Accurate High-N Rest Frequencies for CO+, an ... - ACS Publications

Jul 1, 2013 - I. Physikalisches Institut, Universität zu Köln, 50937 Köln, Germany. ‡ Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121...
0 downloads 0 Views 859KB Size
Article pubs.acs.org/JPCA

Accurate High‑N Rest Frequencies for CO+, an Ideal Tracer of PhotonDominated Regions Silvia Spezzano,*,†,‡ Sandra Brünken,† Holger S. P. Müller,† Gabriele Klapper,† Frank Lewen,† Karl M. Menten,‡ and Stephan Schlemmer† †

I. Physikalisches Institut, Universität zu Köln, 50937 Köln, Germany Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany



ABSTRACT: The submillimeter-wave rotational spectra of CO+, 13CO+ and C18O+ in the v = 0 and 1 vibrational states were measured through a hollow cathode dc discharge in a cryogenic cell cooled to liquid nitrogen temperature. In addition, a few transitions of the main isotopic species have been measured between 1.1 and 1.3 THz. An updated isotopically invariant fit, including Born−Oppenheimer breakdown corrections, is presented: the derived set of independent molecular parameters, valid for all the isotopologues of the molecule included in the fit, allows to predict the rotational spectrum with calculated 1σ uncertainty of 280 kHz at 2 THz.



INTRODUCTION CO is the cation of the second most abundant molecule in space, carbon monoxide. However this small, open-shell cation is not as widespread as its neutral counterpart. CO+ is in fact easily destroyed as it reacts almost on every collision with H2 and H. Its abundance becomes important only in hot layers of photodissociation regions (PDRs), where hydrogen is mostly atomic.1 CO+ was successfully detected toward several PDRs, some of them associated with planetary nebulae,2−5 where it appears to be a good tracer of the hot gas. Thus far, CO+ emission lines have been observed with a rotational excitation temperature of 10 K, much lower than the temperature of the surrounding medium in a PDR, ranging from hundreds to thousands of Kelvin.2 The processes that are responsible for this anomalous excitation are still poorly understood: Stäuber and Bruderer6 suggest that CO+ may be excited upon formation. To test their model, observations of high-N rotational transitions in the terahertz region are needed, probing the population of higher energy levels. Up to date, CO+ has been observed only in the lowest rotational transitions, normally N = 2 → 1 and 3 → 2 since N = 1 → 0 is not observable because of atmospheric O2. The recent development of facilities such as Herschel (HIFI operates between 480 and 1910 GHz), SOFIA (GREAT operates in the 1.25−1.5 and 1.82−1.92 THz range), and ALMA (in full science it will operate in all the atmospheric windows accessible from the ground between 80 and 900 GHz), has opened the far-infrared region to radioastronomy: those facilities are well suited to study the rotational excitation of CO+. Prior to this work, precise THz measurements of CO+ were missing: Dixon et al.7 measured for the first time the two components of the N = 0 → 1 transition, and in the work of Sastry et al.8 the measurements are extended up to 470 GHz.

Isotopically substituted species and vibrational excited states have also been studied:9−11 13CO+ and C18O+ in the lowest rotational transition for v = 0, and the main isotopic species in v = 0−4, C18O+ v = 0,1, 13CO+ v = 0,1, and 13C18O+ v = 0. Savage and Ziurys12 measured transition frequencies of the main isotopic species in the 200−590 GHz range with a newly built velocity modulation spectrometer. The submillimeter spectrum of 13CO+ has been extensively studied by Klapper.21 The only far-infrared (supra-THz) measurements have been made by van den Heuvel et al.,13 who reported the detection of a single transition of CO+ with a far-infrared (FIR) sideband spectrometer with an accuracy of 1 MHz at 1.06 THz. Motivated by the lack of accurate data at high frequencies, this paper combines the new data set on CO+ (see Table 1) with all previous high resolution data to perform a

+

© 2013 American Chemical Society

Table 1. Summary of New Laboratory Measurements isotopic species 12 16

+

C O

v

spectral range (GHz)

rotational transition

0

353−943 1110−1300 350−351 700 336−337 338−901

10 4 2 1 2 19

1 12 18

C O+ C O+

13 16

0 0

Note: Including the data of Klapper21 Special Issue: Oka Festschrift: Celebrating 45 Years of Astrochemistry Received: December 21, 2012 Revised: May 31, 2013 Published: July 1, 2013 9814

dx.doi.org/10.1021/jp312616u | J. Phys. Chem. A 2013, 117, 9814−9818

The Journal of Physical Chemistry A

Article

kHz, and then 2f-demodulated by a digital lock-in amplifier, leading to the observed 2nd derivative line shape. At 1.1−1.3 THz, the radiation is generated by a commercial multiplier chain (Virginia Diodes, Inc.), which generates the 72nd harmonic of a 18 GHz synthesizer signal (Rhode and Schwarz). The measured lines are reported in Table 2. The accuracy of the lines was estimated on the basis of signal-to-noise ratio and line shape.

global analysis. The aim is to obtain spectroscopic parameters for predictions at high frequencies (THz) with the accuracy needed for astronomical identification.



EXPERIMENTAL DETAILS For the production of CO+, a discharge cell as described in the work of Gendriesch et al.14 was used, modified with a long hollow cathode to optimize the production of positive ions.13 Typically, 80 mA discharge current was used, through pure CO at a pressure of about 0.3 mbar. The experimental conditions were optimized for the production of CO+, and a compromise had to be made between low pressure conditions and the production of the ions in a detectable amount. Although pressure broadening was observed in the CO precursor (J = 3 − 2) line, no frequency shift was detectable. C18O+ was measured in natural abundance, and 13CO+ was measured in natural abundance in the new measurements, while Klapper21 used a 99% enriched sample. The cell walls were cooled to liquid nitrogen temperature for all measurements. Due to the electric field in the dc discharge, molecular ions are accelerated, and this drift produces a frequency shift that is not trivial to take into account. In order to cancel this shift, all new measurements were taken in a double pass configuration: by letting the radiation counterpropagate in the cell, the effect of the drift velocity of the ions in one direction cancels out. It was found that in a single-pass arrangement, the shifts are on the order of 30 kHz at 350 GHz, well within the measurement uncertainties, so that single-pass measurements by Klapper21 were not needed to be corrected prior to the analysis. The radiation sources employed in the 300−900 GHz range are phase-locked backward wave oscillators (BWO). A part of the BWO radiation is mixed with an appropriate harmonic of a commercial synthesizer in a Schottky diode multiplier mixer to produce an intermediate frequency signal (IF). The phase-lock loop (PLL) compares the IF with a stable reference, in our case the output of a Rubidium atomic clock, to derive a phase error signal which is processed by a loop filter and sent to the power supply of the BWO.15 An example of a line measured with a BWO is shown in Figure 1. The signal, detected by a liquid Helium cooled InSb bolometer, is frequency modulated at f = 8.7



ANALYSIS The open shell CO+ (2Σ+) has a considerably large dipole moment of 2.61 D.16 Owing to the presence of an unpaired electron, CO+ has an electronic spin of 1/2 so each rotational level N is split in two, J = N ± 1/2. The allowed transitions follow the ΔJ = 0, ± 1 selection rule, but the transitions with ΔJ = 0 are much weaker than the other two, and have intensities below the noise level for most of the transitions. In the case of 13CO+ there is further splitting due to the presence of the nuclear spin of I = 1/2 of 13C, so the quantum numbers are expessed in this way: J = N ± 1/2, and F = J ± 1/2. The approach introduced by Dunham17 to describe the potential of a diatomic uses the Born−Oppenheimer approximation. The ro-vibrational energy levels are expressed in the following way: E(v , J )/h =



∑ Yij⎝v + ⎜

ij

1 ⎞⎟i j J (J + 1) j 2⎠

with Yij being the so-called Dunham coefficients. According to the Born−Oppenheimer approximation, it is possible to separate the electronic and nuclear wave functions assuming that the electrons instantaneously follow the motion of the nuclei, given the big difference in their respective masses. The electronic potential is assumed to be independent of the masses of the nuclei and their motion, so it is the same for every isotopologue. Therefore, within this approximation, the equilibrium bond lengths do not change in different isotopic species of a given diatomic molecule. Watson18 studied the isotopic dependence of the Dunham coefficients, and derived the following way to express those coefficients in terms of mass-independent parameters Uij and corrections Δij to the Born−Oppenheimer approximation ⎛ meΔijA meΔijB ⎞ −(i + 2j)/2 ⎜ ⎟μ Yij = Uij⎜1 + + MA MB ⎟⎠ ⎝

where Uij is an isotopic invariant coefficient, me is the electron mass, MA and MB are the atomic masses, μ is the charge modified reduced mass introduced by Watson, and Δij are the Born− Oppenheimer breakdown correction terms. All masses used in this work are taken from the 2003 atomic mass evaluation (AME), except for 18O, which is taken from a more recent Penning-trap experiment.19 An isotopically invariant fit has been performed according to Watson’s formula with SPFIT and SPCAT20 using the data from this work, and from previous works.8,11,13,21 All measurements included in the fit are summarized in Table 2. The molecular parameters derived from this fit are reported in Table 3, where Yij are the Dunham coefficients, Δij are the corrections to the Born− Oppenheimer approximation, γij are the spin-rotation interaction constants, and bF and c are the isotropic (Fermi contact interaction) and anisotropic parts of the electron spin-nuclear spin coupling constants, respectively. Instead of Y01, U01/μ was determined in the fit because of the Born−Oppenheimer

Figure 1. Spectrum of the NJ = 25/2 → 37/2 transition of 12C16O+. The line shown is the sum of two scans, with total integration time of less than 3 min. 9815

dx.doi.org/10.1021/jp312616u | J. Phys. Chem. A 2013, 117, 9814−9818

The Journal of Physical Chemistry A

Article

Table 2. Measured Rotational Transition Frequencies of CO+ [MHz] with Uncertainties, Quantum Numbers, and Residuals O−C [kHz] between Observed Frequencies and Those Calculated from the Final Set of Spectrocopic Parameters CO+ v

N → N′

0

0→1 0→1 1→2 1→2 1→2 2→3 2→3 3→4 3→4 4→5 4→5 5→6 5→6 6→7 6→7 7→8 7→8 8→9 9 → 10 9 → 10 10 → 11 10 → 11 0→1 0→1 1→2 1→2 2→3 2→3 5→6 0→1 0→1 1→1 1→1 0→1 0→1

1

2

3 4 0

1

J → J′ 1/2 → 1/2 1/2 → 3/2 3/2 → 3/2 1/2 → 3/2 3/2 → 5/2 3/2 → 5/2 5/2 → 7/2 5/2 → 7/2 7/2 → 9/2 7/2 → 9/2 9/2 → 11/2 9/2 → 11/2 11/2 → 13/2 11/2 → 13/2 13/2 → 15/2 13/2 → 15/2 15/2 → 17/2 15/2 → 17/2 17/2 → 19/2 19/2 → 21/2 19/2 → 21/2 21/2 → 23/2 1/2 → 1/2 1/2 → 3/2 1/2 → 3/2 3/2 → 5/2 3/2 → 5/2 5/2 → 7/2 6/2 → 7/2 1/2 → 1/2 1/2 → 3/2 1/2 → 3/2 3/2 → 5/2 1/2 → 3/2 1/2 → 3/2 C18O+ 1/2 →1/2 1/2 → 3/2 3/2 → 3/2 1/2 → 3/2 3/2 → 5/2 3/2 → 5/2 5/2 → 7/2 1/2 → 3/2 1/2 → 3/2 3/2 → 5/2

0→1 0→1 1→2 1→2 1→2 2→3 2→3 0→1 1→2 1→2

13

CO+

Frequency

O−C

v

N → N′

J → J′

F → F′

Frequency

O−C

117692.360(30)a 118101.812(50)a 235380.046(150)b 235789.641(30)b 236062.553(20)b 353741.223(30)c 354014.257(20)c 471679.213(120)b 471952.343(100)b 589599.236(100)d 589872.224(100)d 707496.506(100)d 707769.401(100)d 825366.363(200)d 825639.665(200)d 943204.603(250)d 943477.836(200)d 1061005.9(10)e 1178767.451(200)c 1179040.392(100)c 1296756.174(100)c 1296483.005(200)c 116553.376(80)a 116960.305(80)a 233509.032(70)a 233780.342(80)a 350320.005(50)c 350591.300(50)c 700924.571(150)d 115411.284(80)a 115814.790(80)a 231221.423(180)a 231490.470(180)a 114665.212(120)a 113511.230(100)a

34 15 −66 57 −11 −49 4 −129 20 −15 −11 37 −48 −81 240 −35 216 −611 −69 −108 74 −114 111 16 18 22 −5 −60 91 −7 82 −36 66 159 −94

0

0→1 0→1 0→1 0→1 1→ 2 1→ 2 1→ 2 2→3 2→3 2→3 2→3 3→4 3→4 3→4 4→5 4→ 5 4→5 4→5 5→6 5→6 5→6 6→ 7 6→7 6→7 7→8 7→8 0→ 1 0→ 1 1→ 2

1/2 → 1/2 1/2 → 1/2 1/2 → 3/2 1/2 → 3/2 1/2 → 3/2 3/2 → 3/2 3/2 → 5/2 3/2 → 5/2 3/2 → 5/2 5/2 → 7/2 5/2 → 7/2 5/2 → 7/2 7/2 → 9/2 7/2 → 9/2 7/2 → 9/2 7/2 → 9/2 9/2 → 11/2 9/2 → 11/2 9/2 → 11/2 9/2 → 11/2 11/2 → 13/2 11/2 → 13/2 11/2 → 13/2 13/2 → 15/2 13/2 → 15/2 15/2 → 17/2 1/2 → 3/2 1/2 → 3/2 3/2 → 5/2

1→0 1→1 1→1 1→2 0→1 1→2 2→3 1→2 2→3 2→3 3→4 2→3 3→4 4→5 3→4 4→5 4→5 5→6 4→5 5→6 5→6 5→6 6→7 7→8 6→7 7→8 0→1 1→2 2→3

112465.938(120)a 112695.175(80)a 112753.480(40)a 112902.557(40)a 225444.382(200)a 225504.854(70)a 225678.183(160)a 338189.096(150)c 338251.478(50)c 338377.449(50)c 338443.745(50)c 450937.978(250)d 451143.635(250)d 451196.212(100)d 563672.549(100)d 563713.663(100)d 563890.319(100)d 563931.766(250)d 676387.165(200)d 676420.085(100)d 676613.565(100)d 789076.889(200)d 789103.863(200)d 789336.965(200)d 901737.536(100)d 901975.512(100)d 111687.554(70)a 111835.716(100)a 223545.426(80)a

−11 32 37 80 −55 −17 −13 143 47 31 33 −231 −210 −51 −231 −78 −74 256 34 59 6 97 194 33 29 −39 −146 15 26

0→ 1 0→ 1 0→ 1

1/2 → 1/2 1/2 → 3/2 1/2 → 3/2

107081.520(80)a 107137.965(60)a 107278.209(60)a

70 18 89

112088.491(40)a 112478.502(40)a 224172.862(180)a 224562.753(60)a 224822.772(90)a 336898.495(100)c 337158.761(100)c 111417.700(60)a 222443.650(140)a 222702.180(70)a

30 91 55 −2 51 −338 55 −36 −5 56

1

13 18

C O+ 1→1 0→ 1 1→ 2

Note: Numbers in parentheses are 1σ uncertainties in kHz. aBogey et al., 1983. bSastry et al., 1981. cThis work. dKlapper, 2001. evan den Heuvel, 1982.

first vibrational correction to the Fermi contact constant, but the value of the parameter was not well determined.

correction. Also the Δ01 were not determined directly, but δX01 = ΔX01 × (me/MX) × (U01/μ) are used in the fit. In the fit, the value of Y03 could not be well determined, so it was calculated in the following way: Y03 = He =

2De 3ωe2



DISCUSSION As can be seen in Table 3, the spectroscopic parameters determined in the present study agree to within their experimental uncertainty with those obtained by Bogey and co-workers,11 but are considerably more accurate. The weighted root-mean-square (rms) of our final global fit including all data is

(12Be2 − ωeαe) = 1.2905 × 10−7 MHz

using the constants given in Bogey et al.,11 and kept fixed during the fit. We tried to extract also the value of bF10 from the fit, the 9816

dx.doi.org/10.1021/jp312616u | J. Phys. Chem. A 2013, 117, 9814−9818

The Journal of Physical Chemistry A

Article

Table 3. Spectroscopic Parameters of 12C16O+ (MHz) U01/μ Y01 ΔC01 ΔO01 Y11 Y21 Y02 Y03 × 10−7 γ00 γ10 γ20 c00 bF00

present work

Bogey et al.11

59270.227(76) 59267.5148(80)b −0.226(14)b,c −1.033(31)b,c −568.503(18) −0.9216(19) −0.189136(17) 1.2905d 273.506(49) −0.856(104) −0.378(37) 150.32(27) 1526.92(74)

59270.406(165)a 59267.5205(103) −0.269(36)a −1.060(28)a −568.492(20) −0.9267(71) −0.18965(21) − 273.550(43) −0.936(91) −0.362(31) − 1525.8(12)

Table 5. Equilibrium Bond Length for CO+ in the Born− Oppenheimer Approximation, and for Several Isotopic Species Considering the Deviation from the Born− Oppenheimer Approximation

0.94. The weighted rms for the partial fit, which includes just the measurements from this work, is 0.99, while the weighted rms for the partial fit, which includes just the data sets of Bogey et al.11 and Sastry et al.,8 are 0.78 and 1.03, respectively. The accuracies of the lines in Table 2 were estimated based on the line shape and signal-to-noise ratio, and were treated by spfit as absolute values. Table 4 compares the Born−Oppenheimer breakdown terms Δij Table 4. Comparison of Born−Oppenheimer Breakdown Terms CO+ (present work)

CO (George et al.29)

CN (Salek et al.30)

−0.226(14) −1.033(31)

−2.05603(23) −2.09934(28)

−0.895(11) −1.056(73)

re (pm) 111.517760(64) 111.5203128(67) 111.5200927(65) 111.5202683(86) 111.5200482(86)

± 0.0034) amu MHz Å2 is a conversion factor, and μ is the charge corrected reduced mass. The small difference between Be and Y01 could not be determined with confidence for the present parameters plus those of the A − X electronic spectrum,26 mainly because of the large uncertainty in Y20, so we used Y01 = Be. In the case of the Born−Oppenheimer approximation, the bond distance has been calculated as re = (X/(U01))1/2, and is independent of the isotopic species. The difference in the bond length of the isotopic species, determined by taking into account the breakdown of the BO approximation, are in the oder of 10−4 pm, and are thus very small as one might expect from the small BOB parameters, but determined with significance. On the other hand, the BO equilibrium bond length differs by a much greater value, around 2.4 × 10−3 pm. The error introduced by the assumption Y01 = Be may lead to errors outside the uncertainties of the BO-corrected bond length, but is most probably within the uncertainty of the equilibrium bond length. Accurate terahertz measurements have been performed on CO+, leading to a more accurate fit and improved catalogue entries at high frequencies: for example, the previously predicted value for the NJ = 1021/2 → 1123/2 line at 1.29 THz in the CDMS catalogue was off by 1.5 MHz from the observed value, while the actual catalogue has instead a calculated uncertainty of only 560 kHz at 2.5 THz. The measured and predicted high-N transitions of CO+ will be useful for future astronomical observations: ideal test candidates would be planetary nebulae and proto-planetary nebulae.2,27 Table 6 lists THz lines of possible astrophysical

Note: Numbers in parentheses are 1σ uncertainties in units of the least significant figures. aCalculated from U01 and μ. bDerived value. c Dimensionless parameter. dKept fixed during the fit.

ΔC01 ΔO/N 01

species CO+(BO) e 12 16 + C O 13 16 + C O 12 18 + C O 13 18 + C O

derived for CO+ with the Δij of CO and of the isoelectronic CN. The Δ01 for CO and CO+ are significantly different because the two species do not have the same number of valence electrons. On the contrary, with CN being isoelectronic with CO+ the values for these two molecules are expected to be comparable. What we observe is that the ΔC01 are quite different, while the ΔO/N are almost the same. A possible explanation for this 01 behavior could be the fact that in both molecules, the unpaired electron is largely localized on the carbon atom.22 Further investigations on the Born−Oppenheimer breakdown parameters for isoelectronic molecules would be fruitful to check whether it is possible to spot a trend in the behavior of these parameters. Boron monoxide, BO, is also isoelectronic with CO+ and CN but a comparison is not possible, as to our knowledge there is no isotopically invariant fit available in the literature and no accurate data is available for species containing 18O, therefore there are no Δ01 values available, and it was not possible for us to set up an isotopic invariant fit with the data available from the literature. Gauss and Puzzarini23 have demonstrated that it is possible to use quantum-chemical calculations to determine Born−Oppenheimer breakdown parameters for the rotational constants of a diatomic. Recently they have extended to openshell molecules their protocol for the determination of the Born−Oppenheimer breakdown parameters, and the values on CO+ reported in this work have been used for testing their calculations.24 The equilibrium bond lengths for CO+ are calculated for several isotopic species as re=(X/(Be × μ))1/2, as in Müller et al.,25 and are reported in Table 5. X = (505379.0094

Table 6. 12C16O+ THz Lines of Astrophysical Interest N → N′

J → J′

frequency (MHz)

8→9 8→9 10 → 11 10 → 11 11 → 12 11 → 12 15 → 16 15 → 16

17/2 → 19/2 15/2 → 17/2 21/2 → 23/2 19/2 → 21/2 23/2 → 25/2 21/2 → 23/2 31/2 → 33/2 29/2 → 31/2

1061006.518(38) 1061279.494(36) 1296483.126(67) 1296756.102(66) 1414148.772(89) 1414421.749(89) 1884220.905(232) 1884493.880(233)

HIFI HIFI APEX APEX GREAT GREAT GREAT GREAT

interest, reporting only the two stronger components of each rotational transition. Predictions based on the present isotopically invariant fit will be available online via the Cologne Database for Molecular Spectroscopy28 at http://www.cdms.de.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. 9817

dx.doi.org/10.1021/jp312616u | J. Phys. Chem. A 2013, 117, 9814−9818

The Journal of Physical Chemistry A



Article

(21) Klapper, G. Präzisionsmessungen des Reinen Rotationsspektrums von CO, CO+ und ihren Isotopomeren. 2001. Ph.D. Thesis. Universität zu Köln, Germany. (22) Carrington, A.; Sarre, P. Electronic Absorption Spectrum of CO+ in an Ion Beam. Mol. Phys. 1977, 33, 1495−1497. (23) Gauss, J.; Puzzarini, C. Quantum-Chemical Calculation of BornOppenheimer Breakdown Parameters to Rotational Constants. Mol. Phys. 2010, 108, 269−367. (24) Puzzarini, C.; Gauss, J. Quantum-Chemical Determination of Born-Oppenheimer Breakdown Parameters for Rotational Constants: The Open Shell Species CN, CO+ and BO. Mol. Phys. 2013, DOI: 10.1080/00268976.2013.797614. (25) Müller, H. S. P.; McCarthy, M. C.; Bizzocchi, L.; Gupta, H.; Esser, S.; Lichau, H.; Caris, M.; Lewen, F.; Hahn, J.; Degli Esposti, C.; et al. Rotational Spectroscopy of the Isotopic Species of Silicon Monosulfide, SiS. Phys. Chem. Chem. Phys. 2007, 9, 1579−1586. (26) Kepa, R.; Kocan, A.; Ostrowska-Kopeć, M.; PiotrowskaDomagała, I.; Zachwieja, M. New Spectroscopic Studies of the Comet Tail (A2Πi X2Σ+) System of the CO+ Molecule. J. Mol. Spectrosc. 2004, 228, 66−75. (27) Bell, T. A.; Whyatt, W.; Viti, S.; Redman, M. P. A Search for CO+ in Planetary Nebulae. Mon. Not. R. Astron. Soc. 2004, 382, 1139−1144. (28) Müller, H. S. P.; Schlöder, F.; Stutzki, J.; Winnewisser, G. The Cologne Database for Molecular Spectroscopy, CDMS: A Useful Tool for Astronomers and Spectroscopists. J. Mol. Struct. 2005, 742, 215− 227. (29) George, T.; Urban, W.; Le Floch, A. Improved Mass-Independent Dunham Parameters for the Ground State of CO and Calibration Frequencies for the Fundamental Band. J. Mol. Spectrosc. 1994, 165, 500−505. (30) Saleck, A. H.; Simon, R.; Winnewisser, G. Interstellar CN Rotational Spectra: 12C15N. Astrophys. J. 1994, 436, 176−182.

ACKNOWLEDGMENTS This work has been supported by SFB956. S. Spezzano has been supported in her research with a stipend from the International Max-Planck Research School (IMPRS) for Astronomy and Astrophysics at the Universities of Bonn and Cologne. H.S.P.M. acknowledges support by the Bundesministerium für Bildung und Forschung (BMBF) through project FKZ 50OF0901 (ICC HIFI Herschel). S. Schlemmer and S.B. acknowledge support by the Deutsche Forschungsgemeinschaft (DFG) through project BR 4287/1-1.



REFERENCES

(1) Sternberg, A.; Dalgarno, A. Chemistry in Dense PhotonDominated Regions. Astrophys. J., Suppl. Ser. 1995, 99, 565−607. (2) Latter, W. B.; Walker, C. K.; Maloney, P. R. Detection of the Carbon Monoxide Ion (CO+) in the Interstellar Medium and a Planetary Nebula. Astrophys. J. 1993, 419, L97−L100. (3) Fuente, A.; Martin-Pintado, J. Detection of CO+ toward the Reflection Nebula NGC 7023. Astrophys. J. 1997, 477, L107−L109. (4) Savage, C.; Ziurys, L. M. Ion Chemistry in Photon-Dominated Regions: Examining the [HCO+]/[HOC+]/[CO+] Chemical Network. Astrophys. J. 2004, 616, 966−975. (5) Rizzo, J. R.; Fuente, A.; Rodrguez-Franco, A.; García-Burillo, S. Detection of Reactive Ions in the Ultracompact H II Regions Monoceros R2 and G29.96−0.02. Astrophys. J. 2003, 597, L153−L156. (6) Stäuber, P.; Bruderer, S. Excitation and Abundance Study of CO+ in the Interstellar Medium. Astron. Astrophys. 2009, 505, 195−203. (7) Dixon, T. A.; Woods, R. C. Microwave Absorption Spectrum of the CO+ Ion. Phys. Rev. Lett. 1975, 34, 61−63. (8) Sastry, K. V. L .N .; Helminger, P.; Herbst, E.; De Lucia, F. C. Laboratory Millimeter and Submillimeter Spectra of CO+. Astrophys. J. 1981, 250, L91−L92. (9) Piltch, N. D.; Szanto, P. G.; Anderson, T. G.; Gudeman, C. S.; Dixon, T. A.; Woods, R. C. The Microwave Spectrum of Isotopically Substituted CO+ Ion. J. Chem. Phys. 1982, 76, 3385−3388. (10) Bogey, M.; Demuynck, C.; Destombes, J. L. Millimeter Wave Spectrum of CO+ in Excited Vibrational States. Mol. Phys. 1982, 46, 679−681. (11) Bogey, M.; Demuynck, C.; Destombes, J. L. Equilibrium Structure of CO+ from Its Millimeter Wave Spectrum - Breakdown of the BornOppenheimer Approximation. J. Chem. Phys. 1983, 79, 4704−4707. (12) Savage, C.; Ziurys, L. M. A Millimeter/Submillimeter Velocity Modulation Spectrometer for Studies of Molecular Ions. Rev. Sci. Instrum. 2005, 76, 043106−043111. (13) van den Heuvel, F. C.; Dymanus, A. Observation of Far-Infrared Transitions of HCO+, CO+ and HN+2 . Chem. Phys. Lett. 1982, 92, 219− 222. (14) Gendriesch, R.; Pehl, K.; Giesen, T.; Winnewisser, G.; Lewen, F. Terahertz Spectroscopy of Linear Triatomic CCC: High Precision Laboratory Measurement and Analysis of the Ro-vibrational Bending Transitions. Z. Naturforsch. 2003, 58a, 129−138. (15) Lewen, F.; Gendriesch, R.; Pak, I.; Paveliev, D. G.; Hepp, M.; Schieder, R.; Winnewisser, G. Phase Locked Backward Wave Oscillator Pulsed Beam Spectrometer in the Submillimeter Wave Range. Rev. Sci. Instrum. 1998, 69, 32−39. (16) Thompson, J. K.; Rainville, S.; Pritchard, D. E. Cyclotron Frequency Shifts Arising from Polarization Forces. Nature 2004, 430, 58−61. (17) Dunham, J. L. The Energy Levels of a Rotating Vibrator. Phys. Rev. 1932, 41, 721−731. (18) Watson, J. K. G. The Isotope Dependence of Diatomic Dunham Coefficients. J. Mol. Spectrosc. 1980, 80, 411−421. (19) Redshaw, M.; Mount, B. J.; Myers, E. G. Penning-trap Measurement of the Atomic Masses of 18O and 19F with Uncertainties