OPTICAL PROPERTIES OF NUCLEIC ACIDS 547 



formly mixed with the haUde salt before pressing, and its vokime concentra- 

 tion in the resulting "sample plate" can be readily controlled. The spectra 

 obtained from samples prepared in this way are free from solvent absorption 

 and scattering errors and are of quantitative value, so that this technique 

 promises to be a very real advance in infrared analytical technique. 



Nucleic acid samples can be examined as cast films, prepared by evap- 

 orating aqueous solutions on silver chloride plates.'''^ Oriented samples can 

 be obtained by shearing viscous gels or by stretching fibers ;^°' such samples 

 are usually too small to be examined directly with conventional spectrom- 

 eters and require some form of reflecting micro.scope to form an enlarged 

 image of the specimen on the entrance slit.^®*^ 



The infrared absorption spectra of the pyrimidines and purines found in 

 nucleic acids and of xanthine, hypoxanthine, and of some methylated 

 xanthines have been measured by Blout and Fields,"' '^"'^^^ mainly on 

 evaporated films. Several features in the spectra that are of value for identi- 

 fication and analysis are noted and some bond assignments proposed. 

 Brownlie^" has examined twenty-five pyrimidines, containing two or more 

 substituents, as mulls over the 2-1 5-m range and has attempted a large 

 number of possible bond assignments. Some of his conclusions are debated 

 jby Short and Thompson'^- in a study of no less than eighty pyrimidines 

 ncluding many important mono-substituted derivatives, over the range 

 2-25 M- These workers used lithium fluoride, sodium chloride, and potas- 

 sium bromide prisms to cover this wide range of wavelengths, and treated 

 some of their compounds with deuterium oxide to assist the recognition of 

 vibrations involving hydrogen atoms. Their discussion of this large volume 

 of data is an excellent example of the partly empirical, partly theoretical 

 approach that has to be employed in the interpretation of such complex 

 spectra. All the spectra are highly characteristic and contain many strong 

 sharp bands suitable for identification and analysis, even when precise 

 structural correlations are difficult. Interpretations based on previously 

 established group correlations favor a ketonic structure for the 2-hydroxy- 

 and 4-hydroxypyrimidines and probably a diketonic structure for the 

 2 , 4-dihydroxy derivatives. Amino groups appear to exist in the non-tauto- 

 merized form. Both the tautomerism and the hydrogen-bonding in the 

 solid state are influenced by the various substituents present in the com- 

 pounds studied. Short and Thompson conclude that frequency assignments 

 in pyrimidine spectra must be made with considerable reserve, in view of 

 the meager X-ray structure data that can be utilized to provide confirma- 

 tion. Even so, their tentative findings are obviously of great interest in 



»«» R. D. B. Fraser, Discussions Faradaxj Sor. No. 9, 378 (1950). 



i6» I. A. Brownlie, J. Chem. Soc. 1950, 3062. 



'»2 L. N. Short and H. W. Thompson, J. Chem. Soc. 1952, 168. 



