b}\rho_{N_b}\int_0^\infty dx\,x e^{-x}\sigma_0^{(q)}(xk_{\mathrm{B}}T;N_b), \tag*{(2)}

and their leading speed dependence [46]

Γ2+iΔ2=12πcpkBTvb∗πe−y2∑NbρNb∫0∞dx[2xcosh⁡(2xy)−(1y+2y)sinh⁡(2xy)]x2e−x2σ0(q)(μ(xvb∗)2/2;Nb),(3)\Gamma_2+i\Delta_2=\frac{1}{2\pi c}\frac{p}{k_{\mathrm{B}}T}\frac{v_b^*}{\sqrt{\pi}}e^{-y^2}\sum_{N_b}\rho_{N_b}\int_0^\infty dx\left[2x\cosh(2xy)-\left(\frac{1}{y}+2y\right)\sinh(2xy)\right]x^2e^{-x^2}\sigma_0^{(q)}\left(\mu\left(xv_b^*\right)^2/2;N_b\right), \tag*{(3)}

where pp is the pressure, TT is the temperature, ρNb\rho_{N_b} is the Boltzmann distribution of rotational states of the perturbing N2_2, vˉr\bar{v}_r is the average relative O2_2-N2_2 speed, vb∗v_b^* is the most probable N2_2 speed, and y2=mb/may^2=m_b/m_a is the perturber-to-absorber mass ratio. On the other hand, λ=1\lambda=1 cross-sections were used for calculating the Dicke parameter [41, 44, 46]

νopt≡νoptr+iνopti=12πcpkBTmbma+mbvˉr∑NbρNb∫0∞dx[23xσ1(q)(xkBT;Nb)−σ0(q)(xkBT;Nb)]xe−x.(4)\nu_{\mathrm{opt}}\equiv\nu_{\mathrm{opt}}^r+i\nu_{\mathrm{opt}}^i=\frac{1}{2\pi c}\frac{p}{k_{\mathrm{B}}T}\frac{m_b}{m_a+m_b}\bar{v}_r\sum_{N_b}\rho_{N_b}\int_0^\infty dx\left[\frac{2}{3}x\sigma_1^{(q)}(xk_{\mathrm{B}}T;N_b)-\sigma_0^{(q)}(xk_{\mathrm{B}}T;N_b)\right]xe^{-x}. \tag*{(4)}

In Figure 2 (c) shows the per-unit-pressure ab initio lineshape parameters (denoted γ0\gamma_0, δ0\delta_0, γ2\gamma_2, δ2\delta_2, ν~optr\tilde{\nu}_{\mathrm{opt}}^r and ν~opti\tilde{\nu}_{\mathrm{opt}}^i) as a function of temperature, where the values at the experimental temperature, 296.8 K, are reported in Table I (a). The collisional effects are dominated by γ0\gamma_0, an expected consequence of the two-orders-of-magnitude difference between the real and imaginary parts of σ0(q)\sigma_0^{(q)} (cf. Figure 2 (b) and Eq. (2)). While one might expect ν~optr\tilde{\nu}_{\mathrm{opt}}^r to be similarly large, Eq. (4) involves the difference between two cross-sections of comparable magnitude therefore yielding a substantially smaller value than γ0\gamma_0. The calculated collisional parameters were used to generate the fully ab initio mHT lineshape profiles, shown in Figure 3 (a) as red curves for eight pressures spanning from the low-pressure Doppler limit to the collision-dominated high-pressure regime.

Figure 3. Experimental validation of our quantum ab initio calculations of the lineshapes. (a) Experimental (black points) and theoretical (red lines) shapes of the N2_2-perturbed target line at T=296.8T = 296.8 K and eight pressures. (b) Differences between the experimental and fully ab initio theoretical spectra. (c) The same as panel (b) but the collisional line shift parameter in theoretical spectra is adjusted to experimental spectra. (d) The same as panel (b) but in the ab initio calculations the potential energy surface for the b1Σg+b^1\Sigma_g^+ state was multiplied by 1.0182. The relative root-mean-square error (rRMSE), shown above each curve in panels (b)–(d), was calculated within ±1\pm1 full width at half maximum (FWHM) and normalized to the profile amplitude. The pressure-averaged rRMSE are reported in the right side of the figure.

Measurements— To benchmark the ab initio calculations, high-precision measurements of the N2_2-perturbed target transition were made using a frequency-stabilized cavity-ring down spectroscopy system, previously described in Refs. [24, 47]. This system provided enough sensitivity to enable use of a dilute oxygen sample (0.25% of O2_2 in N2_2) and yielded a nearly distortion-free measurement of lineshapes [22]. These experimental attributes resulted in high fidelity spectra against which the ab initio calculations were compared.

Comparison of experiment and theory— The experimental N2_2-perturbed lineshapes are shown in Figure 3 (a) as black points. The differences between the experimental and theoretical spectra are shown in Figure 3 (b). Also shown are the corresponding values of the relative root-mean-square error (rRMSE), calculated within ±1\pm1 full width at half maximum (FWHM) and normalized to the profile amplitude, which quantify the magnitude of the differences between theory and experiment [48]. The pressure-averaged rRMSE is 3.3%. However, the differences between the experimental and theoretical lineshapes systematically exhibit odd symmetry about line center and increase with pressure, which indicates that the dominant source of the discrepancy is the collisional shift parameter, δ0\delta_0. To confirm the impact of pressure shift, we simultaneously fit the mHT profile to all experimental spectra adjusting the value of δ0\delta_0 while fixing all other collisional lineshape parameters to the ab initio theoretical values, see Table I (c). In Figure 3 (c), we show the corresponding residuals, which confirm the differences in Figure 3 (b) can mostly be attributed to inaccuracy in the calculated collisional line-shift parameter. Noting that the collision-induced line shift only displaces the line center as a function of pressure, a key result is that the present methodology can predict the collision-induced shapes of a molecular line from fully ab initio quantum calculations at the sub-percent level (a mean rRMSE = 0.57%, see Figure 3 (d)).

TheoryExperiment
(a)(b)(c)(d)(e)
γ0\gamma_059.159.359.160.33(2)60.33(–)
δ0\delta_02.2−4.0−4.0 (*)−3.58(2)−3.48(2)
γ2\gamma_23.33.43.34.35(8)3.46(–)
δ2\delta_20.3−0.80.3−0.69(8)−0.34(11)
ν~optr\tilde{\nu}^{r}_{\mathrm{opt}}5.15.05.12.1(2)5.54(12)
ν~opti\tilde{\nu}^{i}_{\mathrm{opt}}−1.2−1.5−1.20.7(3)−0.5(4)

Table 1. Pressure-normalized collisional lineshape parameters for the N2_2-perturbed target transition at T=296.8T = 296.8 K. All reported quantities have units of 10−3 cm−1/atm10^{-3}\,\mathrm{cm}^{-1}/\mathrm{atm} (1 cm−1/atm=0.295 872 15\mathrm{cm}^{-1}/\mathrm{atm} = 0.295\,872\,15 MHz/Pa). (a) Fully ab initio calculations. (b) Ab initio calculations with excited-state VbV_b PES multiplied by 1.0182. (c) Ab initio calculations (same as column (a)) with collisional shift δ0\delta_0 adjusted to experimental spectra, denoted by (*). (d) Results from least-squares multi-spectrum fit of the mHT profile to measured spectra giving an rRMSE = 0.06%. (e) Same as (d) with two constraints: γ0\gamma_0 equal to the experimental value in column (d), and γ2/γ0=0.057\gamma_2/\gamma_0 = 0.057 given by the ratio of corresponding theoretical values in column (b). Varying the remaining four fitted parameters gives an rRMSE = 0.08%. Values in parentheses are the statistical uncertainties in the last reported digit. Constrained parameters do not report a statistical uncertainty.

To extend the analysis, we consider the origin of the pressure shift discrepancy. The value of δ0\delta_0 comes from the imaginary part of the generalized spectroscopic cross-section, see Eq. (2). We note that this component is related to the difference between scattering phase shifts [49], which may change sign [50] and which depends on the number of closed channels included in the basis [51]. Consequently, Im⁡σ0(q)\operatorname{Im}\sigma_0^{(q)} converges more slowly in numerical quantum scattering calculations than Re⁡σ0(q)\operatorname{Re}\sigma_0^{(q)}. The pressure shift is also much more sensitive to the details of the PES, which we observe in the quantum-mechanical case and which was realized already in the context of semi-classical scattering calculations [52, 53]. To illustrate the sensitivity of δ0\delta_0 to the PES, we modified the PES of the excited state by 1.82% (i.e., we scaled VbV_b by a factor of 1.0182) and repeated the ab initio calculations. The results are shown in Table I (b) and in Figure 3 (d). The small modification of the PES considerably changed the calculated δ0\delta_0, yielding a result close to the experimental value, while introducing negligible changes to the other lineshape parameters. An exception is the δ2\delta_2 parameter whose value is directly correlated with δ0\delta_0, but for which we found had a negligible influence on the lineshape.

Parameter correlation— A long-standing challenge in spectroscopic reference data is the physical interpretation of beyond-Voigt parameters obtained from fits of parameterized profiles to measured spectra [1, 22, 43, 54, 55, 56, 57]. When fitting beyond-Voigt profiles to experimental spectra, substantial numerical correlation between parameters occurs because of their similar influence on the profile [57]. In practice, various combinations of correlated parameters yield statistically indistinguishable profiles when residuals are smaller than the background noise. Therefore, unconstrained analyses typically lead to unphysical or biased fitted parameters. Typically, parameter correlation is reduced by employing a multi-spectrum fitting approach over a range of pressure and temperature [58, 59]. Nevertheless, the ability to completely disentangle parameter correlation is limited by the finite signal-to-noise ratio and the lack of more general models that constrain the relative magnitudes of the relevant beyond-Voigt lineshape terms [22]. The determination of fitted mHT parameters is further complicated by other higher-order effects, such as line mixing and intensity depletion [23, 60, 61], that introduce additional model parameters.

Estimating bias in calculated lineshape parameters and implementing advanced constraints— Our measurements of the speed-independent pressure broadening parameter have combined relative uncertainties of less than 0.1%, as shown in Table I (d). This result is a consequence of the high signal-to-noise ratio of the measurements and the fact that this fit-derived quantity is weakly correlated with the other higher-order mHT parameters. Therefore, we expect the difference between the calculated and measured speed-independent and speed-dependent collisional broadening parameters to be dominated by bias in our calculation of Re⁡σ0(q)\operatorname{Re}\sigma_0^{(q)} (Eqs. (2)–(3)). Comparing the γ0\gamma_0 values in columns (b) and (d) of Table I, we find that the ratio of measured-to-calculated values is 1.017. If we attribute this discrepancy to systematic error in Re⁡σ0(q)\operatorname{Re}\sigma_0^{(q)}, then the calculated value of γ2\gamma_2 would also be underestimated by 1.7%.

With γ0\gamma_0 and γ2\gamma_2 constrained by comparison to experiment and theory (Table I (e)), the remaining mHT parameters were obtained from multi-spectrum fits to the measured spectra. This analysis yielded an average rRMSE value of 0.08%, which is comparable to that obtained with the unconstrained fit results (0.06%), shown in Table I (d). The resulting ν~optr\tilde{\nu}_{\mathrm{opt}}^{r} value is 10.8% greater than the theoretical value. Recall that ν~optr\tilde{\nu}_{\mathrm{opt}}^{r} depends on the difference between two cross-section terms of similar magnitude, see Eq. (4), making its prediction more uncertain than that of γ2\gamma_2. As expected, this constrained mHT fit has a much larger impact on δ2\delta_2 and ν~optr\tilde{\nu}_{\mathrm{opt}}^{r}, yielding a negative value for the latter more consistent with the theoretical prediction, albeit with a relatively large uncertainty. To the best of our knowledge, this represents the most reliable determination of collisional lineshape parameters for O2_2 to date, achieving exceptionally low fit residuals while preserving the physical consistency of the retrieved parameters.

In addition to providing accurate ab initio lineshape parameters in the absence of experimental results, the present theory provides additional constraints for fitting the mHT profile to experimental spectra. Here, this methodology has been demonstrated for the P1P1 transition, but it can be applied to other AA-band lines, other O2_2 bands and other diatomic molecules perturbed by O2_2 and N2_2. In future experimental analyses, by constraining γ2\gamma_2 as discussed above, its numerical correlation with the other mHT parameters, especially ν~optr\tilde{\nu}_{\mathrm{opt}}^{r}, can be substantially reduced. This approach, which leverages new theoretical constraints, will yield more physically consistent mHT parameter retrievals.

Conclusions— The present study demonstrates an exceptional level of agreement between ab initio and measured collision-induced shapes of a rovibronic molecular transition. The strength of the first-principles approach is seen in Figure 2 (c). The calculations not only provide the beyond-Voigt lineshape parameters, but also cover a wide range of temperatures which are difficult to experimentally realize. The future of accurate molecular lineshape data will be heavily based on the interplay between ab initio quantum calculations and accurate spectroscopic measurements. The calculations can provide accurate temperature-dependent lineshape parameters originating from the real parts of the spectroscopic cross-sections. Experiments can be used to refine the calculations of the complex-valued cross-sections to more accurately account for the relative magnitudes of the lineshape terms. By coupling ab initio calculations with high-resolution measurement benchmarking, it will be possible to generate comprehensive and reference-grade spectroscopic databases, which are necessary for atmospheric studies of Earth and other planetary systems [62]. Beyond the context of atmospheric studies, the present methodology has basic-science applications as it provides a sensitive tool for evaluating ab initio PESs by probing the imaginary part of the cross-sections shown in this work, challenging the present state-of-the-art electronic structure calculations [48, 63, 64].

Acknowledgments

The authors would like to thank Daniel Barker (NIST) and Eisen Gross (NIST) for commenting on the manuscript. The research was supported by the National Science Centre, Poland, Project Nos.: 2021/43/O/ST2/00211 (MG) and 2022/46/E/ST2/00282 (HJ, NS, PW). RD and EQS. are supported by the United States Department of Energy (DOE), Grant No. DE-SC0025420. We gratefully acknowledge Polish high-performance computing infrastructure PLGrid (HPC Center: ACK Cyfronet AGH) for providing computer facilities and support within computational grant no. PLG/2026/019877. Created using resources provided by Wroclaw Centre for Networking and Supercomputing (http://wcss.pl). The research is part of the program of the National Laboratory FAMO in Toruń, Poland.

Data availability— Experimental spectra are available at https://data.nist.gov/. Additional data are available from the authors upon reasonable request.

References

End Matter

Additional calculation details— Our calculations revealed that the dominating real part of the spectroscopic cross-sections, Re⁡σλ(q)\operatorname{Re}\sigma_\lambda^{(q)}, follows a power law in EkinE_{\rm kin} for large kinetic energies. We therefore performed the ab initio calculations up to Ekin=300 cm−1E_{\rm kin}=300\ \mathrm{cm}^{-1} and used the high-energy “tail” of the calculated Re⁡σλ(q)\operatorname{Re}\sigma_\lambda^{(q)} to fit a power-law dependence and extrapolate Re⁡σλ(q)\operatorname{Re}\sigma_\lambda^{(q)} to higher kinetic energies. We observed no such dependence on EkinE_{\rm kin} for the imaginary part, Im⁡σλ(q)\operatorname{Im}\sigma_\lambda^{(q)}, two orders of magnitude smaller than the real part. Instead, the calculated Im⁡σλ(q)\operatorname{Im}\sigma_\lambda^{(q)} stabilize at high EkinE_{\rm kin}, approaching small constant values near zero. Due to lack of a more definite functional dependence on EkinE_{\rm kin}, we extrapolated Im⁡σλ(q)\operatorname{Im}\sigma_\lambda^{(q)} to higher kinetic energies using the stabilized constant values. We verified that due to the two-orders-of-magnitude difference between the real and imaginary parts, the extrapolation scheme does not influence the accuracy of the ab initio lineshape; the primary factor is the quality of the O2_2-N2_2 interaction PES, as demonstrated in Table I and Figure 3.

Furthermore, for all calculated σλ(q)\sigma_\lambda^{(q)}, we observed that the cross-sections become almost independent of the rotational state of the perturbing N2_2 already for Nb≳4N_b\gtrsim4 (Figure 2 (b)). Consequently, we performed the ab initio calculations up to Nb=9N_b=9 and then—depending on the spin isomer of N2_2 (ortho vs. para)—we extrapolated σλ(q)\sigma_\lambda^{(q)} to higher rotational quantum numbers NbN_b using the ab initio cross-sections calculated for Nb=8,9N_b=8,9.

Finally, we emphasize that, when calculating the lineshape parameters (Eqs. (2)–(4)), we use all ab initio values of σλ(q)\sigma_\lambda^{(q)} and switch to the extrapolated cross-sections only when the ab initio values reach their EkinE_{\rm kin} and NbN_b limits.

Additional measurement details— This work reports high-precision measurements of the N2_2-perturbed target transition made using a frequency-stabilized cavity-ring down spectroscopy system, previously described in Refs. [24, 47]. The optical cavity, HeNe length-stabilization servo, frequency agile rapid scanning, and temperature measurement scheme remain the same as those detailed in Ref. [47]. The current measurements included optical frequency comb (OFC) referencing of the frequency axis and flowed samples, which differ from the previous descriptions Refs. [24, 47].

The current measurements referenced the cavity mode axis at each pressure to the visible output of a self-referenced, femtosecond optical frequency comb (OFC) (referenced to a global positioning satellite disciplined Rubidium oscillator) using a heterodyne beat signal. A free-space optical band pass filter and RF band pass filters were used to ensure the laser carrier frequency was only beat with one comb tooth. Heterodyne beat note measurements were made at discrete cavity resonant frequencies across the O2_2 A-band, providing absolute frequency anchors to the broadband cavity mode axis, with an absolute frequency uncertainty less than 300 kHz.

The 0.25% of O2_2 in N2_2 sample measured in this work was under flow where the intra-cavity pressure was stabilized (≈1\approx1 Pa over broadband spectrum) using a proportional servo operating on a voltage-controlled valve placed between the output of the optical cavity and a vacuum pump. The volumetric flow-rate was between 10 and 70 std. cm3^3 min−1^{-1}. The pressure was measured using a resonant silicon manometer with a full-scale range of 130 kPa, calibrated against a NIST primary pressure transfer standard, and corrected to account for flow-induced pressure drop between the manometer and the sample volume, where the maximum magnitude of the correction was less than 20 Pa.

References

  1. [1]D. A. Long and J. T. Hodges, Journal of Geophysical Research: Atmospheres 117, D12309 (2012).
  2. [2]G. Yamamoto and D. Q. Wark, Journal of Geophysical Research 66, 3596 (1961).
  3. [3]I. J. Barton and J. C. Scott, Applied Optics 25, 3502 (1986).
  4. [4]A. Kuze and K. V. Chance, Journal of Geophysical Research 99, 14481 (1994).
  5. [5]D. Crisp, R. M. Atlas, F.-M. Breon, L. R. Brown, J. P. Burrows, P. Ciais, B. J. Connor, S. C. Doney, I. Y. Fung, D. J. Jacob, C. E. Miller, D. O’Brien, S. Pawson, J. T. Randerson, P. Rayner, R. J. Salawitch, S. P. Sander, B. Sen, G. L. Stephens, P. P. T. n, G. Toon, P. O. Wennberg, S. C. Wofsy, Y. L. Yung, Z. Kuang, B. Chudasama, G. Sprague, B. Weiss, R. Pollock, D. Kenyon, and S. Schroll, Advances in Space Research 34, 700 (2004).
  6. [6]B. van Diedenhoven, O. P. Hasekamp, and I. Aben, Atmospheric Chemistry and Physics 5, 2109 (2005).
  7. [7]S. Corradini and M. Cervino, Journal of Quantitative Spectroscopy and Radiative Transfer 97, 354 (2006).
  8. [8]L. Guanter, L. Alonso, L. Gómez-Chova, M. Meroni, R. Preusker, J. Fischer, and J. Moreno, Journal of Geophysical Research 115, D19303 (2010).
  9. [9]Y. Zhu and M. Kaufmann, Geophysical Research Letters 46, 8536 (2019).
  10. [10]J. K. Ritter and T. D. Wilkerson, Journal of Molecular Spectroscopy 121, 1 (1987).
  11. [11]B. Ray and P. N. Ghosh, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 53, 537 (1997).
  12. [12]B. Ray, D. Biswas, and P. N. Ghosh, Journal of Molecular Structure 407, 39 (1997).
  13. [13]R. Schermaul and R. C. M. Learner, Journal of Quantitative Spectroscopy and Radiative Transfer 61, 781 (1999).
  14. [14]L. R. Brown and C. Plymate, Journal of Molecular Spectroscopy 199, 166 (2000).
  15. [15]H. Tran, C. Boulet, and J.-M. Hartmann, Journal of Geophysical Research: Atmospheres 111, D15210 (2006).
  16. [16]H. Tran and J.-M. Hartmann, Journal of Geophysical Research: Atmospheres 113, D18104 (2008).
  17. [17]D. J. Robichaud, J. T. Hodges, D. Lisak, C. E. Miller, and M. Okumura, Journal of Quantitative Spectroscopy and Radiative Transfer 109, 435 (2008).
  18. [18]A. Predoi-Cross, C. Holladay, H. Heung, J.-P. Bounachir, G. C. Mellau, R. Keller, and D. R. Hurtmans, Journal of Molecular Spectroscopy 251, 159 (2008).
  19. [19]A. Predoi-Cross, K. Hambrook, R. Keller, C. Povey, I. Schofield, D. Hurtmans, H. Over, and G. C. Mellau, Journal of Molecular Spectroscopy 248, 85 (2008).
  20. [20]B. J. Drouin, D. C. Benner, L. R. Brown, M. J. Cich, T. J. Crawford, V. M. Devi, A. Guillaume, J. T. Hodges, E. J. Mlawer, D. J. Robichaud, F. Oyafuso, V. H. Payne, K. Sung, E. H. Wishnow, and S. Yu, Journal of Quantitative Spectroscopy and Radiative Transfer 186, 118 (2017).
  21. [21]J. Mendonca, K. Strong, D. Wunch, G. C. Toon, D. A. Long, J. T. Hodges, V. T. Sironneau, and J. E. Franklin, Atmospheric Measurement Techniques 12, 35 (2019).
  22. [22]H. Fleurbaey, Z. D. Reed, E. M. Adkins, D. A. Long, and J. T. Hodges, Journal of Quantitative Spectroscopy and Radiative Transfer 270, 107684 (2021).
  23. [23]E. M. Adkins, J. T. Hodges, K. Bielska, A. Campargue, R. Ciuryło, J. Domysławska, R. P. Fernandez, H. Fleurbaey, M. Ganczewski, H. Jóźwiak, S. Kassi, D. Lisak, D. Mondelain, G. G. Palançar, W. Somogyi, O. G. Tomazzeli, H. Tran, P. Wcisło, S. Wójtewicz, S. N. Yurchenko, and I. E. Gordon, Journal of Quantitative Spectroscopy and Radiative Transfer 347, 109629 (2025).
  24. [24]L. E. Stevenson, J. L. Laughner, M. Okumura, J. T. Hodges, and E. M. Adkins, Journal of Quantitative Spectroscopy and Radiative Transfer 342, 109480 (2025).
  25. [25]J. Klemm, A. Campargue, H. Fleurbaey, S. Kassi, D. Romannini, and D. Mondelain, Journal of Quantitative Spectroscopy and Radiative Transfer 342, 109448 (2025).
  26. [26]T. Nakazawa, T. Yamanouchi, and M. Tanaka, Journal of Quantitative Spectroscopy and Radiative Transfer 27, 615 (1982).
  27. [27]R. S. Pope, P. J. Wolf, and G. P. Perram, Journal of Quantitative Spectroscopy and Radiative Transfer 64, 363 (2000).
  28. [28]S. Yang, M. R. Canagaratna, S. K. Witonsky, S. L. Coy, J. I. Steinfeld, R. W. Field, and A. K. Kachnov, Journal of Molecular Spectroscopy 201, 188 (2000).
  29. [29]C. Hill, D. A. Newnham, and J. M. Brown, Journal of Molecular Spectroscopy 219, 65 (2003).
  30. [30]D. P. Merrell, A. T. Theorx, C. L. Strand, and R. K. Hanson, Journal of Quantitative Spectroscopy and Radiative Transfer 348, 109683 (2026).
  31. [31]J.-M. Hartmann, V. Sironneau, C. Boulet, T. Svensson, J. T. Hodges, and C. T. Xu, Physical Review A 87, 032510 (2013).
  32. [32]D. D. Tran, V. T. Sironneau, J. T. Hodges, R. Armante, J. Cuesta, and H. Tran, Journal of Quantitative Spectroscopy and Radiative Transfer 222-223, 108 (2019).
  33. [33]D. L. A. G. Grimmink, F. R. Spiering, L. M. C. Janssen, A. van der Voird, W. J. van der Zande, and G. C. Groenenboom, The Journal of Chemical Physics 140, 204314 (2014).
  34. [34]G. Knizia, T. B. Adler, and H. J. Werner, The Journal of Chemical Physics 130, 054104 (2009).
  35. [35]M. Gancewski, H. Jóźwiak, E. Quintas-Sánchez, R. Dawes, F. Thibault, and P. Wcisło, The Journal of Chemical Physics 155, 124307 (2021).
  36. [36]H. Jóźwiak, F. Thibault, A. Viel, P. Wcisło, and F. Lique, Astronomy and Astrophysics 685, A113 (2024).
  37. [37]H. Jóźwiak, The SCATTERING code adjusted for diatom-atom calculations (2024).DOI
  38. [38]S. Green, The Journal of Chemical Physics 62, 2271 (1975).
  39. [39]A. R. Offer, M. C. van Hemert, and E. F. van Dishoeck, The Journal of Chemical Physics 100, 362 (1994).
  40. [40]H. Jóźwiak and P. Wcisło, Ab initio calculations of lineshape parameters, in High-Resolution Molecular Spectroscopy 1B (John Wiley & Sons, Ltd, 2026) Chap. 5, pp. 153–198.
  41. [41]L. Monchick and L. W. Hunter, The Journal of Chemical Physics 85, 713 (1986).
  42. [42]A. Olejnik, H. Jóźwiak, M. Gancewski, E. Quintas-Sánchez, R. Dawes, and P. Wcisło, The Journal of Chemical Physics 159, 134301 (2023).
  43. [43]P. Wcisło, N. Stolarczyk, M. Słowiński, H. Jóźwiak, D. Lisak, R. Ciuryło, A. Cygan, F. Schreier, C. D. Boone, A. Castrillo, L. Gianfrani, Y. Tan, S.-M. Hu, E. M. Adkins, J. T. Hodges, H. Tran, H. N. Ngo, J.-M. Hartmann, S. Beguier, A. Campargue, R. J. Hargreaves, L. S. Rothman, and I. E. Gordon, Journal of Quantitative Spectroscopy and Radiative Transfer 347, 109596 (2025).
  44. [44]J. Schaefer and L. Monchick, Astronomy and Astrophysics 265, 859 (1992).
  45. [45]P. Wcisło, F. Thibault, M. Zaborowski, S. Wójtewicz, A. Cygan, G. Kowzan, P. Masłowski, J. Komasa, M. Puchalski, K. Pachucki, R. Ciuryło, and D. Lisak, Journal of Quantitative Spectroscopy and Radiative Transfer 213, 41 (2018).
  46. [46]P. Wcisło, F. Thibault, N. Stolarczyk, H. Jóźwiak, M. Słowiński, M. Gancewski, K. Stankiewicz, M. Konefał, S. Kassi, A. Campargue, Y. Tan, J. Wang, K. Patkowski, R. Ciuryło, D. Lisak, R. Kochanov, L. S. Rothman, and I. E. Gordon, Journal of Quantitative Spectroscopy and Radiative Transfer 260, 107477 (2021).
  47. [47]E. M. Adkins, S. N. Yurchenko, W. Somogyi, and J. T. Hodges, Journal of Quantitative Spectroscopy and Radiative Transfer 338, 109412 (2025).
  48. [48]M. Słowiński, F. Thibault, Y. Tan, J. Wang, A.-W. Liu, S.-M. Hu, S. Kassi, A. Campargue, M. Konefał, H. Jóźwiak, K. Patkowski, P. S. Żuchowski, R. Ciuryło, D. Lisak, and P. Wcisło, Physical Review A 101, 052705 (2020).
  49. [49]H. Jóźwiak, F. Thibault, N. Stolarczyk, and P. Wcisło, Journal of Quantitative Spectroscopy and Radiative Transfer 219, 313 (2018).
  50. [50]K. Stankiewicz, H. Jóźwiak, M. Gancewski, N. Stolarczyk, F. Thibault, and P. Wcisło, Journal of Quantitative Spectroscopy and Radiative Transfer 254, 107194 (2020).
  51. [51]H. Jóźwiak, F. Thibault, H. Cybulski, and P. Wcisło, The Journal of Chemical Physics 154, 054314 (2021).
  52. [52]J. Lamouroux, R. R. Gamache, A. L. Laraia, J.-M. Hartmann, and C. Boulet, Journal of Quantitative Spectroscopy and Radiative Transfer 113, 991 (2012).
  53. [53]J. Lamouroux, R. R. Gamache, A. L. Laraia, J.-M. Hartmann, and C. Boulet, Journal of Quantitative Spectroscopy and Radiative Transfer 113, 1536 (2012).
  54. [54]A. Cygan, D. Lisak, S. Wójtewicz, J. Domysławska, J. T. Hodges, R. S. Trawiński, and R. Ciuryło, Physical Review A 85, 022508 (2012).
  55. [55]A. Cygan, S. Wójtewicz, J. Domysławska, P. Masłowski, K. Bielska, M. Piwiński, K. Stec, R. S. Trawiński, F. Ożimek, C. Radzewicz, H. Abe, T. Ido, J. T. Hodges, D. Lisak, and R. Ciuryło, The European Physical Journal Special Topics 222, 2119 (2013).
  56. [56]P. Wcisło, I. E. Gordon, H. Tran, Y. Tan, S.-M. Hu, A. Campargue, S. Kassi, D. Romanini, C. Hill, R. V. Kochanov, and L. S. Rothman, Journal of Quantitative Spectroscopy and Radiative Transfer 177, 75 (2016).
  57. [57]E. M. Adkins and J. T. Hodges, Journal of Quantitative Spectroscopy and Radiative Transfer 280, 108100 (2022).
  58. [58]D. C. Benner, C. P. Rinsland, V. M. Devi, M. A. H. Smith, and D. Atkins, Journal of Quantitative Spectroscopy and Radiative Transfer 53, 705 (1995).
  59. [59]A. S. Pine and R. Ciurylo, Journal of Molecular Spectroscopy 208, 180 (2001).
  60. [60]V. H. Payne, B. J. Drouin, F. Oyafuso, L. Kuai, B. M. Fisher, K. Sung, D. Nemchick, T. J. Crawford, M. Smyth, D. Crisp, E. Adkins, J. T. Hodges, D. A. Long, E. J. Mawer, A. Merrelli, E. Lunny, and C. W. O’Dell, Journal of Quantitative Spectroscopy and Radiative Transfer 255, 107217 (2020).
  61. [61]H. Tran, J. T. Hodges, E. M. Adkins, E. C. Gross, J. Klemm, H. Fleurbaey, A. Campargue, and D. Mondelain, The Journal of Chemical Physics 164 (2026).
  62. [62]I. E. Gordon, L. S. Rothman, R. J. Hargreaves, F. M. Gomez, T. Bertin, C. Hill, R. V. Kochanov, Y. Tan, P. Wcisło, V. Y. Makhnev, P. F. Bernath, M. Birk, V. Boudon, A. Campargue, A. Coustenis, B. J. Drouin, R. R. Gamache, J. T. Hodges, D. Jacquemart, E. J. Mawer, A. V. Nikitin, V. I. Perevalov, M. Rotger, S. Robert, J. Tennyson, G. C. Toon, H. Tran, V. G. Tyuterev, E. M. Adkins, A. Barbe, D. M. Bailey, K. Bielska, L. Bizzocchi, T. A. Blake, C. A. Bowesman, P. Cacciani, P. Čermák, A. G. Császár, I. Denis, S. C. Egbert, O. Egorov, A. Y. Ermilov, A. J. Fleisher, H. Fleurbaey, A. Foltynowicz, T. Furtenbacher, M. Germann, E. R. Guest, J. J. Harrison, J.-M. Hartmann, A. Hijlťen, S.-M. Hu, X. Huang, T. J. Johnson, H. Jóźwiak, S. Kassi, M. V. Khan, F. Kwabia-Tchana, T. J. Lee, D. Lisak, A.-W. Liu, O. Lyulin, N. A. Malarich, L. Manceron, A. A. Marinina, S. T. Massie, J. Mascio, E. S. Medvedev, V. V. Meshkov, G. C. Mellau, M. Melosso, S. N. Mikhailenko, D. Mondelain, H. S. P. Müller, M. O’Donnell, A. Owens, A. Perrin, O. L. Polyansky, P. L. Raston, Z. D. Reed, M. Rey, C. Richard, G. B. Rieker, C. Röske, S. W. Sharpe, E. Starikova, N. Stolarczyk, A. V. Stolyarov, K. Sung, F. Tamasia, J. Terragni, V. G. Ushakov, S. Vasilchenko, B. Vispoel, K. L. Vodopyanov, G. Wagner, S. Wójtewicz, S. N. Yurchenko, and N. F. Zobov, Journal of Quantitative Spectroscopy and Radiative Transfer 353, 109807 (2026).
  63. [63]M. Słowiński, H. Jóźwiak, M. Gancewski, K. Stankiewicz, N. Stolarczyk, Y. Tan, J. Wang, A.-W. Liu, S.-M. Hu, S. Kassi, A. Campargue, K. Patkowski, P. S. Żuchowski, R. Ciuryło, F. Thibault, and P. Wcisło, Journal of Quantitative Spectroscopy and Radiative Transfer 277, 107951 (2022).
  64. [64]H. Jóźwiak, N. Stolarczyk, K. Stankiewicz, M. Zaborowski, D. Lisak, S. Wójtewicz, P. Jankowski, K. Patkowski, K. Szalewicz, F. Thibault, I. E. Gordon, and P. Wcisło, Astronomy and Astrophysics 687, A69 (2024).

Paper details

vlnc:W_kk9nj7qj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents