First identification of the a1Deltag-X3Sigmagminus electric quadrupole transitions of oxygen in solar and laboratory spectraIntroductionFirst identification in atmospheric solar spectraLaboratory measurementsTheoretical background and line list calculationDiscussionConclusionAcknowledgmentsSupplementary materialReferencesFirst identification of the a1Dg2X3Sgelectric quadrupole transitionsof oxygen in solar and laboratory spectraIouli E. Gordona,, Samir Kassib, Alain Campargueb, Geoffrey C. TooncaHarvard-Smithsonian Center for Astrophysics, Atomic and Molecular Physics Division, Cambridge, MA 02138, USAbUniversite´Joseph Fourier/CNRS, Laboratoire de Spectrome´trie Physique, 38402 Saint–Martin–d’Heres, FrancecJet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USAarticle infoArticle history:Received 19 November 2009Received in revised form6 January 2010Accepted 7 January 2010Keywords:Electric quadrupoleSinglet deltaOxygenSolar spectrumCavity ring down spectroscopyabstractElectric quadrupole transitions in the a1Dg2X3Sgband of16O2near 1.27 mmarereported for the first time. They were first detected in atmospheric solar spectraacquired with a ground-based Fourier transform spectrometer (FTS) in Park Falls, WI.Subsequently high-sensitivity CW—cavity ring down spectroscopy (CW-CRDS) experi-ments were carried out at Grenoble University in the 7717– 7917 cm1region in orderto provide quantitative intensity information for the electric quadrupole transitions.Measured intensities were used as input data for the calculation of the complete list ofelectric quadrupole transitions withDJ= 7 2, 71 and 0. The calculation was carried outfor the intermediate coupling case and assuming that these transitions are possible onlythrough mixing of theO=0 compon ent of the ground electronic state and b1Sþgstateinduced by spin–orbit coupling. The calculated line list agrees well with experimentalmeasurements and was used to improve the residuals of the fitted solar atmosphericspectrum. Emission probability for the electric quadrupole band was determined to be(1.027 0.10) 106s1.& 2010 Elsevier Ltd. All rights reserved.1. IntroductionElectronic transitions of oxygen have been a subject ofover a hundred theoretical, laboratory and field (atmo-spheric) studies for over a century. Of particular interestare transitions that involve X3Sg, a1Dgand b1Sþgelectro-nic states that arise from the ground electron configura-tion of oxygen. Since all three of these states have geradesymmetry, only magnetic dipole (M1) and electric quad-rupole transitions (E2) are possible between them. Themajority of the previous works have reported themagnetic dipole components with the exception ofstudies of the so-called Noxon band ðb1Sþg2a1DgÞ near1.91mm [1] that is purely electric quadrupole in nature.Also electric quadrupole transitions within the groundelectronic state have been known for a long time [2–4]and are already tabulated in the HITRAN database [5].Studies of the singlet–triplet electric quadrupole transi-tions are relatively limited. The b1Sþg2X3Sgelectricquadrupole transitions near 760 nm (henceforth termedthe A-band) were first observed by Brault [6] whoidentified eight lines in the solar spectrum. The firstlaboratory observation of a single line was reported in thework of Naus et al. [7], and very recently Long et al. [8]have carried out high-quality quantitative measurementsof nine E2 transitions in the A-band using frequency-stabilized cavity ring-down spectroscopy (FS-CRDS) tech-nique. The latter study allowed determination of the bandstrength of the electric quadrupole component of the A-band which was reported to be equal to only 8 106of the magnetic dipole component intensity. No electricquadrupole lines in the a1Dg2X3Sgband near 1.27mm(hereon referred to as the 1Delta band) were identifiedbefore. Even sensitive CRDS studies such as those byContents lists available at ScienceDirectjournal homepage: www.elsevier.com/locate/jqsrtJournal of Quantitative Spectroscopy &Radiative TransferARTICLE IN PRESS0022-4073/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.doi:10.1016/j.jqsrt.2010.01.008Corresponding author. Tel.: +1 617 496 2259.E-mail address: [email protected] (I.E. Gordon).Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 1174–1183Newman et al. [9,10] did not report the E2 contributionsdue to their anticipated weakness. Indeed, the theoreticalpredictions of the emission probability for the E2component in the classic work of Klotz et al. [11] givesvalue of 5 107s1which is significantly lower than thevalue measured for the magnetic dipole component2.19 104s1[9]. However, even this value (which isunderestimated as will be shown in the discussionsection) is still larger than the one predicted in the samepaper for the quadrupole transitions for the A-band(1.55 107s1). Moreover, Sveshnikova and Minaev[12] have predicted even larger value of 2.3 106s1,although in the later publication Minaev and Agren [13],corroborated the value from Klotz et al. [11]. Thus,considering that 1Delta transitions are located at a lowerwavenumber, one would expect their E2 band intensity tobe at least an order of magnitude stronger than that of theA-band (see Eq. (19) of Ref. [14] ).In accordance with this logic, we report the identifica-tion of electric quadrupole transitions in the 1Delta bandin solar atmospheric spectra and their subsequent con-firmation in the laboratory using cavity ring downspectroscopy. In addition, an extensive line list of all E2lines with intensities larger than 1030cm/molecule wasgenerated assuming intermediate coupling Hund’s case(a–b). This list was used in combination with a magneticdipole line list in order to simulate the solar atmosphericspectrum and yielded improved residuals.It is our pleasure to contribute this article to the specialissue to celebrate Laurence Rothman’s jubilee. His con-tributions to the spectroscopy of oxygen [2,3,15–19] aremany and invaluable as reference data for atmospheric,planetary and industrial applications. It makes this paperparticularly appropriate for the special issue that theaddition of the oxygen molecule to the AFGL spectraldatabase was Laurence Rothman’s first task when hestarted to work on that project.2. First identification in atmospheric solar spectraSolar spectra were