COMMUNICATIONS T O T H E EDITOR SPECTROSCOPIC EVIDENCE
FOR GASEOUS
bands in both the 240 and 350 mM regions indicated that its pressure was considerably above the equiSir: librium vapor pressure. Spectra taken 200 psec. We wish to report detection of a volatile pre- after initiation of photolysis revealed biphenylene cursor of biphenylene from flash photolysis oi long-wave-length bands. in addition to intense abbenzenediazonium-?-carboxylate(I). The short- sorption in the region below 270 mp by a species alllk-ed intermediate is tentatively identified as parently different from biphenylene. At 10 psec., benzyne. This assignment follows a consideration with relatively large samples (20 p moles), the longof the source of the species, its ultraviolet spectrum, wave-length bands of biphenylene were barely and the rapidity with which it forms gaseous bi- detectable, yet the region below 270 mp was virtually opaque and strong absorption extended beyond 300 mp. Use of smaller samples allowed a study of the spectrum of the intermediate 10 psec. after initiation of photolysis. I I1 111 The spectrum of the intermediate divides itself into several regions: (1) very weak absorption a t phenylene (111). I t owes its credibility to the wave lengths greater than 270 mp; (2) slightly previous accumulation of evidence' that such an stronger absorption from 270 to about 255 or 250 intermediate is involved in certain aromatic sub- mp; ( 3 ) distinctly stronger absorption in the region stitution reactions, and in the formation of bi- 250-235 mp, and (4) very strong absorption bephenylene and triphenylene from monocyclic ma- yond 230 mp. Thus far, our efforts have been diterials. rected principally a t studying the 250-235 mp Films of solid salt (I) were deposited in a trans- region. We find a broad maximum between 238.5 parent silica tube (60 cm. long, 1.6 cm. diam.) and 241.5 mp. Table I gives plate densities for which was evacuated and closed off. Samples comparison flash and photolyzed sample, as well were subjected t o photolysis by light from a dis- as raw differences and differences multiplied by charge through a silica tube filled with 2-6 cm. of scale factors to correct for plate response, for a argon. The photolytic flash discharged '730 joules typical experiment. It is difficult to estimate in approximately 20 psec.; 60 pf. capacitance the extinction coefficient a t Xmax because instancharged to 5 kv. supplied the energy. When taneous concentrations are not known. h minirelatively large samples were photolyzed (0.1- mum value can be set, however, a t about 600. 0.2 mmole), one could see that immediately following photolysis the tube was filled with a cloud of TABLE I fine yellow dust which settled in about a minute. DESSITOMETER READINGS O F PLATE DESSITYI N THE 240 Quantities of carbon dioxide and nitrogen proM p REGION duced indicated that approximately SOY0 reaction Plate Denpity----Difference Corrected' Blank A , mp Photolysis occurred under these conditions. The yellow ether fO.060 $0.060 0.520 0.460 249.6 extract of the product was separated from consider130 135 ,439 able insoluble material which was not examined 246.6 .309 143 ,165 ,340 243.3 ,197 further. Chromatography of the soluble material ,155 ,185 ,335 on alumina with benzene-petroleum ether (30241.9 ,180 160 ,200 241.2 ,160 ,320 60') as eluent furnished biphenylene (m.p. 112160 197 113' after sublimation) and triphenylene (m.p. ,314 ,154 240.4 ,148 , 190 198-199"), which were identified by their character,295 239.8 ,147 ,138 ,184 istic ultraviolet spectra in ethanol, in a ratio of ,138 ,276 239.1 ,119 168 ,265 approximately A : 1. 238.4 ,146 Light from a quartz capillary (190 joules in ca. 112 162 ,255 237.7 ,143 20 psec.) was passed the length of the sample tube, ,105 161 ,148 ,263 237.0 093 140 into a Hilger medium quartz spectrograph. The 234.8 ,152 ,245 032 083 spectral flash was triggered by the output of a ,123 ,155 231.3 variable delay circuit. In spectroscopic experia Difference corrected for plate response; spectrutn taken ments, small samples (0.4 to 20 pmoles) were used on Ilford Zenith plate; delay of 10.8 psec. and no precipitation was visible. Since the broad absorption maximum of the inSpectra taken 5 sec. or more after photolysis indicated biphenylene only, when compared with the termediate is so near the much sharper peak of spectrum of a gaseous sample prepared from crystal- gaseous biphenylene a t 237.5 m p , we should enterline biphenylene. A delay of 1 msec. likewise tain the possibility that the transient spectrum is revealed only biphenylene) but the intensity of the due t o excited biphenylene. At times 1 msec. or less after photolysis, the stronger long-wave(1) ?if. Stiles and K. G. Miller. THISJOCRNAL, 82, 3802 (1960). length band of biphenylene appears broad, with ( 2 ) See G. W i t t i g , Aflgew Chem., 69, 245 (1957); J. D. Roberts, Xmax a t 358.7 mp. After two minutes, this band Chemical Society (London) Special Publication No. I'l, 115 (1958); R.Huisgen and I.Sauer, Angezu. Chem., 72, 91 (1960). seemsbarrower, and Xmas shifts t o 354.3 mp. This
BENZYNE
COMMUNICATIONS TO THE EDITOR
Oct. 5, 1960
seems t o indicate that biphenylene is formed in vibrationally-excited states. Nonetheless, persistence of the absorption maximum a t about 240 mp for times up to 200 psec., together with the fact that the long-wave-length bands of biphenylene are exceedingly weak a t times when absorption in regions (1) and (2) is quite strong, indicates to US that the transient spectrum is not from excited biphenylene. Drawing an analogy with the pyridine spectrum, one can make a speculative assignment of bands consistent with the benzyne structure. Regions (1) and (2) would be assigned t o n ?r* excitation, the band being broad (rather than sharp as in pyridine) because such a transition in benzyne would be accompanied by a large change in equilibrium nuclear configuration. Absorption in the 240 mp region could be assigned t o a T + ?r* transition, like that of pyridine in the 245-270 region. Further work will be directed principally toward studying other spectral regions more closely and obtaining a detailed theoretical interpretation of the structure and spectrum of the transient intermediate. We are grateful to Mr. Roy G. Miller for preparing the benzenediazonium-2-carboxylatej and to Mr. Wayne G. Warren for his part in constructing the apparatus. This work was supported in part by a contract with the United States Air Force Office of Scientific Research.
+.
5241
0.0130M NMA
._- _. --’ lOlM NM4 IIOcmI
70
1
1400
I500
mu
1600
i’ 1700
/I 2 66 M NMA 200MH20
Po 1400
150b-I mp -~
1600
1700
R. STEPHEN BERRY DEPARTMENT OF CHEMISTRY THEUNIVERSITY OF MICHIGAN G. NEIL SPOKES Fig. l.-Near infrared absorption spectra of N-methylR. MARTINSTILES acetamide in carbon tetrachloride and in water. Water ANN ARBOR, MICHIGAX solutions contain dimethylacetamide in reference cell to RECEIVED AUGUST24, 1960 balance all absorption except that of N-H of methylacetaT H E STABILITY OF INTERPEPTIDE HYDROGEN mide. Cell thickness is marked for each solution. BONDS I N AQUEOUS SOLUTION
solutions of N-methylacetamide in the region of X-ray diffraction and infrared spectra provide 1.4-1.7 p (Fig. 1). I n this range glass vessels can direct evidence for interpeptide hydrogen bonds1 be used, and optical densities are not too intense, for substances (model peptides or proteins) in the so concentrated solutions can be examined in cells solid state where the bond strength is high. I n with precisely matched thicknesses. The absorpaqueous solution, however, where water provides tions of N-H and 0-H groups still overlap, howcompeting groups, the stabilizing effect of such ever, and hence the absorption of water must be bonds is much more uncertain. The intrinsic compensated for. This was accomplished by the stability of the N-H.O=C bond in water use of reference cuvettes which contained aqueous actually never has been measured despite the fact dimethylacetamide solutions a t concentrations that it is a fundamental quantity in any analysis carefully selected to place the same amount of of the configuration of proteins in aqueous solution. water in the light path as was present in the An attempt to estimate its free energy of formation sample cuvette containing aqueous monomethylwas made by Schellman,2 based on an analysis of acetamide. Liquid dimethylacetamide is essenliterature data for the thermodynamic behavior tially transparent in the range 1.45-1.60 p and of urea solutions, but this computation is based shows peaks very comparable to those of monoon the risky assumption that deviations from methylacetamide on both flanks of this range. In Fig. 1 are illustrated spectra in relatively ideal behavior may be attributed to the formadilute and concentrated solutions, respectively, tion of aggregated species of urea. This note provides direct experimental informa- in each of two solvents. I n carbon tetrachloride, tion on the stability in water of a model inter- N-methylacetamide exists as a monomer in dilute amide hydrogen bond. Of the several approaches solution (-0.01 M ) and as such shows a sharp In more concentried, the successful one was an examination peak a t 1.47 p (6800 of the infrared spectra3 of concentrated aqueous trated solutions aggregation sets in
Sir :
(1) G. Pimentel and A. L. McClellan, “The Hydrogen Bond,” W. H. Freeman and Co.,San Francisco. 1960. (2) J. A. Schellman, Comfit. rend Ira?. lab. Cavlsberg, Ser. Chim., 29, 223 (1955). (3) Compare J. Hermans, Jr., and H. A. Scheraga, THISJOURNAL. 82, 5156 (1960), for another application of near infrared spectra to problems of protein structure.
CHz
I c=o 1 H-N--cH~
H-N-cH~ * *
I H-N-CHa
