o. naumenko et al- cavity ring-down spectroscopy of h2^17-o in the range 16570–17125 cm^-1

8/3/2019 O. Naumenko et al- Cavity ring-down spectroscopy of H2^17-O in the range 1657017125 cm^-1

1/7

Cavity ring-down spectroscopy of H217O

in the range 1657017125 cm1

O. Naumenko a, M. Sneep b, M. Tanaka c, S.V. Shirin d, W. Ubachs b,*, J. Tennyson c

a Russian Academy of Science, Institute of Atmospheric Optics, Tomsk, 634055, Russiab Laser Centre, Department of Physics and Astronomy, Vrije Universiteit, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands

c Department of Physics and Astronomy, University College London, London WC1E 6BT, UKd Russian Academy of Science, Institute of Applied Physics, Nizhnii Novgorod 603950, Russia

Received 21 September 2004; in revised form 16 February 2006Available online 29 March 2006

Abstract

Following previous investigations on H216O and H2

18O by cavity ring-down spectroscopy, this method has now been applied to inves-tigate the energy region of the 5m polyad in the absorption spectrum of H2

17O. In the range 1657017125 cm1, the highest energy rangeinvestigated for the H2

17O isotopologue so far, 516 lines are attributed to H217O and assigned from a newly generated line list.

2006 Elsevier Inc. All rights reserved.

PACS: 33.20.Kf; 33.15.Mt; 92.60.Jq

Keywords: Water vapor; Visible spectra; Isotopologues; Cavity ring-down; Water vapor spectrum; Atmosphere

1. Introduction

The absorption spectrum of water vapor in the 1657017125 cm1 energy range covers the entire 5m polyad,which lies in a window used for remote sensing of this mol-ecule, which is considered the most important greenhousegas. Retrieval of water vapor column densities, using thisrather weak absorption feature on a global scale has beendemonstrated with data obtained by the satellite instru-ment SCIAMACHY [1,2]. Naus et al. [3] studied the absorptionspectrum of natural water (containing 99.732% H2

16O,

0.200% H218

O, 0.037% H217

O, and 0.031% HDO, see also[4]) in the frequency region 1655518000 cm1 employingthe laser-based technique of cavity ring-down (CRD) spec-troscopy. Due to the good signal-to-noise ratio provided bythe highly sensitive CRD method 1830 lines were found ofwhich 800 were not included in HITRAN 96, the databaseprevailing at the time. Of the set of weak lines 111 were giv-en an assignment from a comparison with first principles

calculations. Subsequently from a 18O isotopically enrichedwater sample a narrower energy region, still covering themost intense part of the 5m polyad, was investigated forH2

18O, yielding 596 lines belonging to H218O, of which

375 could be assigned based on the production of a newtheoretical line list [5]. Here, in a continuation of this work,we report of an investigation of the H2

17O spectrum, againcovering the 5m polyad, using the same CRD setup at theLaser Centre VU Amsterdam. From a comparison withspectra of H2

16O and H218O 516 lines are attributed to

H217O and assigned on the basis of a newly produced line

list. After previous investigations in the mid-infraredreporting on H217O (000) and (010) rotational states [6],

overtone rovibrational transitions were investigated in theenergy regions 66007640 cm1 [7], 971111335 cm1 [8],and 1133514 520 cm1 [9]. The present observations arein the highest energy region investigated so far for H2

17O.

2. Experimental

The experimental proceduresand conditions are very sim-ilar tothe onesused for the H2

16OandH218O measurements

0022-2852/$ - see front matter 2006 Elsevier Inc. All rights reserved.

doi:10.1016/j.jms.2006.02.012

* Corresponding author.E-mail address: [email protected] (W. Ubachs).

www.elsevier.com/locate/jms

Journal of Molecular Spectroscopy 237 (2006) 6369
mailto:[email protected]:[email protected]


2/7

described previously [3,5]. In short, a cavity ring-downexperiment was performed using a cell of length %85 cmsealed withmirrors of reflectivity%99.99% and radius of cur-vature of 50 cm. Tunable laser radiation in pulses of 5 ns andbandwidth %0.06 cm1, obtained from a Nd:YAG pumpeddye laser system (Quanta-Ray PDL-3), was employed to

induce the ring-down transients. The observed exponentialdecays were converted into an absorption spectrum follow-ing well-established methods [10]. In case ofvery strong linesthe obtained decay transients are almost fully collapsed,causing saturation in the absorption profiles. Line positionsand intensities were obtained from the resulting absorptionspectrum by fitting Voigt profiles to the resonances.

Isotopic composition of the 17O-enriched water sample(CAMPRO Scientific) was: 83.8% atom 17O, 8.3% atom16O, and 7.9% atom 18O. The contamination of H2

16Owas estimated to be about 10% and that of H2

18O is rough-ly the same because of the residuals from the sample celland mirrors, which were used for the previous H2

18O

measurements. This leaves about 80% of sample waterbelonging to H2

17O, with an estimated error of about5%. The total pressure during the measurements was atthe saturation point of water at 294 1 K, correspondingto 25 mbar. As for the frequency positions the H2

17O lineswere calibrated against an I2 reference spectrum [11] as well

as against the known line positions of H216

O and H218

Olines, resulting in an estimated accuracy of 0.05 cm1.To eliminate the H2

16O and H218O lines and to identify

the lines belonging to H217O, the spectrum of natural abun-

dance water vapor by Naus et al. [3] and the H218O spec-

trum by Tanaka et al. [5] were plotted with the H217O

spectrum in a single graph. Fig. 1 shows a portion of thegraph with the natural abundance (top), H2

17O (middle),and H2

18O spectrum (bottom). The H216O lines are nar-

rower than the H217O and H2

18O lines because of the differ-ent pressure conditions used in the experiments. Each linein the H2

17O spectrum was manually attributed to H216O,

H217O o r H2

18O. After this procedure, 516 lines are

1.2

0.8

0.4

0.0lossinH2

18O/106 cm

-1

1670416702167001669816696

Wavenumber /cm-1

2.0

1.5

1.0

0.5

0.0lossinH2

16O/106

cm

-1

2.0

1.5

1.0

0.5

0.0lossinH2

17O/106

cm

-1

Fig. 1. A portion of the measured spectrum of H217O (middle) compared with natural abundance spectrum by Naus et al. [3] (top) and H2

18O spectrum by

Tanaka et al. [5] (bottom). The markers indicate line positions of the different isotopologuesH216

O (+), H217

O (,), and H218

O (s).

64 O. Naumenko et al. / Journal of Molecular Spectroscopy 237 (2006) 6369


3/7

identified as belonging to H217O. Of these 516 lines, 32 and

27 lines are found to be overlapped by lines pertaining toH2

16O a n d H218O, respectively. All lines attributed to

H217O are presented in the Appendix to this paper, which

is deposited in the electronic archive of the Journal ofMolecular Spectroscopy. The resulting line list contains

576 transitions taking into account unresolved multiplets.

3. Assignment results

A theoretical line list for H217O, at a temperature of

296 K, was calculated using the DVR3D program suite

[12] with potential energy surfaces (PES) by Shirin et al.[13]. This adiabatic potential energy surface for three majorisotopologues of water, H2

16O, H217O, and H2

18O, wasconstructed by fitting to a large set of vibrationrotationenergy levels of the three isotopologues known at the time.The fit was based on 1788 experimental energy levels with

rotational quantum numbers J= 0, 2, and 5 and repro-duced these observed levels with a standard deviation of0.079 cm1.

This line list for H217O is built upon energy levels up to

26000 cm1 and J6 10. The DVR3D [12] was used with 29radial grid points for Morse oscillator-like basis functionsand 40 angular grid points based on (associated) Legendrepolynomials. Vibrational Hamiltonian matrices of finaldimension 1500 were diagonalised and for rotational prob-lems these matrices had dimension 300 (J+ 1 p), whereJ is the rotational quantum number and p, the parity.Nuclear masses of 17O and H have been used to generatethe line list. For the intensity calculations the best available

dipole moment surface (DMS) of Schwenke and Partridge[14] was used.

Assignments were made relying on matches betweenobserved and calculated frequencies and intensities. Forthis purpose values for absolute intensities (in cm mole-cule1) were estimated based on sample pressure and rela-tive abundance of H2

17O in the gaseous sample. Theintensities are put on an absolute scale for H2

17O in naturalabundance by adopting 0.0372% of H2

17O, normalizing to296 K as in HITRAN [15] and using partition functionsQ (294) = 1037.253 and Q (296) = 1047.931 (from cfa-ftp.harvard.edu). The uncertainties in the line intensities,

listed in the Appendix, are at least 15% for medium inten-sity and well isolated lines, and may be significantly largerfor the weakest lines. This uncertainty also includes an

0

20

10

16700 16800 16900 17000 17 100

20

10

Wavenumber / cm1

Intensity/1

06cm2atm1

Observed

Calculated

Fig. 2. Overview of the H217O experimental spectrum and the calculated line list in the 5m polyad region. For making a good comparison the experimental

line positions and derived intensities are converted into a stick spectrum. Only 451 weakest lines of 516 observed are included in the comparison (see text).The intensities of the experimental spectrum are not scaled to its natural abundance as is common in the HITRAN database; we have chosen to present

here the line strengths pertaining to the pure H217O isotopologue.

Table 1Summary of H2

17O vibrational energy levels determined in this study

Band Origin (cm1) Numberof levels

Number oftransitions

(340) or 30+4 14 19(241) or 304 9 10

(043) or 21

4 5 9(142) or 21+4 16 19(420) or 40+2 16 798.83 (10) 66 86(321) or 402 16 797.167 71 135(222) or 31+2 1 1(123) or 312 2 2(500) or 50+0 16 875.27 (2) 51 92(401) or 500 16 875.620 61 135(302) or 41+0 2 1(260) or 20+6 3 5(161) or 206 1 1(081) or 108 1 1

Total 303 516

Bands are labelled using normal mode (left) and local mode (right)

notations. Number of levels shows the number of newly determined energylevels. Number of transitions shows the number of transitions to thevibrational bands.

O. Naumenko et al. / Journal of Molecular Spectroscopy 237 (2006) 6369 65


4/7

Table 2H2

17O energy levels in cm1 for the (321), (420), (500), and (401) vibrational states

J Ka Kc (321) (420) (401) (500)

0 0 0 16 797.167 1 16 875.620 11 0 1 16 819.528 1 16 897.698 19.1 2 16 897.357 11 1 1 16 834.715 7.3 2 16 908.774 0.7 2 16 908.474 11 1 0 16 840.429 5.7 2 16 913.851 5.7 2 16 913.535 1

2 0 2 16 862.895 9.4 2 16 940.629 9.4 2 16 940.235 13.7 22 1 2 16 873.744 1 16 878.026 1 16 947.859 8.1 2 16 947.579 14.0 22 1 1 16 890.803 2.0 2 16 963.130 4.8 22 2 1 16 934.743 1 16 997.232 4.1 2 16 997.158 12 2 0 16 935.897 2.6 2 16 933.302 1 16 998.753 4.2 2 16 998.550 13 0 3 16 924.975 3.9 2 16 927.232 9.8 2 17 001.077 2.9 2 17 000.660 7.2 33 1 3 16 931.569 9.4 2 16 936.017 21.0 2 17 006.098 0.0 2 17 005.395 6.2 33 1 2 16 965.318 7.4 2 16 971.751 14.2 2 17 035.922 3.6 3 17 035.598 1.2 33 2 2 17 001.662 1.7 3 16 999.058 7.4 2 17 063.768 4.8 3 17 063.848 13 2 1 17 008.435 8.8 3 17 005.404 9.9 2 17 070.088 7.6 3 17 070.022 4.9 33 3 1 17 083.619 7.0 2 17 082.457 1 17 138.736 5.9 2 17 137.421 13 3 0 17 083.856 0.3 2 17 082.702 1 17 138.966 3.9 2 17 137.628 14 0 4 17 004.235 10.8 2 17 010.364 2.4 2 17 079.356 1.8 2 17 079.173 9.4 24 1 4 17 006.961 2.0 2 17 012.284 3.2 3 17 081.528 3.4 2 17 081.378 5.8 34 1 3 17 062.430 11.5 2 17 071.209 1 17 130.655 8.0 3 17 130.388 16.6 3

4 2 3 17 089.543 3.1 3 17 087.030 0.6 2 17 151.318 6.1 2 17 151.545 8.2 34 2 2 17 106.187 2.9 3 17 103.291 2.2 2 17 167.451 6.0 4 17 167.557 14 3 2 17 174.188 2.1 3 17 173.159 2.6 2 17 230.588 0.3 3 17 229.229 14 3 1 17 175.674 0.2 2 17 174.701 19.4 2 17 232.052 6.4 3 17 230.650 14 4 1 17 273.011 1 17 272.421 2.8 2 17 345.023 1 17 343.784 14 4 0 17 273.297 8.3 3 17 272.502 1 17 344.996 1 17 343.786 15 0 5 17 098.881 1 17 102.719 1 17 174.909 8.6 2 17 173.823 7.6 35 1 5 17 099.444 1 17 105.902 1 17 174.993 2.3 2 17 173.936 13.2 25 1 4 17 179.875 1.1 2 17 174.210 1 17 245.123 1.6 2 17 244.972 15 2 4 17 197.436 3.2 2 17 195.067 1 17 259.003 6.2 3 17 259.530 9.0 25 2 3 17 229.592 5.6 2 17 226.505 12.7 2 17 290.418 10.3 3 17 290.819 7.6 35 3 3 17 286.838 9.8 4 17 285.892 1.1 2 17 345.313 4.3 4 17 343.838 15 3 2 17 292.615 6.3 3 17 291.805 1 17 351.051 4.2 3 17 348.871 8.7 25 4 2 17 386.151 4.4 2 17 385.659 1 17 461.631 15 4 1 17 386.501 18.3 2 17 386.022 6.8 2 17 460.559 15 5 1 17 481.582 1 17 481.351 15 5 0 17 481.579 0.6 2 17 481.324 16 0 6 17 210.005 1 17 215.015 1 17 285.756 9.8 3 17 285.087 16 1 6 17 217.608 1 17 285.765 4.6 2 17 285.053 16 1 5 17 315.417 26.7 2 17 311.782 1 17 377.466 13.0 3 17 377.483 16 2 5 17 324.302 1 17 321.993 6.4 3 17 385.936 4.8 2 17 386.175 6.7 46 2 4 17 376.552 4.5 3 17 373.026 1 17 435.707 14.7 2 17 436.723 16 3 4 17 421.752 3.4 3 17 420.684 1.3 2 17 482.344 0.9 3 17 480.675 5.0 26 3 3 17 434.884 9.9 2 17 521.950 1 17 494.398 1.5 2 17 493.452 16 4 3 17 523.657 1 17 601.601 1 17 600.387 16 4 2 17 524.307 5.6 2 17 523.413 16 5 2 17 617.891 5.4 2 17 617.681 1 17 773.448 16 5 1 17 617.906 8.9 3 17 617.775 1 17 773.452 16 6 1 17 750.873 1 17 745.243 16 6 0 17 750.873 1 17 745.239 17 0 7 17 328.143 1 17 343.962 1 17 412.671 1.6 27 1 7 17 345.830 1 17 413.344 5.7 3 17 413.039 17 1 6 17 466.893 1 17 527.176 17 2 6 17 469.558 6.2 2 17 531.411 1 17 533.271 17 2 5 17 544.033 0.7 2 17 539.962 1 17 610.338 1 17 609.082 3.1 27 3 5 17 576.927 16.3 2 17 575.992 1 17 640.767 1 17 638.869 17 3 4 17 606.965 1 17 606.418 1 17 666.841 1 17 664.428 17 4 4 17 680.687 1 17 679.648 1 17 765.072 17 4 3 17 685.263 11.9 2 17 686.703 17 5 3 17 776.715 17 5 2 17 777.023 17 6 2 17 908.538 14.9 2 17 903.357 17 6 1 17 908.536 17.7 2 17 903.321 17 7 1 18 054.558 1



5/7

underestimate of the line strength in typical pulsed CRDexperiments, caused by multi-exponential contributions tothe decays for situations where the laser line width

approaches the widths of the resonances [10]. For the weaklines, as well as for blended lines, the uncertainty may reach100300%, an estimate based on the tendency of(obs. calc.)/obs. in the intensities, and on the knownquality of the SP dipole moment surface [14]. For the stron-gest lines the experimental values seem to be greatly under-estimated due to saturation effects in the CRDexperiments.

Comparisons of line positions and intensities were usedto suggest possible assignments; additional ingredients forachieving reliable line assignments are combination differ-ences, as well as ratios of new energy levels of H2

17O toknown levels of H2

16O and H218O. Identification of the

four strongest bands formed by transitions involving rota-tional sublevels of two local mode pairs: (321)(420) and(401)(500) was facilitated by the use of the fact that therotational structure of the components of local mode pairhas to be identical. The assignment procedures have ledto positive identification of all 516 lines attributed toH2

17O. All the transitions originate in the (000) vibrationalground level and are assigned to 14 different upper levelsincluding (321), (420), (500), and (401). A summary ofvibrational energy levels included in the present analysisis given in Table 1.

Fig. 2 shows a comparison of the experimental spectrum

with the calculated line list. Of the 516 observed lines, the

65 strongest ones were excluded from the comparison sincetheir intensities were greatly underestimated due to satura-tion. The assignment of the strongest lines is not so diffi-

cult, even if they are saturated: as usual, they correspondto transitions involving low J values, and are part ofwell-behaved combination differences (CD). The mainproblem is assignment of medium intensity and, especially,the weakest lines, which often are not included into CDrelations. As it is seen from the figure, observed and calcu-lated intensities agree well throughout the entire range,confirming a high accuracy of the calculated data and theassignments performed. In the resulting identification list,the calculated intensities are presented along with theobserved values. For the strongest observed lines, corre-sponding to calculated intensities larger than2.2 1028 cm molecule1 (natural abundance), the calcu-lated data are more precise, in view of the saturation inthe experimental data.

The band origins for (321) and (401) are newly deter-mined at 16797.167 and 16875.620 cm1, respectively.For (500) and (420) band origins were not directlyobserved experimentally, thus they are estimated at16875.27 0.02 and 16798.83 0.1 cm1, respectivelyfrom the obs. calc. tendency for the energy levels[J0J] (in conventional [J, Ka, Kc] notation). The band ori-gin for (500) has not been observed for either H2

16O orH2

18O. The Obs. Calc. differences were reasonably sys-tematic for energy levels of all four above-mentioned

states, while for the other ten states the obs. calc. ten-

Table 2 (continued)

J Ka Kc (321) (420) (401) (500)

7 7 0 18 054.558 18 0 8 17 480.218 1 17 489.375 1 17 557.813 1 17 557.151 18 1 8 17 495.650 1 17 557.769 1 17 557.548 18 1 7 17 629.496 1 17 693.568 18 2 7 17 614.643 1 17 700.215 1

8 2 6 17 729.099 1 17 786.345 18 3 6 17 750.816 1 17 817.599 18 3 5 17 800.129 1 17 798.843 1 17 859.471 18 4 4 17 873.015 1 17 871.827 1 17 960.256 18 5 4 17 957.879 18 5 3 17 959.150 18 6 3 18 083.555 18 6 2 18 083.623 18 7 2 18 233.506 18 7 1 18 233.501 19 0 9 17 719.223 19 1 9 17 717.897 19 1 8 17 881.549 19 2 8 17 803.428 19 2 7 17 944.381 1

9 3 7 18 020.385 19 3 6 1 18 013.991 19 4 6 18 061.642 1 18 166.120 1

10 0 10 17 808.943 1 17 897.202 110 1 10 17 811.656 110 2 9 18 002.266 1 18 069.909 110 5 5 18 396.659 1

Also given are the root mean square experimental uncertainties in 103 cm1 for energy levels derived from two and more lines and the number oftransitions used in the level derivation.



6/7

dencies were less obvious since only separate energy lev-els were derived for these states from the perturbed tran-sitions borrowing their intensity from stronger bands.The level energies for all considered states are tabulatedin Tables 2 and 3. The averaged absolute deviationbetween level energies as derived from experiment andfrom calculation is 0.08 cm1, whereas the maximal devi-ation is 0.33 cm1. These values represent the criteriaon the accuracy of the theoretical models. As the pres-ently observed levels were not included in the fitting pro-cedure in [13] the standard deviation shows goodextrapolation quality of the PES used for generating the-oretical line list.

4. Discussion and conclusion

Fewer lines were attributed to H217O in this spectrum

compared to the H218O lines in the same spectral region[5]. This is because the sample used in this experiment con-tained a lower concentration of H2

17O and also the spec-trum was significantly contaminated by H2

16O andH2

18O. We decided to keep all the H217O lines overlapped

by lines pertaining to other isotope species in the line list, ifthe contribution from the H2

17O was more than 3050%.However, some relatively strong predicted H2

17O lines werenot found in the spectrum since the corresponding experi-mental lines were unrecoverably blended by near-by muchstronger H2

16O, H218O, or other H2

17O lines. For thestrongest vibrational level in the considered polyad region,

the (401) vibrational mode, many energy levels are con-firmed by combination differences. However, for the lessstrong modes like (321), (500), and others, energy levelswere determined from a single transition (60% in total).The energy levels determined by a single weak transitionare less reliable than those confirmed by combinationdifferences.

Many energy levels of the highly excited states like (340),(142), and (241) (see Table 3) were derived from the transi-tions borrowing intensity from resonance partners belong-ing to the four strongest bands. To perform reasonablerovibrational labelling of these highly excited levels wasrather difficult. As it was mentioned above, the ratios of

Table 3H2

17O energy levels in cm1 for (340), (260), (302), (142), (222), (123),(043), (241), (161), and (081) vibrational states

(340)4 3 2 17 003.351 12.3 24 4 1 17 180.383 11.2 24 4 0 17 180.333 1

5 3 2 17 120.958 15 4 2 17 297.767 0.7 26 4 3 17 439.310 5.4 26 5 2 17 663.621 16 5 1 17 663.654 17 3 5 17 421.573 17 3 4 17 430.236 17 4 4 17 599.878 17 5 3 17 827.467 17 5 2 17 827.712 19 2 7 17 725.894 1

(241)4 3 2 17 004.696 6.0 24 4 1 17 182.849 1

4 4 0 17 182.822 15 4 2 17 300.006 15 5 0 17 523.875 16 3 3 17 270.407 16 4 3 17 440.738 26.2 26 4 2 17 442.315 20.2 27 4 4 17 601.678 1

(302)4 4 1 17 862.098 14 4 0 17 862.094 1

(123)6 0 6 17 701.228 16 1 5 17 809.266 1

(222)4 1 4 17 414.118 1

(142)3 0 3 16 901.708 13 1 2 16 954.765 14 1 4 16 993.400 14 2 3 17 128.444 6.8 24 3 2 17 272.364 15 0 5 17 082.689 15 1 5 17 087.879 15 1 4 17 193.189 15 2 3 17 267.082 16 1 6 17 199.139 1

6 2 5 17 367.961 17 2 5 17 594.981 18 1 8 17 478.691 18 2 7 17 684.477 19 0 9 17 652.142 19 1 9 17 649.025 19 1 9 17 649.025 1

(043)3 0 3 17 068.650 0.1 23 1 3 17 081.924 15 1 4 17 349.543 2.6 35 3 3 17 522.873 16 1 5 17 496.478 14.7 3

(continued on next page)

Table 3 (continued)

(260)6 5 2 17 521.583 2.0 26 5 1 17 520.658 17 5 3 17 683.552 1

(161)

6 4 2 17 215.931 1

(081)8 5 3 17 691.982 1

The three quantum numbers preceding the level energies are J, Ka and Kc.Also given are the root mean square experimental uncertainties in103 cm1 for energy levels derived from two and more lines and thenumber of transitions used in the level derivation.



7/7

new energy levels of H217O to known levels of H2

16O andH2

18O were used not only for assignment, but for labellingpurposes also. However, a number of newly observedH2

17O energy levels were not available for the H216O and

H218O molecules. In this case, the labelling was determined

on the basis of calculations within the effective Hamiltoni-

an approach.As it was anticipated, the line intensities generated usingthe DMS of [14] and the PES of [13] were found to be, onaverage, very close to those obtained with the same DMSof[14] and the PES by Partridge and Schwenke [16]. How-ever, in some cases, these intensities differed strongly, inparticular for transitions involving upper energy levels inclose (less than 1 cm1) resonance. Thus, the SP intensitiesfor transitions (500)[717][808] at 16670.641 cm1, and(500)[717][606] at 16967.364 cm1 were predicted to be6.6 1030 and 2.0 1030 cm molecule1, respectively,while the experimental values were of 2.27 1029 and4.55 1029 cm molecule1, close to the ones evaluated in

[13,14]: 2.3 1029 and 3.9 1029 cm molecule1. Betteragreement of the latter intensities with the observed valuesis a result of using more correct wave functions as providedby [13].

The H217O water vapor spectrum in the 5m polyad mea-

sured by CRD spectroscopy has been analysed and all 516observed lines are now assigned yielding an importantnumber of 303 new highly excited energy levels. The com-plete assignment of the observed H2

17O lines has beenmade possible due to high-quality synthetic spectrum pro-vided by [13,14], while the well-known calculation bySchwenke and Partridge [14,16] may be inaccurate in the

considered spectral range up to 5 cm1

. This is the highestfrequency region to be studied for H2

17O, and as such itadds to the understanding of the entire water spectrum.

Acknowledgments

The authors acknowlegde the contribution by N.F. Zo-bov and O.L. Polyansky (Nizhnii Novgorod) to the gener-ation of the line list for H2

17O, that will be publishedindependently [13]. This research forms part of an effortby a Task Group of the International Union of Pure andApplied Chemistry (IUPAC, Project No. 2004-035-1-100)

on A database of water transitions from experiment and

theory. This work is supported by the Space ResearchOrganization Netherlands (SRON) with a Project Grant(EO 036), and by the European CommunityAccess toResearch Infrastructures action of the Improving HumanPotential program, Contract No. HPRICT199900064,as well as the INTAS foundation and the Russian Founda-

tion for Basic Research.

Appendix A. Supplementary data

Supplementary data for this article are available onScienceDirect (www.sciencedirect.com) and as part of theOhio State University Molecular Spectroscopy Archives(http://msa.lib.ohio-state.edu/jmsa_hp.htm).

References

[1] A.N. Maurellis, R. Lang, J.E. Williams, W.J. van der Zande, K.Smith, J. Tennyson, R.N. Tolchenov, The impact of new water vaporspectroscopy on satellite retrievals, NATO SCIENCE series: IV.Earth and Environmental Sciences, 27, Kluwer Academic Publishers,2003, pp. 259270.

[2] R. Lang, A.N. Maurellis, W.J. van der Zande, I. Aben, J. Landgraf,W. Ubachs, J. Geophys. Res. 107 (D16) (2002) 4300.

[3] H. Naus, W. Ubachs, P.F. Levelt, O.L. Polyansky, N.F. Zobov, J.Tennyson, J. Mol. Spectrosc. 205 (2001) 117121.

[4] S. Kassi, P. Macko, O. Naumenko, A. Campargue, Phys. Chem.Chem. Phys. 7 (2006) 24602467.

[5] M. Tanaka, M. Sneep, W. Ubachs, J. Tennyson, J. Mol. Spectrosc.226 (2004) 16.

[6] R.A. Toth, J. Opt. Soc. Am. B 9 (1992) 462482.[7] R.A. Toth, Appl. Opt. 33 (1994) 48684879.[8] C. Camy-Peyret, J.-M. Flaud, J.-Y. Mandin, A. Bykov, O. Nau-

menko, L. Sinitsa, B. Voronin, J. Quant. Spectrosc. RA 61 (1999)795812.

[9] M. Tanaka, O. Naumenko, J.W. Brault, J. Tennyson, J. Mol.Spectrosc. 234 (2005) 19.

[10] H. Naus, I.H.M. van Stokkum, W. Hogervorst, W. Ubachs, Appl.Opt. 40 (2001) 44164426.

[11] S. Gerstenkorn, P. Luc, Atlas du Spectroscopie dAbsorption de laMolecule de lIode Entre 1480020000 cm1, Presses du CNRS,1978.

[12] J. Tennyson, M.A. Kostin, P. Barletta, G.J. Harris, J. Ramanlal, O.L.Polyansky, N.F. Zobov, Comp. Phys. Commun. 163 (2004) 85116.

[13] S.V. Shirin, O.L. Polyansky, N.F. Zobov, A.G. Csaszar, R.I.Ovsyannikov, J. Tennyson, J. Mol. Spectrosc. 236 (2006) 216223.

[14] D.W. Schwenke, H. Partridge, J. Chem. Phys. 113 (2000) 65926597.[15] L.S. Rothman et al., J. Quant. Spectrosc. RA 96 (2005) 139204.[16] H. Partridge, D.W. Schwenke, J. Chem. Phys. 106 (1997) 46184639.

http://www.sciencedirect.com/http://msa.lib.ohio-state.edu/jmsa_hp.htmhttp://msa.lib.ohio-state.edu/jmsa_hp.htmhttp://www.sciencedirect.com/

o. naumenko et al- cavity ring-down spectroscopy of h2^17-o in the range 16570–17125 cm^-1

Documents