SEPARATION BY PAPER CHROMATOGRAPHY 251 



tures remain the most valuable. In Table I are listed the reported Rf values 

 of some purine and pyrimidine bases and nucleosides in a number of solvent 

 systems selected to include (a) those which, in the writer's estimation, have 

 proved the most broadly useful, and (b) representatives of different types 

 of mixture which have been tried with some degree of success. It is ob- 

 vious that for some purposes other solvents will be preferable to those 

 listed. For completeness, the Rp values reported by the original authors 

 have been supplemented with some determined by the writer. 



Aqueous n-butanol, with and without added ammonia, was one of the 

 first solvents tested for separation of nucleic acid derivatives^ • ^^ and remains 

 among the most useful (Table I, solvents a, h, and c). Butanol saturated 

 with water has usually been used; however, the solution is most conveni- 

 ently made up at a fixed percentage composition slightly under-saturated, 

 thus making its composition independent of tempera ture.^i Since Rp values 

 are rather low, it is advantageous to use the descending method and to cut 

 the end of the paper to a number of teeth, from which the solvent is allowed 

 to drip.2^ Ammonia may be added either to the solvent irrigating the paper 

 or to that in the bottom of the tank; owing to the volatility of ammonia, 

 the latter practice affords the more constant conditions. It has the effect of 

 slowing the movement of substances with acidic substituents, and the 

 results given by two different concentrations are illustrated in Table I 

 (solvents h and c). With butanol-water-ammonia, all of the purine and pyrim- 

 idine bases known to occur in nucleic acids (except 5-hydroxymethylcy- 

 tosine, which runs close to guanine) may be resolved from one another. In 

 addition to the authors cited in the Table, Chargaff, el al.^'^- ^^^ Marshak 

 and Vogel,*' and others have used butanol-ammonia mixtures for quanti- 

 tative separation of nucleic acid components. 



Addition of formic acid to aqueous butanol (solvent d) results in more rapid move- 

 ment of acidic substances such as uracil, thymine, xanthine, and hypoxanthine. 



If butanol is saturated with a saturated solution of boric acid instead of with 

 water, ribosides, by virtue of their cj's-diol configuration, form borate complexes and 

 do not move. Complete separation of free bases from ribosides may thus be obtained.*^ 



Admixture of various other substances with butanol-water systems has been tried, 

 generally without much advantage. Butanol may be saturated with a 10% solution 

 of urea, instead of water, with similar results. ^^ Mixing morpholine or diethylene 



^' n-Butanol satd. with water at 20° contains 84% by vol. of butanol, calcd. from 



data given by A. Seidell, "Solubilities of Organic Compounds," Vol. 2, p. 266. 



Van Nostrand Co., New York, 1941. 

 " E. Chargaff, R. Lipshitz, C. Green, and M. E. Hodes, /. Biol. Chem. 192, 223 



(1951). 

 "» C. Tamm, H. S. Shapiro, R. Lipshitz, and E. Chargaff, J. Biol. Chcyn. 203, 673 



(1953). 

 " A. Marshak and H. J. Vogel, J. Biol. Chem. 189, 597 (1951). 

 " I. A. Rose and B. S. Schweigert, J. Ain. Chem. Soc. 73, 5903 (1951). 



