The Infrared Absorption Spectrum of Carbon Sub Oxide
Total Page:16
File Type:pdf, Size:1020Kb
AUGUST, 1937 JOURNAL OF CHEMICAL PHVSICS VOLUME 5 The Infrared Absorption Spectrum of Carbon Suboxide R. C. LORD, JR.* AND NORMAN WRIGHT Departments of Chemistry and Physics, University of Michigan, A nn Arbor, Michigan (Received "May 24, 1937) The infrared absorption spectrum of carbon suboxide vapor has been studied under low dispersion in the region from 2).1 to 25).1. The vibrational bands observed may be satisfactorily accounted for on the basis of a linear symmetric structure for the carbon suboxide molecule, in agreement with evidence obtained from Raman spectra and from electron diffraction investigations. ECENT studies of the structure of the material was prepared by the pyrolysis of R carbon suboxide molecule by electron dif diacetyl tartaric anhydride8 and was purified by fractiont ,2 and by ultraviolet absorption and two distillations in an all-glass system. Measure 3 Raman spectra - 5 have made desirable an in ments of the vapor pressure of the suboxide over vestigation of the molecule's absorption spectrum a range of temperatures up to O°C agreed in the infrared region. The information to be satisfactorily with those in the literature. 9 The gained from such an investigation should be absorption cell was filled by distilling the middle particularly helpful since the spectroscopic and fraction of the particular sampJe at hand into electron diffraction evidence indicates that the the evacuated ce1l, the desired pressure being most probable form of the carbon suboxide obtained by surrounding the distilling vessel molecule is one possessing a symmetry center. with a bath of acetone of the proper temperature. For such a structure the Raman and infrared While in every instance we were satisfied absorption spectra are complementary, the befote making the absorption measurements fundamental frequencies active in either type of that the carbon suboxide in the cell was of spectrum being inactive in the other. suitable purity, each absorption curve was ex We have studied the absorption by carbon amined for the presence of bands due to foreign suboxide vapor under various pressures in the TABLE 1. Absorption bands of carbon suboxide. region between 2/.1 and 25/.1. The absorption curves were made with the KBr prism, vacuum, PERCENT ABSORPTION 6 recording spectrometer of Randall and Strong WAVE-LENGTH WAVE-NUMBER operated in conjunction with the periodic ampli IN I' IN CM-t p=14mm 67 mm I 200mm 7 2.18 4590 5 fier described by Firestone. The cell containing 2.64 3790 30 2.96 3380 5 the absorbing gas was 12 em in length, with 3.17 3150 - 30 40 4.15 2410t plane polished KBr windows cemented to each 4.37 2290 80 90 95 4.57 2190) end. The customary Nernst glowe~ furnished 5.05 1980 - 10 unresolved 5.41 1850 - 5 unresolved the infrared radiation. 5.68 1760 25* 40 80 5.98 1670 40 sO Professor J. O. Halford kindly supplied several 6.37 1570 25 75 95 6.80 1470 5 10 25 samples of liquid carbon suboxide, of two or 7.21 1387 20 30 70 8.16 1225 - 2 15 three cubic centimeters each, for our use. The 8.88 1126 - 10 35 9.76 1024 - 5 20 * National Research Council Fellow. 11.00 909 10 30 65 11.24 889 60 1 Brockway and Pauling, Proc. Nat. Acad. Sci. 19, 860 - 25 12.84 779 - 20 40 (1933). 15.70 637 15 60 75 3 Boersch, Naturwiss. 22, 172 (1934); Wien Ber. 144 17.94 5S7 75 (lIb) 1 (1935). 18.18 550 50 65t 90 18.40 544 70 3 Badger and Barton, Proc. Nat. Acad. Sci. 20, 166 (1934). 4 Thompson and Healey, Proc. Roy. Soc. (London)A1S7, Spectral range: 14 mm: 31'-201'; 67 mm: 2..-251'; 200 mm: 31'-251" * Water vapor band on a blank rnn with source intensity comparable 331 (1936). to that used for the 14 mm run absorbs some 20 percent at this point. 6 Engler and Kohlrausch, Zeits. f. physik. Chemie B34, Most of this band in the 14 mm curve is doubtless due to water. 214 (1936). t Central minimum of this band resolved only in 61 mm curve. 6 H. M. Randall and J. Strong, Rev. Sci. Inst. 2, 585 (1931), g Hurd and Pilgrim, J. Am. Chern. Soc. SS, 757 (1933). 7 F. A. Firestone, Rev. Sci. Inst. 3, 163 (1932). • E.g., reference 4, p. 332. 642 SPECTRUM OF CARBON SUBOXIDE 643 100 r-------------------------------------------------------~I~ ~ ~80 80 -4.. ~ .:~ 60 ~ 40 ""~ "'0 ~ v ~ 20 zo ~ 0 L-__~ ____~----~~~~----~~--~--------~~--------~ 0 0 5 10 15 20 25 Wave-Ieflsfh i'l P. FIG. 1. Infrared absorption of carbon suboxide vapor. Cell length, 12 cm; pressure, 67 mm. gases. Absorption bands in the region 3.2-3.5~ always led to the value, within a millimeter or were either extremely weak or nonexistent. Since two, of the initial pressure with which the cell compounds with the C - H linkage exhibit rather had been filled. When the cell was filled for the powerful absorption in this region, the concen first time (at about 700 mm pressure), a very tration of such possible impurities as ketene, faint fogging of the windows was observed after acetic acid and other pyrolytic products with a few hours. This cloudiness did not increase C - H groups must have been very small. The with time, however. The windows were not weak carbon dioxide bands present in the ab visibly less transparent after the cell had been sorption curves were of the intensity to be filled and emptied several times than at the expected from the amount of carbon dioxide in time when the fogging was first observed. The the air path of the incident radiation. The absorption of the clouded windows was tested fundamental band of carbon monoxide, if present with the cell exhausted to an oil pump vacuum. in the absorption curves, would have been Only those absorption bands due to the carbon masked by the strong and rather wide absorption dioxide and water vapor in the 50 cm airpath band found at 4.4~. outside the spectrometer were observed. No In the paper of Thompson and Healey,4 trace of the strong bands of carbon suboxide attention is called to the possibility that carbon appeared. suboxide prepared by the pyrolytic method may Absorption curves were made with the carbon be contaminated with a small amount of sulfur suboxide under pressures of 14, 67, 200 and 700 dioxide. There is no trace in our absorption mm. The absorption of the gas under 700 mm curves of the strong sulfur dioxide fundamental was not studied over the whole available range at 8.67/10. The other strong fundamental, at of the spectrograph because in the region first 7.40/10, might be partially obscured by the 7.21/10 explored, from 10 to 20/10, it was found that band of the carbon suboxide spectrum, but the absorption was too strong to allow satisfactory shape of the latter band gives no indication of a resolution of the individual bands. Table I lists satellite on the long wave-length side. The remaining sulfur dioxide bands are of lower the wave-lengths and wave numbers of the intensity, and of these, the ones which do not observed bands, together with the approximate coincide with the bands of the suboxide make percentages of absorption under the several no appearance on the absorption records. pressures. Of these pressures, 67 mm was found It might be of interest to remark that at no to be the most favorable for resolution of the time did we experience difficulty due to polymeri structures of the various bands. The curve for zation of the carbon su box ide vapor in the absorp 67 mm is reproduced in Fig. 1. As the figure tion cell. Measurement of the pressure of the gas indica tes, no absorption bands were found be after it had remained in the cell for several days tween 20 and 25j.1. 644 R. C. LORD, JR., AND N. WRIGHT o c c c 0 symmetry has seven vibration frequencies, three of which are doubly degenerate. The vibrational ..... .. .. ~ modes are shown in Fig. 2. Our frequency enumeration follows that of Engler and Kohl .. .. •• )I, .. 5 rausch. These authors, as well as Thompson and Healey,4 give both the vibration forms and .. .. .. .. .. ~ also equations for the frequencies based on simple potential systems of the valence force .. .. .. .. .. )/4 type. , , , , Ys The four nondegenerate frequencies, which f preserve the figure axis as an element of sym , , metry, may be classified as symmetrical (VI and V6 f t t V2) and antisymmetrical ("3 and V4) to the , symmetry center. The totally-symmetrical VI • 1 ~ ~ t and V2 may appear in the Raman effect, while FIG. 2. Modes of vibration for linear carbon suboxide V3 and V4, which produce temporary electric molecule. moments along the figure axis, should be active in infrared absorption. By estimating force DISCUSSION OF RESULTS constants from analogous frequencies in carbon To interpret the infrared absorption data, one dioxide and allene, Engler and Kohlrausch may make profitable use of some preconceptions calculate that V1 and V2 should appear, respec about the structure of the carbon suboxide tively, at about 2200 cm-1 and 800 cm-1. They molecule. From the electron diffraction studies,!. 2 are thus able to identify the Raman frequencies it appears very likely that all the atoms in observed at 2200 cm-1 and 843 cm-1 with these carbon suboxide are on a line, and furthermore vibrations. Having accomplished this assign that the oxygen atoms and two of the carbon ment, they can then compute accurate values of atoms are symmetrically disposed about the the force constants and predict that V3 should lie 1 third carbon atom.