546 G. H. BEAVEN, E. R. HOLIDAY, AND E. A. JOHNSON 



of infrared absorption spectra on an essentially empirical basis and no 

 fundamental interpretation of the spectrum is really required. The other 

 application is for structural studies, and in this case very detailed analysis 

 of the spectrum is necessary. This may take the form of the correlation of 

 band frequencies with particular structural elements of the sample, i.e., 

 the use of "group correlations," for which there is an extensive empirical 

 and theoretical literature (see Fraser^*" for references). For more funda- 

 mental indications of structure, e.g., the detection of hydrogen-bonding 

 in polynucleotide chains and other detailed aspects of macromolecular 

 structure, reliable assignments of bands to known modes of vibration are 

 required. The progress which has been made in this direction has also been 

 critically discussed by Fraser^*" for biological macromolecules, and by 

 Thompson et aZ.^" and Sutherland^ ^'^ for large molecules in general. 



Problems of sample preparation and of solvent selection are prominent 

 in the infrared spectroscopy of nucleic acids and the related simple com- 

 pounds. Because of its heavy absorption water is virtually useless as a 

 solvent, quite apart from its solvent power for the rock-salt windows nor- 

 mally used in absorption cells. The available solvents with useful trans- 

 mission windows in the infrared^*^ are of little value; thus Lacher and co- 

 workers'^^ went to the trouble of using antimony trichloride at 100° C. to 

 obtain absorption data in the 1-2. 5-m overtone region on some pyrimidines. 

 Pyrimidines and purines have therefore mainly been examined as mulls, 

 in which a finely powdered solid is mixed with a suitable nonvolatile liquid. 

 The spectrum of a mulled sample will be partly obscured by any absorption 

 bands of the mulling liquid, and the observed band intensities are subject 

 to uncertainties arising from scattering losses and the difficulties in con- 

 trolling sample concentration and film thickness. The fact that the spec- 

 trum refers to the solid state must also be borne in mind when theoretical 

 interpretation is attempted. Many pyrimidines and purines are sufficiently 

 stable to be sublimed in vacuo, however, and thin-film samples prepared 

 in this way have been used by many workers for both infrared''*^ •'"■'** and 

 ultraviolet^^ ■^'* absorption studies. A new method'^* for preparing solid 

 samples for infrared absorption studies makes use of the fact that powdered 

 potassium bromide can be converted under moderate pressure into clear 

 plates with good infrared transmission. The finely ground sample is uni- 



'" H. W. Thompson, D. Nicholson, L. N. Short, Discn^sions Faraday Soc. No. 9, 222 



(1950). 

 '" G. B. B. M. Sutherland, Discussions Faraday Soc. No. 9, 274 (1950). 

 1" P. Torkington and H. W. Thompson, Trans. Faraday Soc. 41, 184 (1945). 

 i6« J. R. Lacher, D. E. Campion, and J. D. Park, Science 110, 300 (1949). 

 1" E. R. Blout and M. Fields, J. Biol. Chem. 178, 335 (1949). 



168 E. R. Blout and M. Fields, /. Am. Chem. Soc. 72, 479 (1950). 



169 M. M. Stimson, J. Am. Chem. Soc. 74, 1805 (1952); U. Schiedt, Z. Naturforsch 76, 

 270 (1952) ; Appl. Spectroscopy 7, 75 (1953). 



