INDUSTRIAL AND ENGINEERING CHEMISTRY
758
is not necessary, as excess amounts interfere with the sensitivity of the method. The yellow color of the nitroso R salt interferes with the determination of ferrous iron with an ordinary colorimeter without the aid of light filters. This difficulty may be overcome if more than one standard is employed. The green color formed by the compound iron-nitroso R salt in aqueous solutions is very stable and does not fade in the course of 48 hours under ordinary room conditions.
Interfering Elements Cobalt, if present, interferes with the determination of iron through the formation of a wine-red color. Copper and nickel form with nitroso R salt a yellowish brown color a t concentrations as low as 1 microgram per ml. of copper or nickel. However, the yellowish brown color forms only a t p H values lower than 7.0 and almost disappears a t p H values most suitable for the development of the green color between nitroso R salt and iron. The per cent of deviation in the colorimeter readings resulting from contaminations of 1 to 5 micrograms of either copper or nickel per milliliter of solution is reported in Table 11.
Vol. 14, No. 9
The data in Table I1 show that the per cent deviation of the colorimeter readings of nitroso R salt and iron solutions, when admixed with copper and nickel as contaminants, ranged from 0 to 7.5, excepting in solutions 12 and 13 where the deviation values reached 11.1and 11.4, respectively. The color produced by cobalt cannot be eliminated under the same conditions. However, the writer has not so far encountered cobalt in the tissues of Ananas comosus (L.) Merr., in amounts sufficiently high to cause interference in the determination of iron by this method. The sensitivity of the nitroso R salt method for iron is as good as that of the o-phenanthroline or a-a-dipyridyl methods. The green color of the nitroso R salt method is better suited for colorimetric measurements than the pink color of the other two methods.
Literature Cited H. S. van, J. Am. Chem. SOC.,43, 746 (1921). (2) Sideris, C. P., IND. ENG.CHEM.,ANAL.ED.,9, 145 (1937). (1) Klooster,
PUBLISHED with t h e approval of t h e acting director as Technical P a p e r No. 141 of t h e Pineapple Research I n s t i t u t e of Hawaii, University of Hawaii.
The Microanalysis of Gases J
Nitric Oxide-Hydrogen, Nitric Oxide-Nitrogen, and Nitrous Oxide-Ammonia Mixtures R. NELSON SMITH
AND
PHILIP A. LEIGHTON, Stanford University, Calif.
In the microanalysis of mixtures of nitric oxide with hydrogen and nitrogen, the nitric oxide may be removed by addition of oxygen in the presence of potassium hydroxide and determined by difference from analysis of the remaining gas. Ainmonia may be separated from nitrous oxide by absorption on monochloroacetic acid. A further improvement in combustion coil construction is described.
P
ROCEDURES have been devised for the determination of nitric oxide-hydrogen, nitric oxide-nitrogenl and nitrous oxide-ammonia mixtures b y the dry method of gas microanalysis described in earlier papers (2, 3).
Nitric Oxide The method for the determination of mixtures containing nitric oxide is based on the macromethod of Baudisch and Klinger (1). When adapted to a micro scale, i t becomes indirect and somewhat unorthodox, but if performed correctly i t works well.
A moist potassium hydroxide bead is inserted into a measured volume of the sample containing nitric oxide. [Nitric oxide was prepared by allom-ing solid potassium iodide and sodium nitrite to react with 6 N sulfuric acid (with dissolved air removed) in an evacuated vessel. The water present and the resulting iodine were frozen out with dry ice. The nitric oxide was removed from the system by a Toepler pump, transferred to a gas cup, and stored over mercury. To test its purity an excess of oxygen was added; the fact that only oxygen remained after removal of the nitrogen dioxide so formed showed the purity to be 100 per cent.] , The gas cup together with the potassium hydroxide bead and
sample is lowered over the buret tip until the latter is just below the mercury surface. Oxygen is then slowly added from the buret. Any nitrogen dioxide or intermediates in its formation are immediately removed by the moist potassium hydroxide bead. If the nitric oxide content of the sample is high, a momentary brown coloration due to nitrogen dioxide is noticed; otherwise there is none. It is highly important to use a moist bead. Unless performed in the presence of the potassium hydroxide bead, this reaction cannot be used because the resulting nitrogen dioxide reacts with mercury. The amount of oxygen added need not be known, since, contrary to the work of Baudisch and Klinger, there is no quantitative relationship between the amount of oxygen used and the amount of nitric oxide originally present. T h s is probably due to the formation of unpredictable relative amounts of potassium nitrate and nitrite. To accommodate both the bead holder and the buret tip, a gas cup with an inside diameter of 9 mm. has been found to be satisfactory. With added care, a cup of 8-mm. inside diameter works equally well. It is immaterial whether oxygen is added to the nitric oxide sample or vice versa, but it is advisable to use a volume of oxygen a t least equal to the volume of nitric oxide plus the volume of any hydrogen which may be present. In the case oi nitric oxide-hydrogen mixtures, after addition of oxygen the potassium hydroxide bead is removed and the resulting volume measured. A combustion coil is then introduced and the hydrogen burned to water, the water removed on another potassium hydroxide bead, and the contraction measured. TABLE I. ANALYSES OF NITRICOXIDE-HYDROGEN ~~.IIXTUREB Determination
Volume of Sample Cu. m m . 22.77 18.90 19.13 25.85 30.17 24.55 19.41 15.05
--
Nitric OxideTheoretical Determined
% 0.00
13.76 25.35 38.88 49.62 58.66 74.86 100.00
% 0.66 14.79 26.61 39.75 50.28 59.02 74.45 99.40
Difference
70 0.66
1.03 1.26 0.87
0.66
0.36 -0.41 -0.60 Av. 0.73
ANALYTICAL EDITION
September 15, 1942
759
The hydrogen is calculated from this contraction, and the nitric oxide determined by difference. Results on known mixtures are shown in Table I.
For the nitric oxide-nitrogen samples (Table 11) the procedure was the same except that after the removal of the nitric oxide a measured excess of hydrogen was added and combustion carried out. From the resulting contraction, the remaining volume of hydrogen and the amount of oxygen present in the sample after the removal of the nitric oxide were calculated; b y difference the amount of nitrogen became known. From this, the volume of nitric oxide originally present was calculated. A sample containing hydrogen, nitrogen, and nitric oxide was analyzed to determine the reproducibility of this method for such a mixture (Table 111). The procedure in this case was the same as for hydrogen-nitric oxide mixtures, except t h a t after determination of the amount of hydrogen present a measured excess of hydrogen was added and a second combustion performed. From the resulting contraction the amount of oxygen present in the sample after removal of the hydrogen and nitric oxide was calculated; b y difference the amount of nitrogen became known. Hydrogen and nitrogen were chosen as diluent gases because these gases occurred with nitric oxide in the problem for which the method was developed. T h e volumes used are smaller than those generally reported in papers of this nature, but again, this was conditioned b y the nature of the problem t o which the method was t o be applied. As a result of the small volumes, the per cent error is rather large, but the absolute error is about the same as that for other methods using the same type of apparatus. Mixtures of nitric oxide and nitrous oxide cannot be determined in this manner because nitrous oxide is absorbed by potassium hydroxide.
_- _. B
C FIGURE 1. COMBUSTION COIL A
A . Complete B. Enlarged detail of t i p C . Operating position
TABLE IV. Determination 1 2 3 4 5 6 7 8 9 10 11
ANALYSESO F Volume of Sample Cu. m m . 18.40 19.42 24.47 24.49 28.35 29.54 36.44 36.55 28.77 28.88 34.23
NITROUS
OXIDE-AMMOK A MIXTURES
--Nitrous Theoretical
OxideDetermined
%
70
100.00 100.00 78.96 77.66 64.90 63.54 50.55 49.25 40.91 23.99 16.45
100.00 99.4s 79.28 17.79 65.43 64,22 51.04 49.44 41.01 24.40 16.51
TABLE11. AXALYSES OF SITRIC OXIDE-SITROGEN LIIXTURES Determination
l-olume of Sample Cu. m m
r-s-itric Theoretical
OxidDetermined
70
%
Difference
70 0.00 -0.52 0.32 0.13 0.53 0.68 0.49 0.19 0.10 0.41 0.06 Av. 0 . 3 0
Difference
70 after fusion. Table IT' shows the results of analyses performed on known mixtures.
An Improved Combustion Coil Av. 0 . 4 9
TABLE 111. Volume of Sample Cu. m m . 30.55 30.25 34.43
k . 4 L Y S E S O F .4 SAMPLE CONTAINING
XITROGEN, ASD XITRICOXIDE
HYDROGEN, NO
Hz
N2
50
%
%
33.45 33.59 32.82
34.70 34.38 34.85
31.85 32.03 32.32
hv. 33.29
34.64
32.07
Determination of Nitrous Oxide i n the Presence of Ammonia Combustion of nitrous oxide with hydrogen (4) cannot be accomplished in the presence of ammonia because i t also takes part to some extent in the reduction. A phosphorus pentoxide bead cannot be used to remove ammonia, as some of the nitrous oxide is also removed. It was found that a fused bead of monochloroacetic acid rapidly, quantitatively, and specifically removes ammonia in the presence of nitrous oxide. T o prevent contamination of the mercury surface inside the gas cup, it is necessary to submerge the bead below the mercury surface just before complete crystallization
The combustion coil mentioned in connection with t h e analyses for nitric oxide was a n improvement over the older type ( 3 ) and warrants a short description. A coil holder (Figure 1) was shaped to fit snugly into the customary absorbent holder on the apparatus. A piece of No. 24 (0.5-mm., 0.02-inch) platinum wire was sealed into the end of the holder at the center so as t o make contact with the mercury with which the coil holder was filled. A second piece was sealed into a little glass arm at one side and near the end. These two pieces of wire were trimmed off about 1 mm. from the end of the holder and joined by a 2.5-mm. piece of 0.05-mm. (0.002inch) platinum wire spot-welded to their ends. A small commercial spot welder may be used for this purpose. The coil operates very satisfactorily on 12 volts, and may be suitably controlled if it is used in series with a slide wire resistance. With reasonable care the coil will not burn out, but if this should happen, it is a matter of but a few rninut'es to spot-weld another platinum thread into place. I t was found that in a gas cup with an inside diameter of 8 mm. the coil could be used for combustion in which the volume remaining at the end of the process was 11 cu. mm. or more.
Literature Cited (1) Baudisch, O., and Klinger, G., Ber., 45, 3231 (1912). ( 2 ) Blacet, F. E.. and Leighton, P. A , IXD.ENQ.CHEM.,ANAL.ED., 3 , 266 (1931). (3) Blacet, F. E., Sellers, A . G., and Blaedel, W '. J., Ibid., 12, 3 5 6 (1940). (4) Blacet, F. E., and Volman, D. H., Ibid., 9, 44 (1937).
