13 16

nd

a1

inu

sso-

Journal of Molecular Spectroscopy203,196–199 (2000)doi:10.1006/jmsp.2000.8166, available online at http://www.idealibrary.com on

First Observation of the Forbidden k–X Transition of C O

Jacob Baker*,1 and Franc¸oise Launay†

*Division of Environmental Health and Risk Management, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom; a†DAMAp et UMR 8588 du CNRS, Observatoire de Paris, Section de Meudon, 92195 Meudon Cedex, France

Received April 21, 2000; in revised form May 24, 2000

A weak rotationally resolved absorption band of13C16O has been identified from photographic spectra at 1078 Å andassigned to the forbidden transition from theX1S1 (v 5 0) ground state to thek3P (v 5 3) valence state. It is similar instructure to the corresponding band of12C16O but due to the fact it was photographed at a relatively low pressure only the twostrongest rotational branches could be positively identified. It is the first time that the forbiddenk–X transition has beenobserved for13C16O. The measured isotopic shift in the band origin provides independent confirmation of the new vibrationalassignment of thek state by G. Berden, R. T. Jongma, D. Van der Zande, and G. Meijer (J. Chem. Phys.107, 8303–8310(1997)). © 2000 Academic Press

INTRODUCTION 12 (4, 5). This perturbation, known as an accidental predi

anndwasingt imph

ngtee

m–ich

bph

ptiotioCon th

ithdhsi

va(

ngtioowiatis

darklence

nsf ls

phic

in

yof

-l vialevelum-thistione

t ontory.

for-enal

de

T ns pec-t pec-t at a.ac

Carbon1 monoxide is believed to be the most abundmolecule after molecular hydrogen in interstellar space awidely used as a tracer for mapping out the density and flomolecules in the interstellar medium. To be used reliablytracer, it is necessary to understand the processes governabundance relative to that of hydrogen. One of the mosportant processes governing its abundance is its rate oftodestruction in the interstellar radiation field. The waveleregion of relevance for photodestruction of CO lies betwthe first dissociation limit of CO (111.782 nm–89 460 cm21)

nd the ionization limit of atomic hydrogen (91.175 n09 679 cm21). This latter boundary marks the point at wh

the radiation field is drastically reduced due to shieldinghydrogen atoms. In the wavelength region of relevance,todissociation of carbon monoxide is dominated by absorinto rotationally resolved bands followed by predissociarather than through direct transition to repulsive states.sequently, radiative transfer and shielding are important ichemical modeling of CO in the interstellar medium.

The effects of self- and mutual shieldings coupled wisotope shifts in transition energy are expected to leaisotopic fractionation (1–3). This is the process by whicsotopic differences in photodissociation rates can lead toificant variations in isotope ratios from net backgroundes. The modeling studies of Van Dishoeck and Black3)

show that theE–X (1–0) band is most effective in contributito isotopic fractionation. The upper state of this absorpband, theE1P (v 5 1) level, has localized perturbations at lrotational quantum numbers that increase their predissocrates. For the normal isotopomer12C16O, these perturbationresult in increased predissociation rates forJ 5 7, 9, 10, and

1 To whom correspondence should be addressed. E-mail: j.baker@bham

1960022-2852/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

tis

ofaits-o-hn

yo-n

nn-e

to

g-l-

n

on

ciation, is caused by an interaction with a predissociatedstate. This dark state has recently been identified as a va3P state and given the labelk (4, 6). Spectroscopic transitiorom the ground state of12C16O to different vibrational leveof thek state have been seen in VUV absorption spectrograstudies and a double-resonance laser study (6–8). Transitionsto this state from the metastablea3P have also been seentwo laser-based studies, by Wan and Langhoff (9) (their jstate), and more recently by Berdenet al.(10). This latter studprepared both12C16O and13C16O in a single quantum levelthe a3P (v 5 1, J 5 1, V 5 1, parity 5 f ) state within acold molecular beam and then measured the 11 1 resonanceenhanced multiphoton ionization from this prepared levethek state. They observed a new lower energy vibrationalof the k state and showed that the previous vibrational nbering of thek state had to be incremented by one unit. Inarticle, we report on a weak band in the VUV absorpspectrum of the isotopomer13C16O, which is assigned to thforbidden transition from the ground state to thev 5 3 vibra-tional level of thek state.

SPECTROSCOPIC DATA

This study is an extension of previous work carried outhe 10.68-m VUV spectrograph at the Meudon ObservaPreviously, we have reported on bands corresponding tobidden transitions from theX1S1 (v 5 0) ground state to thv 5 1, 3, 4, and 6 vibrational levels, using the new vibrationumbering, of thek3P state of normal carbon monoxi12C16O, recorded over the pressure range of 0.1–1.5 Torr (6, 7).

he strongest of these bands is thek–X (3–0) band, which catill clearly be seen relatively strongly in an absorption srum photographed at 0.07 Torr of CO. This absorption srum suggested that this band should also be evident.uk.

thd tt–l

ptiondth

n

n ress

p ynuu

9lythe

r rr-

wetherttw

yt

X1S1 ground state to thef- ande-parity levels, respectively, of theessurednot

o wectrumcture

htedrmed

ttedc

tere

ndo

lues

ld

t fofr

nde-nt of

or

he

wereork.

orbitistor-

lar

rvedper-

ional.4perener-using

TABLE 1

197OBSERVATION OF THE FORBIDDENk–X TRANSITION OF 13C16O

pressure of 0.02 Torr within the spectrograph, which washighest pressure in which we have previously photographeisotopomer 13C16O, although at that time only the singlesinglet transitions were of interest (11, 12). Indeed, a carefuexamination of a photographic plate containing the absorspectrum of13C16O at 0.02 Torr revealed a very weak basimilar to that previously observed and now assigned tok–X (3–0) band of12C16O but displaced to lower transitioenergy. In this article, we report on this weak band.

The experimental details, including calibration, have beeported elsewhere (6, 11). We estimate the relative error to be lthan 0.3 cm21 when comparing13C16O and 12C16O spectral lineositions. For both13C16O and 12C16O, an unidentified impurit

atomic absorption line was observed in the background contivery close to thek–X (3–0) band. For12C16O, it was identifiedwithin the rotational structure of thek–X (3–0) band at 92 790.3cm21. The measured position in the13C16O spectrum differed onby 0.1 cm21 from this value, well within the above estimate forelative error in isotopic line positions. The13C16O gas was pu-

chased from the Commissariat a` l’Energie Atomique with a cetified 99% isotopic purity.

RESULTS/ANALYSIS

Table 1 gives the line positions and assignments for theabsorption band of13C16O observed at 1078 Å and located onlonger wavelength side of the strongE–X (0–0) band. It is similain structure to the corresponding band of12C16O but due to the facit was photographed at a relatively low pressure only thestrongest rotational branches,QQ and RR, could be positivelidentified. These branches correspond to transitions from

Line Positions for the k3P (vv 5 3)–X1S1 (vv 5 0)Transition of 13C16O

Note.All values are in units of cm21. Values between parenthesesare estimated errors. o2c signifies observed2 calculated value.

* Overlapped lines.

Copyright © 2000 by

ehe

n

e

e-

m

ak

o

he

f2 spin–orbit component of thek3P upper state. Although the lincould be clearly identified on the photographic plate and meawith the Meudon photoelectric comparator, the lines didappear clearly on positive prints made with Ilford paper and sdo not reproduce the actual photographic absorption spehere. We shall, however, discuss the rotational intensity struof the band in the next section.

In a preliminary analysis of the observed band, a weigleast-squares fit of the measured lines positions was perfowhere the3P excited state rotational energy levels were fito the second eigenvalue (i.e., thef 2 component) of the effe-ive 3P Hamiltonian matrix given by Brown and Merer (13),while the X1S1 ground state rotational energy levels wfitted simply to the expressionF( J) 5 BJ( J 1 1) 2D( J( J 1 1))2 usingB 5 1.837972 cm21 andD 5 5.59231026 cm21 from Ref. (14). The energy position of the ba

rigin, T30, and the upper state rotational constant,B, werevaried in the fit, while other constants were fixed to the vaof the corresponding level of12C16O (6). The band originobtained was 92 742.76 0.2 cm21. Interestingly, using the ovibrational numbering of thek state and isotopic scaling (15) ofhe vibrational constants of12C16O (7), a band origin o92 762.3 cm21 is predicted, giving a large discrepancyapproximately 20 cm21. Incrementing the vibrational numbe-ing by one unit removes this discrepancy. This provides ipendent confirmation of the recent vibrational reassignmethe k state (Berdenet al. (10)).

Berdenet al. (10), in their 11 1 REMPI study, observed f13C16O six transitions from thea3P (v 5 1, J 5 1, V 5 1,parity 5 f ) state to thek (v 5 3) state, corresponding to tfollowing levels in thek (v 5 3) state:f 1: J( f ) 5 0, J(e) 51, J( f ) 5 2; f 2: J(e) 5 1, J( f ) 5 2; f 3: J( f ) 5 2, wherethe e/f parities are given in parentheses. These levelsincluded in the final fit of the data obtained in the current wInclusion of this data made no difference to theT30 and Bvalues obtained but enabled the determination of the spin–constant and lambda-doubling constant. The centrifugal dtion constant,D, was fixed at 9.03 1026 cm21, the isotopicallyscaled value of the corresponding level of12C16O (6). Theresults of this final fit are given in Table 2. All other molecuparameters were fixed to zero.

DISCUSSION

To explain why some lines of the weak band of13C16O wereobserved while others were not, i.e., to explain the obserotational intensity distribution, some simulations wereformed, similar to those performed for12C16O (6). Figure 1isa simulation of the absorption band assuming a rotattemperature ofT 5 298 K and a Gaussian linewidth of 0cm21 (FWHM). The line positions were calculated from upand lower state term energy differences, where the termgies were determined as described in the previous section

Academic Press

d t

di . It me

, to

d a

Thisgh

t theoto-

Fig.rota-theure 2here

hefrom

ne

tronly re--

to theireynd atnds,

romes.

TABLE 2

thet

gion.pectra.

198 BAKER AND LAUNAY

the upper state molecular constants given in Table 2 anknown ground state constants (14). 3P–1S1 linestrength fac-tors were taken from Kovacs (16), where the3P state is treaten the intermediate Hund’s coupling case (a/b) frameworkhese linestrength formulas, there is an adjustable para

s 5 D/E, whereD andE correspond, in the present casethe effective transition moments arising from mixing of1S1

and 1P character into the3P state, respectively. In Ref. (6), aqualitative simulation for thek–X (3–0) band of12C16O (therecalledk–X (2, 0) in the old vibrational numbering) gaves 520.3. A reexamination of the intensity structure of this banlower pressures suggested that a value ofs 5 20.1 gave a

Molecular Constants for the k3P (vv 5 3)State of 13C16O

Note.All molecular constants are in units of cm21.Values between parentheses are errors, to one standarddeviation, in the least-significant figure. ForT30 theerror given is estimated rather than obtained from thefit since the latter smaller value does not take intoaccount possible calibration errors.s is the standarddeviation of the weighted fit. See text for furtherdetails.

FIG. 1. A simulation of the forbiddenk3P (v 5 3)–X1S1 (v 5 0)absorption band of13C16O. See text for details. Only rotational lines abovehreshold represented by the dashed horizontal line were observed.

Copyright © 2000 by

he

nter

t

better representation, and this value was used for Fig. 1.implies that thek–X (3–0) band gains intensity mainly throua k3P–1P interaction, almost certainly theE1P state.

The observed band was very weak, suggesting thaabsorption was close to the threshold sensitivity of the phgraphic plate. In fact, the majority of the lines predicted in1 must have been below the threshold since only a fewtional lines were clearly visible. In fact, only lines abovedashed horizontal threshold of Fig. 1 were observed. Figshows the spectrum above this threshold more clearly wthe observed lines are marked by an asterisk.

We note that the old (incorrect) vibrational numbering of tkstate was used only as a notational convenience. It originatedthe work of Wan and Langhoff (9) where triplet–triplet absorptiofrom the metastablea3P state of12C16O was measured. In thstudy, thea3P state was produced in a “hot” state by elecbombardment. They observed a new progression of partialsolved red-degraded bands (their “j0–a0” to “ j4–a0” bands) corresponding to thek3P (v 5 1–5)–a3P (v 5 0) transitions. How-ever, at that time they believed the upper state to be thej 3psRydberg state. This state has an expected origin band closej0–a0 band (which is in fact thek–a (1–0) band) and hence thmay not have been aware of the possibility of another balonger wavelengths. This, together with overlap with hot basuch as thec–a (0–1) band, accounts for the actualk3P–a3P(0–0) absorption band being overlooked in that study.

In our spectrographic studies, observation of thek–X for-bidden transition depends on intensity borrowing arising fthe interaction of thek3P state with neighboring singlet statThe k–X (0–0) band of12C16O is expected at 90 334.0 cm21

(110.70 nm) (10) in our spectrographic data for12C16O. Thisregion lies in a spectral window to the red of theB–X (2–0)

FIG. 2. An expanded view of Fig. 1 showing the above-threshold reOnly the lines marked by an asterisk were observed in the photographic s

Academic Press

band at 90 988.2 cm21. An examination of this region foral

chasnk–tio

ndoto

tl

ctio

ndurenfir

the

REFERENCES

.

stas,

1 .

1 d J.

1 ).1 ).1 s,

1 ectra

1 les,”

199OBSERVATION OF THE FORBIDDENk–X TRANSITION OF 13C16O

12C16O at the highest pressures employed of 1.5 Torr reveno band. This implies that thek–X (0–0) band has a mulower cross section than the other observed bands, suchk–X (3–0) band. This is most likely due to poorer FraCondon overlap with the ground state and weaker interacof the k (v 5 0) level with singlet states.

CONCLUSION

A faint absorption band of13C16O has been observed aassigned to the forbiddenk–X (3–0) band. Only the twstrongest branchesQQ andRR were observed correspondingransitions from theX1S1 ground state to thef- ande-parityevels, respectively, of thef 2 spin–orbit component of thek3Pstate. The band gains intensity mainly through the interaof thek (v 5 3) level with a1P state, almost certainly theE1Pstate. The band structure is similar to the analogous ba12C16O except that it is shifted to lower energy. The measisotopic shift in the band origin provides independent comation of the new vibrational assignment of thek state (10).

ACKNOWLEDGMENTS

We are indebted to Franc¸ois Rostas for supporting and encouragingwork. We thank Maurice Benharrous for technical assistance.

Copyright © 2000 by

ed

the

ns

n

ofd-

1. A. E. Glassgold, P. J. Huggins, and W. D. Langer,Astrophys. J.290,615–626 (1985).

2. Y. P. Viala, C. Letzelter, M. Eidelsberg, and F. Rostas,Astron. Astrophys193,265–272 (1988).

3. E. F. Van Dishoeck and J. H. Black,Astrophys. J.334, 771– 802(1988).

4. J. Baker, J. L. Lemaire, S. Couris, A. Vient, D. Malmasson, and F. RoChem. Phys.178,569–579 (1993).

5. P. Cacciani, W. Hogervorst, and W. Ubachs,J. Chem. Phys.102,8308–8320 (1995).

6. J. Baker and F. Launay,J. Mol. Spectrosc.165,75–87 (1994).7. J. Baker,J. Mol. Spectrosc.167,323–333 (1994).8. A. Mellinger and C. R. Vidal,J. Chem. Phys.101,104–110 (1994).9. B. N. Wan and H. Langhoff,Z. Phys. D: At., Mol. Clusters21: 245–249

(1991).0. G. Berden, R. T. Jongma, D. Van der Zande, and G. Meijer,J. Chem

Phys.107,8303–8310 (1997).1. M. Eidelsberg, J.-Y. Roncin, A. Le Floch, F. Launay, C. Letzelter, an

Rostas,J. Mol. Spectrosc.121,309–336 (1987).2. M. Eidelsberg and F. Rostas,Astron. Astrophys.235,472– 489 (19903. J. M. Brown and A. J. Merer,J. Mol. Spectrosc.74, 488 – 494 (19794. G. Guelachvili, D. De Villeneuve, R. Farrenq, W. Urban, and J. VergeJ.

Mol. Spectrosc.98, 64–79 (1983).5. G. Herzberg, “Molecular Spectra and Molecular Structure, Vol. 1. Sp

of Diatomic Molecules,” Van Nostrand, New York, 1950.6. I. Kovacs, “Rotational Structure in the Spectra of Diatomic Molecu

Hilger, London, 1969.

Academic Press