weight unit to obtain the weight of carbon dioxide. Similarly, us? a long capillary to pass chlorine gas from a cylinder through the molecular weight flask (Table I), Table I also lists the titration of chlorine in a 15-ml. round-bottomed flask with both parts of a 10/30 standard-taPer ground joint and also a 2-mm. straight-bore stopcock. Displace the air with chlorine, shake with excess
potassium iodide solution, and then titrate the liberated iodine with standard sodium thiosulfate solution. LITERATURE CITED
(1) Anderson, H. H., ~ A L CHEni. . 20, 1241 (1948); 24, 5 i 9 (1952). (2) Anderson, H. H., J . Am. Chem. SOC. 67, 223 (1945). (3) Ibid., 71, 1801 (1949).
(4) Ibid., 72, 2089 (1950).
( 5 ) Anderson, H. H., Fischer, H , J . Org.
Chem. 19, 1296 (1954). (6) Johnson, E. W.,Xash, L. K., J . Am. Chem. SOC.72, 547 (1950). (7) Kraus, C. A., Flood, E. A, Ibid., 54, 1635 (1932).
RECEIVED for reviex July 30, 1956. A4ccepted January 30, 1957. First Delaware Valley Regional Meeting, ACS, Philadelphia, Pa., February 16, 1956.
Colorimetric Estimation of Milligram Quantities of Inorganic Azides CHARLES E. ROBERSON' and CALVIN M. AUSTIN Quality Evaluation laboratory, U. S. Naval Ammunition Depof, Crane, lnd. ,The azide is acidified and the resulting hydrazoic acid is distilled into an acidified ferric nitrate solution. The reddish brown color due to formation of ferric azide is suitable for quantitative measurement at 460 mp. The color obeys Beer's law over the range of concentration investigated-i.e., 0.5 to 4 mg. of azide ion per 50 ml. of colored solution. The method is applicable to the determination of lead azide in primer mixtures, except those containing thiocyanates. It is not recommended for purity determinations.
It mas felt that the method would be particularly useful for determination of lead azide in the presence of other materials and when only small samples are available for analysis, as in some primer mixtures. Synthetic primer mixtures were prepared and analyzed with favorable results. The method was also employed successfully in this laboratory to investigate the rate and extent of formation of copper azide on small strips of copper foil, which had been exposed to hydrazoic acid vapor ( 5 ) .
C
Very simple distillation equipment is employed. A 10-ml. micro round-bottomed flask with a neck about 4 cm. long is fitted with a one-hole rubber stopper. The latter supports a glass tube drawn out until a very small orifice (approximately 0.4 mm. in inside diameter) is obtained on the receiver end. The tube is bent to extend nearly to the bottom of a 50-ml. volumetric flask. Absorption measurements are made with a Beckman Model DU spectrophotometer using 1-cm. Corex cells.
APPARATUS
methods for the quantitative estimation of inorganic azides have usually been restricted to determining purity of lead azide. The azide is oxidized by a cerium(1V) salt with subsequent measurement of the volume of nitrogen gas evolved or by titration of a n excess of a measured volume of a standard cerium(1V) salt solution with ferrous perchlorate ( I , 3). Because these methods are not wholly satisfactory for the determination of milligram quantities of azides, either alone or in the presence of other substances, the method described herein was developed. The formation of red ferric azide when hydrazoic acid reacts with ferric chloride, is the basis of the qualitative test described by Feigl ( 2 ) . An attempt to apply this test to quantitative measurement was unsuccessful because of the instability of the ferric chloride. However, use of ferric nitrate proved satisfactory in connection with a simple distillation of the hydrazoic acid formed when a n azide solution is acidified. OXVENTIONAL
Present address, Water Resources Division, U. S. Geological Survey, Menlo Park, Calif. 854
ANALYTICAL CHEMISTRY
REAGENTS
All reagents were reagent grade chemicals. An acidified ferric nitrate solution is prepared by dissolving 2.000 grams of ferric nitrate in about 50 ml. of water, and adding 5 ml. of concentrated nitric acid which is relatively free of nitrogen dioxide as determined by observation. This is filtered and the filtrate is diluted to 1 liter. A standard azide solution is prepared from sodium azide (Fisher Scientific Co., Cat. S-227). PREPARATION OF STANDARD CURVE
Sodium azide solution (1 ml. equal to 1 mg. of azide ion) is used in preparing the standard samples containing u p to 4.0 mg. of azide ion. The standard solu-
tion placed in the distilling flask is diluted to about 5 ml. with distilled water. Two or three glass beads are used to help prevent bumping. Twenty milliliters of the ferric nitrate solution are pipetted into a 50-ml. volumetric flask and the stopper assembly is made ready for immediate connection after the addition of 0.6 ml. of 1 to 4 sulfuric acid to the contents of the distilling flask. The orifice of the glass tube is placed well below the surface of the ferric nitrate solution in the receiver. The receiver is cooled in a n ice bath and the distillation is continued for 3.5 minutes after first appearance of the reddish brown ferric azide color. The solution is adjusted in the receiver to approximately 25" C., diluted to the mark with n-ater, and again adjusted to 25" i 1 " C. The transmittance is immediately measured a t 460 mp against a reference solution made by diluting 20 ml. of the ferric nitrate solution to 50 ml. PROCEDURE FOR PRIMER MIXTURES
Place a sample containing 0.5 to 4.0 mg. of NBin the distilling flask, add glass beads and distilled mater until the flask is about half filled. Add 2 drops of a 27, solution of sodium hydroxide and 0.3 ml. of a 30% hydrogen peroxide solution, and then boil the mixture gently for 1 minute to remove the excess hydrogen peroxide. Cool the flask and contents to 25' C. or below and proceed as indicated under preparation of standard curve. DISCUSSION A N D RESULTS
The absorption maximum is broad' and measurement is made at 460 mp. The intensity of the color varies with pH, but this variable is satisfactorily controlled by the distillation technique so long as no acid distills over. For this reason sulfuric acid is used to displace the azide ion. The color is stable for a t least an hour if the volumetric receiver is full and stoppered to prevent,
Table I. Determination of Azide Ion in Synthetic Primer Mixtures
Azide Ion Present, Mg. 0.50
Azide (Ns) Found, Mg. 0.46 1.04 1.01 2.01 3.40 3.49 3.88
1.00
1.00 2.00 3.50 3.60 4.00
Mean Standard deviation
Relative Error,
%
-8.0 +4.0
Nj,-+NOz-+2Hf-tH20
i-1.0 +o 5
The use of ferric sulfate in sulfuric acid to eliminate this difficulty was unsatisfactory, as color development with a sulfate reagent was poor. Azide ion gave, in equivalent concentration, a 57% T in a nitric acid reagent and 86% T in the sulfuric acid media. Samples of synthetic primer mixtures were prepared, using 10 mg. of antimony trisulfide (stibnite), 10 mg. of potassium chlorate (both approximate weights), and accurately measured amounts of sodium azide solution of known concentration. Table I indicates results for determinations on such mixtures. Table I1 shows azide ion values found from analyses of military grade lead azide, the purity of which was 92.05y0 by the Navy titration method ( 3 ) .
-2.9 -3.1 -3.0 -1.6 3.9
Table II. Determinations of N3in Lead Azide, Purity 92.05%
Azide Azide (NI) (No) Calcd., Found, Mg. 2.64 2.38 1.81 2.84 3.16 1.49 1.80 2.77 1.68 3.68
Mg. 2.66 2.42 1.74 2.67 3.20 1.50 1.85 2.64 1.79 3.55
Purity Found,
%
92.7 93.5 88.7 86.6 93.3 92.5 94.6 87.9 98.0 88.7 Mean 9 1 . 6 Standard deviation 3 . 6
to shift the standard curve slightly and it is advisable to prepare a ne\T standard curve when new ferric nitrate solution is prepared. The necessity for preparing a new curve with each batch of reagent is found in the reaction.
Relative Error, % +0.8 +1.7 -3.9 -6.0 +1.3 $0.7 +2.8 -4.7 +6.5 -3.5 -0.4 3.9
‘undue exposure to air. Ricca (4) has shown t h a t aeration causes appreciable fading of the color and suggests that this is due to the loss of hydrazoic acid. He has also presented evidence that the color is due not to undissociated ferric azide but to the complex ion Fe(NS)++. Interference from sulfites, thiosulfates, and sulfides is avoided by preliminary oxidation of the alkaline sample with hydrogen peroxide ( 2 ) . Cyanates and thiocyanates, however, interfere and render the method ineffective. Nitric acid batch variations may tend
+r\‘>+N20
While the relative errors shown are not considered excessive with small amounts of azide, they are too large to permit the use of the method for purity determinations. This is clearly illustrated by the purities calculated from the results and also shotvn in Table 11. LITERATURE CITED
( I ) Davis, T. L., “Chemistry of Powder and Explosives,” p. 430, Wiley, Xe-x York, 1943. (2) Feigl, F., “Spot Tests, Inorganic iipplications,” 4th ed., Vol. 1, p. 268, Elsevier, New York, 1954. (3) Military Specification, MIL-L-3055, Sept. 30, 1949. (4) Ricca, B., Gazz. chim. ilal. 75, 71 (1945).
( 5 ) U. S. Naval Ammunition Depot, Crane, Confidential Rept. QE/C 56-40.
RECEIVED for review August 27, 1956. Accepted December 13, 1956. The opinions and assertions contained in this article are the private ones of the authors and are not to be construed a8 reflecting the views of the Navy Department.
CORRESPONDENCE Determination of Potassium as the Metaperiodate SIR: After publication of “Determination of Potassium as the Metaperiodate” [ANAL. CHEM. 28, 2011-5 (1956)l our attention mas called to “Determination of Potassium in Soap and Mixed Caustic Lye” by W. T. Miller and J. T. R. Andrews [J. Am. Oil Chemists’ SOC.26, 309-12 (1949)]. These authors report increased accuracy and sensitivity of the Willard and Boyle periodate procedure for potas-
sium through (1) hand stirring during precipitation of the potassium periodate, (2) totally reducing the periodate to iodine, and (3) titrating the iodine ivith sodium thiosulfate solution. Because our own paper deals in part with these points, we wish to call this paper to the attention of those interested in this procedure. RALPHE. JENTOFT REX J. ROBINSON
154.
Yttrium Trifluoride, YF,,
155.
Samarium Trifluoride, SmF,, Orthorhombic Form
Orthorhombic Form
156. Ytterbium Trifluoride, Y bF,, Orthorhombic Form EUGENE STARITZKY and L.
B. ASPREY, The
HE trifluorides of yttrium, samaT r i u m , and ytterbium were precipitated with hydrofluoric acid from aqueous solutions of corresponding chlorides.
University of California, Los Alamos Scientific Laboratory, Los Alamos, N. M.
The precipitates were oven-dried a t 110” C., dried under vacuum a t 1000” C.. and then heated under argon to about 100” C. above their melting point.
The yttrium oxide used as starting material for this preparation is believed to be about 99% pure. The samarium was purified by E. I. Onstott of this VOL. 2 9 , NO. 5 , M A Y 1957
855
