ANNUAL REPORT, 1946-47 25 



The Investigation of Agricultural Waste Products. 1. The Chemical Investi- 

 gation of Lignin. (Emmett Bennett.) Continued attempts have been made to 

 increase the nitrogen content of lignin. Oxidations with nitric acid were not 

 particularly fruitful and have been discontinued. Attempts also have been made 

 to unite urea and lignin. With the means at our disposal, the reaction was not 

 successful. 



The most promising procedure tried was the ammonification and simultaneous 

 oxidation of lignin in a medium of concentrated ammonium hydroxide and oxygen 

 at atmospheric pressure. This reaction was improved by previously subjecting 

 the lignin to a mild hydrolysis with citric acid. Chlorinated lignin ammoniated 

 in the above manner yielded the highest percentage of total nitrogen. The max- 

 imum content of total nitrogen obtained was about 3.9 percent. 



Studies on the Quantitative Estimation of Hemicelluloses. (Emmett Bennett.) 

 Procedures relating to the development of a method for the quantitative estima- 

 tion of polyuronide hemicelluloses have been continued. The technique adopted 

 is based on the action of sodium chlorite on plant tissue. After certain preliminarj^ 

 extractions, this reagent removes the lignin to almost any desired extent, leaving 

 behind the hemicelluloses. This group of substances may in turn be removed by 

 extraction with relatively dilute solutions of strong alkali. The extracts are then 

 oxidized in a solution of chromate in sulfuric acid. The reduced chromate is 

 measured spectrophotometrically. By referring to a curve which relates reduced 

 chromate to anhydropentose, the content of hemicellulose as anhj'dropentose 

 may be ascertained. 



In the case of cornstalks, extractions were found to be most efficient when the 

 tissue was mercerized with 14 percent sodium hydroxide at 20° C. for ten niinutes, 

 then an equal volume of water added and allowed to stand for a total of 90 

 minutes. Approximately 85 percent of the furfuraldehyde-j'ielding substances 

 may be removed in this way. It is also quite probable that such a partition 

 separates the polyuronide from the cellulosan hemicellulose. It is felt that more 

 drastic treatment is undesirable since some pentose units might be removed 

 from the cellulose chains. It seems naore desirable to consider such units as 

 constituents of cellulose. 



Pectic compounds and ether-soluble substances should be removed prior to 

 treatment with sodium chlorite. Starch, to some extent at least, is solubilized 

 by the action of this reagent. Protein is not removed completely during the 

 production of holocellulose, but Is extracted almost completely bj' the alkaline 

 medium; hence corrections must be applied in obtaining the final figure for 

 hemicelluloses. 



The procedure has been applied to six different plant tissues with a hemicellu- 

 lose content ranging from about 7 to 25 percent. Consistent results can usually 

 be obtained and variations between samples usually do not exceed 5 percent. 

 It is believed that the simplicity, rapidity, and consistency of the procedure will 

 make It of value. 



The Chemical Investigations of Hemicelluloses. (Emmett Bennett.) In- 

 vestigations dealing with the chemistry of hemicelluloses have been continued. 

 Special attention has been given the hemicelluloses of cornstalks. This fraction 

 was obtained from holocellulose prepared according to a previous report (Jour. 

 Ind. Eng. Chem. Anal. Ed. 19, 215, 1947). To date solubility studies have been 

 made and some of the fractions characterized by certain determinations. All 

 solvents used extracted some hemicellulose, but the bulk was freed by the action 

 of 4 percent sodium hydroxide at room temperature. All fractions tested were 

 optically active and ranged from about —60'' to —85°. The more negative ro- 

 tations were associated with the higher content of pentosans. Methoxyl and 



