348 PER FLODIN AND RARE ASPBERG 



this approach was chosen. A sieve fraction passing through 200 mesh 

 proved satisfactory. 



As seen from Fig. i the volume between the peaks is only between 50 

 and 100 ml. compared with a total bed volume of 2000 ml. Thus, it was 

 necessary to pack the column very carefully, but once packed it could be 

 used for a long time. For example, in three experiments made with the 

 same cellodextrin material the elution patterns were practically identical. 

 There was a slight variation in the positions of the maxima but the form 

 of the peaks and the resolution was the same. As seen in Fig. i the zones 

 overlapped slightly indicating that a complete separation was not obtained. 

 The amount of contamination in the peaks was so small that it was elimina- 

 ted in the crystallizations. This was evidently not the case for the cello- 

 hexaose which showed inhomogeneity by some of the criteria used. 



The volume of the sample must not be larger than l\{Kj) — -/v^,) 

 where F, is the volume of water inside the gel grains and Kj^ and Kj) are 

 the distribution coefficients of the solutes to be separated. For the oligosac- 

 charides in the column used the sample volume must be less than 80 ml. 

 and in practice considerably less. On the other hand the process itself is 

 insensitive to the solute concentration if the viscosity is low. Thus, an 

 optimal amount is separated if the concentration is high and the volume 

 about one-fourth of the calculated value. To isolate a large amount of one 

 of the oligosaccharides a two-step procedure may be preferable in which a 

 preliminary rapid separation is made in a short column packed with, say, 

 the 50-100 mesh sieve fraction. A further fractionation is then made in a 

 high-efficiency column. 



The outlined procedure is in principle applicable to any separation 

 within a series of homologues in which the molecular size difference is at 

 least as large as for the oligosaccharides. With ionized solutes, however, 

 it is necessary to take the effect of the charges into consideration. This 

 means that separations often have to be made in the presence of strong 

 electrolytes. High-efficiency separations may also be made in higher 

 molecular weight ranges. Dextrans on molecular weights up to about 

 40 000 have been fractionated with excellent results in dextran gels with 

 different degrees of cross-linking. 



References 



1. Dickey, E. E., and Wolfrom, M. h.,jf. Aniey. c/ietn. Soc. 71, 825 (1949). 



2. Whittaker, D. R., Arch. Biochefti. Biophys. 53, 439 (1954). 



3. Lathe, G. H., and Ruthven, C. R. J., Biochem. J. 62, 665 (1956). 



4. Porath, J., and Flodin, P., Nature, Loud. 183, 1657 (1959). 



5. Flodin, P., and Granath, K., /;/ "Symposium iiber Makromolekyle", Wies- 

 baden, Oct. 1959. 



6. Flodin, P.,X Chromatography 5, 103 (1961). 



7. Peat, S., Whelan, W. J., and Roberts, J. G.,y. chem. Soc. 2258 (1956). 



