Absolute Line Intensities for the ν6Band of H2O2

Journal of Molecular Spectroscopy 195, 154 –161 (1999) Article ID jmsp.1999.7807, available online at http://www.idealibrary.com on

Absolute Line Intensities for the n6 Band of H2O2 S. Klee,* M. Winnewisser,* A. Perrin,† and J.-M. Flaud† *Justus Liebig Universitaet, Physikalisch-Chemisches Institut, Heinrich-Buff-Ring 58, D-35392 Giessen, Germany; and †Laboratoire de Photophysique Mole´culaire, CNRS, Universite Paris Sud, Bat 210, Campus d’Orsay, F-91405 Orsay Cedex, France Received October 8, 1998; in revised form December 17, 1998

The purpose of this work was to obtain reliable absolute intensities for the n6 band of H2O2. It was undertaken because strong discrepancies exist between the different n6 band intensities which are presently available in the literature (A. Perrin, A. Valentin, J.-M. Flaud, C. Camy-Peyret, L. Schriver, A. Schriver, and P. Arcas, J. Mol. Spectrosc. 1995. 171, 358), (R. May, J. Quant. Radiat. Transfer 1991. 45, 267), and (R. L. Sams, personal communication). The method which was chosen in the present work was to measure simultaneously the far-infrared absorptions and the n6 absorptions of H2O2. Consequently, Fourier transform spectra of H2O2 were recorded at Giessen in a spectral range (370 –1270 cm21) which covers both the R branch of the torsion–rotation band and the P branch of the n6 band which appear at low and high wavenumbers, respectively. From the low wavenumber data, the partial pressure of H2O2 present in the cell during the recording of the spectra was determined by calibrating the observed absorptions in the torsion–rotation band with intensities computed using the permanent H2O2 dipole moment measured by Stark effect (A. Perrin, J.-M. Flaud, C. Camy-Peyret, R. Schermaul, M. Winnewisser, J.-Y. Mandin, V. Dana, M. Badaoui, and J. Koput, J. Mol. Spectrosc. 1996. 176, 287–296) and [E. A. Cohen and H. M. Pickett, J. Mol. Spectrosc. 1981. 87, 582–583). In the high frequency range, this value of the partial pressure of H2O2 was used to measure absolute line intensities in the n6 band. Finally, the line intensities in the n6 band were fitted using the theoretical methods described in detail in our previous works. Using these new results on line intensities together with the line position parameters that we obtained previously, a new synthetic spectra of the n6 band was generated, leading to a total band intensity of 0.185 3 10216 cm21/(molecule.cm22) at 296 K. It has to be pointed out that the new line intensities agree to within the experimental uncertainties with the individual line intensity measurements performed previously by May and by Sams. © 1999 Academic Press INTRODUCTION

Hydrogen peroxide, H2O2, is an important trace atmospheric molecule. It is a reservoir molecule for the HOx species and some of the most important reactions which have been discussed by several authors (1, 2) are OH 1 H2O2 3 H2O 1 HO2

[1]

H2O2 1 hn 3 OH 1 OH

[2]

HO2 1 HO2 3 H2O2 1 O2,

[3]

where (1) dominates reaction (2) and is a sink for OH, and where reaction (3) is the reaction leading to the generation of H2O2 in the stratosphere. Furthermore, the stratospheric NOx and ClOx cycles are strongly affected by the abundance of the OH radical and the related species HO2 and H2O2. Given the influence of H2O2 in stratospheric chemistry, it is important to get accurate altitude concentration profiles for H2O2. Unfortunately, the atmospheric measurements of H2O2 are extremely difficult due to its low concentration and its spectroscopic properties. Two spectral ranges have been used for optical detection of this molecule in the stratosphere: the far infrared

region (3, 4) and the 7.9-mm region, which according to ab initio studies (5, 6) is predicted to correspond to the strongest infrared band of this molecule. Several attempts have been made to measure H2O2 in the stratosphere using this spectral domain. However, to our knowledge only upper limits of H2O2 could be determined up to now (7, 8). The determination of precise column densities and altitude concentration profiles in atmospheric measurements requires the most accurate spectroscopic data possible on line positions, absolute intensities, and pressure broadening parameters. These spectral data are difficult to derive for H2O2. Indeed this molecule exhibits internal rotation (9 –11): the OH torsion around the O-O bond. This torsional effect and the existence of vibration torsion rotation resonances imply that a sophisticated model has to be used both for the line position and line intensity calculations. Furthermore, from an experimental point of view H2O2 is not easy to handle because it decomposes in the cell and it is then extremely difficult to know precisely its concentration. Line Positions

Supplementary data for this article may be found on the journal home page (http://www.academicpress.com/jms).

Far infrared and 11-mm region. From high resolution studies performed in the far-infrared and 11-mm regions (10, 12), accurate torsion–rotation energy levels were obtained both for the ground vibrational state (hereafter denoted by v 5 0) and the v3 5 1 vibrational state. For the observed torsional

154 0022-2852/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

ABSOLUTE INTENSITIES FOR THE n6 BAND OF H2O2

subbands,1 which involve for v 5 0 the t 5 1, 2 and t 5 3, 4 torsional quantum numbers up to n # 3, and (n 5 0, t 5 1) and (n 5 0, t 5 3) for v3 5 1, the energy level calculations (12) were performed using an Hamiltonian matrix which takes into account (i) the strong Coriolis-type interaction which couples the levels of the v 5 0 vibrational state with those of the v3 5 1, state, (ii) the (Dn 5 61, DK a 5 72) torsional– rotational resonances within each vibrational state (v 5 0 and v3 5 1), and (iii) the “staggering” effect in v 5 0. This last effect, which was observed for the first time by Flaud et al. (10), is due to the tunnelling through the rather high cis barrier (V cis 5 2562.8 cm21 in the ground vibrational state). The 7.9-mm region. The first rotational analysis of the n6 band was performed at medium resolution by Hillman (13). More recently, from high resolution Fourier transform spectra (14, 15) an extended analysis of the Dn 5 0 torsion rotation subbands of the n6 band (v6 5 1 4 v 5 0 vibrational transition) was performed for the torsional subbands involving the (n, t ) 5 (0, 1), (1, 1), (2, 1), (0, 3), and (1, 3) torsional substates.1 In addition to the (Dn 5 61, DK a 5 72) torsional–rotational resonances within the v6 5 1 vibrational state which are usually observed for H2O2, the Hamiltonian model takes into account the v2 5 1 7 v6 5 1, v3 5 1 7 v6 5 1, and v 5 0 7 v6 5 1 vibration torsion rotation resonances. On the other hand, the cis-staggering effect was not considered since it is negligible for the torsional subbands analyzed at 7.9 mm. Line-Broadening Parameters To our knowledge the only existing air-broadening parameters for H2O2 were measured by Malathy et al. (16) for eight lines in the n6 band using a diode laser technique. Obviously this work should be completed by further studies devoted to line-broadening parameters. Line Intensities Far-infrared region. Individual relative line intensities measurements were performed in the far infrared spectrum of H2O2 using Fourier transform spectra (17). These intensities were introduced into a least squares fit calculation in order to obtain the expansion parameters of the pure torsional–rotation transition moment of H2O2. For the intensity calculation, the theoretical model takes into account the cos g-type dependence of the dipole moment due to the large amplitude motion of the H atoms relative to the O-O bond, where 2g is the torsion angle. The value of the permanent part of the dipole moment obtained from the fit on the relative experimental intensities was then scaled to the value obtained from Stark effect measurements (18). 1 Following previously used notations (9, 15), the torsional energy levels are labeled using the torsional quantum numbers (n, t ). As the cis-barrier staggering is negligible for the observed torsional subbands at 7.9 mm (14), the torsional levels with t 5 1, 2 (respectively t 5 3, 4) form degenerate pairs which will be simply identified by t 5 1 (respectively t 5 3).

155

The 7.9-mm region. The theoretical band intensity of the n6 band was calculated by Rogers and Hillman (5, 6). Also the total band intensity was measured at medium resolution by Niki et al. (19) and later by Valero et al. (20), who reported a value of the total band intensity about two times stronger than Niki’s band intensity. More recently, two types of individual line intensities measurements in the n6 band were performed at high resolution. May (21) and Sams (22) used a diode laser technique, and the measurements were performed using a flowing gas system of purified H2O2. In the case of May, the quantity of H2O2 present in a flowing gas mixture was determined through simultaneous measurements performed in the ultraviolet at 254 nm. Perrin et al. (14) used Fourier transform spectra to measure individual line intensities. In this case the H2O2 sample was prepared by low pressure distillation of a commercial sample of hydrogen peroxide, and the quality of the distillated H2O2 sample was checked, before its introduction in the cell through the analysis of its vibrational spectrum recorded at 11 K by matrix technique. In this procedure it was then assumed that there was no decomposition of the H2O2 sample during the recording of the high resolution FTS spectrum. It is not always easy to draw a global conclusion from measurements obtained by differing techniques and at either low or high resolution, but one may say that the existing data parameters for the n6 band are in strong contradiction. The total band intensity reported by Valero et al. (20) is about two times stronger than Niki et al.’s (19) band intensity. May’s measurements (21) and Sams’ measurements (22) lead to intensities about 1.6 stronger than the ones from Valero et al. (20) and two times stronger than Perrin et al. (14). Finally, let us point out that in the 1996 version of the HITRAN database (23) the line list present for the n6 band involves line positions deduced from the medium resolution analysis performed by Hillman (13) and line intensities which were calibrated relative to the results obtained by May (21). Because of the existence of these strong discrepancies, the present work was undertaken to try to obtain reliable absolute intensities for the n6 band of H2O2. The method was to use Fourier transform spectra of H2O2 recorded in a spectral range (370 –1270 cm21) which covers simultaneously the R branch of the torsion–rotation band (low wavenumber range) and the P branch of the n6 band (high wavenumber range). From the low wavenumber data, the partial pressure of H2O2 present in the cell during the recording of the spectra was determined by calibrating the observed absorptions in the torsion–rotation band with intensities computed using the permanent H2O2 dipole moment measured by Stark effect. In the high frequency range, this value of the partial pressure of H2O2 was used to measure absolute line intensities in the n6 band. Finally, the line intensities in the n6 band were fitted using the theoretical methods described in detail in our previous works (15).

Copyright © 1999 by Academic Press

156

KLEE ET AL.

EXPERIMENTAL DETAILS

LINE INTENSITIES MEASUREMENTS

As was pointed out previously in the text, the primary goal of the present study was to evaluate the intensities of spectral lines belonging to the most intense H2O2 bands, namely the torsion–rotation and the n6 bands. Since a satisfactory reproducibility of the experimental conditions is very difficult to attain with such a reactive and rapidly decomposing sample like highly concentrated hydrogen peroxide, the bands were not recorded independently, but simultaneously. Therefore, the joint spectrum had to cover lines from the torsional absorption in the far-infrared spectral region as well as those belonging to the mid-infrared n6 band. This method required a broad-band operation of the employed Bruker IFS120HR interferometer with the folding limits being set at 0 and 1316 cm21, respectively. To keep the signal-to-noise ratio still as high as possible, a double-band-pass optical filter was applied that was transparent only between 330 –500 and 880 –1250 cm21. Electronic filtering was active for the range 138 –1383 cm21. As the light source a SiC globar was used. The detector consisted of a Ge:Cu-type detector operating at liquid-helium temperature provided with a CsI window. The cell windows of the 3mPyrex glass cell with an inner diameter of 10 cm were also made of CsI and were additionally coated with thin polyethylene foils to protect them against corrosion and to avoid substantial heterogeneous decomposition of hydrogen peroxide at the alkali halogenide surfaces as has been observed before (17). The linearized H2O2 decomposition rate could be suppressed below 0.5% per scan which allowed the record of blocks of 20 scans at a time before the sample of hydrogen peroxide under a static pressure of 40 Pa was replenished. An alternative experiment under constant flow conditions turned out to be disadvantageous due to then emerging pressure fluctuations. Up to seven individual blocks were coadded to yield the final transmittance spectrum after normalization to the spectrum of the evacuated cell. A second experiment was carried out at the same partial pressure of H2O2, but this time adding 680 Pa of nitrogen buffer gas to weakly pressurebroaden the spectral lines. In a former study (17) this procedure has been proven advantageous to increase the quality of the line intensity fit using the INTBAT program (24 –26). All spectra were recorded at a stabilized room temperature of 297 K and without apodization at an interferometric maximum optical path difference of 346 cm which corresponds to an instrumental resolution of 0.0029 cm21. This value was fixed by the limited amount of memory on the data acquisition processor board for the chosen broad spectral range. Thereby the instrumental resolution was the principal contribution to the observed line width. However, this resolution turned out to be sufficient to fully resolve both bands of interest and to perform a reliable line intensity analysis. Hydrogen peroxide was prepared by vacuum distillation of a 30% solution leading to approximately 89% H2O2. Further purification was carried out by pumping on this solution until almost complete evaporation.

The line parameters, namely linestrengths, line positions, and line broadening coefficients, of 455 selected transitions covering absorption peaks in the two spectral ranges under study were evaluated using the line fitting program INTBAT (24 –26) which performs a line by line non linear least squares fit to the experimental profiles. Because the H2O2 spectrum is rather dense, the line parameters of the various transitions selected in a spectral window have to be fitted simultaneously. During the iterative fitting of the lines to a Voigt profile, the Doppler width was constrained to its theoretical value. On the other hand, depending on the density of the lines in the spectral window, the broadening coefficients (which account for the overall broadenings due to all the gas present in the cell, i.e., molecular nitrogen, H2O2, water, molecular oxygen, . . . ) were either fixed to a mean value g 5 0.14 cm21 atm21 or were determined by the line-fitting program to improve somehow the accuracy on the intensity measurements. In this later case the values which were obtained for the line-broadening parameter g (between 0.1 cm21 atm21 6 20% and 0.2 cm21 atm21 6 20%) could not be significantly linked to any rotational dependency. Finally, 273 and 182 individual line intensities2 were measured for the far-infrared region and the 7.9-mm region, respectively, with, because of different signal to noise ratios depending on the spectral range, corresponding estimated accuracy of about 5–15% and of about 5–20%. Actually, most of these measurements were performed on the second spectrum (i.e., the H2O2 spectrum recorded with adding 680 Pa of nitrogen buffer gas to weakly pressure-broaden the spectral lines) because, as it was already pointed out in the text, a former study has proved that the quality of the line intensity fit using the INTBAT program is better (24 –26). Also, these experimental results were yielded by evaluation of line intensities performed on the recorded spectrum of pure H2O2. LINE INTENSITY CALCULATIONS

Theoretical Considerations The intensity of a line (27–28) for a pure isotopic sample of H216O2 is given (in cm21/(molecule.cm22) by k nN˜ 5

S S D

S DD

8 p 3n˜ EL EU exp 2 2 exp 2 3hc4 pe 0 kT kT

gL R U, Z~T! L

[4]

where L and U are, respectively, the lower and upper levels of the transition, n˜ 5 (E U 2 E L )/hc is the wavenumber of the transition, g L is the nuclear spin degeneracy of the lower level L ( g L 5 1 (respectively gL 5 3) for levels of total symmetry (9, 10) A gs or A us (respectively B gs or B us)), and Z(T) is the 2

The complete list of measured intensities is available upon request from Agne`s Perrin or as supplementary data on the journal home page.

Copyright © 1999 by Academic Press

ABSOLUTE INTENSITIES FOR THE n6 BAND OF H2O2

partition function. As in Ref. (15, 17), we used Z(294 K) 5 9751.7, Z(295 K) 5 9801.5, Z(296 K) 5 9851.4 and Z(297.2 K) 5 9911.4; these values taking into account both the vibrational and the torsion–rotation contributions. One has R UL 5 u^V Un Ut U, J UK Ua K Uc u m 9ZuV Ln Lt L, J LK La K Lc &u 2,

157

TABLE 1 C- and A-type Transition Moment Constants for the Torsion– Rotation Band and for the n6 Band of H2O2, Respectively

[5]

where the notation uVn t & is used for the upper and lower vibration-torsional states (with a U and L superscript respectively). One should point out the following: 1. The far-infrared region involves lines belonging to the pure torsion rotation band of H2O2 with, due to symmetry considerations, only C-type transitions: t U 5 3 and t L 5 1, t U 5 4 and t L 5 2, t U 5 1 and t L 5 3, or t U 5 2 and t L 5 4. 2. The 7.9-mm region involves (except for a very few resonating transitions from the n2 or n3 bands or within the ground vibrational state) only n6 lines. Actually, only A-type transitions were observed with n U 5 n L and, depending on the line considered, t U 5 t L 5 1 or t U 5 t L 5 3. In Eq. (5), m 9Z is the transformed dipole moment operator (27, 28). The calculation of the matrix elements of m 9Z is extensively described in Refs. (27, 28, 15, 17), and only the relevant details will be given. The upper and lower vibro– torsion–rotation wavefunctions are expanded as follows:

O OC

uVU nU tU , JU KUa KUc & 5

K9 V9n9 t 9

uV9n9t9&uJ9K9g9&

[6.a]

K0 V0n0 t 0

uV0n0t0&uJ0K0g0&,

[6.b]

V9n9 t 9[B9 K9

uVL nL tL , JL KLa KLc & 5

O OC

Note. The results are in Debye (1 Debye 5 3.33564 3 10230 Coulomb z m). a Values fixed to those of Table 2 of Ref. (17).

and 0n0t0,6n9t9

where the summation is performed on the blocks B9 and B0 of vibration–torsion interacting states. B9 and B0 include the v6 5 1, v2 5 1, v3 5 1, and the ground vibrational states (15), and the ground v 5 0 and the v3 5 1 vibrational states (12), respectively. m 9Z is expanded (27, 28, 15, 17) as

O O

O

0n0t0,6n9t9

m 9Aj A Aj ,

[8.b]

j

V0n0 t 0[B0 K0

m 9Z 5

m 9Z 5

uV0n0 t 0& V0n0t0,V9n9t9m Z^V9n9 t 9u.

[7]

A where the A C j and A j are C- and A-type rotational operators (Table 1) which have uDKu 5 odd and uDKu 5 even nonzero matrix elements, respectively. The transition moment constants 0n0 t 0,0n9 t 9 m 9Cj and 0n0 t 0,6n9 t 9 m 9Aj are numerical coefficients which, in principle, are to be determined through least squares fit performed on experimental line intensities. However, from the intensity study performed in Refs. (17) and (15), it appears that these parameters can be expressed as

V9n9t9[B9 V0n0t0[B0 0n0t0,0n9t9

Since the n3 and n2 bands are very weak compared to the far infrared absorption or to the n6 band, their transition moments have been set to zero. As pointed out previously in the text, only C- and A-type transitions were observed in the far-infrared and in the 7.9-mm spectral ranges, respectively. Consequently, the transition moment operator takes the following form in each spectral range: 0n0t0,0n9t9

m 9Z 5

O

0n0t0,0n9t9

m 9Cj A Cj

m 9Cj 5 ^ 0,Torsn9 t 9u 0,0O Torsu 0,Torsn0 t 0& 0,0m 9ROTC j

0n0t0,6n9t9

m 9Aj 5 ^ 6,Torsn9 t 9u 0,6O Torsu0,Torsn0 t 0& 0,6m 9ROTA . j

[9.a] [9.b]

In these expressions, *u0,Torsn t & and u6,Torsn t & are the torsional wavefunctions obtained from the diagonalization of the V 5 0 and V 5 6 (v 6 5 1) hydrogen peroxide torsional Hamiltonians

[8.a]

j

Copyright © 1999 by Academic Press

V

H Tors 5 $ J g2 , VB gg% 1 VV~ g !, 0

[10]

158

KLEE ET AL.

where g represents half of the HOOH dihedral angle and J g 5 2i­/­ g . In Equation (10) V B gg and V V( g ) are the zeroth order torsional constant and the torsional potential of the vibrational state V. Their numerical values are given in Tables VI and III of Ref. (10) and in Table III of Ref. (14). 0,0 Tors O and 0,6OTors are torsional operators. Taking into account the symmetry considerations which are given in Ref. (9 –11) these operators may be expanded up to the first order as follows: 0,0

O Tors 5 cos g 1 · · ·

[11.a]

O Tors 5 1 1 · · · .

[11.b]

and 0,6

The matrix elements on the torsional wavefunctions of these first order terms operators are easily computable. For the torsion–rotation band, the ^0,Torsn9 t 9ucos g u 0,Torsn0 t 0& matrix elements are given in Table 3 of Ref. (17). For the n6 band, one has (15) ^ 6,Torsn9 t 9u1u 0,Torsn0 t 0& > d ~n9, n0! d ~ t 9, t 0!.

[12]

C A As a consequence only the 0,0 m 9ROT and 0,6 m 9ROT transij j tion moment constants have to be determined from experimental intensities.

RESULTS

In our previous studies, the following results were obtained. For the torsion–rotation bands (17) an extensive set of precise relative intensities were fitted and the transition moment parameters derived from this fit were scaled to the dipole moment constant measured by Stark effect (18). For the n6 band (15) also numerous experimental line intensities were fitted satisfactorily showing that on a relative basis the measured intensities are correct. However, as far as absolute intensities were concerned, there is some doubt since we assumed that the sample in the cell was pure H2O2, which is not obvious. In the present paper, using the line intensities calculated in Ref. (17) and the 273 line intensities measured in this work for the torsion–rotation band, it was possible to determine the partial pressure of H2O2 in the sample P H2O2 5 0.094671(34) hPa. This pressure was then used to derived the intensities of the 182 lines belonging to the n6 band. The n6 line intensities calculated in Ref. (15) were then calibrated using the 182 line intensities measured in this work showing that the absolute intensities given in (15) were underestimated by a factor 1.8395(10). Table 1 gives the new values of the transition moment constants of the n6 band of H2O2 together with the transition moment constants of the ground state. Also a statis-

tical analysis of the intensity calculations is given showing the quality of the fits. SYNTHETIC SPECTRUM

Using, together with the transition moment constants quoted in Table 1, the band centers and the rotational and coupling constants given in Tables I and IV of Ref. (15) for the upper resonating states and in Tables II,a and II,b of Ref. (12) for the ground state, we have generated a comprehensive list of line positions and intensities (synthetic spectrum) for the n6 band of H2O2. The calculations were performed using an intensity cutoff of 0.1 3 10224 cm21/(molecule.cm22) at 296 K, maximum values of 40 for J and 12 for K a , maximum upper and lower state energies of 4300 and 3000 cm21, respectively. It is worth noticing that because the energy level calculation performed in Ref. (15) for n6 the band did not allow to reproduce some energy levels to within their experimental accuracy, we have used, whenever possible, the observed energy levels instead of the calculated ones for the upper states. The total band intensity of the n6 band (i.e., the sum of the line intensities in the 1170 –1380 cm21 spectral region) was found at 296 K to be equal to S y 6(296 K) 5 0.185 3 10216 cm21/(molecule.cm22) 5 458 cm22 atm21 (details on the intensity repartition within the various torsional subbands of the n6 band can be found in Table VII of Ref. (15)). We estimate the uncertainty for this band intensity to be about 10%, when accounting for all the possible sources of errors which involve the average uncertainty on the 273 and 182 line intensity measurements, the calculations, and the definition of the experimental parameters for the recording of the spectrum (temperature etc. . . . ). DISCUSSION

As already pointed out in the introduction, both low-resolution and high-resolution line intensities measurements were performed for the 7.9-mm band of H2O2. For those performed at low resolution (total band intensity measurements) the contributions from hot bands and from different isotopic species of hydrogen peroxide were estimated using the usual method (29) which leads to S TOT(296 K) < Z VIB(296 K) 3 (Sn6(296 K)),

[13]

which, with Z vib(296 K) 5 1.0184 for the vibrational partition function,3 leads to an estimation of S TOT(296 K) ' 467 cm22 atm21 (610%). This band intensity is significantly stronger than the band intensity given by Niki et al. (19) (S 7.9mm(298 K) 3 Contrary to Ref. (21) this value of the vibrational partition function Z vib(296 K) 5 1.0184 does not account for the n4 large amplitude torsional mode. The so-called “n4 1 n6 2 n4 torsional hot bands” are actually directly accounted for in our model by the n $ 1 torsional subbands and therefore included in Z rot.

Copyright © 1999 by Academic Press

ABSOLUTE INTENSITIES FOR THE n6 BAND OF H2O2

TABLE 2 Comparison Between the Results of the Present Line-Intensity Calculations and the Line-Intensity Measurements Performed by May (21) at 295 K

Note. J9K9a K9c V9n9 t 9 (respectively, J0K 0a K 0c V0n0 t 0): upper level (respectively lower level) rotational and vibration–torsion assignments. y˜ and Int: calculated (this work) and measured (Ref. 21) linepositions and line intensities. R: ratios of the calculated (this work) to measured (21) line intensities.

Copyright © 1999 by Academic Press

159

160

KLEE ET AL.

TABLE 3 Comparison Between the Results of the Present Line-Intensity Calculations and the Line-Intensity Measurements Performed by Sams (22) at 294 K

Note. J9 K9a K9c V9n9 t 9 (respectively J0K 0a K 0c V0n0 t 0): upper level (respectively lower level) rotational and vibration–torsion assignments. y˜ and Int: calculated (this work) and measured (Ref. 22) line positions and line intensities. R: ratios of the calculated (this work) to measured (22) line intensities.

5 200 cm22 atm21) and by Valero et al. (20) (S 7.9mm 5 375 6 17 cm22 amagat21 which leads to S 7.9mm(296 K) 5 346 cm22 atm21 (65%) when using the conversion factor given in Ref. (30)). We also compared our linelist to the individual line intensities measurements performed by May (21) and by Sams (22) at T 5 295 K and T 5 294 K, respectively: the results of these comparisons are given in Tables 2 and 3, respectively. The measurements of Sams (22) lead to almost no significant discrepancy on the average with the present linelist (I THIS WORK/I SAMS 5 1.051(32)). On the other hand the present lines intensities are on the average slightly weaker (I THIS WORK/I MAY 5 0.928(11)) than those of May (21). The same ratio is obtained when comparing our results to those of the HITRAN database (23), but this is normal since the HITRAN intensities were scaled to those of May. On the other hand, one should recall that the HITRAN linelist was generated from the line positions given in Ref. (13) and then involves only the torsional subbands with n 5 0 with significant deficiencies for some spectral regions (see Figs. 1– 4 in Ref. (15).

Conclusion High resolution Fourier transform spectra were recorded in a spectral region covering both the R branch of the torsion rotation band at low wavenumbers and the P branch of the n6 band at high wavenumbers. Because an absolute calibration of the line intensities in the torsion rotation band was performed using the H2O2 dipole moment measured by Stark effect, it has been possible, by comparing H2O2 absorption features for individual lines in both spectral regions, to determine in absolute the line intensities in the n6 band. These intensities are in satisfactory agreement with other recent high resolution measurements (21, 22).

ACKNOWLEDGMENTS The experimental work at Giessen was financially supported by the Deutsche Forschungsgemeinschaft (DFG). The authors are also very grateful to Dr R. L. Sams for the communication of his results on line intensities measurements for the n6 band of H2O2.

Copyright © 1999 by Academic Press

ABSOLUTE INTENSITIES FOR THE n6 BAND OF H2O2

REFERENCES 1. D. D. Davies, Can. J. Chem. 52, 1405–1414 (1974). 2. P. S. Connel, D. J. Wuebbles, and J. S. Chang, J. Geophys. Res. 90, 10 726 –10 732 (1985). 3. K. V. Chance, D. G. Johnson, W. A. Traub, and K. W. Jucks, Geophys. Res. Lett. 18, 1003–1006 (1991). 4. J. H. Park and B. Carli, J. Geophys. Res. D96, 22 535–22 541 (1991). 5. J. D. Rodgers and J. J. Hillman, J. Chem. Phys. 75, 1085–1090 (1981). 6. J. D. Rodgers and J. J. Hillman, J. Chem. Phys. 76, 4046 – 4055 (1982). 7. R. D. May and C. D. Webster, J. Geophys. Res. 94, 16 343–16 350 (1989). 8. J. C. Larsen, C. P. Rinsland, A. Goldman, D. G. Murcray, and F. J. Murcray, Geophys. Res. Lett. 10, 663– 666 (1985). 9. J. T. Hougen, Can. J. Phys. 62, 1392–1402 (1984). 10. J.-M. Flaud, C. Camy-Peyret, J. W. C. Johns, and B. Carli, J. Chem. Phys. 91, 1504 –1510 (1989). 11. J.-M. Flaud and A. Perrin, in “Vibration-Rotational Spectroscopy & Molecular Dynamics,” “Advanced Series in Physical Chemistry” (D. Papousek, Ed.), World Scientific Publishing Company, Singapore, 1997. 12. C. Camy-Peyret, J.-M. Flaud, J. W. C. Johns, and M. Noel, J. Mol. Spectrosc. 155, 84 –104 (1992). 13. J. J. Hillman, D. E. Jennings, W. B. Olson, and A. Goldman, J. Mol. Spectrosc. 117, 46 –59 (1986). 14. A. Perrin, J.-M. Flaud, C. Camy-Peyret, A. Goldman, F. J. Murcray, and R. D. Blatherwick, J. Mol. Spectrosc. 142, 129 –147 (1990). 15. A. Perrin, A. Valentin, J.-M. Flaud, C. Camy-Peyret, L. Schriver, A. Schriver, and Ph. Arcas, J. Mol. Spectrosc. 171, 358 –373 (1995). 16. V. Malathy Devi, C. P. Rinsland, M. A. H. Smith, D. Ch. Benner, and B. Fridovich, Appl. Opt. 25, 1844 –1847 (1986). 17. A. Perrin, J.-M. Flaud, C. Camy-Peyret, R. Schermaul, M. Winnewisser,

18. 19. 20. 21. 22. 23.

24. 25. 26. 27.

28.

29. 30.

161

J.-Y. Mandin, V. Dana, M. Badaoui, and J. Koput, J. Mol. Spectrosc. 176, 287–296 (1996). E. A. Cohen, and H. M. Pickett, J. Mol. Spectrosc. 87, 582–583 (1981). H. Niki, P. D. Maker, C. M. Savage, and L. P. Breitenbach, Chem. Phys. Lett. 73, 43– 46 (1980). F. P. J. Valero, D. Goorvitch, F. S. Bonomo, and R. W. Boese, Appl. Opt. 20, 4097– 4101 (1981). R. D. May, J. Quant. Spectrosc. Radiat. Transfer 45, 267–272 (1991). R. L. Sams, personal communication (1995). L. S. Rothman, C. P. Rinsland, A. Goldman, S. T. Massie, D. P. Edwards, J.-M. Flaud, A. Perrin, V. Dana, J.-Y. Mandin, J. Schroeder, A. McCann, R. R. Gamache, R. B. Wattson, K. Yoshimo, K. Chance, K. Jucks, L. R. Brown, V. Nemtchinov, and P. Varanasi, J. Quant. Spectrosc. Radiat. Transfer 60, 665–710 (1998). J. W. C. Johns, Microchim. Acta 3, 171–188 (1987). M. Birk, D. Hausamann, G. Wagner, and J. W. Johns, Appl. Opt. 35, 2971–2985 (1996). M. Birk, G. Wagner, J.-M. Flaud, and D. Hausamann, J. Mol. Spectrosc. 163, 262–275 (1994). J.-M. Flaud, C. Camy-Peyret, and R. A. Toth, “Water Vapour Line Parameters from Microwave to Medium Infrared,” Pergamon Press, Oxford, 1981. C. Camy-Peyret, and J.-M. Flaud in “Molecular Spectroscopy Modern Research” (K. Narahari Rao, Ed.), pp. 69 –110 Vol. 3, Academic Press, New York, 1985. P. M. Chu, S. J. Wetzel, W. J. Lafferty, A. Perrin, J.-M. Flaud, Ph. Arcas, and G. Guelachvilli, J. Mol. Spectrosc. 189, 55– 63 (1998). M. A. Smith, C. P. Rinsland, V. Malathy Devi, L. S. Rothman, and K. Narahari Rao, in “Spectroscopy of the Earth’s Atmosphere and Insterstellar Medium” (K. Narahari Rao and A. Weber, Eds.), Academic Press, New York, 1991.

Copyright © 1999 by Academic Press