INDUSTRIAL AND ENGINEERING CHEMISTRY
58
TABLEIV. DUPLICATE FLUORINE DETERMINATIONS [Mg(CzHaOz)z,4HzO as ashing agent] Food Sample Fluorine Found So. -4liauant I Aliauant I1 Average Variation
1 2 3
4 5 6 7 S
9 IO 11 12 13 14 15 16 17
IS 15
20 21 22
0.129 0.084 0.156 0.139 0.136 0.164 0.068 0.107 0.050 0.139 0.075 0,056 0.087 0.075 0.092 0.066 0.048
0.082 0.117 0.116 0.092
0.083 Av 0 . 0 9 7
0.112 0.064 0.142 0.132 0.140 0.174 0,067 0.107 0.04i 0.139 o.09n 0.053 0.084 0.073 0.106 0.070 0,047 0,076 0.094 0.120 0.100 0.082 0 097
0,120 0.074 0.149 0.133 0.138 0.169 0 . ntji 0.10i 0,048 0.138 0 . 082 0.054 0.08n 0 . Oil 0 . OO!) 0 ,068 0.04T n OGY 0.10ci 0.122 0 096 0.082 0 097
0.017 0.020 0.014 0 .O O i 0.004
0.010 0.001 0,000 0.003 0.000 0.015 0.003 0,003 0 . on2 0.014 0.004 0.001 0.014 0.023
n.oix
0.OOh 0.001 0 008
14.1 36.9 5.4 5.1 2.5 5.9 1.4 0.0 6.3 0.0 18.3 5.6 3 . ;r 2.7 14.1 5.9 2.1 20.0 21.3 10.6 8.3 1.2
Method A duplicate of the food consumed in each 24-hour period is collected in glass-topped, quart-size, fruit jars. The composite sample is ground in a household electric mixer with grinding attachment, the met weight is determined, and the ashing agent is added. If magnesium peroxide is employed, 1.0 gram per 100 grams of wet food is added; if magnesium acetate (60 per cent aqueous solution), 5 ml. per 100 grams. Mixing is continued for 0.5 to 1.5 hours, depending upon the size and character of the sample. Two 150-gram aliquants are weighed into 200-ml. nickel evaporating dishes, dried on an electric burner, and then placed in the automatically controlled electric muffle furnace, set at 570’ C. After complete combustion, the ash is weighed and prepared for distillation. The fluoride is separated from the ash by distillation with perchloric acid according to MacIntire and Hammond’s ( 4 ) modification of the method of Willard and Winter (6). Silver sulfate, as recommended by McClure (3),is added to the ash to prevent volatilization of the fluorides. The back-titration method of Dahle, Bonnar, and Wichman (1) is employed from this point on; the unconcentrated distillate is used.
3.2
oxide as the ashing agent (Table 111),and the other likewise with magnesium acetate (Table 117). The results might give the unwarranted inference that determinations made with magnesium peroxide gave higher fluorine values in comparable samples. Actually the two sets of samples differed in their fluorine content, their variability in this respect being in no wise unusual. (Experience has shown that considerable variability occurs from week to week as a n expression both of dietary selection and of differences in tile fluorine content of the same articles obtained at different seasons and from different localities.) It is obvious from Tables 111and IV that the use of magnesium acetate has increased the reproducibility of the results not only on samples of low fluorine content but also over the entire range of quantities dealt with.
Vol. 14, No. 1
Conclusions Comparative results indicate that magnesium acetate can be used advantageously in ashing organic materials in preparation for fluorine analysis. Since the blank is negligible, good results can be obtained when fluorine is present in either high or low concrntrations. The reagent can be handled in solution, thus saving the time required to weigh solid agents.
Literature Cited (1) D a h h D., Bonnar, R. U.,and Wichman, H. J., J . Assoc. Oficial Agr. Chem., 21, 459 (1938). (2) L ~E. J., ~ ~ cHEM., ~ ~ En., 3,~93 (1931). . , , (3) McClure, F.J., I b i d . , 11, 171 (1939). (4) MacIntire, W.H., and Hammond, J. W., J . Assoc. Oficial Chem., 22, 231 (1939). ( 5 ) Willard, H. H., and Tvinter, B., IND. ENQ. CHEM., ANaL. ED,, 5, 7 (1933). (6) Winter, 0. B., J . Assoc. OficinE Agi. Chem., 19, 359 (1936).
Determination of Hydrazine J
A Rapid Ferricyanide-Ceriometric Method C. J. DERNRACH’ WITH J. P. MEHLIG, Oregon State College, Corvallis, Ore.
C
ERIC sulfate as a standard oxidant in titrimetry has
been used more and more in the last few years, partly because of the availability of a n increasing number of good reversible redox indicators (9). The many advantages of this reagent-its extreme stability, simplicity of chemical change, and use in hydrochloric acid solution-make i t far superior to potassium permanganate, and in many respects to potassium dichromate. There are many different methods in the literature for the determination of hydrazine. Ray and Sen (IO)were the first to determine the quantitative relationship between hydrazine and alkaline ferricyanide solutions. The reaction proceeds as follows:
They made this reaction the basis for a gasometric method for the determination of hydrazine. The reaction was found to take place without the formation of ammonia, as is the case when many oxidizing agents react with hydrazine. Cuy and Bray (6) Present address, Electro Metallurgical Sales Corporation, Xiagara Falla, x. Y. 1
showed that the oxygen error is negligible it the alkali is added after the ferricyanide. They determined conclusively that in alkaline solution it is the atmospheric oxidation of the hydrazine that produces a weakening of the solution and not the decomposition of the hydrazine. They also su gested a titrimetric method for hydrazine, in which excess of standard potassium ferricyanide solution is added to the alkaline hydrazine solution and the excess ferricyanide determined iodometricnlly. However, since several more rapid methods were available, they did not recommend this procedure for the determination of hydrazine. Benrath and Ruland (2) state that hydrazine is oxidized by ceric sulfate to nitrogen and ammonia according to the following equation: 2Ce (SO&
+ 2N2H4 +NZ+ (SH4)2SOc + Cez(S04)a
However, Lang (8) determined hydrazine and hydrazoic acid together by adding an excess of standard ceric sulfate solution, treating the excess with an excess of standard arsenious acid, and completing the titration with standard ceric sulfate. The hydrazine was then determined in a separate sample by the iodine cyanide procedure. The purpose of the present work was to develop a simple method for the determination of hydrazine by making use of potassium ferricyanide and ceric sulfate. The method con-
ANALYTICAL EDITION
January 15, 1942
59
Preparation of Solutions
cyanide solution. In KO. 6 the color change was not sufficiently sharp t o indicate the end point. Too large a n excess of ferricyanide imparts a dark red-brown coloration, and the transition from green to red-brown is not distinct enough to serve as the end point. However, in a potentiometric titration, this excess would not interfere with the end point. The normality of the hydrazine sulfate solution obtained b y the Bray and Cuy iodate-thiosulfate method (4) was 0.1062 N .
Hydrazine Sulfate. .4n approximately 0.1 N solution !vat. prepared by dissolving about 3.3 grams per liter, and was standardized by the well-knoTvn iodat,e-thiosulfate method of Bray and Cuy (4). Potassium Ferricyanide, 0.5 JI solution. Ceric Sulfate,. An approximately 0.1 X solution was prepared by dissolving 80 grams of ceric ammonium sulfate in 500 ml. of S N sulfuric acid and diluting to 1 liter. This solution was standardized against pure potassium ferrocyanide, following the directions of Furman and Evans (6). Sodium Hydroxide, 6 M solution. Hydrochloric Acld, 6 .I4 solution.
Table I1 s l i o w the effect of time on the reaction. With 10 ml. of 6 Jf sodium hydroxide solution in 30 ml. of solution, oxidation is complete if the solution is shaken gently for 0.5 minute after mixing the reagents with the hydrazine and then allowed to stand for another 1.5 minutes. Prolonged standing before acidification produces no error. The same results were obtained on a run which stood for 35 minutes as on one that stood for 6 minutes.
sists in the oxidation of hydrazine with a n alkaline ferricyanide solution, and the titration with a standard ceric sulfate solution of the ferrocyanide which is produced. Potassium ferrocyanide is readily titrated wibh standard ceric sulfate, either potentiometrically (1, 6, 12) or with indicators (3, 6, 13).such as o-phenanbhroline ferrous complex and diphenylamine.
Effect of Time
Effect of Final Acidity on Titration CONCENTRATIOX TABLE I. EFFECTOF FERRICYANIDE :IO nil. of 6 M S a O H , time 2 minutes, 30 ml. of 0
100 ml.!
So.
1 2 3 4 3 6
Hydrazine Sulfate MI.
0.3 21 KIFF(CX)I
10.84
J
19.54 19.84
10 1.5
19.84 19.84 19.54
Jfl,
0.10Z4 S Ce(SOdP 411 .
i
19.12 19.80 19.96 19.9s 19.98
...
S HCI, final volume Normality of Hydrazine 0.1016 0.1052 0.1062 0,1063 0.1063
Furman and Evans (6) state t h a t the acid concentration should be between 0.5 and 2.0 M when ferrocyanide is titrated with ceric sulfate. The writers found that when the free hydrochloric acid in the final solution was more than 30 ml. of 6 .If acid per 100 ml. of solution the visual end point was not very decisive, as there mas not a sharp change from green to brown. However, 15 t o 25 ml. of 6 M hydrochloric acid per 100 ml. of solution gave good results.
....
Procedure The procedure consisted of adding to a measured volume of hydrazine sulfate solution an excess of potassium ferricyanide, making alkaline with sodium hydroxide, shaking for 0.5 minute, and letting stand for 2 minutes. The mixture was then acidified with hydrochloric acid, diluted to 100 ml., and titrated with standard ceric sulfate solution. There is sufficient ferric iron present in the ceric ammonium sulfate, and most other sources of ceric ions, so that colloidal ferric ferrocyanide forms during titration of the ferrocyanide. .It the end point the green color, caused by the colloidal ferric ferrocyanide, sharply disappears and the solution has the brownish color of the ferricyanide ions. The writers prefer to add a few drops of 0.5 M ferric chloride near the end of the titration to give a sharper end point: especially in the presence of excess ferricyanide ions. Hexanitrato ammonium cerate may also be used as standard oxidant as indicated by Smith, Sullivan, and Frank (If). However, since there is not enough iron present in this highly purified substance, it will be necessary to add a few drops of ferric chloride near the end point in order to form the green colloidal ferric ferrocyanide. If the ferric chloride is added too soon, a blue precipitate of ferric ferrocyanide is formed and thus some ferrocyanide is lost. Diphenylamine may also be used as indicator (S), but thp solution must be diluted to about 0.01 N ferrocyanide concentration before titration with ceric sulfate. The writers found some difficulty in obtaining a sharp end point with diphenylamine in the presence of a large excess of ferricyanide.
Recommended Procedure To 25 to 35 ml. of approximately 0.1 N hydrazine sulfate solution or approximately 0.1 gram of hydrazine sulfate accurately wei hed and dissolved in 25 ml. of water in a 250-ml. Erlenmeyer flasf, add 10 ml. of 0.5 M potassium ferricyanide solution, followed by 10 ml. of 6 M sodium hydroxide solution. Shake ently for 0.5 minute and let stand for a t least another 2 minutes. I d d 30 ml. of 6 M hydrochloric arid and dilute to 100 ml. Titrate with 0.1 N ceric ammonium sulfate solution until the green color in the solution just disappears and the solution has a brownish color. For a sharper color change, 2 or 3 drops of 0.5 M ferric chloride solution may be added within 4 or 5 ml. of the end point.
Results Table I11 gives the results obtained on another hydrazine sulfate solution b y both the recommended procedure and the iodate-thiosulfate method (4). Table I V gives the results obtained with a sample of Eastman hydrazine sulfate which had been dried for 3 hours at 140" C. A different standard ceric ammonium sulfate solution was used for these determinations.
TABLE 11. EFFECTOF TIME [ l o mi of 6 M NaOH, 10 ml. of 0.5 M &Fe(CN)6, 30 mi. of 6 M HC1, final volume 100 m1.l Hydrazine 0.1054.N Normality of
30.
Sulfate
Time
Ce(S0rh
Hydrazine
Effect of Alkalinity Experiments showed that the oxidation of hydrazine b y ferricyanide mas complete in solutions of either low or high alkalinity.
Effect of Excess Potassium Ferricyanide Theoretically 20 ml. of 0.1 N hydrazine sulfate solution should require 4 ml. of 0.5 M potassium ferricyanide solution for complete oxidation. Table I shows results obtained using varying amounts of ferricyanide, other conditions remaining constant. I n Nos. 1 and 2, containing 1 and 2 ml. more than t h e theoretical amount of potassium ferricyanide, the oxidation was incomplete in the 2-minute interval, and even after a 5-minute interval, on a run containing 5 ml. of ferri-
TABLE111. DETERMINATION OF NORMALITY OF HYDRAZINE SOLUTION XO.
Hydrazine Sulfate
1 2 3 4
21.86 27.55 22.87 24.19
M1.
Normality of Hydrazine Solution B y ceric B y iodatesulfate thiosulfate
0.1054 N Ceric Sulfate MI. 21.76 27.39 22.75 24.04
Av.
0,1049 0.1048 0,1049 0.1048 0.1048
0.1047 0,1046 0.1048
....
0.1047
INDUSTRIAL AND ENGINEERING CHEMISTRY
60
TABLE Iv. PnR SO,
Hydrazine Sulfate
Gram 1
,
3 4
0.1055 0,1072 0.1020 01021
CENT PURITY OF HYDRAZINE SULFATE 0,09858 .\‘ Ceric Sulfate Ml, 3 % .85 33.39 31.75 31.77
13y ceric sulfate
% 99.85 99.88 99.82 99.78 A V . 99.84
Purity By iodatethiosulfate c;
99.81 119.85 99.80 99.88 99 83
Vol. 14, No. 1
point of the titration cannot easily be seen. Ferricyanide should not be added beyond the point at which i t just begins t’o cause a reddish-brown colora,tion. The solution during the final titration should be not more than 1.8M in hydrochloric acid. The method, when properly carried out, is capable of giving results reproducible within less than 0.1 per cent, with a percentage error of about 0.02 to 0.11 per cent compared wit)h the iodate-thiosulfate method (4).
Literature Cited
Summary A new titrimetric method for the determination of hydrazine, in R-hich ceric sulfate is the standard oxidant, gives results which check very closely with those obtained by the Bray-Cuy iodate-thiosulfate method (4, but it is much more rapid and the cost of reagents per determination is less than a third as much. The titration is much more easily carried out than the titration in the Jamieson iodate method (7), in which the end point is dependent upon the disappearance of the purple iodine color in the chloroform layer. This necessitates closing the titration flask and shaking vigorously after the addition of each drop of iodate solution near the end point, a slow, tedious process. The main disadvantage of the method lies in the regulation of the amount of ferricyanide to be used. If the excess over the theoretical amount is too small, the reaction with hydrazine will be incomplete, and if the excess is too great, the end
(1) Atanasiu, J. A , , and Stefanescu,V. Be?., 61, 1343 (1928). (2) Benrath, V. A . , and Ruland, K., Z . anorg. allgem. Chem., 114, 226 (1920). (3) Berry, A. J., Analyst, 54, 461 (1929). (4) Bray, W. C., and Cuy, E. J., J . Am. Chern. SOC.,46, 858 (1924). (5) Cuy, E. J., and Bray, W.C., Ibid., 46, 1786 (1924). (6) Furman, N.H., and Evans, 0. M., Ibid., 51, 1128 (1929). (7) Jamieson, G . S.; Am. J . Sci., 33, 352 (1912). (8) Lang, R., “Newer Methods of Volumetric Analysis”, p. 85, New York, D. Van Nostrand Co., 1938. (9) Oesper, K. E., “Kewer Methods of Volumetric Analysis”, p p . 27-52, New York, D. Van Nostrand Co., 1938. (10) Ray, P. R., and Sen, H. K., Z. anorg. Chem., 76, 380 (1912). (11) Smith, G. F., Sullivan, V. R., and Frank, G., IND. ENG.CHEM., ANAL.ED.,8, 449 (1936). (12) Someya, V. K., Z. anorg. allgem. Chem., 181, 183 (1929). (13) Willard, H. H., and Young, P., J . .Im. Chem. SOC.,55, 3260 (1933). .IBSTR.ACTED from a thesis submitted by C. J. Dernbach t o t h e Graduate School of Oregon State College in partial fulfillment of t h e requirements far the degree of doctor of philosophy.
A Sodium-Hydrogen Geissler Tube FRANK &I. GOYAN College of Pharmacy, University of California, Medical Center, Sail Francisco, Calif.
r HE
need for special glass, preheated electrodes, and a r v a c u u m jacket places the construction of the common sodium vapor arcs (1) beyond the resources of the average worker. Such a n arc requires a separate installation in addition to the equipment needed for the operation of Geissler tubes. In connection with the problem of supplying a suitable light source for a Pulfrich refractometer a very satisfactory Geissler tube was developed in this laboratory. It gives a good sodium line as seen through the telescope of the instrument and, under certain conditions, permits focusing on the C and F lines of hydrogen. Since these three spectral lines are frequently encountered in tabulations of refractive index ( 3 ), this tube serves as a valuable addition to the Geissler tubes used with the refractometer. The design of the tube is such that i t may be constructed a t a nominal cost by anyone who has mastered the a r t of making a small ring seal. Pyrex glass is used and no elaborate technique is required in introducing the sodium because of the presence of a n inner shield which effectively prevents surface impurities on the metal from reaching the glass walls that must maintain a pressure difference. T h e introduction of a stopcock permits the tube to be used while i t is undergoing preliminary pressure fluctuations, thus eliminating the need for extreme care in the initial construction.
Construction of Tube Figure 1shows the tube with the sodium electrodes, S a n d S’, in place With the exception of these electrodes, the glass-wo?l plugs, G, and the rod, D.the diagram represents the tube as it
leaves the glass-blowing table. A mercury manometer (not shown) was also included as an integral part of several of the tubes studied. The glass blowing involved is simple. Each arm of the tube was made as a separate unit, the tungsten electrodes, W and W’, being introduced after the ring seals required to secure the shields, F and F’, were finished. The tungsten wire was prepared by a method suggested by W. D. Kumler. I t was outgassed by heating to a high temperature in the oxygen-gas flame. cleaned by applying a film of sodium nitrite to the hot wire, and washed with water. The bright wire ivas re-oxidized in the Bunsen flame before fusing the sleeve of Pyrex tubing around it t o give the characteristic orange-colored seal. The two completed arms were then connected with a short length of approximately 1-mm. capillary RS shown, and the stopcock and manometer were added.
Charging and Filling The sodium electrodes, S and S‘, are prepared by forcing commercial metallic sodium into lengths of Pyrex tubing. This operation is most conveniently carried out by pressing one end of the tubing, lubricated with a drop of kerosene, into a freshly cut surface of a’large piece of sodium. The completed electrode is immediately lowered into place with the aid of the glass rod, D,which is prepared for this purpose. The point of the rod is so shaped that it, will hold the sodium cylinder but will not break the small tubing when pressure is exerted to force the other end of the sodium cylinder over the tungsten wire. The rod is removed by twisting while pulling very gently. TheTupper Pyrex-wool plugs, G, are added and the tubes, E and E‘, sealed off in the glassblowing torch under ordinary conditions. The tube is evacuated, gently flamed, and filled with hydrogen several times before use. The h drogen inlet is designed to permit displacement of air in tube by passing hydrogen through the capillary, B, before the rubber tubing, C, is forced into place and the stopcock opened. Commercial electrolytic hydrogen was drawn from a tank equipped with a reductmionvalve.