Improved Direct Nesslerization MicroKjeldahl Method for Nitrogen Determination in Organic Compounds and Biological Fluids EDWARD S. WESTAND ARTHURL. BRANDON, Washington University School of Medicine, St. Louis, Mo.
AURO (5) has recently described the use of selenium
as a catalyst in macro-Kjeldahl digestions. He found that it is much superior to either copper or mercury. These results suggested to the writers the possibility of improving the micro-Kjeldahl method as it is applied in determining the nitrogen of biological fluids and of compounds. The phosphoric-sulfuric acid digestion of Folin and Wu (2) has the disadvantage of severely attacking the digestion tubes and materials used to prevent bumping, leading t o breakage and some difficulty in obtaining clear nesslerized solutions. The hydrogen peroxide-sulfuric acid digestion of Koch and McMeekin (4) is rather free from these objections, but there is the danger of spattering and loss of sample upon addition of peroxide to the hot digestion mixture, and there may be appreciable nitrogen in the peroxide which must be determined and corrected for. Also, the necessity of interrupting the digestion and adding the peroxide is an inconvenience. It has been found that blood filtrates and urines are rapidly and completely digested by 1 to 1 sulfuric acid containing approximately 0.2 per cent selenium, and the nitrogen values are similar to those found by the peroxide method of Koch and McMeekin. The selenium does not interfere with nesslerization. By using a digestion mixture of 1 to 1 sulfuric acid saturated with potassium sulfate and 0.2 per cent selenium, the rate of digestion is increased, and such a mixture is preferable in the analysis of substances difficult to oxidize, although quite unnecessary for blood filtrates and urine.
cc. level. To this 1 cc. of the digestion mixture and 2 dry glass beads (2 are better than 1) are added, and the tube covered with a glass bulb or light short-stemmed funnel with the stem sealed. The solution is rapidly evaporated over a microburner until all water is driven off, the mixture chars, and sulfate fumes appear. The flame is then reduced, the mixture gently boiled until it becomes clear and colorless (usually 1 to 2 minutes), and the boiling continued a further 4 minutes. The digest is cooled, diluted with 25 cc. of water, and 15 cc. of Nessler's reagent are added quickly.2 The tube is diluted to the mark, stoppered with a rubber stopper, and mixed. A standard is prepared using 1 cc. of the digestion mixture and 0.2 to 0.3 mg. of nitrogen, nesslerized and diluted to 50 cc. Colorimetric readings and calculations are made in the usual way. RESULTS Table I records data obtained on tungstate filtrates of blood and diluted urines using 1 to 1 HzS04 3 per cent HzOz, 1to 1Hi304 30 per cent HzOz, and 1to 1HzS04 0.2 per cent Se as the digestion mixtures. The values are identical within the limits of error of the methods. I n running duplicates, the selenium method seems to give somewhat better checks than the peroxide method, possibly because of loss by spattering with the latter procedure.
Blood filtrates: Sheep blood Pig blood Urines: Sample 1 Sample 2 Sample 3 Sample 4
REAGENTS.Selenium digestion mixtures :
1 Tubes constricted to about one-third diameter a t the graduation mark were used. This enables more accurate dilution and also tends to minimize the loss of sulfate.
(All determinations in duplicate) 1 TO 1 &SO4 1 TO 1 HzSOa 1 TO 1 H r 8 0 r 3% HrOz 30% HzOr 0.2% Se Mg. N/100 cc. MQ. N/100 cc. Mg. N/100 cc
1. Without potassium sulfate. One cubic centimeter of selenium oxychloride (Eimer & Amend) is added to 500 cc. of 1 to 1 sulfuric acid and mixed. This mixture is recommended for the determination of blood N. P. N., the nitrogen of urine, and other biological fluids. The solution should be kept stop-
Nessler reagent: This was prepared according to Bock and Benedict (3) and then diluted with an equal volume of 10 per cent sodium hydroxide. This formula has been found quite satisfactory. Hydrogen peroxide : Three per cent hydrogen peroxide from Parke Davis, and Merck's Superoxol. This was found to be practically nitrogen-free. METHOD.A quantity of solution representing approximately 0.2 to 0.3 nig. of nitrogen (5 cc. of Folin-Wu blood filtrate or 1 cc. of urine diluted 1 to 40) is pipetted into a Pyrex digestion tube' (25 by 200 mm.) calibrated a t the 50-
FILTRATES OF BLOODAND TABLEI. DATAON TUNGSTATE DILUTEDURINES
pered when not in use to prevent taking up water and precipitation of selenium. 2. With potassium sulfate. One volume of sulfuric acid is mixed with an equal volume of saturated potassium sulfate solution, and to 500 cc. of this mixture 1 cc. of selenium oxychloride is added. A small amount of potassium sulfate separates overnight. The solution is decanted from this.
50.5 37.7 1240 1120 1120 693
50.5 37.0 1220 1200 1120 690
50.5 37.0 1240 1140
TABLE11. NITROGEN VALUES (All determinations in duplicate) NITROQENFOUXD IN!
45.9 33.3 17.0 9.45 7.76 21.70
Urea Uric ,acid Alanine Acetanilide Phenacetin Creatine
46.3 33.3 17.1 10.47 7.30 20.90
45.9 33.3 17.0 10.47 7.70 23.60
46.66 33.33 15.73 Unpurified 10.36 7.82 28.17 29.20% by macroKieldahl usine: 0.2 gram of Se 0.5 gram of sucrose 28.17 2 mg. sucrose in each digestion 37.13 2 mg. sucrose in each digestion 11.66 2 mg. sucrose in each digestion
Table I1 gives nitrogen values on several compounds. Inasmuch as only comparative results were desired, the compounds used, with one or two exceptions, were commercial samples and not specially purified. For these 2 Nessler's reagent must be added quickly after water t o prevent separation of elementary selenium with resulting cloudiness.
INDUSTRIAL AND ENGINEERING CHEMISTRY
July 15, 1932
determinations the KzS04-HzS04-Se and KzSOp-HzS04-peroxide digestion mixtures were used for comparison. Results with the 1 to 1 Se mixture are quite comparable on these compounds to those obtained by using the KzS04HzS04-Se mixture, as are also the 1 to 1 HpSO4-peroxide mixtures, although the digestion may require a few minutes longer. The nitrogen of creatine and creatinine is not given completely by any of the above digestions. However, recovery 2 mg. of is attained by using the K2S04-HzS04-Semixture sucrose. Here the presence of potassium sulfate, by raising the boiling temperature, markedly speeds up the rate of digestion; in fact, without i t and using sucrose, not all of the nitrogen of creatine or creatinine was recovered. Neither 3 nor 30 per cent peroxide combined with sulfuric acid and potassium sulfate sucrose gave all of the nitrogen of either creatine or creatinine. It seems, therefore, that the selenium method combined with sucrose is a little more effective on resistant compounds than the peroxide method. It should be noted that the addition of sugar with KZSO4-HzS04-Se gives the same nitrogen values for urine as HzS04-Se. The creatinine is completely determined by the ordinary methods when admixed with the other substances of urine. One-tenth per cent selenium gives rather satisfactory digestion but not quite as good as 0.2 per cent. Four-
tenths per cent is very little better than 0.2 per cent. The selenium digestion gives correct values in the presence of chlorides. It was noted that there is much less tendency toward bumping in the selenium digestions than the others, probably because of the separation of elementary selenium (which redissolves as the acid becomes concentrated). This work confirms that of Davenport ( I ) that 3 per cent peroxide may satisfactorily, and preferably from the standpoint of handling and cost, be substituted for the Superoxol recommended by Koch and McMeekin. From the experience of the authors, the selenium method seems to be essentially free from the annoyances of the other methods and gives a little more consistent results than the peroxide method. The cost of the catalyst is insignificant, about 4.5 cents for 500 cc. of digestion mixture.
LITERATURE CITED (1) Davenport, H. A,, S. Lab. CZin. Med., 12, 286 (1926). (2) Folin, O., and Wu, H., J. Bid. Chem., 38, 81 (1919). (3) Hawk and Bergeim, “Practical Physiological Chemistry,” p. 893, 10th ed., 1931. (4) Koch, F. C., and McMeekin, T. L., J. Am. Chem. Soc., 56, 2066 (1924). (5) Lauro, M. F., IXD. ENG.C H ~ MAnal. . , Ed., 3, 401 (1931).
RIJCEIVED February 16, 1932.
The Photovoltaic Cell as Indicator in Precise Titrornetry HERMANM. PARTRIDGE, Washington Square College, New York University, New York, N. Y. cell was most sensitive in the N PREVIOUS communicaBecame of its inherent stability and simplicity Yellow-green Portion of the spections @-J$), various methof operating requirements, the photovoltaic cell is trum. It was also noted that ods employing the photoadmirably suited to precise titrometry, especially response was obtained in electric cell for following titrawhere the use O f small samples is necessary* I n tions have been described or the near infra-red and ultramentioned. B e c a u s e of t h e the method herein described, precision of one part v i o l e t . Later publication of t e c h n i c a l data by the manufeeble photoelectric currents obper thousand is obtained with three titrations. facturers giving relative spectral tained under most conditions, Under the most favorable conditions only jfive sensitivity bears out these conamplification h a s b e e n necesreadings for each determination are necessary, clusions, but unfortunately the sary. In order to o b t a i n ingas-filled and the end point is obtained by extrapolation. makers do not state whether or creased not these values are based on cells were usually employed and possibilities of instability and electrical leakage were some- equal-energy considerations or are dependent on the intensity times troublesome to obviate. Obviously a method which distribution of a particular type of source. To gain further knowledge of the electrical characteristics of would retain most of the advantages and a minimum of the disadvantages of former methods would be highly desirable. the cell under convenient working conditions, a 6-volt, 21The dry types of so-called “photronic” cells which have been candle power, concentrated-filament, nitrogen-filled lamp was commercially obtainable for a few months, because of their mounted on an optical bench, and the cell was mounted so that favorable characteristics, have been chosen for use in the it might be moved with respect to the source. The data given in Table I were obtained as follows: presept method. Although these cells are said to be “electronic” in action, D = distance between cell and source, cm. their output is dependent upon the photovoltaic e. m. f. deE = photovoltaic e. m. f., millivolts, measured by conventional Hildebrand potentiometric method. veloped ( I ) . For this reason, the author prefers to call them I = calculated current, microamperes, which would flow “photovoltaic” cells. with cell short-circuited. R = corrected resistance of cell obtained by deducting reCELL CHARACTERISTICS sistance of measuring instruments from total circuit resistance. A few simple experiments conducted with one of these W = calculated power, millimicrowatts, dissipated by cell cells showed that in one respect, a t least, it was far superior when short-circuited, to the photoelectric cell-namely, stability. Also, the fact I n order to understand better the relations between e. m. f., that its resistance under working conditions is comparable with that of medium-sensitivity portable galvanometers current, and power, respectively, with light intensity, their makes possible simple yet sensitive electrical arrangement. logarithmic values were plotted as shown in Figure 1. At the Simple experiments with filters showed that, spectrally, the values of distance from the source corresponding to the lower