1997-McCarthy

THE ASTROPHYSICAL JOURNAL SUPPLEMENT SERIES, 113 :105È120, 1997 November1997. The American Astronomical Society. All rights reserved. Printed in U.S.A.(

EIGHT NEW CARBON CHAIN MOLECULES

M. C. M. J. A. C. A. AND P.MCCARTHY,1,2 TRAVERS,1,2 KOVA� CS,1 GOTTLIEB,1 THADDEUS1,2Received 1997 January 10 ; accepted 1997 June 3

ABSTRACTA summary is given of the laboratory measurements of the rotational spectra of eight new carbon

chains : the cyanopolyynes and the carbon chain radicals andHC11N HC13N; C7H, C8H, C9H, C11H;and the cumulene carbenes and Measured line frequencies and derived spectroscopic con-H2C5 H2C6.stants are listed for all eight, and rest frequencies for the astronomically important transitions that arenot easily calculated are tabulated. With our laboratory measurements, four of the new carbon chainshave already been detected in at least one astronomical source ; the remaining four should be detectablewith existing large centimeter- or millimeter-wave telescopes.Subject headings : ISM: molecules È line : identiÐcation È molecular data È molecular processes È

radio lines : ISM

1. INTRODUCTION

In a series of short papers we recently reported the labor-atory detection and spectroscopic characterization of theeight new carbon chain molecules shown in Mol-Figure 1.ecules of this kind with linear carbon backbones are thedominant structural theme of the 114 molecules that havenow been identiÐed in space, and with our laboratory fre-quencies four of our new chains have already been identiÐedin either the molecular shell of the evolved carbon star IRC]10216 or the rich molecular source TMC-1 in the Taurusdark clouds. Detection of the remainder is probably only amatter of time. The purpose of the present article is toprovide in one place a concise and useful summary of ourlaboratory results and tabulations of measured line fre-quencies that for brevity were omitted from our short pre-vious papers. is a brief overview: a summary of theTable 1laboratory and astronomical references, and the frequencybands that have been covered in the laboratory.

In our discovery papers, particular attention was given toestablishing the identiÐcation of the molecules inquestionÈto demonstrate that no other molecule could bethe carrier of the lines assigned. No argument or informa-tion has come to light to cause us to question any of ouridentiÐcations, and without further discussion the originalassignments will be assumed here. They are all about assecure as molecular identiÐcations can be that are basedlargely on spectroscopic evidence.

As shows, our carbon chains are of three kinds :Figure 1(1) closed shell cyanopolyynes, (2) radicals with one non-bonding electron, and (3) cumulene carbenes with two suchelectrons. Because of the nonbonding electrons or the ter-minal CN group, all of these chains are highly polar (Fig. 2),with intense rotational spectraÈa property of great advan-tage to both laboratory and astronomical detection. In thepresent paper, experimental details are largely omitted,except for brief descriptions of the reactive molecule pro-duction by means of discharges ; a brief description of ourFourier transform microwave (FTM) spectrometer, the

1 Division of Engineering and Applied Sciences, Harvard University, 29Oxford Street, Cambridge, MA 02138.

2 Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,Cambridge, MA 02138.

main laboratory instrument for the present work, is given inAppendix A.

2. CYANOPOLYYNES

Cyanopolyynes are closed-shell linear mol-HC2nCNecules. In the laboratory, the shorter members in the series(HCN, and can be synthesized byHC3N, HC5N, HC7N)standard techniques of organic synthesis, but hasHC9Nbeen observed only in discharges through stable gases. Wedetected both and in a supersonic molecu-HC11N HC13Nlar beam with our FTM spectrometer. In this instrument, asin similar ones, the molecules are irradiated with a shortmicrowave pulse that is timed to coincide with the passageof a gas pulse through the confocal Fabry-Perot cavity andthat is strong enough to saturate an allowed moleculartransition if it lies within the D1 MHz bandwidth of theconfocal Fabry-Perot cavity. The resulting free inductiondecay is then Fourier transformed to obtain the(Fig. 3a)power spectrum and line proÐle. Owing to the Doppler shiftof the molecular beam with respect to the two travelingwaves that compose the cavity mode, the line proÐle of asingle transition has the double-peaked proÐle shown in

In a Mach 2 supersonic beam, the separation ofFigure 3a.the two peaks, as shown, is about 68 kHz at 13.2 GHz, butthe much smaller width of the individual peaks is the reallimit on the spectral resolution. It is only about 7 kHz or 0.2km s~1 in the transition shown inHC11N Figure 3.

The cyanopolyynes were produced in a low-current 1100V gas discharge synchronized with a gas pulse 360 ks long,the gas sample consisting of either (cyanoacetylene)HC3Nor and HCCH (acetylene)Èdiluted in either case withN2Ar (by about 99%). The mixture gave theHC3N/HCCH/Arstrongest linesÈby a factor of roughly 3Èbut the

mixture was crucial to the detection ofN2/HCCH/Arsince 15N-enriched is fairly inexpensive, whileHC1115N, N2the synthesis of 15N-enriched is an expensive exer-HC3Ncise in isotopic chemistry.

2.1. HC11NFor this carbon chain, 20 lines of the normal isotopic

species and seven lines of the 15N species were measured toan accuracy of 0.1 km s~1 or better ; the data are given in

As shown in the level diagram of theTable 2. Figure 4,

105

106 McCARTHY ET AL. Vol. 113

FIG. 1.ÈMolecular geometries of the new carbon chains, showing the characteristic alternating triple and single carbon-carbon bonds of the cyano-polyynes and radicals with an odd number of carbon atoms, and the double bonds of the cumulene carbenes, with their two nonbonded electrons at theC

nH

terminal carbon (double dots). The valence structures of the radicals are approximate ; several resonance structures contribute to the 2% electronicCnH

ground states, but for simplicity, only one is shown.

measured lines arise from a fairly wide range of rotationallevels, with J from 17 to 43. There is no evidence in thenormal isotopic species for quadrupole hyperÐne structure(hfs) from the 14N nucleus, and none is expected, since suchstructure is \eqQ/4J2 & Schawlow or ¹3(Townes 1955)kHz in the high-J transitions observed (if it is assumed thateqQ \ [4 MHz, the approximate value for andHC3NCresswell, & WinnewisserHC5N: Winnewisser, 1978).

The lines were Ðtted to an rms that isHC11N (Table 2)comparable to the measurement uncertainty (3È5 kHz) withthe standard expression l\ 2B(J ] 1) [ 4D(J ] 1)3, whereJ is the rotational angular momentum of the lower level ofthe transition, B is the rotational constant, and D is thecentrifugal distortion constant. Even at 300 GHz (J D 900),

neglect of the next term H in the centrifugal expansionprobably results in an error in the calculated rest fre-quencies of ¹1 km s~1. Thus, the entire radio spectrum of

is accurately calculated for astronomical purposesHC11Nwith the values of B and D in Table 3.

2.2. HC13NTwenty-one rotational transitions of between 5HC13Nand 12 GHz were detected in the same source in(Table 2)

which was observed. Although the number ofHC11Nmolecules per gas pulse is about 7 times less thanHC13Nthat of the spectra are of excellent quality. The lineHC11N,frequencies were again measured to an accuracy of(Table 2)0.1 km s~1 or better, and were analyzed as before, yielding

TABLE 1

SUMMARY OF CARBON CHAIN DETECTIONS

FREQUENCY BANDS (GHz) REFERENCES

ELECTRONIC RARE

MOLECULE STATE ISOTOPE PARENT GASES FTM Free Space Laboratory Astronomical

Cyanopolyynes :HC11N . . . . . . 1&` 15N HCCH/(N2 or HC3N)/Ar 6È15 1 2HC13N . . . . . . 1&` HCCH/(N2 or HC3N)/Ar 5È12 3

Acetylenic radicals :C7H . . . . . . . . . 2%1@2 , 2%3@2 D HCCH/Ar 69È196 4 5

2%1@2 HC4H/Ar 7È19 6C8H . . . . . . . . . 2%1@2 , 2%3@2 HCCH/Ar 80È100 7 8

2%1@2 HC4H/Ar 8È12 6C9H . . . . . . . . . 2%1@2 HC4H/Ar 5È17 9C11H . . . . . . . . 2%1@2 HC4H/Ar 6È12 10

Cumulene carbenes :H2C5 . . . . . . . . 1A D2 HC4H/Ne or HCCH/Ar 13È23 11H2C6 . . . . . . . . 1A D2 HC4H/Ne or HCCH/Ar 8È19 11 12

REFERENCES.È(1) et al. (2) et al. (3) et al. (4) et al. (5) et al. (6) this work ; (7)Travers 1996c ; Bell 1997 ; Travers 1996d; Travers 1996b ; Gue� lin 1997 ;et al. (8) & Gue� lin (9) et al. (10) et al. (11) et al. (12)McCarthy 1996b ; Cernicharo 1996 ; McCarthy 1996a ; McCarthy 1997a ; McCarthy 1997b;

et al.Langer 1997.

No. 1, 1997 EIGHT NEW CARBON CHAIN MOLECULES 107

TABLE 2

MEASURED ROTATIONAL FREQUENCIES OF ANDHC11N, HC1115N, HC13N (in MHz)

HC11N HC1115N HC13N

J@] J Frequency O[C Frequency O[C Frequency O[C

18 ] 17 . . . . . . 6086.259 [0.00119 ] 18 . . . . . . 6424.384 [0.00120 ] 19 . . . . . . 6762.510 0.00021 ] 20 . . . . . . 7100.636 0.00122 ] 21 . . . . . . 7438.763 0.00423 ] 22 . . . . . . 7776.887 0.003 7654.170 0.001 4920.733 [0.00124 ] 23 . . . . . . 8115.007 [0.001 5134.682 0.00325 ] 24 . . . . . . 8453.135 0.003 5348.622 [0.00126 ] 25 . . . . . . 8791.257 0.001 5562.566 [0.00227 ] 26 . . . . . . 9129.376 [0.004 8985.325 0.002 5776.514 0.00228 ] 27 . . . . . . 9467.504 0.000 9318.110 [0.002 5990.458 0.00129 ] 28 . . . . . . 9805.622 [0.005 9650.900 0.00030 ] 29 . . . . . . 10143.750 [0.001 9983.685 [0.002 6418.340 [0.00531 ] 30 . . . . . . 10481.874 0.000 10316.475 0.000 6632.290 0.00132 ] 31 . . . . . . 6846.235 0.00233 ] 32 . . . . . . 7060.178 0.00134 ] 33 . . . . . . 7274.120 [0.00135 ] 34 . . . . . . 11834.366 0.001 11647.625 0.001 7488.066 0.00136 ] 35 . . . . . . 7702.009 0.00037 ] 36 . . . . . . 7915.955 0.00238 ] 37 . . . . . . 12848.728 [0.003 8129.894 [0.00239 ] 38 . . . . . . 13186.853 0.000 8343.839 [0.00140 ] 39 . . . . . . 13524.980 0.00641 ] 40 . . . . . . 8771.727 0.00142 ] 41 . . . . . . 14201.218 0.00243 ] 42 . . . . . . 14539.334 [0.00344 ] 43 . . . . . . 9413.556 0.00048 ] 47 . . . . . . 10269.327 0.00053 ] 52 . . . . . . 11339.040 0.00154 ] 53 . . . . . . 11552.980 [0.001

NOTE.ÈEstimated measurement uncertainty : 3È5 kHz.

the B and D in Detection of in either IRCTable 3. HC13N]10216 or TMC-1 with existing large centimeter-wave tele-scopes will not be easy, but at least the rest frequencies arenow available to undertake such a search.

3. CARBON CHAIN RADICALS

Following its detection in IRC ]10216 by Gue� lin (1986)

FIG. 2.ÈDipole moment vs. chain length for the cumulenes et(Gottliebal. & Botschwina & McLean the1993 ; Oswald 1995 ; Maluendes 1992),acetylenic radicals and the cyanopolyynes (P. Botschwina(Woon 1996),1997, private communication).

and subsequently in the laboratory et al.(Pearson 1988),remained until now the longest radical observedC6H C

nH

both in the laboratory and in space. Recently, &CernicharoGue� lin tentatively identiÐed in the circumstellar(1996) C8Henvelope of IRC ]10216, and its detection in a glow dis-charge with the millimeter-wave absorption spectrometerdescribed in Appendix B conÐrmed the astronomical identi-Ðcation and quickly led to the detection of the similarradical, in the same discharge. The andC7H, C7H C8Hradicals were then detected in a pulsed discharge supersonicmolecular beam with our FTM spectrometer, whichbecause of its high resolution provided precise rest fre-quencies. The sensitivity was so high that detection of C9Hand soon followed, but it has not yet been possible toC11Hdetect these longer radicals at millimeter wavelengths.

The spectra of the radicals are more complex thanCnH

those of the cyanopolyynes, because of the unpaired elec-tron and 2% electronic ground state. Coupling of the spinangular momentum of the electron to both the orbitalangular momentum and the molecular rotation causes the2% state to split into two Ðne-structure components

TABLE 3

ROTATIONAL AND CENTRIFUGAL DISTORTION CONSTANTS OF

ANDHC11N, HC1115N, HC13N (in MHz)

Constant HC11N HC1115N HC13N

B . . . . . . . . . . . . . 169.06294(3) 166.39524(15) 106.97258(4)D] 106 . . . . . . 0.238(11) 0.252(81) 0.092(10)

NOTE.ÈUncertainties (in parentheses) are 1p in the last signiÐ-cant digit.

108 McCARTHY ET AL.

FIG. 3.ÈSample FTM spectra. (a) (top) The free induction decay (FID)of at 13,186.853 MHz and (bottom) its Fourier transform (powerHC11Nspectrum) showing the double-peaked line proÐle resulting from the inter-action of the axial supersonic molecular beam with the two traveling wavesof the Fabry-Perot mode. The FID (b) (top) and power spectrum (bottom)of the J \ 11.5] 10.5 f transition of at 9500.060 MHz, showingC9Hwell-resolved hfs. Asterisks denote weak ghosts of the lower frequency (e)lambda component. (c) The power spectrum of the tran-J \ 20,2] 10,1sition of at 8452.440 MHz, showing partially resolved deute-D2CCCCCrium quadrupole hfs.

FIG. 4.ÈLower rotational levels of and showing theHC11N HC13N,transitions detected in the normal isotopic species (arrows) and the 15Nspecies of (dots).HC11N

& Schawlow The separation of the two is(Townes 1955).equal to the spin-orbit constant A, which is roughly 25cm~1 or 35 K in all four of the new molecules here. Rota-tional transitions in the and Ðne-structure com-2%1@2 2%3@2ponents are further split into two components of equalintensity by lambda doubling. For radicals with anC

nH

even number of carbon atoms, A is negative, and the 2%3@2component is lower in energy ; conversely, A is positive inradicals with an odd number of carbon atoms, and theC

nH

component is lower.2%1@2The millimeter-wave absorption lines of andC7H C8Hwere strongest in a 0.4 A DC discharge through a 4 :1 molarmixture of HCCH and Ar at a temperature of 100 K andtotal pressure of 20 mtorr. All four radicals were detectedwith the FTM spectrometer in a low-current 1000 V DCdischarge synchronized with a gas pulse 280 ks long, the gassample consisting of 1% (diacetylene) in Ar.HC4H

3.1. C7HThe millimeter-wave measurements of andC7H C7Dmade with the free-space laboratory spectrometer are given

in and the centimeter-wave measurements ofTable 4, C7Hin low rotational levels made with the FTM spectrometerare given in Both Ðne-structure components areTable 5.populated in the millimeter-wave absorption spectrum, butonly the ground component is populated in the cold2%1@2(3 K) supersonic molecular beam of the FTM spectrometer.The measured rest frequencies in both wavelength regionsare accurate to about 0.3 ppm. HyperÐne structure was notobserved in the high rotational transitions at millimeterwavelengths because it collapses down to a small fraction ofthe pressure-broadened line width, which is roughly 100

TABLE 4

MEASURED ROTATIONAL FREQUENCIES OF AND IN THE 2% GROUND STATEC7H C7D (in MHz)

C7H C7DTRANSITION e/fJ@ÈJ ) " Comp.a Frequency O[C Frequency O[C

6.5] 5.5 . . . . . . . . . . 1/2 e 10930.149 (05) 0.008H G39.5^ 38.5 . . . . . . . . 3/2 f [0.01769238.956 (27)be [0.003

65.5^ 64.5 . . . . . . . . 1/2 e 114545.484 (17) 0.016f 114547.516 (17) 0.013

68.5^ 67.5 . . . . . . . . 1/2 e 119790.761 (19) 0.010f 119792.818 (23) [0.008

3/2 f 120061.581 (19) [0.006e 120062.070 (20) 0.004

69.5^ 68.5 . . . . . . . . 1/2 e 121539.166 (20) 0.021f 121541.235 (21) 0.001

3/2 f 121813.828 (26) 0.003e 121814.326 (21) 0.008

70.5^ 69.5 . . . . . . . . 1/2 e 123287.525 (21) 0.002f 123289.623 (19) [0.003

3/2 f 123566.041 (19) [0.002 118802.342 (21) 0.033e 123566.548 (19) [0.001 118802.801 (20) [0.049

71.5^ 70.5 . . . . . . . . 1/2 e 125035.840 (22) [0.044 120225.537 (17) 0.001f 125037.980 (19) [0.021 120227.599 (17) 0.008

3/2 f 125318.202 (25) [0.037 120486.975 (22) [0.008e 125318.818 (41) 0.059 120487.552 (23) 0.013

72.5^ 71.5 . . . . . . . . 1/2 e 126784.225 (18) [0.002 121906.623 (20) [0.013f 126786.370 (20) 0.011 121908.626 (22) [0.081

3/2 f 127070.366 (25) [0.047e 127070.980 (23) 0.032

73.5^ 72.5 . . . . . . . . 1/2 e 128532.558 (21) 0.005 123587.735 (27) 0.015f 128534.714 (20) 0.015 123589.830 (21) 0.024

3/2 f 128822.590 (21) 0.025e 128823.137 (20) 0.022

74.5^ 73.5 . . . . . . . . 1/2 e 130280.883 (19) 0.022 125268.794 (33) 0.006f 130283.006 (19) [0.016 125270.918 (25) 0.028

3/2 f 130574.674 (32) [0.022 125540.936 (43) 0.048e 130575.271 (24) 0.011 125541.477 (35) [0.014

80.5^ 79.5 . . . . . . . . 3/2 f 135648.176 (29) 0.034e 135648.815 (22) [0.028

84.5^ 83.5 . . . . . . . . 1/2 e 142078.514 (20) [0.004f 142080.828 (24) 0.040

90.5^ 89.5 . . . . . . . . 1/2 e 152163.417 (17) [0.052f 152165.904 (23) 0.055

3/2 f 152491.721 (71) [0.060e 152492.603 (44) [0.059

91.5^ 90.5 . . . . . . . . 1/2 e 159999.105 (26) [0.0273/2 f 160357.189 (25) [0.043

e 160358.125 (28) 0.05192.5^ 91.5 . . . . . . . . 1/2 e 155524.973 (31) 0.013

f 155527.389 (27) 0.0103/2 f 155860.287 (27) 0.065

e 155861.131 (27) [0.01093.5^ 92.5 . . . . . . . . 1/2 f 163497.467 (19) [0.007

3/2 f 163860.550 (26) [0.015e 163861.470 (27) 0.027

95.5^ 94.5 . . . . . . . . 1/2 e 166990.723 (23) [0.053f 166993.258 (17) [0.030

3/2 f 167363.752 (23) [0.034e 167364.706 (29) 0.005

96.5^ 95.5 . . . . . . . . 1/2 e 168738.651 (22) 0.021f 168741.138 (25) [0.022

3/2 f 169115.337 (27) [0.017e 169116.259 (26) [0.028

108.5^ 107.5 . . . . . . 1/2 e 189711.069 (26) 0.0993/2 f 190131.863 (26) 0.038

e 190133.014 (30) 0.020111.5^ 110.5 . . . . . . 3/2 f 195385.245 (37) 0.012

e 195386.442 (27) [0.022

NOTE.ÈUncertainties (in parentheses) are 1 p experimental errors in units of the last signiÐcant digit.a Designation of e and f levels on the assumption that hyperÐne coupling constant d is positive.b These partially blended lines were entered into the Ðt as two lines with their centroid at the measured

frequency and their splitting set to a value calculated from other lines.

110 McCARTHY ET AL.

TABLE 5

HYPERFINE SPLIT ROTATIONAL TRANSITIONS (in MHz) OF IN THEC7HSTATE2%1@2TRANSITION e LEVELS f LEVELS

J@] J F@] F Frequency O[C Frequency O[C

4.5] 3.5 . . . . . . . 5] 4 7869.429 0.000 7871.093 0.0004 ] 3 7869.698 0.001 7871.398 [0.001

5.5] 4.5 . . . . . . . 6] 5 9618.440 0.000 9620.086 0.0015 ] 4 9618.626 0.000 9620.291 [0.001

6.5] 5.5 . . . . . . . 7] 6 11367.433 0.000 11369.069 0.0016 ] 5 11367.573 0.000 11369.219 0.000

7.5] 6.5 . . . . . . . 8] 7 13116.414 [0.002 13118.045 0.0017 ] 6 13116.524 [0.002 13118.162 0.000

8.5] 7.5 . . . . . . . 9] 8 14865.391 [0.001 14867.015 [0.0018 ] 7 14865.483 0.001 14867.113 0.001

9.5] 8.5 . . . . . . . 10] 9 16614.362 0.000 16615.986 0.0029 ] 8 16614.438 [0.001 16616.066 0.001

10.5] 9.5 . . . . . . 11] 10 18363.327 [0.001 18364.950 0.00110 ] 9 18363.394 [0.001 18365.022 0.003

NOTE.ÈEstimated measurement uncertainty : 5È10 kHz.

times greater than the line width in the supersonic beam.Owing to the high resolution of the FTM technique,however, hfs was resolved in the lower rotational transitions(Table 5).

The spectroscopic constants were derived by(Table 6)least-squares Ðtting the standard Hamiltonian for a 2%molecule (Appendix C) to the combined millimeter-waveand FTM measurements in Tables and Nine free4 5.parameters, a reasonable number of spectroscopic con-stants for a spectrum of such complexity, were required toreproduce more than 75 lines between 7.9 and 195.4 GHz towithin the measurement uncertainties.

To facilitate future observations of the astronomi-C7H,cally most interesting transitions are listed in Tran-Table 7.sitions below about 25 GHz in which hfs is greater than 0.5km s~1 are tabulated only in the ground Ðne-structure com-ponent, because the upper Ðne-structure component ishighly unlikely to be populated in a cold source likeTMC-1. In sources with higher excitation temperaturessuch as IRC ]10216, both Ðne structure components willbe populated. Calculated rest frequencies in both Ðne-structure components but without hfs are therefore alsolisted in Table 7.

FIG. 5.ÈLower rotational levels of and showing the tran-C9H C11Hsitions detected in the laboratory.

3.2. C8HLambda doubling in both Ðne-structure components of

is large enough to be resolved in the millimeter-waveC8Hspectrum but is so small in the low rotational(Table 8)transitions of the ground state observed with the2%3@2FTM spectrometer that it is barely resolved in spite of thehigh resolution a†orded by the molecular beam. Hfs in the

component is proportional to the coupling constant2%3@2a ] (b ] c)/2. In the next smaller member in theC6H, CnH

series with an even number of carbon atoms, a ] (b ] c)/2,is less than a [ (b ] c)/2, the corresponding coupling con-stant in the component of As can be seen in2%1@2 C11H.

the magnitude of the hyperÐne constants decreaseTable 6,with increasing chain length and are much smaller in C

nH

radicals with an even number of carbon atoms. Although

TABLE 6

SPECTROSCOPIC CONSTANTS OF AND IN THE X 2% STATEC7H, C7D, C8H, C9H, C11H (in MHz)

Constant C7H C7D C8H C9H C11H

Aeff . . . . . . . . . . . . . . . . 784700(2100)a 784046(27) [579900(24) 750000b 750000b(26.17 cm~1) (26.15 cm~1) ([19.34 cm~1) (25 cm~1) (25 cm~1)

c . . . . . . . . . . . . . . . . . . . [16.6(23) [15.91cB . . . . . . . . . . . . . . . . . . . 875.48409(5) 841.76161(13) 587.26379(10) 413.25758(3) 226.90025(3)D] 106 . . . . . . . . . . . 11.290(3) 10.26(1) 4.72(1) 1.92(6) 0.543(36)p ] 2q . . . . . . . . . . . . . 1.597(4) 1.499(9) 10.259(15)d 0.774(1) 0.426(1)eq . . . . . . . . . . . . . . . . . . . [0.0163(1) [0.0178(3) [0.0841(3)a [ (b ] c)/2 . . . . . . 10.2(20) 5.99(99) 4.87(42)b . . . . . . . . . . . . . . . . . . . [19.2(35) [13.7(17)d . . . . . . . . . . . . . . . . . . . 6.51(51)f 2.94(34)f

NOTE.ÈThe 1p uncertainties (in parentheses) are in the units of the last signiÐcant digits.a Aeff ] c.b Assumed.c Estimated from (this work) assuming c is proportional to B.C7Hd Sign assumed to be the same as in et al.C6H (Pearson 1988).e Sign assumed to be the same as in C9H.f Sign assumed to be positive.

TABLE 7

CALCULATED ROTATIONAL FREQUENCIES OF C7H (in MHz)

2%1@2 (with hfs)

e Levels f LevelsTRANSITION

J@ÈJ F\ J ] 1/2 F\ J [ 1/2 F\ J ] 1/2 F\ J [ 1/2

1.5È0.5 . . . . . . . 2621.753(22) 2624.988(121) 2623.784(22) 2628.755(121)2.5È1.5 . . . . . . . 4371.238(8) 4372.107(22) 4373.021(8) 4374.139(22)3.5È2.5 . . . . . . . 6120.380(4) 6120.814(8) 6122.082(4) 6122.598(8)4.5È3.5 . . . . . . . 7869.429(2) 7869.697(4) 7871.093(2) 7871.399(4)5.5È4.5 . . . . . . . 9618.440(2) 9618.626(2) 9620.085(2) 9620.292(2)6.5È5.5 . . . . . . . 11367.433(2) 11367.572(2) 11369.068(2) 11369.219(2)7.5È6.5 . . . . . . . 13116.416(2) 13116.526(2) 13118.044(2) 13118.162(2)8.5È7.5 . . . . . . . 14865.392(2) 14865.482(2) 14867.016(2) 14867.112(2)9.5È8.5 . . . . . . . 16614.362(2) 16614.439(2) 16615.984(2) 16616.065(2)

10.5È9.5 . . . . . . . 18363.328(2) 18363.395(2) 18364.949(2) 18365.019(2)11.5È10.5 . . . . . . 20112.290(2) 20112.349(2) 20113.911(2) 20113.973(3)12.5È11.5 . . . . . . 21861.248(2) 21861.301(3) 21862.870(2) 21862.925(3)13.5È12.5 . . . . . . 23610.202(3) 23610.251(3) 23611.826(3) 23611.876(3)14.5È13.5 . . . . . . 25359.152(3) 25359.197(3) 25360.778(3) 25360.824(3)

2%1@2 (no hfs) 2%3@2 (no hfs)

e Level f Level f Level e Level

1.5È0.5 . . . . . . . 2623.371(72) 2626.270(72)2.5È1.5 . . . . . . . 4371.673(15) 4373.580(15) 4382.407(250) 4382.408(250)3.5È2.5 . . . . . . . 6120.597(6) 6122.340(6) 6135.369(118) 6135.370(118)4.5È3.5 . . . . . . . 7869.563(3) 7871.246(3) 7888.329(70) 7888.331(70)5.5È4.5 . . . . . . . 9618.533(2) 9620.189(2) 9641.288(46) 9641.291(46)6.5È5.5 . . . . . . . 11367.503(2) 11369.144(2) 11394.245(34) 11394.250(34)7.5È6.5 . . . . . . . 13116.471(2) 13118.103(2) 13147.200(26) 13147.206(26)8.5È7.5 . . . . . . . 14865.437(2) 14867.064(2) 14900.153(20) 14900.161(20)9.5È8.5 . . . . . . . 16614.401(2) 16616.025(2) 16653.103(17) 16653.113(17)

10.5È9.5 . . . . . . . 18363.362(2) 18364.984(2) 18406.050(14) 18406.062(14)11.5È10.5 . . . . . . 20112.320(2) 20113.942(2) 20158.994(12) 20159.008(12)12.5È11.5 . . . . . . 21861.275(3) 21862.898(3) 21911.935(11) 21911.951(11)13.5È12.5 . . . . . . 23610.227(3) 23611.851(3) 23664.871(9) 23664.890(9)14.5È13.5 . . . . . . 25359.175(3) 25360.801(3) 25417.803(8) 25417.825(8)15.5È14.5 . . . . . . 27108.120(3) 27109.748(3) 27170.731(7) 27170.756(7)16.5È15.5 . . . . . . 28857.061(3) 28858.692(3) 28923.654(7) 28923.682(7)17.5È16.5 . . . . . . 30605.998(3) 30607.632(3) 30676.571(7) 30676.603(7)18.5È17.5 . . . . . . 32354.930(3) 32356.568(3) 32429.484(6) 32429.520(6)19.5È18.5 . . . . . . 34103.858(3) 34105.499(3) 34182.391(6) 34182.430(6)20.5È19.5 . . . . . . 35852.782(3) 35854.426(3) 35935.291(6) 35935.335(6)21.5È20.5 . . . . . . 37601.700(3) 37603.348(3) 37688.186(6) 37688.234(6)22.5È21.5 . . . . . . 39350.613(3) 39352.265(3) 39441.073(6) 39441.126(6)23.5È22.5 . . . . . . 41099.520(3) 41101.177(3) 41193.954(6) 41194.012(6)24.5È23.5 . . . . . . 42848.422(3) 42850.084(3) 42946.828(4) 42946.891(4)25.5È24.5 . . . . . . 44597.318(3) 44598.984(3) 44699.694(4) 44699.762(4)26.5È25.5 . . . . . . 46346.207(3) 46347.879(3) 46452.553(4) 46452.626(4)27.5È26.5 . . . . . . 48095.090(3) 48096.768(3) 48205.403(4) 48205.482(4)28.5È27.5 . . . . . . 49843.966(3) 49845.649(3) 49958.245(4) 49958.329(4)

< < < < <40.5È39.5 . . . . . . 70829.873(4) 70831.641(4) 70991.580(6) 70991.751(6)41.5È40.5 . . . . . . 72578.641(4) 72580.417(4) 72744.285(6) 72744.464(6)42.5È41.5 . . . . . . 74327.398(4) 74329.183(4) 74496.978(6) 74497.165(6)43.5È42.5 . . . . . . 76076.145(4) 76077.939(4) 76249.657(6) 76249.853(6)44.5È43.5 . . . . . . 77824.882(4) 77826.684(4) 78002.324(6) 78002.529(6)45.5È44.5 . . . . . . 79573.608(4) 79575.419(4) 79754.976(6) 79755.191(6)46.5È45.5 . . . . . . 81322.323(4) 81324.143(4) 81507.615(6) 81507.839(6)47.5È46.5 . . . . . . 83071.027(4) 83072.857(4) 83260.240(6) 83260.473(6)48.5È47.5 . . . . . . 84819.719(4) 84821.559(4) 85012.850(6) 85013.093(6)49.5È48.5 . . . . . . 86568.399(4) 86570.249(4) 86765.446(6) 86765.699(6)50.5È49.5 . . . . . . 88317.067(4) 88318.927(4) 88518.026(6) 88518.290(6)51.5È50.5 . . . . . . 90065.723(4) 90067.594(4) 90270.592(6) 90270.865(6)52.5È51.5 . . . . . . 91814.367(4) 91816.248(4) 92023.141(6) 92023.425(6)53.5È52.5 . . . . . . 93562.998(4) 93564.890(4) 93775.675(6) 93775.970(6)54.5È53.5 . . . . . . 95311.616(4) 95313.519(4) 95528.192(6) 95528.498(6)55.5È54.5 . . . . . . 97060.222(4) 97062.135(4) 97280.693(6) 97281.010(6)56.5È55.5 . . . . . . 98808.812(4) 98810.738(4) 99033.177(6) 99033.506(6)

NOTE.ÈFrequencies calculated from constants in Estimated uncertainties (inTable 6.parentheses) are 1p in the last signiÐcant digit from the least-squares Ðt. Transitions between50 and 70 GHz are omitted owing to high atmospheric opacity. HyperÐne structure (hfs) is lessthan 0.5 km s~1 above 25 GHz. The spin component is approximately 38 K higher than2%3@2the component.2%1@2

112 McCARTHY ET AL.

TABLE 8

MEASURED ROTATIONAL FREQUENCIES OF IN THE(in MHz) C8H2% GROUND STATE

Transition e/fJ@ÈJ ) " Comp.a Frequency O[C

7.5] 6.5 . . . . . . . 3/2 e, f 8800.043(5)b [0.0048.5] 7.5 . . . . . . . 3/2 e, f 9973.391(5)b 0.0079.5] 8.5 . . . . . . . 3/2 e, f 11146.724(5)b 0.003

10.5] 9.5 . . . . . . . 3/2 e, f 12320.053(5)b [0.00468.5^ 67.5 . . . . . . 3/2 e 80367.224(24) [0.06871.5^ 70.5 . . . . . . 3/2 e 83886.508(28) 0.030

f 83889.245(22) 0.04872.5^ 71.5 . . . . . . 3/2 e 85059.531(18) 0.007

f 85062.314(19) [0.00573.5^ 72.5 . . . . . . 3/2 e 86232.571(21) 0.008

f 86235.410(21) [0.02475.5^ 74.5 . . . . . . 3/2 e 88578.640(22) 0.022

f 88581.622(27) [0.02276.5^ 75.5 . . . . . . 3/2 e 89751.636(21) 0.002

f 89754.746(20) 0.00778.5^ 77.5 . . . . . . 1/2 e 92279.781(21) 0.002

f 92286.773(26) 0.00279.5^ 78.5 . . . . . . 1/2 e 93455.122(21) [0.013

f 93462.038(21) [0.00880.5^ 79.5 . . . . . . 1/2 e 94630.469(21) [0.013

f 94637.270(21) [0.04081.5^ 80.5 . . . . . . 3/2 e 95616.577(21) [0.011

f 95620.109(23) 0.0061/2 e 95805.803(19) [0.017

f 95812.528(19) [0.03582.5^ 81.5 . . . . . . 3/2 e 96789.638(21) 0.085

f 96793.161(18) 0.0081/2 e 96981.193(22) 0.046

f 96987.806(19) 0.00083.5^ 82.5 . . . . . . 3/2 e 97962.497(18) [0.013

f 97966.175(18) [0.0211/2 e 98156.478(18) 0.014

f 98163.044(22) 0.00784.5^ 83.5 . . . . . . 3/2 e 99135.362(22) [0.096

f 99139.267(24) 0.0371/2 e 99331.750(23) [0.021

f 99338.311(17) 0.053

NOTE.ÈUncertainties (in parentheses) are 1p experimental errors inunits of the last signiÐcant digit.

a Designation of e and f levels on the assumption that the sign oflambda-type doubling constant p ] 2q is the same as in C6H (Pearsonet al. 1988).

b Lambda doubling was not determined because it is comparable tothe instrumental line width in the FTM spectrometer (see Appendix C).

lambda doubling and hfs was not resolved in the FTMspectrum, it may be possible to resolve it in TMC-1 if C8Hcan be detected there. To aid future observations in this andother astronomical sources, the calculated spectrum in bothÐne-structure components is given in Table 9.

and3.3. C9H C11HAs illustrated in the radical has well-Figure 3, C9Hresolved hfs even in rotational transitions as high as

J \ 19.5] 18.5 in hfs was observed up(Table 10) ; C11H,through the J \ 16.5] 15.5 transition. Indicated on theenergy level diagrams of and are theC9H C11H (Fig. 5)transitions measured in each radical. The leading indepen-dent spectroscopic constants B, D, and p ] 2q were derivedfrom 12 rotational transitions in the ground Ðne-(2%1@2)structure component in both and HfsC9H C11H (Table 10).in was analyzed in the same fashion as that ofC9H C7H,allowing three hyperÐne coupling constants to be deter-mined. Only the leading coupling constant could beobtained in because hfs is much smaller than inC11H, C9H.

The astronomically most interesting transitions in thecomponent of are listed in either (1)2%1@2 C9H Table 11

with hfs if the calculated splittings are greater than about0.5 km s~1 or (2) without hfs if the splittings are less thanthe Doppler width in the narrow-line source TMC-1. Cal-culated frequencies of the astronomically interesting tran-sitions of are listed inC11H Table 12.

4. CUMULENE CARBENES

The cumulene carbenes and are readilyH2C3 H2C4observable in the laboratory and in the astronomicalsources IRC ]10216 and TMC-1, so the next members inthe series are excellent candidates for astronomical detec-tion. The structure of the rotational spectra of these C2vmolecules (i.e., selection rules, intensity ratios, etc.) is similarto that of (formaldehyde). As shown in theH2CO Figure 6,K \ ^1 rotational ladders lie about 14 K above ground,but they are metastable owing to the ortho/para nuclearspin statistics. In and each *J \ 1 rotationalH2C5 H2C6,transition consists of fairly tight triplets with intensity ratiosof 3 :2 :3. The triplets correspond to the high- and low-frequency K \ ^1 transitions and the K \ 0 transitioncentered approximately midway between. In the iso-D2topic species, the intensities of the K \ ^1 lines are 4 timesless than the K \ 0 lines, a result of the nuclear spin sta-tistics for two equivalent bosons with spin 1.

Both new cumulene carbenes were produced by a 1050 Vdischarge synchronized with a 390 ks gas pulse, the gasconsisting of a dilute (1%) mixture of either orHC4HHCCH in Ar. For both, yielded stronger signals thanHC4HHCCH by about a factor of 2, much as observed with the

radicals.CnH

FIG. 6.ÈRotational energy levels of and Owing to ortho/H2C5 H2C6.para spin statistics, the K \ ^1 levels are metastable. The rotational tran-sitions occur as closely spaced triplets, each originating from a level withthe same J but with K \ 0 or ^1. Dots indicate those measured in thedoubly deuterated isotopic species.

TABLE 9

CALCULATED ROTATIONAL FREQUENCIES OF C8H (in MHz)

2%3@2 2%1@2TRANSITION

J@ÈJ e Level f Level e Level f Level

1.5È0.5 . . . . . . . 1758.443(8) 1768.701(8)2.5È1.5 . . . . . . . 2933.349(1) 2933.353(1) 2934.158(8) 2944.414(8)3.5È2.5 . . . . . . . 4106.688(1) 4106.694(1) 4109.874(8) 4120.126(8)4.5È3.5 . . . . . . . 5280.026(1) 5280.036(1) 5285.590(8) 5295.838(8)5.5È4.5 . . . . . . . 6453.362(1) 6453.378(1) 6461.305(8) 6471.548(8)6.5È5.5 . . . . . . . 7626.698(1) 7626.720(1) 7637.021(8) 7647.257(8)7.5È6.5 . . . . . . . 8800.032(1) 8800.062(1) 8812.736(8) 8822.965(8)8.5È7.5 . . . . . . . 9973.365(2) 9973.404(2) 9988.451(8) 9998.671(8)9.5È8.5 . . . . . . . 11146.697(2) 11146.745(2) 11164.165(8) 11174.375(8)

10.5È9.5 . . . . . . . 12320.027(2) 12320.086(2) 12339.879(8) 12350.078(8)11.5È10.5 . . . . . . 13493.356(2) 13493.427(2) 13515.592(8) 13525.779(8)12.5È11.5 . . . . . . 14666.682(2) 14666.767(2) 14691.303(8) 14701.478(8)13.5È12.5 . . . . . . 15840.008(2) 15840.106(2) 15867.014(8) 15877.174(8)14.5È13.5 . . . . . . 17013.331(3) 17013.445(3) 17042.724(8) 17052.869(8)15.5È14.5 . . . . . . 18186.652(3) 18186.782(3) 18218.432(8) 18228.561(8)16.5È15.5 . . . . . . 19359.972(3) 19360.119(3) 19394.139(8) 19404.250(8)17.5È16.5 . . . . . . 20533.289(3) 20533.454(3) 20569.844(8) 20579.937(8)18.5È17.5 . . . . . . 21706.604(3) 21706.789(3) 21745.548(8) 21755.621(8)19.5È18.5 . . . . . . 22879.916(3) 22880.122(3) 22921.249(8) 22931.303(8)20.5È19.5 . . . . . . 24053.226(3) 24053.453(4) 24096.949(8) 24106.981(8)21.5È20.5 . . . . . . 25226.534(4) 25226.783(4) 25272.647(9) 25282.656(9)22.5È21.5 . . . . . . 26399.838(4) 26400.112(4) 26448.342(9) 26458.327(9)23.5È22.5 . . . . . . 27573.141(4) 27573.439(4) 27624.035(9) 27633.996(9)24.5È23.5 . . . . . . 28746.440(4) 28746.764(4) 28799.726(9) 28809.660(9)25.5È24.5 . . . . . . 29919.736(4) 29920.087(4) 29975.414(9) 29985.321(9)26.5È25.5 . . . . . . 31093.029(4) 31093.409(4) 31151.099(9) 31160.979(9)27.5È26.5 . . . . . . 32266.319(4) 32266.728(4) 32326.781(9) 32336.632(9)28.5È27.5 . . . . . . 33439.606(5) 33440.045(5) 33502.461(9) 33512.281(9)29.5È28.5 . . . . . . 34612.890(5) 34613.359(5) 34678.137(9) 34687.926(9)30.5È29.5 . . . . . . 35786.170(5) 35786.672(5) 35853.810(9) 35863.567(9)31.5È30.5 . . . . . . 36959.446(5) 36959.982(5) 37029.480(9) 37039.203(9)32.5È31.5 . . . . . . 38132.719(5) 38133.289(5) 38205.146(9) 38214.835(9)33.5È32.5 . . . . . . 39305.988(5) 39306.594(5) 39380.808(9) 39390.462(9)34.5È33.5 . . . . . . 40479.254(5) 40479.895(5) 40556.467(9) 40566.084(9)35.5È34.5 . . . . . . 41652.515(5) 41653.194(5) 41732.122(9) 41741.701(9)36.5È35.5 . . . . . . 42825.772(5) 42826.490(5) 42907.773(9) 42917.314(10)37.5È36.5 . . . . . . 43999.026(5) 43999.783(5) 44083.420(9) 44092.921(10)38.5È37.5 . . . . . . 45172.275(5) 45173.073(6) 45259.062(10) 45268.522(10)39.5È38.5 . . . . . . 46345.519(6) 46346.360(6) 46434.700(10) 46444.119(10)40.5È39.5 . . . . . . 47518.760(6) 47519.643(6) 47610.333(10) 47619.709(10)41.5È40.5 . . . . . . 48691.995(6) 48692.922(6) 48785.962(10) 48795.294(10)42.5È41.5 . . . . . . 49865.226(6) 49866.198(6) 49961.586(10) 49970.873(10)

< < < < <60.5È59.5 . . . . . . 70982.485(6) 70984.442(6) 71121.853(9) 71130.155(9)61.5È60.5 . . . . . . 72155.609(6) 72157.631(6) 72297.362(9) 72305.599(9)62.5È61.5 . . . . . . 73328.728(6) 73330.814(6) 73472.863(9) 73481.035(9)63.5È62.5 . . . . . . 74501.839(6) 74503.992(6) 74648.357(9) 74656.462(9)64.5È63.5 . . . . . . 75674.944(6) 75677.164(6) 75823.843(9) 75831.881(9)65.5È64.5 . . . . . . 76848.041(6) 76850.330(6) 76999.321(9) 77007.291(9)66.5È65.5 . . . . . . 78021.132(6) 78023.490(6) 78174.792(9) 78182.693(9)67.5È66.5 . . . . . . 79194.216(6) 79196.644(6) 79350.255(9) 79358.085(9)68.5È67.5 . . . . . . 80367.292(6) 80369.792(6) 80525.709(8) 80533.468(8)69.5È68.5 . . . . . . 81540.361(6) 81542.933(6) 81701.155(8) 81708.842(8)70.5È69.5 . . . . . . 82713.423(6) 82716.069(6) 82876.593(8) 82884.206(8)71.5È70.5 . . . . . . 83886.478(6) 83889.197(6) 84052.022(8) 84059.561(8)72.5È71.5 . . . . . . 85059.524(6) 85062.319(6) 85227.443(8) 85234.907(8)73.5È72.5 . . . . . . 86232.563(6) 86235.434(6) 86402.855(8) 86410.243(8)74.5È73.5 . . . . . . 87405.595(6) 87408.543(6) 87578.258(8) 87585.568(8)75.5È74.5 . . . . . . 88578.618(6) 88581.644(6) 88753.652(8) 88760.884(8)76.5È75.5 . . . . . . 89751.634(6) 89754.739(7) 89929.037(8) 89936.190(8)77.5È76.5 . . . . . . 90924.641(7) 90927.826(7) 91104.412(8) 91111.486(8)78.5È77.5 . . . . . . 92097.640(7) 92100.907(7) 92279.779(8) 92286.771(8)79.5È78.5 . . . . . . 93270.631(7) 93273.980(7) 93455.135(8) 93462.046(8)80.5È79.5 . . . . . . 94443.614(7) 94447.045(7) 94630.482(8) 94637.310(8)81.5È80.5 . . . . . . 95616.588(7) 95620.103(8) 95805.820(8) 95812.563(8)82.5È81.5 . . . . . . 96789.553(8) 96793.153(8) 96981.147(8) 96987.806(8)83.5È82.5 . . . . . . 97962.510(8) 97966.196(8) 98156.464(8) 98163.037(8)84.5È83.5 . . . . . . 99135.458(8) 99139.230(8) 99331.771(8) 99338.258(8)

NOTE.ÈFrequencies calculated from constants in Estimated uncertaintiesTable 6.(in parentheses) are 1p in the last signiÐcant digit from the least-squares Ðt. Transitionsbetween 50 and 70 GHz are omitted owing to high atmospheric opacity. The spin2%1@2component is approximately 28 K higher than the component.2%3@2

114 McCARTHY ET AL.

TABLE 10

MEASURED ROTATIONAL FREQUENCIES OF AND(in MHz) C9H C11H

TRANSITION C9H C11He/fJ@] J F@] F " Comp.a Frequency O[C Frequency O[C

6.5] 5.5 . . . . . . . 7] 6 e 5368.954 0.0006 ] 5 e 5369.030 [0.0017 ] 6 f 5369.743 0.0006 ] 5 f 5369.826 0.000

7.5] 6.5 . . . . . . . 8] 7 e 6195.019 [0.0027 ] 6 e 6195.081 0.0008 ] 7 f 6195.808 0.0017 ] 6 f 6195.870 0.000

8.5] 7.5 . . . . . . . 9] 8 e 7021.085 0.0008 ] 7 e 7021.137 0.0049 ] 8 f 7021.867 [0.0018 ] 7 f 7021.918 [0.001

11.5] 10.5 . . . . . . 12 ] 11 e 9499.267 0.00111 ] 10 e 9499.295 [0.00112 ] 11 f 9500.045 0.00011 ] 10 f 9500.075 [0.001

13.5] 12.5 . . . . . . 14 ] 13 e 11151.382 0.001 6124.228 0.00013 ] 12 e 11151.406 0.001 6124.244 0.00314 ] 13 f 11152.160 0.001 6124.654 0.00013 ] 12 f 11152.185 0.002 6124.668 0.001

14.5] 13.5 . . . . . . 15 ] 14 e 11977.437 0.000 6577.890 0.00014 ] 13 e 11977.459 0.000 6577.900 [0.00215 ] 14 f 11978.214 0.000 6578.316 0.00014 ] 13 f 11978.236 0.000 6578.330 0.002

15.5] 14.5 . . . . . . 16 ] 15 e 12803.492 0.000 7031.552 [0.00115 ] 14 e 12803.513 0.001 7031.565 0.00216 ] 15 f 12804.270 0.001 7031.978 0.00015 ] 14 f 12804.288 [0.002 7031.988 [0.001

16.5] 15.5 . . . . . . 17 ] 16 e 13629.546 [0.001 7485.212 [0.00316 ] 15 e 13629.566 0.001 7485.223 [0.00117 ] 16 f 13630.322 [0.001 7485.644 [0.00116 ] 15 f 13630.341 [0.001 7485.644 [0.001

17.5] 16.5 . . . . . . 18 ] 17 e 14455.600 0.00017 ] 16 e 14455.617 [0.00118 ] 17 f 14456.377 0.00017 ] 16 f 14456.394 0.000

18.5] 17.5 . . . . . . 19 ] 18 e 15281.653 0.00018 ] 17 e 15281.668 [0.00119 ] 18 f 15282.429 0.00018 ] 17 f 15282.447 0.002

19.5] 18.5 . . . . . . 20 ] 19 e 16107.705 0.000 8846.202 0.00019 ] 18 e 16107.720 0.000 8846.202 0.00020 ] 19 f 16108.481 0.00119 ] 18 f 16108.496 0.000

20.5] 19.5 . . . . . . 21 ] 20 e 9299.857 [0.00320 ] 19 e 9299.868 0.002

f 9300.291 0.00321.5] 20.5 . . . . . . e 9753.520 [0.003

f 9753.950 0.00122.5] 21.5 . . . . . . e 10207.187 0.004

f 10207.605 [0.00423.5] 22.5 . . . . . . e 10660.841 [0.002

f 10661.265 [0.00324.5] 23.5 . . . . . . e 11114.504 0.002

f 11114.928 0.00025.5] 24.5 . . . . . . e 11568.164 0.003

f 11568.590 0.00326.5] 25.5 . . . . . . e 12021.818 [0.002

f 12022.245 [0.001

NOTE.ÈAll transitions are in the ground Ðne structure ladder. In both molecules, the2%1@2 2%3@2ladder, an estimated 35 K higher, remains undetected. Measurement uncertainties are 2È5 kHz.a Designation of e and f levels is based on the assumption that the hyperÐne coupling constant d is

positive in and the sign of lambda-type doubling constant p ] 2q in is the same as inC9H C11H C9H.

4.1. H2C5Three sets of triplets were measured in and inH2C5 D2C5The spacing between lines in the triplet pattern is(Table 13).

determined by the asymmetry splitting of the K \ ^1 tran-

sitions. In the triplets are split by ¹50 MHz forH2C5,transitions up to 23 GHz. As discussed in Appendix C, thespectroscopic constants were derived by (1)(Table 14)Ðtting two rotational and two centrifugal distortion con-

TABLE 11

CALCULATED ROTATIONAL FREQUENCIES OF IN THE STATEC9H 2%1@2 (in MHz)

2%1@2 (with hfs)

e Levels f LevelsTRANSITION

J@ÈJ F\ J ] 1/2 F\ J [ 1/2 F\ J ] 1/2 F\ J [ 1/2

1.5È0.5 . . . . . . . 1238.200(16) 1240.214(80) 1239.170(16) 1241.968(80)2.5È1.5 . . . . . . . 2064.544(6) 2065.066(16) 2065.402(6) 2066.037(16)3.5È2.5 . . . . . . . 2890.698(3) 2890.953(6) 2891.518(3) 2891.811(6)4.5È3.5 . . . . . . . 3716.800(2) 3716.954(3) 3717.603(2) 3717.775(3)5.5È4.5 . . . . . . . 4542.882(1) 4542.987(2) 4543.676(1) 4543.791(2)6.5È5.5 . . . . . . . 5368.954(1) 5369.031(1) 5369.743(1) 5369.826(1)7.5È6.5 . . . . . . . 6195.021(1) 6195.081(1) 6195.807(1) 6195.870(1)8.5È7.5 . . . . . . . 7021.085(1) 7021.133(1) 7021.868(1) 7021.919(1)9.5È8.5 . . . . . . . 7847.147(1) 7847.187(1) 7847.928(1) 7847.970(1)

10.5È9.5 . . . . . . . 8673.207(1) 8673.241(1) 8673.987(1) 8674.023(1)11.5È10.5 . . . . . . 9499.266(1) 9499.296(1) 9500.045(1) 9500.076(1)12.5È11.5 . . . . . . 10325.324(1) 10325.351(1) 10326.102(1) 10326.130(1)13.5È12.5 . . . . . . 11151.381(1) 11151.405(1) 11152.159(1) 11152.183(1)14.5È13.5 . . . . . . 11977.437(1) 11977.459(1) 11978.214(1) 11978.237(1)15.5È14.5 . . . . . . 12803.492(1) 12803.512(1) 12804.269(1) 12804.290(1)16.5È15.5 . . . . . . 13629.547(1) 13629.565(1) 13630.323(1) 13630.342(1)17.5È16.5 . . . . . . 14455.600(1) 14455.618(1) 14456.376(1) 14456.394(1)18.5È17.5 . . . . . . 15281.653(1) 15281.669(1) 15282.429(1) 15282.445(1)

2%1@2 (no hfs) 2%1@2 (no hfs)

J@ÈJ e Level f Level J@ÈJ e Level f level

19.5È18.5 . . . . . . 16107.713(1) 16108.487(1) 60.5È59.5 49974.562(47) 49975.333(47)20.5È19.5 . . . . . . 16933.764(1) 16934.537(1) . . .21.5È20.5 . . . . . . 17759.813(1) 17760.587(1) . . .22.5È21.5 . . . . . . 18585.862(1) 18586.635(1) . . .23.5È22.5 . . . . . . 19411.909(2) 19412.683(2) . . .24.5È23.5 . . . . . . 20237.956(2) 20238.729(2) 85.5È84.5 70623.062(138) 70623.831(138)25.5È24.5 . . . . . . 21064.001(2) 21064.775(2) 86.5È85.5 71448.957(144) 71449.726(144)26.5È25.5 . . . . . . 21890.045(3) 21890.819(3) 87.5È86.5 72274.849(149) 72275.617(149)27.5È26.5 . . . . . . 22716.088(3) 22716.862(3) 88.5È87.5 73100.736(154) 73101.505(154)28.5È27.5 . . . . . . 23542.130(4) 23542.904(4) 89.5È88.5 73926.620(159) 73927.389(159)29.5È28.5 . . . . . . 24368.171(4) 24368.944(4) 90.5È89.5 74752.500(165) 74753.268(165)30.5È29.5 . . . . . . 25194.210(5) 25194.984(5) 91.5È90.5 75578.376(171) 75579.144(171)31.5È30.5 . . . . . . 26020.248(5) 26021.021(5) 92.5È91.5 76404.247(176) 76405.015(176)32.5È31.5 . . . . . . 26846.285(6) 26847.058(6) 93.5È92.5 77230.115(182) 77230.883(182)33.5È32.5 . . . . . . 27672.320(7) 27673.093(7) 94.5È93.5 78055.978(188) 78056.746(188)34.5È33.5 . . . . . . 28498.353(7) 28499.127(7) 95.5È94.5 78881.837(195) 78882.605(195)35.5È34.5 . . . . . . 29324.385(8) 29325.159(8) 96.5È95.5 79707.692(201) 79708.460(201)36.5È35.5 . . . . . . 30150.416(9) 30151.189(9) 97.5È96.5 80533.543(207) 80534.310(207)37.5È36.5 . . . . . . 30976.445(10) 30977.218(10) 98.5È97.5 81359.389(214) 81360.157(214)38.5È37.5 . . . . . . 31802.472(11) 31803.245(11) 99.5È98.5 82185.231(220) 82185.999(220)39.5È38.5 . . . . . . 32628.498(12) 32629.271(12) 100.5È99.5 83011.069(227) 83011.836(227)40.5È39.5 . . . . . . 33454.522(13) 33455.294(13) 101.5È100.5 83836.902(234) 83837.669(234)41.5È40.5 . . . . . . 34280.544(14) 34281.316(14) 102.5È101.5 84662.731(241) 84663.498(241)42.5È41.5 . . . . . . 35106.564(15) 35107.336(15) 103.5È102.5 85488.555(249) 85489.321(249)43.5È42.5 . . . . . . 35932.582(16) 35933.355(16) 104.5È103.5 86314.374(256) 86315.141(256)44.5È43.5 . . . . . . 36758.598(18) 36759.371(18) 105.5È104.5 87140.189(264) 87140.956(264)45.5È44.5 . . . . . . 37584.613(19) 37585.385(19) 106.5È105.5 87965.999(271) 87966.766(271)46.5È45.5 . . . . . . 38410.625(20) 38411.398(20) 107.5È106.5 88791.805(279) 88792.571(279)47.5È46.5 . . . . . . 39236.635(22) 39237.408(22) 108.5È107.5 89617.606(287) 89618.371(287)48.5È47.5 . . . . . . 40062.644(23) 40063.416(23) 109.5È108.5 90443.401(295) 90444.167(295)49.5È48.5 . . . . . . 40888.650(25) 40889.422(25) 110.5È109.5 91269.192(304) 91269.958(304)50.5È49.5 . . . . . . 41714.654(27) 41715.426(27) 111.5È110.5 92094.978(312) 92095.744(312)51.5È50.5 . . . . . . 42540.655(28) 42541.427(28) 112.5È111.5 92920.760(321) 92921.525(321)52.5È51.5 . . . . . . 43366.654(30) 43367.427(30) 113.5È112.5 93746.536(329) 93747.301(329)53.5È52.5 . . . . . . 44192.651(32) 44193.424(32) 114.5È113.5 94572.307(338) 94573.072(338)54.5È53.5 . . . . . . 45018.646(34) 45019.418(34) 115.5È114.5 95398.073(347) 95398.838(347)55.5È54.5 . . . . . . 45844.638(36) 45845.410(36) 116.5È115.5 96223.834(356) 96224.599(356)56.5È55.5 . . . . . . 46670.628(38) 46671.400(38) 117.5È116.5 97049.590(366) 97050.354(366)57.5È56.5 . . . . . . 47496.615(40) 47497.387(40) 118.5È117.5 97875.340(375) 97876.105(375)58.5È57.5 . . . . . . 48322.600(42) 48323.372(42) 119.5È118.5 98701.085(385) 98701.850(385)59.5È58.5 . . . . . . 49148.582(45) 49149.354(45) 120.5È119.5 99526.825(395) 99527.589(395)

NOTE.ÈFrequencies calculated from constants in Estimated uncertainties (in parentheses) are 1pTable 6.in the last signiÐcant digit from the least-squares Ðt. Transitions between 50 and 70 GHz are omitted owing tohigh atmospheric opacity. HyperÐne structure (hfs) is less than 0.5 km s~1 above 15 GHz. The spin2%3@2component, an estimated 35 K higher in energy, remains undetected.

116 McCARTHY ET AL.

stants in the asymmetric top Hamiltonian to the measure-ments ; and (2) Ðtting an e†ective B and D in the expressionfor a linear molecule separately to the K \ 0 and each(° 2.1)K \ ^1 ladder. The e†ective B and D in the K ladders arerelated to the asymmetric top constants (see footnote c in

Astronomers can readily calculate the entireTable 14).radio spectra of and with the e†ective B and DH2C5 D2C5values in Table 14.

4.2. H2C6The FTM spectrum of was approximately 7 timesH2C6weaker than that of Four sets of triplets wereH2C5.observed in but only the K \ 0 lines were observedH2C6,in The triplet pattern is tighter inD2C6 (Table 13). H2C6than in with a spacing of ¹30 MHz below 25 GHz.H2C5,Because of the nuclear spin statistics & Schawlow(Townes

the K \ ^1 lines in the isotopic species were too1955), D2weak to detect with the present sensitivity. The radio spec-trum of is readily calculated with the e†ective B andH2C6D in Table 14.

5. ASTRONOMICAL DETECTIONS

As mentioned in the Introduction, four of the eight

carbon chains discussed here have already been identiÐed inastronomical sources. The and radicals wereC7H C8Hdetected with the IRAM 30 m telescope in IRC ]10216

et al. & Gue� lin the(Gue� lin 1997 ; Cernicharo 1996) ; H2C6carbene has very recently been detected in TMC-1 with oneof the NASA 70 m telescopes of the Deep Space Network

et al. and there is a positive detection of(Langer 1997) ;in TMC-1 et al. which conÐrms thisHC11N (Bell 1997),

cyanopolyyne as the largest interstellar molecule. All ofthese detections have required very long integrations inexcess of 10 hr with very sensitive receivers, and it is likelythat detection of the remaining four molecules here willrequire at least a comparable devotion of time on the mostpowerful existing telescopes.

6. SUMMARY

A summary of recent laboratory measurements of therotational spectra of eight highly polar carbon chains ofastronomical interest is reported here. The measurementswere made primarily at centimeter wavelengths with a newFourier transform microwave spectrometer. Two of themolecules were also detected with a free space millimeter-

TABLE 12

CALCULATED ROTATIONAL FREQUENCIES OF IN THE STATEC11H 2%1@2 (in MHz)

2%1@2 (with hfs)

e Levels f LevelsTRANSITION

J@ÈJ F\ J ] 1/2 F\ J [ 1/2 F\ J ] 1/2 F\ J [ 1/2

1.5È0.5 . . . . . . . 679.957(28) 681.904(140) 680.383(28) 682.330(140)2.5È1.5 . . . . . . . 1133.806(12) 1134.269(28) 1134.232(12) 1134.695(28)3.5È2.5 . . . . . . . 1587.531(7) 1587.747(12) 1587.957(7) 1588.173(12)4.5È3.5 . . . . . . . 2041.222(4) 2041.348(7) 2041.648(4) 2041.774(7)5.5È4.5 . . . . . . . 2494.900(3) 2494.983(4) 2495.326(3) 2495.409(4)6.5È5.5 . . . . . . . 2948.572(2) 2948.631(3) 2948.998(2) 2949.057(3)7.5È6.5 . . . . . . . 3402.240(2) 3402.284(2) 3402.666(2) 3402.710(2)8.5È7.5 . . . . . . . 3855.907(1) 3855.941(2) 3856.333(1) 3856.367(2)9.5È8.5 . . . . . . . 4309.573(1) 4309.600(1) 4309.999(1) 4310.026(1)

10.5È9.5 . . . . . . . 4763.237(1) 4763.259(1) 4763.663(1) 4763.685(1)11.5È10.5 . . . . . . 5216.901(1) 5216.920(1) 5217.327(1) 5217.346(1)12.5È11.5 . . . . . . 5670.565(1) 5670.580(1) 5670.990(1) 5671.006(1)13.5È12.5 . . . . . . 6124.228(1) 6124.241(1) 6124.654(1) 6124.667(1)14.5È13.5 . . . . . . 6577.890(1) 6577.902(1) 6578.316(1) 6578.328(1)15.5È14.5 . . . . . . 7031.553(1) 7031.563(1) 7031.978(1) 7031.989(1)16.5È15.5 . . . . . . 7485.215(1) 7485.224(1) 7485.640(1) 7485.649(1)

2%1@2 (no hfs) 2%1@2 (no hfs)

J@ÈJ e Level f Level J@ÈJ e Level f level

17.5È16.5 . . . . . . 7938.880(1) 7939.306(1) 71.5È70.5 32435.916(48) 32436.341(48)18.5È17.5 . . . . . . 8392.541(1) 8392.967(1) 72.5È71.5 32889.545(50) 32889.971(50)19.5È18.5 . . . . . . 8846.202(1) 8846.628(1) 73.5È72.5 33343.174(52) 33343.600(52)20.5È19.5 . . . . . . 9299.862(1) 9300.288(1) 74.5È73.5 33796.802(55) 33797.227(55)21.5È20.5 . . . . . . 9753.523(1) 9753.949(1) 75.5È74.5 34250.429(57) 34250.854(57)22.5È21.5 . . . . . . 10207.183(1) 10207.609(1) 76.5È75.5 34704.055(59) 34704.480(60)23.5È22.5 . . . . . . 10660.842(1) 10661.268(1) 77.5È76.5 35157.680(62) 35158.105(62)24.5È23.5 . . . . . . 11114.502(1) 11114.928(1) 78.5È77.5 35611.304(64) 35611.729(65)25.5È24.5 . . . . . . 11568.161(1) 11568.587(1) 79.5È78.5 36064.927(67) 36065.352(67)26.5È25.5 . . . . . . 12021.820(1) 12022.246(1) 80.5È79.5 36518.549(70) 36518.974(70)27.5È26.5 . . . . . . 12475.478(2) 12475.904(2) 81.5È80.5 36972.170(73) 36972.595(73)28.5È27.5 . . . . . . 12929.136(2) 12929.562(2) 82.5È81.5 37425.790(75) 37426.215(75)29.5È28.5 . . . . . . 13382.794(2) 13383.220(2) 83.5È82.5 37879.408(78) 37879.833(78)30.5È29.5 . . . . . . 13836.451(3) 13836.877(3) 84.5È83.5 38333.026(81) 38333.451(81)31.5È30.5 . . . . . . 14290.108(3) 14290.534(3) 85.5È84.5 38786.643(84) 38787.068(84)32.5È31.5 . . . . . . 14743.765(3) 14744.190(3) 86.5È85.5 39240.258(87) 39240.683(87)33.5È32.5 . . . . . . 15197.421(4) 15197.847(4) 87.5È86.5 39693.873(90) 39694.298(90)34.5È33.5 . . . . . . 15651.076(4) 15651.502(4) 88.5È87.5 40147.486(94) 40147.911(94)

TABLE 12ÈContinued

2%1@2 (no hfs) 2%1@2 (no hfs)

J@ÈJ e Level f Level J@ÈJ e Level f level

35.5È34.5 . . . . . . 16104.732(5) 16105.157(5) 89.5È88.5 40601.098(97) 40601.523(97)36.5È35.5 . . . . . . 16558.386(5) 16558.812(5) 90.5È89.5 41054.709(100) 41055.134(100)37.5È36.5 . . . . . . 17012.041(6) 17012.467(6) 91.5È90.5 41508.319(104) 41508.744(104)38.5È37.5 . . . . . . 17465.695(6) 17466.120(6) 92.5È91.5 41961.927(107) 41962.352(107)39.5È38.5 . . . . . . 17919.348(7) 17919.774(7) 93.5È92.5 42415.535(111) 42415.960(111)40.5È39.5 . . . . . . 18373.001(7) 18373.426(7) 94.5È93.5 42869.141(115) 42869.566(115)41.5È40.5 . . . . . . 18826.653(8) 18827.079(8) 95.5È94.5 43322.746(119) 43323.171(119)42.5È41.5 . . . . . . 19280.305(9) 19280.731(9) 96.5È95.5 43776.350(122) 43776.775(122)43.5È42.5 . . . . . . 19733.956(9) 19734.382(9) 97.5È96.5 44229.952(126) 44230.377(126)44.5È43.5 . . . . . . 20187.607(10) 20188.032(10) 98.5È97.5 44683.554(130) 44683.979(130)45.5È44.5 . . . . . . 20641.257(11) 20641.682(11) 99.5È98.5 45137.154(135) 45137.579(135)46.5È45.5 . . . . . . 21094.906(12) 21095.332(12) 100.5È99.5 45590.753(139) 45591.177(139)47.5È46.5 . . . . . . 21548.555(13) 21548.981(13) 101.5È100.5 46044.350(143) 46044.775(143)48.5È47.5 . . . . . . 22002.203(14) 22002.629(14) 102.5È101.5 46497.946(147) 46498.371(147)49.5È48.5 . . . . . . 22455.851(15) 22456.277(15) 103.5È102.5 46951.541(152) 46951.966(152)50.5È49.5 . . . . . . 22909.498(16) 22909.924(16) 104.5È103.5 47405.134(156) 47405.559(156)51.5È50.5 . . . . . . 23363.145(17) 23363.570(17) 105.5È104.5 47858.727(161) 47859.151(161)52.5È51.5 . . . . . . 23816.790(18) 23817.216(18) 106.5È105.5 48312.317(166) 48312.742(166)53.5È52.5 . . . . . . 24270.435(19) 24270.861(19) 107.5È106.5 48765.907(171) 48766.331(171)54.5È53.5 . . . . . . 24724.080(20) 24724.505(20) 108.5È107.5 49219.495(176) 49219.919(176)55.5È54.5 . . . . . . 25177.723(21) 25178.149(21) 109.5È108.5 49673.082(181) 49673.506(181)56.5È55.5 . . . . . . 25631.366(23) 25631.792(23) . . .57.5È56.5 . . . . . . 26085.008(24) 26085.434(24) . . .58.5È57.5 . . . . . . 26538.650(25) 26539.075(25) . . .59.5È58.5 . . . . . . 26992.291(27) 26992.716(27) 154.5È153.5 70082.826(515) 70083.249(515)60.5È59.5 . . . . . . 27445.931(28) 27446.356(28) 155.5È154.5 70536.334(525) 70536.757(525)61.5È60.5 . . . . . . 27899.570(30) 27899.995(30) 156.5È155.5 70989.841(536) 70990.264(536)62.5È61.5 . . . . . . 28353.208(31) 28353.634(31) 157.5È156.5 71443.345(546) 71443.768(546)63.5È62.5 . . . . . . 28806.846(33) 28807.271(33) 158.5È157.5 71896.847(557) 71897.270(557)64.5È63.5 . . . . . . 29260.482(35) 29260.908(35) 159.5È158.5 72350.348(567) 72350.771(567)65.5È64.5 . . . . . . 29714.118(36) 29714.544(36) 160.5È159.5 72803.846(578) 72804.269(578)66.5È65.5 . . . . . . 30167.753(38) 30168.179(38) 161.5È160.5 73257.342(589) 73257.765(589)67.5È66.5 . . . . . . 30621.388(40) 30621.813(40) 162.5È161.5 73710.837(600) 73711.259(600)68.5È67.5 . . . . . . 31075.021(42) 31075.446(42) 163.5È162.5 74164.329(612) 74164.751(612)69.5È68.5 . . . . . . 31528.653(44) 31529.079(44) 164.5È163.5 74617.819(623) 74618.241(623)70.5È69.5 . . . . . . 31982.285(46) 31982.710(46) 165.5È164.5 75071.306(635) 75071.729(635)

NOTE.ÈFrequencies calculated from constants in Estimated uncertainties (in parentheses) are 1pTable 6.in the last signiÐcant digit from the least-squares Ðt. Transitions between 50 and 70 GHz are omitted owing tohigh atmospheric opacity. HyperÐne structure (hfs) is less than 0.5 km s~1 above 7.5 GHz. The spin2%3@2component, an estimated 35 K higher in energy, remains undetected.

TABLE 13

MEASURED MICROWAVE FREQUENCIES OF AND AND THEIR DEUTERIUM ISOTOPIC SPECIESH2C5 H2C6 (in MHz)

H2C5 D2C5 H2C6 D2C6JK{a,K{c@ ] J

Ka,KcFrequency O[C Frequency O[C Frequency O[C Frequency O[C

21,2 ] 11,1 . . . . . . 8420.444 0.00120,2 ] 10,1 . . . . . . 8452.440 0.00021,1 ] 11,0 . . . . . . 8484.105 [0.00131,3 ] 21,2 . . . . . . 13743.013 0.003 12630.657 0.00030,3 ] 20,2 . . . . . . 13771.755 0.001 12678.640 [0.001 8068.319 [0.001 7521.720 0.00131,2 ] 21,1 . . . . . . 13799.945 [0.003 12726.150 0.00041,4 ] 31,3 . . . . . . 18324.000 [0.001 16840.859 [0.001 10744.150 [0.00140,4 ] 30,3 . . . . . . 18362.323 0.000 16904.820 0.000 10757.756 0.000 10028.952 0.00041,3 ] 31,2 . . . . . . 18399.918 0.000 16968.185 0.000 10771.100 0.00051,5 ] 41,4 . . . . . . 22904.980 [0.001 13430.184 0.00150,5 ] 40,4 . . . . . . 22952.877 [0.001 13447.188 [0.001 12536.180 0.00051,4 ] 41,3 . . . . . . 22999.878 0.001 13463.870 0.00161,6 ] 51,5 . . . . . . 16116.214 0.00260,6 ] 50,5 . . . . . . 16136.617 0.000 15043.402 0.00061,5 ] 51,4 . . . . . . 16156.634 [0.00171,7 ] 61,6 . . . . . . 18802.235 [0.00270,7 ] 60,6 . . . . . . 18826.042 0.00171,6 ] 61,5 . . . . . . 18849.398 0.001

NOTE.ÈEstimated measurement uncertainty : 2È5 kHz.

118 McCARTHY ET AL. Vol. 113

TABLE 14

SPECTROSCOPIC CONSTANTS OF ANDH2C5, D2C5, H2C6, D2C6 (in MHz)

Constant H2C5 D2C5 H2C6 D2C6Asymmetric-top rotational and centrifugal distortion constants :a

A . . . . . . . . . . . . . . . . . . . 277600b 140300b 268400bB . . . . . . . . . . . . . . . . . . . 2304.7845(3) 2129.0276(2) 1348.0891(1)C . . . . . . . . . . . . . . . . . . . 2285.8053(3) 2097.1963(2) 1341.3519(1)D

J] 106 . . . . . . . . . . 104(6) 82(5) 28(1)

DJK

. . . . . . . . . . . . . . . . 0.04642(17) 0.04268(9) 0.01642(6)E†ective rotational and centrifugal distortion constants :c

BK/1L . . . . . . . . . . . . . 2290.5042(5) 2105.1122(5) 1343.0199(3)

DK/1L ] 106 . . . . . . 126(11) 151(20) 31(4)

BK/0 . . . . . . . . . . . . . . . 2295.2952(5) 2113.1123(5) 1344.7205(2) 1253.6209(3)

DK/0 ] 106 . . . . . . . 150(11) 307(21) 32(3) 57(6)

BK/1U . . . . . . . . . . . . . 2299.9927(5) 2121.0273(5) 1346.3884(3)DK/1U ] 106 . . . . . . 98(11) 131(21) 29(4)

NOTE.ÈUncertainties (1 p) are in units of the last signiÐcant digit.a Derived from a least-squares Ðt of WatsonÏs S-reduced Hamiltonian (Appendix C) to

the data in Table 13.b Derived assuming a planar structure (i.e., 1/C[ 1/A [ 1/B\ 0).c The entire radio spectrum in the K \ 0 and ^1 rotational ladders can be calculated

with the expression where the indices 1L and 1UlJ`1?J

\ 2Beff(J ] 1) [ 4Deff(J ] 1)3,refer to the lower and upper frequency K \ ^1 components ; andBeff \ (B] C)/2 ; Deff Bwhere c is 1/8 for K \ 0 and 1/32 for K \ ^1.D

J] c(B[C)2/4(A[Beff),

wave absorption spectrometer.Precise values of the rotational and centrifugal distortion

constants of and are given.HC11N, HC13N, H2C5, H2C6Owing to the unpaired electron in the carbon chain rad-icals, the spectra of and are moreC7H, C8H, C9H, C11Hcomplex ; lambda doubling and hyperÐne coupling con-stants were determined for all four, and the spin-orbit con-stant was also obtained from the millimeter-wavemeasurements of andC7H C8H.

The entire radio spectrum of andHC11N, HC13N, H2C5,can be calculated to an accuracy of 1 km s~1 or betterH2C6with only two spectroscopic constants in the standardexpression for a rotational transition in a linear molecule.Because of the resolved Ðne and hyperÐne structure presentin the radio spectra of the carbon chain radicals, the calcu-

lated frequencies of the astronomically interesting tran-sitions are tabulated.

Four of the molecules described here have been identiÐedin either the molecular shell of IRC ]10216 or the richmolecular source TMC-1. Now that accurate rest fre-quencies are available, deep searches for HC13N, H2C5,and can be undertaken with existing large tele-C9H, C11Hscopes.

We are indebted to our coauthors on some of the pre-vious papers P. Kalmus, S. E. Novick, and W. Chen. Wealso wish to thank E. S. Palmer for assistance with themicrowave electronics and E. W. Gottlieb for computa-tional assistance.

APPENDIX A

THE FOURIER TRANSFORM MICROWAVE SPECTROMETER

The Ðrst long carbon chain molecule of astrophysical interest observed by Fourier transform microwave (FTM) spectros-copy & Flygare was the cyanopolyyne Ohshima, & Endo demonstrated that reactive(Balle 1981) HC9N. Iida, (1991)hydrocarbon molecules observed in glow discharges by millimeter-wave absorption spectroscopy (e.g., andC3N C4H,

et al. et al. can be generated in a discharge nozzle Ðrst described by et al.Gottlieb 1983b ; C6H, Pearson 1988) SchlachtaCompared with the initial detection, which required about 8 minutes integration, can now be detected in a few(1991). HC9Nseconds integration with our new FTM spectrometer. This increase in sensitivity for long chain molecules results from (1)

orienting the supersonic molecular beam parallel, rather than perpendicular, to the Fabry-Perot axis ; (2) using certainmoderately large organic precursors (e.g., cyanoacetylene and diacetylene), rather than commercially available gases such asacetylene and nitrogen ; and (3) experimenting with the dimensions and electrical characteristics of the discharge nozzle.

A block diagram of our FTM spectrometer is shown in In this instrument, a pulsed supersonic molecular beam ofFigure 7.dilute (1%) organic vapor in an inert gas is produced by a 1 mm diameter commercial solenoid valve (General Valve). As thebeam expands through a nozzle positioned in one of the mirrors parallel to the axis of the high-Q confocal Fabry-Perot cavity(QB 104), a low-current pulsed discharge is applied synchronously with the gas pulse. Typically, 1000 V DC is applied to theelectrodes as a 300 ks long gas pulse expands from a backing pressure of about 2 atm through a discharge nozzle consisting ofalternating layers of copper electrodes and teÑon insulators, each a few millimeters thick. The initial organic vapors areheavily diluted in either argon or neon to maintain a steady gas discharge and to produce a very low rotational temperature ofabout 3 K in the free expansion. As the molecular beam traverses the large cavity (70 cm long, 36 cm diameter), the moleculesare irradiated with a short (1 ks) microwave pulse at frequency following which the free-induction decay (FID) is detectedl0,with a superheterodyne receiver. The Fourier transform of the FID is displayed as a frequency o†set from the pump frequency

No. 1, 1997 EIGHT NEW CARBON CHAIN MOLECULES 119

FIG. 7.ÈBlock diagram of the FTM spectrometer. A detailed description of the spectrometer is given in Appendix A.

The vacuum chamber was maintained at a background pressure of 2] 10~6 torr with a large 14 inch (35 cm) diameterl0.di†usion pump (Varian) backed by a dual-stage mechanical pump; typical peak pressures were 5] 10~5 torr at the 2 Hzrepetition rate of the nozzle. Helmholtz coils were used to cancel EarthÏs magnetic Ðeld to within about 2%.

APPENDIX B

THE MILLIMETER-WAVE ABSORPTION SPECTROMETER

The free space millimeter-wave absorption spectrometer Gottlieb, & Thaddeus et al.(Mollaaghababa, 1993 ; McCarthyis essentially the same device used for most of the molecules discovered in this laboratory since 1983 including1995) C2HGottlieb, & Thaddeus and et al. cyclic Vrtilek, & Gottlieb(Gottlieb, 1983a) ; C3N C4H (Gottlieb 1983b) ; C3H2 (Thaddeus,and et al. The reactive hydrocarbon species are produced in low-pressure DC glow discharges1985) ; H2CCC (Vrtilek 1990).

through Ñowing mixtures of gases such as HCCH and Ar. Long absorption path lengths of about 6 m are required fordetection of carbon chain molecules such as and by millimeter-wave spectroscopy, because the mole fractions areC7H C8Hvery low: D10~9.

The spectrometer is a nonresonant device that operates between 70 and 400 GHz and has excellent frequency agility. TheGunn oscillator millimeter-wave source is capable of scanning continuously over a wide frequency range of up to 400 MHz atthe 70È150 GHz fundamental. The output frequency of the Gunn oscillator can be doubled, tripled, or quadrupled, providingfrequency coverage of up to 1.6 GHz in a single scan. Although the instrumental standing waves are usually greater than theintensity of the absorption lines, they can be removed by a numerical Ðltering technique et al. Though the(McCarthy 1995).present investigation began with the free space device, it has turned out to be more than an order of magnitude less sensitivethan the FTM spectrometer used to detect most of the molecules in Table 1.

120 McCARTHY ET AL.

APPENDIX C

ANALYSIS

The radicals were analyzed with the standard Hamiltonian for a 2% molecule (Brown et al. TheCnH 1978, 1979).

millimeter-wave spectra of and could be adequately reproduced with Ðve free parameters : theC7H (Table 4) C8H (Table 8)rotation constant B, centrifugal distortion constant D, the spin-orbit constant A, and the lambda-type doubling constantsp ] 2q and q. In the rms of the Ðt to the combined millimeter-wave and FTM data is slightly smallerC7H, (Table 4) (Table 5)if the spin-rotation interaction c is included in the Hamiltonian. Owing to the high correlation of A with c, we determinedA] c rather than A separately ; c is in very good agreement with the value estimated from et al. on theC5H (Gottlieb 1986)assumption that c is proportional to B. The spin-orbit constant and one lambda-type doubling constant cannot be deter-mined when only one Ðne-structure component is observed, as in the case of andC9H C11H (Table 6).

Three hyperÐne coupling constants [a [ (b ] c)/2, b, and d] were derived from the FTM measurements in the 2%1@2Ðne-structure component of and but hfs was so small in that just one constant [a [ (b ] c)/2] could beC7H C9H, C11Hdetermined.The spectroscopic constants B, D, p ] 2q, q, a [ (b ] c)/2, and d in radicals decrease with increasing chain length. OurC

nH

measurements of and combined with earlier measurements of and conÐrm that the constants canC7H, C9H, C11H, C3H C5H,be predicted to within a few percent by simple extrapolation from one member in the series with an odd number ofC

nH

carbon atoms to the next. As noted by et al. the spin-orbit constants are only slightly less than for anTravers (1996b),unpaired electron in a 2p orbital on carbonÈi.e., except for et al. there is no evidence for strongC3H (Gottlieb 1985),Renner-Teller interactions between low-lying bending modes and the ground state up through C8H.

The cumulene carbenes were analyzed with WatsonÏs S-reduced asymmetric top Hamiltonian. Two rotational constants (Band C) and two centrifugal distortion constants and were least-squares Ðtted to the and lines in(D

JD

JK) H2C5, D2C5, H2C6with the same computer code used to analyze and (see et al. Another approachTable 13 H2CCC H2CCCC Travers 1996a).

that can be adopted is to perform a least-squares Ðt of the approximate analytic expression appropriate for a near prolate topFrequencies calculated with the expression containing only the leading terms in the power series expansion(Polo 1957).

lJ`1?J

\C(B]C)^ 1

2(B[C)d

K,1 [ 2DJK

K2D(J ] 1) [

G4D

J] (B[C)2

c[A[ (B]C)/2]H(J ] 1)3 ,

where c (a numerical constant) is 32 for K \ ^1 and 8 for K \ 0, agree with those calculated with the asymmetric topHamiltonian to within 0.4 km s~1 up to 300 GHz. Astronomers will probably Ðnd it most convenient, however, to use thelinear molecule expression and values of the e†ective B and D in to calculate thel

J`1?J\ 2B(J] 1)[ 4D(J ] 1)3 Table 14

radio spectra in the K \ 0 and ^1 rotational ladders of the cumulene carbenes.

REFERENCEST. J., & Flygare, W. H. 1981, Rev. Sci. Instrum., 52,Balle, 33M. B., Feldman, P. A., Travers, M. J., McCarthy, M. C., Gottlieb,Bell,

C. A., & Thaddeus, P. 1997, ApJ, 483, L61J. M., Colbourn, E. A., Watson, J. K. G., & Wayne, F. D. 1979, J.Brown,

Mol. Spectrosc., 74, 294J. M., Kaise, M., Kerr, C. M. L., & Milton, D. J. 1978, Mol. Phys.,Brown,

36, 553J., & Gue� lin, M. 1996, A&A, 309,Cernicharo, L27

C. A., Gottlieb, E. W., & Thaddeus, P. 1983a, ApJ, 264,Gottlieb, 7401986, A&A, 164,ÈÈÈ. L5

C. A., Gottlieb, E. W., Thaddeus, P., & Kawamura, H. 1983b,Gottlieb,ApJ, 275, 916

C. A., Killian, T. C., Thaddeus, P., Botschwina, P., Flu� gge, J., &Gottlieb,Oswald, M. 1993, J. Chem. Phys., 98, 4478

C. A., Vrtilek, J. M., Gottlieb, E. W., Thaddeus, P., & Hjalmar-Gottlieb,son, 1985, ApJ, 294,A� . L55

M. 1986, in Molecules in Physics, Chemistry, and Biology, ed.Gue� lin,J. Maruani (Dordrecht : Reidel)

M., et al. 1997, A&A, 317,Gue� lin, L1M., Ohshima, Y., & Endo, Y. 1991, ApJ, 371,Iida, L45

W. D., et al. 1997, ApJ, 480,Langer, L63S. A., & McLean, A. D. 1992, Chem. Phys. Lett., 200,Maluendes, 511M. C., Gottlieb, C. A., Cooksy, A. L., & Thaddeus, P. 1995, J.McCarthy,

Chem. Phys., 103, 7779M. C., Travers, M. J., Kalmus, P., Gottlieb, C. A., & Thaddeus,McCarthy,

P. 1996a, ApJ, 467, L1251997a, Chem. Phys. Lett., 264,ÈÈÈ. 252

M. C., Travers, M. J., Kova� cs, A., Chen, W., Novick, S. E.,McCarthy,Gottlieb, C. A., & Thaddeus, P. 1997b, Science, 275, 518

M. C., Travers, M. J., Kova� cs, A., Gottlieb, C. A., & Thaddeus,McCarthy,P. 1996b, A&A, 309, L31

R., Gottlieb, C. A., & Thaddeus, P. 1993, J. Chem. Phys.,Mollaaghababa,98, 968

M., & Botschwina, P. 1995, J. Mol. Spectrosc., 169,Oswald, 181J. C., Gottlieb, C. A., Woodward, D. R., & Thaddeus, P. 1988,Pearson,

A&A, 189, L13S. R. 1957, Can. J. Phys., 35,Polo, 880

R., Lask, G., Tsay, S. H., & Bondybey, V. E. 1991, Chem. Phys.,Schlachta,155, 267

P., Vrtilek, J. M., & Gottlieb, C. A. 1985, ApJ, 299,Thaddeus, L63C. H., & Schawlow, A. L. 1955, Microwave Spectroscopy (NewTownes,

York : McGraw-Hill)M. J., Chen, W., Novick, S. E., Vrtilek, J. M., Gottlieb, C. A., &Travers,

Thaddeus, P. 1996a, J. Mol. Spectrosc., 180, 75M. J., McCarthy, M. C., Gottlieb, C. A., & Thaddeus, P. 1996b,Travers,

ApJ, 465, L77M. J., McCarthy, M. C., Kalmus, P., Gottlieb, C. A., & Thaddeus,Travers,

P. 1996c, ApJ, 469, L651996d, ApJ, 472,ÈÈÈ. L61J. M., Gottlieb, C. A., Gottlieb, E. W., Killian, T. C., & Thaddeus,Vrtilek,

P. 1990, ApJ, 364, L53G., Creswell, R. A., & Winnewisser, M. 1978, Z. Naturforsch,Winnewisser,

33a, 1169D. E. 1996, Chem. Phys. Lett., 244,Woon, 45