38 METHYL DIACETYLENE THE MICROWAVE SPECTRUM AND STRUCTURE OF METHYL DIACETYLENE BY G. A. HEATH, L. F. THOMAS, E. I. SHERRARD AND J. SHERIDAN Dept. of Chemistry, The University, Birmingham Received 3 1 st January, 1955 Pure rotation spectra of H3C5H, HsC~D, D3C5H and D3CsD show that they are symmetric-top molecules with strictly linear C5H (or C5D) arrangements. The spectro- scopic constants Bo and distortion constants DJ and DJK are evaluated for these molecules. If the structural parameters concerned with the hydrogen and deuterium atoms are the same as the analogous parameters in methyl acetylene, preliminary carbon-carbon distances computed are : C (methyl)-C = 1.459 A, CE C (both assumed equal) = 1.212 A, C-C = 1.366A. Spectra are also measured for these molecules in the first excited level of a degenerate vibrational mode.These spectra are interpreted, and spectroscopic constants assigned for these states, on the basis of the theoryof I-type doubling in symmetric-top molecules. The methyl diacetylene (penta-1 : 3-diyne) structure, H3C-C=C-CfC-H, has been attributed to at least two different substances. The evidence, reviewed by Jones and Whiting,l favours allocation of this structure to the substance made by Armitage, Jones and Whiting.;! The truth of this is conclusively proved by the microwave spectrum of the substance made by this method.3 The inter- nuclear distances in this molecule are of interest, and the object of the work of which this paper forms part is to determine as fully as possible the structural parameters from microwave spectra of various isotopically substituted species.In the present communication, data are given for the molecules H3CCCCCH, H3CCCCCD, D3CCCCCH and D&CCCCD, from which preliminary deductions concerning the structure of the conjugated carbon chain can be made. With the extension of measurements to species containing carbon-13 it is expected that alI the atoms can be accurately located, with the probable exception of the central carbon atom, which is very close to the centre of gravity of the molecule. Rotation spectra of molecules in the first excited level of a degenerate vibrational mode have also been measured, and interpreted to give molecular information concerning this vibrational state. EXPERIMENTAL Methyl diacetylene was made by the method of Armitage, Jones and Whiting,z and purified by vacuum fractionation.The D3CCCCCH was obtained similarly, highly deuterated methyl iodide being used in the methylation stage. Samples of H3CCCCCD and D3CCCCCD were prepared from HJCCCCCH and D3CCCCCH respectively by treatment with an excess of 99.7 % D2O containing NaOD, for two days at room tem- perature, with frequent mixing. The spectra of samples obtained on subsequent fraction- ation showed that a large proportion of the acetylenic hydrogen was exchanged by this treatment.G . A . HEATH, L. F. THOMAS, E. I . SHERRARD AND J . SHERIDAN 39 The spectra were measured with a sweep spectrometer and frequency standard which have been described previously.4 Gas pressures of about 5 x 10-3 mm and temperatures of - 70" C were used.RESULTS The frequencies measured in the spectra of molecules in their ground states are listed, with their assignments, in table 1, The spectroscopic constants and moments of inertia, ZB, are given in table 2. The IB values are calculated from the relationship ZB (g cm2) = [839102/Bo(Mc/s)] x 10-40, derived from constants listed by DuMond and Cohen.5 K 3 2 3 4 5 6 7 9 K 0 1 2 3 4 5 6 7 K 0 1 2 3 4 5 6 7 8 9 TABLE ME MEASURED FREQUENCIES FOR GROUND STATES (MC/S rt 0.1 MC/S) HjCCCCCH J = 4 + 5 J = 5+6 J = 8 4 9 24,428.82 3 6,643.08 20,357.38 24,428.60 3 6,642-77 20,35656 24,427.85 36,641 -70 20,35555 24,42669 36,639.90 20,354.1 8 24,425.03 36,637-49 - 24,422.8 3 36,634-20 - - 36,630.24 - - - - - - H3CCCCCD J=9-+10 40,7 14.56 40,714.14 40,71296 40,7 10.96 40,708.20 40,704.62 40,700.28 40,695.1 0 - J - lo-+ 1 1 44,78592 44,785.48 44,784.16 44,782.02 44,778-98 44,770-20 44,764.52 44,75052 44,775.04 DsCCCCCH J = 4 4 5 J=5+6 19,297.70 23,157.21 19,297.52 23,156-99 19,296.98 23,15634 19,29608 23,155.25 - 23,153.71 - 23,151-67 - - - - - DjCCCCCH (ctd.) J = 8 - + 9 J=11+12 33,027.09 44,03580 33,026-86 44,035.48 33,026.08 44,034-42 33,02477 44,03264 33,022-94 44,030.18 33,020-60 44,027-08 33,017.74 44,023.24 33,014-30 - 33,010.35 - - - J = 1 0 4 11 42,454-66 42,454-25 42,45 3 -05 42,451.03 42,448.24 42,444.61 42,440.20 42,434.89 J=6+7 24,39045 24,390-68 24,390.12 24,389.17 24,387.83 24,386.10 I - - - J = 11412 J = 6 4 7 46,31411 25,687-84 46,313.66 25,687.66 46,3 12.36 25,687.03 46,3 10.1 6 25,686.05 46,307.07 25,68457 46,303-13 - 46,298.32 - - - DjCCCCCD J=9-+10 34,843.89 34,843-65 34,842.86 34,841 -5 3 34,839.63 34,837-21 34,834-26 34,830.70 34,82667 34,82204 J = 12-t 13 45,296-66 45,296.40 45,295.30 45,29356 45,29 1-10 45,287438 45,284.10 - - - TABLE 2.-sPECTROSCOPIC CONSTANTS FOR GROUND STATES molecule Bo DJ DJK IB (kc/s f 0.1) (g X 10-40 (Mcls f 0.02) (kcls f 0.1) f 0.004) H3CCCCCH 2,03 5.741 0.07 19.84 412.185 H3CCCCCD 1,929.772 0.06 18-30 434820 D3CCCCCH 1,834.856 0.1 14-54 457.3 11 D3CCCCCD 1,742.21 5 0.1 13-54 481 *628 IPossible error in Planck's constant is not included in the errors quoted for IB.The errors quoted for the values of DJK are thought to be conservatively estimated. The frequencies calculated from the constants given agree with those measured with an average deviation of 0.03 Mc/s, a discrepancy of 0.1 Mc/s being found only in a few weak lines.40 METHYL DIACETYLENE TABLE 3.-oBSERVED AND CALCULATED FREQUENCIES FOR VIBRATIONAL STATES (MC/S) H3CCCCCH K I r t l z t l r t l r t l rt2 r t l 0 *l r t l 71 *3 r t l &2 7 1 rt4 f l &3 F 1 rt5 f l f 4 F 1 rt5 7 1 unassigned r t l &l rt1 r t l +2 *1 0 ztl &3 -+1 &l F1 f 2 T l 1 4 f 1 f 3 T l rt5 r t l rt5 r l &7 r t l rt6 rl rt8 rtl unassigned J = 4 4 5 obs.calc.* 20,411.95 412.00 20,390.87 -390.96 J = 5 + 6 obs. calc. 24,494.38 -49440 24,469.1 1 469.16 -481.52 24,481'52 481.55 247480'78 -480.75 480.88 24,479.62 i Tz",:",:z 24,475.95 -476.00 H3CCCCCD J = 8-9 obs. calc. 36,74 1-44 -741 -48 36,703'62 -703.60 -722.23 36,722*30{ -722.14 36,713'83 -713.86 36,710.05 -709.98 36,70740 J=4+5 obs.calc. 19,348.50 -348.49 19,328-94 -328.93 -337.92 -337.10 19,335.80 -335.83 K *l f l f l & l f 2 &l 0 f l & l r l +3 *l s 2 r l rt4 &l zt3 r l f 5 &1 zt4 F l &5 r l +8 &l unassigned J = 5 4 6 obs. calc. 23,218.1 7 -21 8.1 9 23,194-67 -194.71 -206.22 -206.21 23,206-23 .( 2 3,202.9 6 -202 97 D3CCCCCH J = 8 4 9 obs. calc. 33,112.67 -1 12.64 33,080.20 -080.16 33,096.27 -096.26 33,096.06 -096.01 -095.36 -094.12 -492.34 -092.04 33,092-23 { 33,089.98 -090.04 33,087.11 -087.19 33,083.30 -083.21 J = 1 1 4 12 obs. calc. 46,436.04 -436.03 46,389.10 -389'09 46,412.42 -412.38 46,411.82 -41 1.84 405.57 46,405.40( -405.20 46,396.90 .-396*75 46,396'24 -396.05 46,390.82 -391.00 (4h,389*94 46,386.78 -390.14 46,387.80 D3CCCCCD J = 9 4 10 obs.calc. 34,932.59 -932.61 34,898.93 -898.93 34,915'70 -915.67 34,9 15.35 -9 15.3 1 34,911.46 L-9 11.44 - 34,909.06 -908.98 34,900.65 34,901.48 * The dash before the calculated figures indicates that the number of thousands has been omitted, for brevity.G. A . HEATH, L . F. THOMAS, E. r. SHERRARD AND J . SHERIDAN 41 The observed frequencies listed in table 3 are due to molecules in the first excited level of a degenerate vibrational mode. The form of these spectra is that predicted by Nielsen’s theory of I-type doubling in symmetric-top molecules.6 Each transition consists of two widely spaced lines (K = I = 4 1) about a central group of lines which become more widely spaced towards lower frequencies. The frequencies predicted theoretically for given values of J, K and I have been expressed 7 by general formulae involving also the rotational constants BV and A v for the vibrational state, the distortion constants BJ and DJK for that state, the Coriolis coupling coefficient for the vibration, 5, and the quantity q, equal to 2aBe2/w, where Be is the equilibrium rotational constant, w is the frequency of the vibration, and a is a factor not far removed from unity.In fitting the spectra it is convenient to assign values to the expression denoted by X in table 4, which lists values of this and other constants giving the best fit with observation. The small constants DJ are taken to be the same as for the ground states. While accurate values of Bvand q are obtainable, and DJK is shown to have values close to those for the ground states,( and X are less precisely determined.This is partly because accurate measurements are somewhat hindered by near-coincidences among the TABLE 4.-sPECTROSCOPIC CONSTANTS FOR VIBRATIONAL STATES 4 t- molecule BV DJK b U a (Mc/s) (Icc/s f 10) (kcls) (kcls) (cm-9 (Mcls) H3C5H 2,040.14 2104 20.0 0.9 0.15 151 -4.40 H3C5D 1,933.86 1956 18.7 0.92 0.20 146 -4.09 D3C5H 1,838.69 1804 14.6 0.9 0.23 143 -3.84 DSsD 1,745.80 1684 14.0 0.9 0.23 138 -3.58 central lines, especially with the heavily deuterated molecules, the lines for which are broader than those for the light species. The most probable values for f are slightly under + 1, and are thus in the theoretically permitted range for the Coriolis coupling coefficient. It is noteworthy that infra-red spectra indicate that the Coriolis coupling factors for vibrations, similar to the present one, in methyl acetylene,s.9 methyl cyanide 109 11 and methyl isocyanide 11 are also slightly under + 1.Some weak unassigned lines are noted also in table 3. They may be due to molecules in other excited vibrational states, some of which may well be appreciably populated for such a molecule under the experimental conditions. We do not, however, regard the present assignment of the spectra as final, and hope later to reexamine them using higher sensitivity. Also included in table 4 are the constants a, for the vibration concerned, which occur in the formulae 12 connecting Bv or Bo with Be. Under the heading 1.15w{a in table 4 are the frequencies of the vibration obtained from q, if a is taken as 1-15, as proposed for methyl acetylene in a similar vibrational state ; 13 Bo is used as an approximation to Be in deriving these frequencies.Since a is uncertain, their absolute values are approximate, but their relative values should be more reliable. The relative intensities of the spectra of the vibrational and ground states are in rough accord with vibration frequencies as low as those listed. Insuflicient is known of the vibration spectra of methyl diacetylene for the vibration concerned to be identified, but its low frequency and degeneracy accord with its involving bending of the C5H (or C5D) chain. Further close groups of lines due to vibrationally excited molecules are observed at frequencies higher, for a given transition, than those in table 3 ; the separations of these further groups from the spectra of the ground states are roughly twice the corresponding separations for the first vibrational state.This fact, and the relative intensities of the spectra, suggest that these further lines are due to molecules in the second excited level of the vibration discussed above. DISCUSSION The spectra of the ground states are those of strictly symmetric-top molecules. This indicates beyond doubt that, as expected, all five carbon atoms and the acetylenic hydrogen atom are located linearly on the figure axis of the molecule.42 METHYL DIACETYLENE As discussed in detail below, the moments of inertia are in excellent agreement with those expected for the methyl diacetylene structure.The other substance, or substances, to which this structure has been attributed 1 must presumablybe unsaturated five-carbon compounds, containing more than four hydrogen atoms, and will be asymmetric-top molecules. A preliminary analysis of the structural parameters can be made by assuming those associated with the hydrogen atoms, and computing information about the carbon skeleton. It is reasonable to suppose that the parameters associated with the H or D atoms will closely resemble the analogous parameters in methyl acetylene.13~ 14 Accordingly we assume the following : 14 d& (methyl) = 1.1 12 A, d c ~ (methyl) = 19108A, d c ~ (acetylenic) = 1.060 A, d c ~ (acetylenic) = 1.058 A, LHCH = 108" 25', L DCD = 108" 32'. Any three IB values can now be combined to give the total length of the Cs chain, a value of 5248A being consistently obtained from each of the four possible combinations of moments of inertia. This length is 0-23A shorter than the sum of two normal single and two normal triple bonds, and a shortening of this order of magnitude, due to conjugation, is expected from a knowledge of related structures.To proceed further, it is convenient to assume that the two triple bonds are equal in length. This is suggested by the equality of such bond lengths found crystallographically in dimethyl triacetylene,ls and by the relatively small variations found in the lengths of C r C bonds in different molecules. Any three IB values can then be combined to obtain the following carbon chain structure (distances in A) : 1'459 1'366 H 3 c - c ~ C-CE C--H 1'212 1'212 Internally consistent results are obtained from the four possible combinations of IB values.This preliminary structure is supported by the similarity of the distances to those found in related molecules. The bond length of 1.459A is close to the lengths of the analogous bonds in methyl acetylene,l3, 14 and in dimethyl tri- acetylene.15 The lengths of the triple bonds resemble those found in various acetylenic substances, and the bond length of 1.366A is similar to that found for other "single" C-C bonds located between two triple bonds, as in dia- cetylene,l6 cyanoacetylene 17 and dimethyl triacetylene.15 The influence of variations in the assumed parameters on the computed chain structure was examined.If the reasonable variations of f 1" in LHCH and of f 0.01 A in the CH and CD distances are allowed, the resulting probable uncer- tainty in the length of the CJ chain is about f 0.02A. A similar uncertainty results for the distance C (methyl)-C. The remaining parameters, however, are individually uncertain by at least twice this amount, which makes desirable the determination of spectra of other isotopic species. It appears, none the less, that considerable shortening of bonds below their normal lengths occurs in the conjugated chain. This is presumably due to contributions of structures which give double bond character to all the carbon- carbon bonds, similar to those proposed for diacetylene and methyl acetylene.16 In keeping with such a similarity, the dipole moment of methyl diacetylene appears, from the intensities of its spectra, to be of the same order of magnitude as that of methyl acetylene (0075D).We hope to determine the dipole moment of methyl diacetylene from measurement of the Stark effect on its rotational transitions of low 3. One of us (G. A. H.) is indebted to the Department of Scientific and Industrial Research for a Maintenance Grant.G . A. HEATH, L. F. THOMAS, E . I . SHERRARD AND J . SHERIDAN 43 1 Jones and Whiting, J. Chem. SOC., 1953,3317. 2 Armitage, Jones and Whiting, J. Chem. SOC., 1952, 1993. 3 Heath, Thomas and Sheridan, Nature, 1953, 172,771. 4 Heath, Thomas and Sheridan, Trans. Faraday SOC., 1954, 50,779. 5 DuMond and Cohen, Rev. Mud. Physics, 1953, 25, 691. 6 Nielsen, Physic. Rev., 1950, 77, 130. 7 Anderson, Trambarulo, Sheridan and Gordy, Physic. Rev., 1951, 82,58. 8 Boyd and Thompson, Trans. Faraday SOC., 1952,48,493. 9 Grisenthwaite and Thompson, Trans. Faraday Soc., 1954, 50,212. 10 Venkateswarlu, J. Chem. Physics, 1951, 19, 293. 11 Thompson and Williams, Trans. Farad~y SOC., 1952,48, 502. 12 Gordy, Smith and Trambarulo, Microwave Spectroscopy (J. Wiley and Sons, Inc., 13 Trambarulo and Gordy, J. Chem. Physics, 1950,18,1613. 14 Thomas, Sherrard and Sheridan, Trans. Faraday SOC. (in press). 15 Jeffrey and Rollett, Proc. Roy. SOC. A, 1952, 213, 86. 16 Pauling, Springall and Palmer, J. Amer. Chem. Suc., 1939, 61, 927. 17 Westenberg and Wilson, J. Amer. Chem. SOC., 1950, 72, 199. 1953), p. 103.