High sensitivity cavity ring down spectroscopy of N2O near 1.22 µm: (II) 14N216O line intensity modeling and global fit of 14N218O line positions

Author’s Accepted Manuscript High sensitivity Cavity Ring Down Spectroscopy of N2O near 1.22µm: (II) 14N216O line intensity modeling and global fit of 14N218O line positions S.A. Tashkun, V.I. Perevalov, E.V. Karlovets, S. Kassi, A. Campargue www.elsevier.com/locate/jqsrt

PII: DOI: Reference:

S0022-4073(16)30016-4 http://dx.doi.org/10.1016/j.jqsrt.2016.02.020 JQSRT5233

To appear in: Journal of Quantitative Spectroscopy and Radiative Transfer Received date: 14 January 2016 Revised date: 15 February 2016 Accepted date: 15 February 2016 Cite this article as: S.A. Tashkun, V.I. Perevalov, E.V. Karlovets, S. Kassi and A. Campargue, High sensitivity Cavity Ring Down Spectroscopy of N2O near 1.22µm: (II) 14N216O line intensity modeling and global fit of 14N218O line positions, Journal of Quantitative Spectroscopy and Radiative Transfer, http://dx.doi.org/10.1016/j.jqsrt.2016.02.020 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

High sensitivity Cavity Ring Down Spectroscopy of N2O near 1.22 µm: (II) 14N216O line intensity modeling and global fit of 14N218O line positions

S.A. Tashkun a,b, V.I. Perevalov a*, E.V. Karlovets b, S. Kassi c,d, A. Campargue c,d

a

Laboratory of Theoretical Spectroscopy, V.E. Zuev Institute of Atmospheric Optics, Siberian Branch, Russian Academy of Sciences,1, Academician Zuev Square, 634055 Tomsk, Russia b Laboratory of Quantum Mechanics of Molecules and Radiative Processes, Tomsk State University, 36, Lenina avenue, 634050 Tomsk, Russia c Univ. Grenoble Alpes, LIPhy, F-38000 Grenoble, France d CNRS, LIPhy, F-38000 Grenoble, France

Number of pages: 20 Number of tables: 7 Number of figures: 3

Keywords: Nitrous oxide; Line intensities

14

N216O;

14

N218O; Vibration-rotational transitions; Line positions;

* Corresponding author: E-mail address: [email protected] (Valery I. Perevalov). Tel.: +7 3822491794 fax: +7 3822492086

1

Abstract In a recent work (E.V. Karlovets et al. J Quant Spectrosc Rad Transfer 2016;169:36– 48), we reported the measurement and rovibrational assignments of more than 3300 transitions belonging to 64 bands of five nitrous oxide isotopologues (14N216O, 15

14

N15N16O,

N14N16O, 14N218O and 14N217O) in the high sensitivity CRDS spectrum recorded in the 7914-

8231 cm-1 spectral range. The assignments were performed by comparison with predictions of the effective Hamiltonian models developed for each isotopologue. In the present paper, the large amount of measurements from our previous work mentioned above and literature data are gathered to refine the modeling of the nitrous oxide spectrum in two ways: (i) improvement of the intensity modeling for the principal isotopologue, 14N216O, near 8000 cm1

from a new fit of the relevant effective dipole moment parameters, (ii) global modeling of

14

N218O line positions from a new fit of the parameters of the global effective Hamiltonian

using an exhaustive input dataset collected in the literature in the 12-8231 cm-1 region. The fitted set of 81 parameters allowed reproducing near 5,800 measured line positions with an RMS deviation of 0.0016 cm-1. The dimensionless weighted standard deviation of the fit is 1.22. As an illustration of the improvement of the predictive capabilities of the obtained effective Hamiltonian, two new 14N218O bands could be assigned in the CRDS spectrum in the 7915-8334 cm-1 spectral range. A line list at 296 K has been generated in the 0-10700 cm-1 range for 14N218O in natural abundance with a 10-30 cm/molecule intensity cutoff.

2

1. Introduction This contribution follows a recent work devoted to the measurement and rovibrational assignments of the high sensitivity absorption spectrum of nitrous oxide recorded by Cavity Ring Down Spectroscopy (CRDS) between 7915 and 8334 cm-1 [1]. The high sensitivity of the recordings (typical noise equivalent absorption of the spectra on the order of αmin ~ 2×10-11 cm-1) allowed for the detection of lines with intensities on the order of 5×10-29 cm/molecule. The rovibrational assignments of the resulting large number of new observations were performed on the basis of the predictions of the effective Hamiltonian (Heff) models developed for each isotopologue. Overall, more than 3300 transitions belonging to 64 bands of five isotopologues (14N216O,

14

N15N16O,

15

N14N16O,

14

N218O and

14

N217O)

could be assigned. In the case of the principal isotopologue,

14

N216O, 41 new bands belonging to the

ΔP=13 and ΔP=14 series of transitions were identified (the pseudo quantum number, P, is given by P= 2V1 + V2 + 4V3 where V1, V2, and V3 denote the normal mode vibrational quantum numbers). The comparison of their measured line intensities to the corresponding values predicted with existing sets of the effective dipole moment (Deff) parameters [2] showed important deviations indicating that the ΔP= 13 and ΔP= 14 Deff parameters have to be improved. In the present paper, we have performed new fittings of the line intensities for the ΔP=13 and ΔP=14 series of transitions using the CRDS line intensities of Ref. [1] and those collected from the literature [2-4]. A second issue evidenced from our analysis of the CRDS spectrum near 8000 cm-1 concerns the insufficient quality of the line positions of the

14

N218O isotopologue obtained

using the available Heff parameters from Ref. [5]. Due to the lack of experimental data, some important parameters were not determined and some predicted line positions deviated by more than 1 cm-1 from the CRDS values of Ref. [1]. The inclusion of the line positions of two newly measured

14

N218O bands is expected to improve significantly the set of the Heff

parameters of this isotopologue. In the present work, we have performed a new fit of the Heff parameters of

14

N218O to the line positions collected from the literature including the line

positions of these two new bands. The new set of parameters has allowed us to assign two additional

14

N218O bands in the CRDS spectrum of Ref. [1]. In the final fit, all four bands

assigned in the CRDS spectrum in the 7915-8334 cm-1 spectral range have been involved. Using this new set of Heff parameters of

14

N218O, the Deff parameters of this

isotopologue were fitted to the measured line intensities for several series of transitions with 3

ΔP up to 14. Finally, the Heff and Deff parameters have been used for the generation of a room temperature line list of the 14N218O isotopologue up to 10700 cm-1. 2. Effective operators approach The polyad model of Heff has been used for the global modeling of the line positions and intensities of the different isotopologues of the nitrous oxide molecule. The N2O polyad model has been suggested by Pliva [6] and developed by Teffo and Chedin [7]. A reduced model of this Hamiltonian was elaborated by Teffo et al. [8] and extended up to the sixthorder terms by Perevalov et al. [9]. The polyad model of Heff is based on the polyad structure of the N2O vibrational states resulting from the approximate relations between harmonic frequencies

ω1≈2ω2 and ω3≈4ω2.

(1)

As a result, the vibrational states are grouped into vibrational polyads corresponding to different pseudo quantum numbers P = 2V1 + V2 + 4V3. The Heff matrix elements are calculated in the basis of harmonic oscillator and rigid symmetric top rotor eigenfunctions VV 1 2l2V3 J = VV 1 2l2V3 JK = l2 ,

(2)

where J is the total angular momentum quantum number, K is quantum number of the projection of the total angular momentum on the molecular-fixed z-axis, and l2 is the vibrational angular momentum quantum number. In order to reduce the size of the effective Hamiltonian matrix, Wang-type basis functions are generally used:

VV 1 2l2V3 J e =

1 ( VV 1 2l2V3 JK = l2 + e VV 1 2 - l2V3 JK = -l2 ) 2

VV 1 2 0V3 J e = 1 = VV 1 2 0V3 JK = 0

(l2 = 0) ,

( l2 ¹ 0) ,

(3a)

(3b)

where ε= 1 and -1 correspond to the so-called e and f levels, respectively. In this basis, Heff matrix splits into independent blocks, each block being defined by three labels: P, the Wang parity, C = {e, f}, and J. Thus, the eigenvalues of Heff can be unambiguously labeled by four labels (P, C, J, i), where i is the ranking index of the eigenvalues in a (P, C, J) block. These labels are called the generalized nomenclature of an energy level. Our model takes into account explicitly all intrapolyad resonance anharmonic and anharmonic+l-type interactions up to the sixth order of perturbation theory. All Coriolis interactions are interpolyad interactions within the framework of this model. They are 4

assumed to be accounted for by the effective values of the Heff parameters. The detailed presentation of the used effective Hamiltonian together with its matrix elements and notations for the parameters are given in our paper [9]. In the process of the least-squares fits of

14

N216O data [9] and analysis of the

experimental spectra [1,3,10-20], several interpolyad resonance Coriolis and anharmonic perturbations were evidenced. An analysis of these interactions showed that they become pronounced for high-excited vibrational states and affect rather weak bands. Perturbed lines affected by these interactions were removed from the input data file. The polyad model was found to be accurate for the majority of non-weak measured 14N216O bands in the 0-9700 cm-1 range [9], i.e. for the bands with line intensities larger or equal to 10-26 cm/molecule at 296 K. Line intensities can be calculated using the eigenfunctions of the effective Hamiltonian but, for this purpose, it is necessary to use an effective dipole moment operator instead of a dipole moment operator. The former is obtained from the dipole moment operator by the same unitary transformation as the effective Hamiltonian from the vibration-rotation Hamiltonian. This approach is presented in Refs. [21,22]. Within the framework of polyad model, all vibrational bands are distributed among different series of transitions, each series corresponding to a different variation of the polyad number, ΔP=P’- P’’, where P’ and P’’ are the polyad numbers of upper and lower states, respectively. Consequently, for each DP series, a set of effective dipole moment parameters has to be determined from a fit of the available experimental line intensities. 3. The DP= 13 and 14 effective dipole moment parameters of 14N216O The CRDS line intensities of the 14N216O bands belonging to the ΔP= 13 and ΔP= 14 series of transitions near 8000 cm-1 [1] were gathered with literature data [2-4] and used as input data to improve the corresponding sets of Deff parameters. The computer code used for the fit is described in Ref. [23]. The aim of the fit is to minimize the value of the dimensionless weighted standard deviation, χ, defined as

å (( S N

c=

i =1

obs i

- Sicalc ) / d i

N -n

)

2

,

(4)

where Siobs and Sicalc are, respectively, measured and calculated intensity values for the ith line, d i = ( Siobss i ) /100% is the absolute measurement error of ith line intensity, σi is the measurement error in %, N is the number of fitted line intensities, and n is the number of

5

adjusted effective dipole moment parameters. Another value used to measure the quality of a fit is the root mean squares deviation defined according to the equation

å (( S N

RMS =

obs i

i =1

- Sicalc ) / Siobs N

)

2

´100%.

(5)

The characteristics of the input data and the results of the line intensity fits are given in Table 1. The partition function was taken from Ref. [24]. The measurement errors of the line intensities of very weak bands studied in our previous paper [1] can exceed the 20% uncertainty given in Ref. [1]. Consequently, about 26% of the line intensities showing very large residuals were excluded from the fits. The majority of the excluded intensities belong to extremely weak hot bands originating from the third polyad. The fitted sets of the effective dipole moment parameters are presented in Table 2. In Fig. 1, the relative differences between measured and calculated line intensities are plotted versus the intensity for the ΔP=14 series of transitions included in the fit. This figure shows that Toth’s line intensities [3] for the strongest 1003-0000 band are about 10% underestimated. A similar conclusion concerning Toth’s line intensities [3] was drawn in Refs. [4,17] in other spectral regions. Table 1. Experimental data and results of the 14N216O line intensity fits.

DP

Wavenumber range, cm-1

13

7018-8287

14

7572-8468

a) b)

Number of lines a)

Number of bands

Jmax

Experimental uncertainty

16(18) 436(500) 44(54) 21(21) 432(456) 49(51) 1896(2578)

1 4 2 1 10 1 29

25 50 42 33 50 64 76

20%-45% Ref. [4] 3%-40% Ref. [2] 20% Ref. [1] 5% Ref. [3] 1%-99% Ref. [4] 4%-7% Ref. [2] 5%-20% Ref. [1]

c

1.91

1.17

RMS (%) 12.0 12.0 34.1 10.5 12.0 8.9 16.4

n b)

6

14

Number of fitted line intensities (number of measured line intensities). n is the number of adjusted parameters.

6

Fig.1. Relative differences between measured intensities (Toth [3] - green full circles, Daumont et al. [4] – red stars, CRDS [1,2] – blue open squares) and calculated line intensities for the ΔP=14 series of transitions in 14N216O. Table 2. Effective dipole moment parameters for ΔP=13 and ΔP=14 series of transitions of 14 N216O. Parameter ΔP= 13 M κ2 aK bJ M M ΔP= 14 M κ1 κ2 bJ M M κ1 κ2 M

ΔV1 ΔV2 ΔV3 Δl2

Value a

0 0 0 0 6 3

1 1 1 1 1 7

3 3 3 3 0 0

1 1 1 1 1 1

0.6628(26) × 10-5 0.299(55) × 10-1 0.208(21) × 10-1 -0.934(15) × 10-2 0.548(19) × 10-7 -0.600(43) × 10-8

1 1 1 1 0 3 3 3 2

0 0 0 0 2 0 0 0 2

3 3 3 3 3 2 2 2 2

0 0 0 0 0 0 0 0 0

0.11519(10) × 10-3 -0.834(16) × 10-1 0.140(55) × 10-1 -0.65(31) × 10-4 -0.6234(33) × 10-5 -0.14498(61) × 10-4 -0.462(26) -0.709(97) × 10-1 0.113(10) × 10-5 7

M M M bJ M

2 1 5 5 4

2 4 0 0 2

2 2 1 1 1

0.126(37) × 10-8 0.430(30) × 10-6 0.3046(19) × 10-5 0.73(16) × 10-3 -0.131(20) × 10-6

2 0 0 0 0

a

The parameters M are given in Debye while the κ1, κ2, aK and bJ parameters are dimensionless. Only relative signs of the M parameters within a given series of transitions are determined. The numbers in parentheses correspond to one standard deviation in units of the last quoted digit.

4. 14N218O line position fit In Ref. [1], the

14

N218O predictions used for the assignment of the CRDS spectrum

were obtained using the effective Hamiltonian parameters published by Vlasova et al. [5], ten years ago. Vlasova et al. [5] have determined a set of 59 parameters from a fit of 4265 measured line positions taken from Refs. [18, 25-27]. In the present fit we included in addition to the file of the input data used in Ref. [5] the line positions published in Refs. [1,19-20,28-33]. Using the results of this preliminary fit two additional

14

N218O bands have

been assigned in the CRDS spectrum of Ref. [1] in the 7915-8334 cm-1 range. The band parameters of these two new bands are presented in the next Section 5. The line positions of these two new bands were included into the final fit. The resulting data file contains 5798 entries of 13 sources and covers the 12 - 8231 cm-1 spectral range (see Table 3). Heff parameters were fitted to the measured line positions with the GIP computer code [34]. The goal of the fitting is to minimize the dimensionless standard deviation defined according to the usual formula 2

æ obs - calci ö c = åç i ÷ / ( N - n), di i è ø

(6)

where obsi and calci are observed and calculated values of the i-th line position, respectively, δi is the measurement uncertainty, N is the number of fitted values, and n is the number of adjusted Heff parameters. With 81 parameters we were able to reach χ= 1.22. The reached χ value indicates that the global fitting of vibrational-rotational line positions has been performed near experimental accuracy. This result is confirmed by the value of 0.0016 cm-1 obtained for the root mean-squares deviation between calculated and experimental positions. In Fig. 2 the (obsi - calci) residuals are plotted versus the wavenumber.

8

Table 3. Experimental data and statistics of the global fit of 14N218O line positions. Reference Andreev et al [27] Morino et al. [30] Drouin et al. [31] Toth [25] Toth [28,29] Krell and Sams [26] Liu et al. [18] Liu et al. [19] Liu et al. [32] Liu et al. [20] Lu et al. [33] Karlovets et al. [1] This work

Setup a MW MW MW FTS FTS FTS CRDS CRDS CRDS CRDS CRDS CRDS CRDS

Spectral domain (cm-1) 12.7-18.2 22.1 30.0-39.5 542.1-4671.9 1108.3-2257.1 2266.9-2435.0 6132.1-6722.3 6007.7-6796.1 6969.5-8230.8 7652.0-7911.1 6952.5-7652.0 7918.2-8150.2 7915.9-8230.8

Jmax b 23 28 50 77 74 59 51 54 37 55 57 41 46

Uncertainty c (10-3cm-1) 0.001 0.001 0.003 1.0 0.5 5.0 2.0 2.0 1.0 1.5 1.0 1.0 1.0

Nfit d 20 1 8 4238 295 130 184 311 49 172 210 105 75

RMS e (10-3cm-1) 0.002 0.003 0.004 1.0 0.32 2.4 2.0 3.7 1.5 2.5 1.3 4.6 3.6

a

FTS - Fourier transform spectrometer, MW - Microwave spectrometer, CRDS - cavity ring-down spectrometer. b Maximal value of the rotational quantum number. c Uncertainties of measured line positions as given in references. d Number of fitted lines. e RMS of residuals.

Fig.2. Residuals between observed and calculated line positions of 14N218O versus wavenumber.

9

One can see from Fig. 2 that the majority of the residuals are randomly scattered around the Y=0 axis. The exceptions are several bands from the line list of Toth [25] which show residuals with systematic parabolic trends but with small amplitude. The above cited line list contains the calculated values of the line positions. The calculations were performed using the spectroscopic constants for a given band fitted to the observed line positions. Those calculations did not take into account any resonance interaction but they were used for the extrapolation calculations of the line positions. Fig. 2 shows that among the

14

N218O bands

observed up to now, there is no manifestation of interpolyad resonance perturbations of the energy levels and that the polyad model of Heff is adequate to the fitted data. This is consistent with the fact that, in the case of the main isotopologue,

14

interpolyad interaction are all weak and then unobserved for

N216O, the bands affected by

14

N218O. The percentage of the

residuals being within ±0.0001 cm-1, ±0.0005 cm-1, ±0.001 cm-1, and ±0.005 cm-1 bounds is 11%, 46%, 68%, and 98%, respectively. The statistics of the global fit of

14

N218O line

positions are included in Table 3. The values of the fitted Heff parameters are presented as the Supplementary Material I. 5. New 14N218O bands assigned near 8000 cm-1 Using the improved set of

14

N218O effective Hamiltonian parameters, new spectra

predictions were performed in the 7915-8334 cm-1 spectral range studied in Ref. [1]. The improved quality of the predictions allowed us assigning two very weak new bands, 50010000 ((14 0 14)-(0 0 1) in cluster labeling notation) and 2312-0110 ((15 1 9)-(1 1 1)), among the set of 1989 lines which were left unassigned in Ref. [1]. The fractions of the corresponding upper states in the normal mode basis are given in Table 4. The spectroscopic constants of the upper states of the newly observed bands were fitted to the observed line positions using the following expression for the energy levels Fv ( J ) = Gv + Bv J ( J + 1) - Dv J 2 ( J + 1) 2 + H v J 3 ( J + 1) 3

,

(8)

where Gv is the vibrational term value, Bv is the rotational constant, Dv and Hv are centrifugal distortion constants, J is the angular momentum quantum number. The spectroscopic constants for the lower states were taken from Ref. [3]. The fitted spectroscopic constants are listed in Table 5. The measured line positions and intensities of these two bands are presented in Supplementary Material II.

10

Table 4. Vibrational assignment and fractions of the upper states in the normal mode basis for the newly assigned bands of 14N218O. ∆P 14 14

Band 5001-0000 2312-0110

(P,l2,i) a (14 0 14) (15 1 9)

Gv (cm-1) 8214.32 8550.19

Basis states b 5001/2601 2312/3112

% Fraction c 31/19 38/25

a

Upper state cluster labeling notation: (P=2V1+V2+4V3, l2,i): i is the order number within the cluster increasing with the energy. b Basis states with modulo of expansion coefficients larger than 0.4. c Squares of the expansion coefficients of the vibrational state for the dominant basis states appearing in the preceding column.

11

Bv 0.395577895 0.395759697 0.396471172 Bv 0.3867352(13) 0.3858630(18) 0.3872609(31)

Gv 0.0 584.22466 584.22466 Gv 8214.31835(21) 8550.19257(52) 8550.19165(66)

Dv´107 0.901(18) 1.052(11) 0.799(29)

1.583456 1.600230 1.610773

Dv×107

Hv´1012 6.52(61) ν0 c 8214.31835(21) 7965.96791(52) 7965.96699(66)

Notes a V1V2 l2 V3 corresponding to the maximum value of the modulo of the expansion coefficients of the eigenfunction. b Cluster labeling notation: (P=2V1+V2+4V3, l2,i): i is the order number within the cluster increasing with the energy. c ν0 is the band center d Observed branch with maximum value of the total angular momentum quantum number. e n is the number of transitions included in the fit; N is the number of assigned rotational transitions. f Root Mean Squares of residuals of the spectroscopic constants fit is given in 10-3 cm-1.

Lower state constants [3] State V1V2 l2V3 a (P l2 i ) b 0000e (0 0 1) 0110e (1 1 1) 0110f Upper state constants V1V2 l2V3 (P l2 i ) 5001e-0000e (14 0 14) 2312e-0110e (15 1 9) 2312f-0110f

Table 5. Spectroscopic constants (in cm-1) of the newly assigned bands of 14N218O.

observed lines d P46/R33 P42/R28 P33/R22

n/N e 42/42 18/19 15/16

12

rms f 0.63 0.93 1.00

6. Room temperature line list for 14N218O The natural abundance of 14N218O isotopologue is 1.986×10-3. It is the fourth most abundant isotopologue of N2O after

14

N216O,

14

N15N16O and

15

N14N16O. The

14

N218O data included in the

HITRAN [35] and GEISA [36] databases were taken from Toth’s SISAM list of N2O line parameters [25]. SISAM contains parameters of 4250

14

N218O lines obtained by FTS in the 542-

4672 cm-1 spectral range. The minimal intensity is about 2.0×10-25 cm/molecule. Most of Toth's line parameters were calculated using fitted spectroscopic constants, band intensities and parameters of Herman-Wallis factors. According to the HITRAN uncertainty codes, the uncertainties of the SISAM line positions and line intensities are declared to be equal or better than 0.001 cm-1 and 5%, respectively. Table 6. Summary of the 14N218O line intensity fits. Experimental data were taken from Ref. [25] and Ref. [1] for the ΔP= 1-6, 8 and ΔP= 14 series, respectively. DP

1 2 3 4 5 6 8 14

Approximate spectral Nline a Nband b Jmax c Npar d RMS e, % range (cm-1) 542-631 515 3 59 2 5.1 1111-1296 969 7 70 5 4.1 1811-1858 81 1 36 2 1.4 2139-2540 1580 10 77 4 3.3 2749-2815 127 1 46 1 2.5 3291-3764 781 7 63 4 2.3 4351-4671 195 3 47 3 2.4 7915-8230 102 4 41 3 12.7

Notes Nline - number of fitted line intensities. b Nband – number involved bands. c Jmax - maximum value of the angular momentum quantum number. d Npar - number of adjusted parameters. e RMS – root mean squares of the residuals between input and fitted line intensities. a

In this section, we present the calculated line list for

14

N218O computed within the

framework of the method of effective operators. This line list contains position, 296 K intensity, lower state energy and rovibrational assignment. The line positions and lower state energies were calculated using the Heff parameters obtained in Section 4. In order to calculate the line intensities, one needs the Deff parameters for each ΔP series of bands. To our knowledge, there are only two sources of

14

N218O line intensities in the literature. Toth's SISAM line list contains

14

N218O

intensities of lines belonging to the ΔP= 1-6,8 series [25]. Karlovets et al. [1] reported CRDS intensities for two

14

N218O bands belonging to the ΔP= 14 series and two additional bands of this

series have been measured in the present work from the CRDS spectra of Ref. [1]. Using the

13

available experimental line intensities and the eigenfunctions of the effective Hamiltonian described in Section 4, the principal effective dipole moment parameters have been fitted for the ΔP= 1-6,8, and 14 series. The statistics of these fits is given in Table 6. The sets of the effective dipole moment parameters are presented in Table 7. Table 7. Effective dipole moment parameters of 14N218O. Parameter ΔP=1 M bJ ΔP=2 M κ1 κ2 M M ΔP=3 M bJ ΔP=4 M M M M ΔP=5 M ΔP=6 M M M M ΔP=8 M M M ΔP=14 M M M

ΔV1

ΔV2

ΔV3

Δl2

Value a

0 0

1 1

0 0

1 1

0.4786(5)×10-1 0.247(5)×10-2

1 1 1 0 2

0 0 0 2 -2

0 0 0 0 0

0 0 0 0 0

0.1398(2) -0.268(3) 0.14(1)×10-1 0.395(3)×10-2 0.63(2)×10-2

1 1

1 1

0 0

1 1

0.4135(3)×10-2 -0.895(6)×10-2

0 2 1 0

0 0 2 4

1 0 0 0

0 0 0 0

0.2581(1) -0.1152(3)×10-1 0.25(1)×10-3 -0.81(3)×10-4

0

1

1

1

-0.6171(7)×10-2

1 0 3 2

0 2 0 2

1 1 0 0

0 0 0 0

-0.3861(2)×10-1 -0.243(5)×10-3 0.1148(4)×10-2 -0.119(3)×10-3

2 1 0

0 2 0

1 1 2

0 0 0

-0.1877(4)×10-2 -0.217(3)×10-3 0.5554(6)×10-2

3 5 4

0 0 2

2 1 1

0 0 0

-0.180(3)×10-4 -0.1123(6)×10-4 0.200(9)×10-5

Note a The parameters M are given in Debye while the κ1, κ2, and bJ parameters are dimensionless. Only relative signs of the M parameters within a given series of transitions are determined. The numbers in parentheses correspond to one standard deviation in units of the last quoted digit.

For the generation of the 14N218O line list in the 1-10,700 cm-1 range, we need in addition the effective dipole moment parameters for the ΔP =0,7,9-13,16, and 18 series of transitions. The line intensities of the bands belonging to the ΔP =15 and 17 series are below the chosen intensity cutoff (see below). For the ΔP series of 14N218O bands for the missing Deff parameters, the Deff parameters of the principal isotopologue were adopted. This approximation was found to be valid for the fitted 14

values of the vibrational transition dipole moments squared for the 0006-0000 band of four nitrous oxide isotopologues [37,38]. These references demonstrate that if an isotopic substitution does not change permutation symmetry of nitrous oxide molecule then the effective dipole moment parameters do not change significantly. Similar situation occurs for the lower order Deff parameters of CO2 which are practically independent on isotopologue in the case of a symmetric substitution [39]. The needed Deff parameters of 14N216O were obtained in Refs. [4,40-42]. Daumont et al. [4,40] have fitted the effective dipole moment parameters to the measured line intensities of the ΔP=79,10-14,16,18 series of transitions. For the ΔP=0 series, the fitted parameters of the effective dipole moment operator are reported in Ref. [41]. As a result, a

14

N218O line list was generated at 296 K with an intensity cut off of 10-30

cm/molecule (including the 1.986×10-3 isotopic abundance factor). The partition function was taken from Ref. [43]. A graphical overview of the line list is given in Fig. 3. The overall number of the lines is 214869. We estimate the uncertainties of the line list positions to be between 0.001 cm-1 and 0.02 cm-1 and the uncertainties of the line list intensities of the strongest bands of each ΔP series to be about 20%. The uncertainties of the line intensities of weak bands can be considerably larger. The

14

N218O line list is provided as Supplementary Material III. The

14

N218O data provided by the

HITRAN database are included on Fig. 3. They correspond to bands with intensity larger than 2×10-25 cm/molecule. Note that the rotational transitions (ΔP=0) are missing in the HITRAN database.

Fig. 3. Graphical overview of the database are highlighted.

14

N218O line list. The

14

N218O data provided by the HITRAN

15

7. Conclusion In the present work, the large amount of new data derived from the CRDS spectrum in the 7915-8334 cm-1 spectral range [1] has been gathered with literature data to improve the modeling of the nitrous oxide spectrum using the global effective operator approach: (i) a new fit of the ΔP=13 and ΔP=14 effective dipole moment parameters was performed for the principal isotopologue, 14

N216O, near 8000 cm-1, (ii) the global modeling of the line positions of the

14

N218O minor

isotopologue was improved by a new fit of the parameters of the effective Hamiltonian using 5,800 measured line positions collected in the literature in the 12-8231 cm-1 region. The fitted set of 81 parameters allowed reproducing the input dataset with an RMS deviation of 0.0016 cm-1. The dimensionless weighted standard deviation of the fit is 1.22. This means that the used polyad model of effective Hamiltonian satisfactory reproduces all known measured line positions. But it is necessary to emphasize that contrary to the main isotopologue for which weak bands affected by interpolyad interaction were observed, only strong bands of the

14

N218O isotopologue have been

detected up to now. The improvement of the predictive abilities of the new model allowed assigning two new

14

N218O bands in the CRDS spectrum of Ref. [1]. The corresponding line parameters are

reported in the present paper. A line list at 296 K including spectroscopic labeling of the energy levels has been generated for

14

N218O in natural abundance in the 0-10700 cm-1 range. The intensity cutoff is 10-30

cm/molecule at 296 K. The line list is based on the polyad model of effective Hamiltonian fitted in this work and the parameters of the effective dipole moment operators of

14

N218O or 14N216O. The

accuracy of the line positions are estimated to be between 0.001 cm-1 and 0.02 cm-1, whereas the accuracy of the line intensities of strong bands is estimated to be about 20%. The uncertainties of the line parameters of weak bands can be considerably larger.

Acknowledgement This work is jointly supported by CNRS (France) and RFBR (Russia) in the frame of the International Associated Laboratory SAMIA. This project takes place in the frame of the LabexOSUG@2020 (ANR10 LABX56).

16

References 1. Karlovets EV, Campargue A, Kassi S, Perevalov VI, Tashkun SA. High sensitivity Cavity Ring Down Spectroscopy of N2O near 1.22 µm: (I) Rovibrational assignments and band-byband analysis. J Quant Spectrosc Radiat Transfer 2016;169:36–48. 2. Karlovets EV, Lu Y, Mondelain D, Kassi S, Campargue, Tashkun SA, Perevalov VI, High sensitivity CW-Cavity Ring Down Spectroscopy of N2O between 6950 and 7653 cm−1 (1.44–1.31 μm): II. Line intensities. J Quant Spectrosc Radiat Transfer 2013;117:81–7. 3. Toth RA. Line positions and strengths of N2O between 3515 and 7800 cm-1. J Mol Spectrosc 1999;197:158-87. 4. Daumont L, Vander Auwera J, Teffo JL, Perevalov VI, Tashkun SA. Line intensity measurements in 14N216O and their treatment using the effective dipole moment approach. II. The 5400–11,000 cm-1 region. J Quant Spectrosc Radiat Transfer 2007;104:342–56. 5. Vlasova AV, Perevalov BV, Tashkun SA, Perevalov VI. Global fittings of the line positions of the rare isotopic species of the nitrous oxide molecule. In: Proceedings of the XVth symposium on high-resolution molecular spectroscopy, Nizhny Novgorod, Russia, vol. 6580. SPIE; 2006. p. 658007. 6. Pliva J. Molecular constants of nitrous oxide, 14N216O. J Mol Spectrosc 1968;27:461–88. 7. Teffo JL and Chedin A, Internuclear Potential and Equilibrium Structure of the Nitrous Oxide Molecule from Rovibrational Data. J Mol Spectrosc 1989;135:389-409. 8. Teffo JL, Perevalov VI, Lyulin OM, Reduced effective Hamiltonian for global treatment of rovibrational energy levels of nitrous oxide. J Mol Spectrosc 1994;168:390–403. 9. Perevalov VI, Tashkun SA, Kochanov RV, Liu AW, Campargue A, Global modeling of the line positions of

14

N216O within the framework of the polyad model of effective

Hamiltonian. J Quant Spectrosc Radiat Transf 2012;113:1004-12. 10. Amiot C, Guelachvili G. Vibration-rotation bands of 14N216O: 1.2 micron-3.3 micron region. J Mol Spectrosc 1974;51:475-91. 11. Amiot C. Vibration-rotation bands of

14

N15N16O and

15

N14N16O: 1.6–5.7 μm region. J Mol

Spectrosc 1976;59:191-208. 12. Toth RA. Line strengths (900-3600 cm-1), self-broadening linewidths, and frequency shifts (1800-2360 cm-1) of N2O. Appl Opt 1993;32:7326-65. 13. Campargue A, Weirauch G, Tashkun SA, Perevalov VI, Teffo JL. Overtone spectroscopy of N2O between 10000 and 12000 cm-1: A test of the polyad approach. J Mol Spectrosc 2001:209;198-206. 17

14. Bertseva E, Kachanov AA, Campargue A. Intracavity laser absorption spectroscopy of N2O with a vertical external cavity surface emitting laser. Chem Phys Lett 2002:351;18-26. 15. Ding Y, Perevalov VI, Tashkun SA, Teffo JL, Bertseva E, Campargue A. Weak overtone transitions of N2O around 1.05 μm by ICLAS-VECSEL. J Mol Spectrosc 2003:220;80-6. 16. Bertseva E, Campargue A, Perevalov VI, Tashkun S.A. New observations of weak overtone transitions of N2O by ICLAS-VeCSEL near 1.07 μm. J Mol Spectrosc 2004:226;196-200. 17. Wang L, Perevalov VI, Tashkun SA, Gao B, Hao LY, Hu SM. Fourier transform spectroscopy of N2O weak overtone transitions in the 1–2 μm region. J Mol Spectrosc 2006:237;129-36. 18. Liu AW, Kassi S, Malara P, Romanini D, Perevalov VI, Tashkun SA, Hu SM, Campargue A. High sensitivity CW-cavity ring down spectroscopy of N2O near 1.5l μm (I). J Mol Spectrosc 2007:244;33-47. 19. Liu AW, Kassi S, Perevalov VI, Tashkun SA, Campargue A. High sensitivity CW-cavity ring down spectroscopy of N2O near 1.5l μm (II). J Mol Spectrosc 2007:244;48-62. 20. Liu AW, Kassi S, Perevalov VI, Tashkun SA, Campargue A. High sensitivity CW-Cavity Ring Down Spectroscopy of N2O near 1.28 μm. J Mol Spectrosc 2011:267;191-9. 21. Lyulin OM, Perevalov VI, Teffo JL. Effective dipole moment and band intensities of nitrous oxide. J Mol Spectrosc 1995;174:566–80. 22. Lyulin OM, Perevalov VI, Teffo JL. Fitting of line intensities using the effective operator approach: The 4 µm region of 14N216O. J Mol Spectrosc 1996;180:72–74. 23. Tashkun SA, Perevalov VI, Teffo JL, Tyuterev VlG. Global fit of

12 16

C O2 vibrational-

rotational line intensities using the effective operator approach. J Quant Spectrosc Radiat Transfer 1999;62:571–98. 24. Fischer J, Gamache RR, Goldman A, Rothman LS, Perrin A. Total internal partition sums for molecular species in the 2000 edition of the HITRAN database. J Quant Spectrosc Radiat Transfer 2003;82:401-12. 25. R.A. Toth, http://mark4sun.jpl.nasa.gov/n2o.html 26. Krell JM, Sams RL. Vibration-rotation bands of nitrous oxide: 4.1 micron region. J Mol Spectrosc 1974;51:492–507. 27. Andreev BA, Burenin AV, Karyakin EN, Krupnov AF, Shapin SM. Submillimeter wave spectrum and molecular constants of N2O. J Mol Spectrosc 1976;62:125–148. 28. Toth RA. Frequences of N2O in the 1100- to 1440 cm-1 region. J Opt Soc Am B 1986;3:1263-81.

18

29. Toth RA. N2O vibration-rotation parameters derived from measurements in the 900-1090and 1580-2380- cm-1 regions. J Opt Soc Am B 1987;4:357-74. 30. Morino I, Fabian M, Takeo H, Yamada KMT. High-J rotational transitions of NNO measured with the NAIR terahertz spectrometer. J Mol Spectrosc 1997;185:142–6. 31. Drouin BJ, Maiwald FW. Extended THz measurements of nitrous oxide, N2O. J Mol Spectrosc 2006;236:260–2. 32. Liu AW, Kassi S, Perevalov VI, Hu SM, Campargue A. High sensitivity CW-cavity ring down spectroscopy of N2O near 1.5 μm (III). J Mol Spectrosc 2009;254:20–7. 33. Lu Y, Mondelain D, Liu AW, Perevalov VI, Kassi S, Campargue A. High sensitivity CWCavity Ring Down Spectroscopy of N2O between 6950 and 7653 cm-1 (1.44–1.31μm): I. Linepositions. J Quant Spectrosc Radiat Transfer 2012;113:749-62. 34. Tashkun SA, Perevalov VI, Teffo JL, Rothman LS, Tyuterev VlG. Global fitting of 12C16O2 vibrational-rotational line positions using the effective Hamiltonian approach. J Quant Spectrosc Radiat Transfer 1998;60:785–801. 35. Rothman LS, Gordon IE, Babikov Y, Barbe A, Benner DC, Bernath PF, et al. The HITRAN2012 molecular spectroscopic database. J Quant Spectrosc Radiat Transfer 2013;130:4–50. 36. Jacquinet-Husson N, Crepeau L, Armante R, Boutammine C, Chedin A, Scott NA, et al. The 2009 edition of the GEISA spectroscopic database. J Quant Spectrosc Radiat Transfer 2011;112:2395–445. 37. Song K-F, Gao B, Liu A-W, Perevalov VI, Tashkun SA, Hu S-M. Cavity ring-down spectroscopy of the 6ν3 bands of

15

N substituted N2O. J Quant Spectrosc Radiat Transfer

2010;111:2370-81. 38. Milloud R, Perevalov VI, Tashkun SA, Campargue A, Rotational analysis of 6ν3 and 6ν3+ν2ν2 bands of 14N216O from ICLAS spectra between 12760 and 12900 cm-1. J Quant Spectrosc Radiat Transfer 2011;112:553–7. 39. Karlovets EV, Perevalov VI. Calculation of the carbon dioxide effective dipole moment parameters of the qJ and q2J types for rare isotopologues Atmos Ocean Opt 2011;24:101–6. 40. Daumont L, Vander Auwera J, Teffo JL, Perevalov VI, Tashkun SA. Line intensity measurements in 14N216O and their treatment using the effective dipole moment approach. I. The 4300 - to 5200 - cm-1 region. J Mol Spectrosc 2001;208:281–91. 41. Aenchbacher W, Naftaly M, Dudley R. Line strengths and self-broadening of pure rotational lines of nitrous oxide measured by teraherz time-domain dpectroscopy. J Opt Soc Am B 2010;27:1717-21. 19

42. Tashkun SA, Perevalov VI, Lavrentieva NN. NOSD-1000, the high-temperature nitrous oxide

spectroscopic

databank.

J

Quant

Spectrosc

Radiat

Transfer

(2015),

http://dx.doi.org/10.1016/j.jqsrt.2015.11.014i. 43. Laraia AL, Gamache RR, Lamouroux J, Gordon IE, Rothman LS. Total internal partition sums to support planetary remote sensing. Icarus 2011;215:391–400.

20

Highlights Line parameters of two new 14N218O bands centered at 7966 cm-1 and at 8214 cm-1. Refined sets of the 14N216O effective dipole moment parameters for ΔP=13,14 series. Global modeling of 14N218O line positions and intensities in the 12-8231 cm-1 range. 5,800 observed of 14N218O line positions reproduced with RMS=0.0016 cm-1. List of 14N218O line parameters in the 0-10700 cm-1 spectral range.

21