INDUXTRIAL A N D ENGINEERING CHEMIXTRY
144
noted that there is an accumulation of sulfur in the first over and especially in the last over-namely, at the higher temperature; also there is a distinct loss of sulfur, 21 per cent as compared with the original amount of sulfur in the gasoline (0.05 per cent). There is also a large accumulation of sulfur in the distilling flask in the form of a sticky residue, as indiTable IV-Highest Crude
Temperatures Attained i n the Oil in Distilling At end of wax distillate At end of gas oil cut cut O
c.
304.5 340.5 288 343 288 264.5 268.5 393 204.5 149
s
O F. 680 645 550 660 550 490 515 740 400a 3000
c.
’F.
304.5
580
Low red heat
... ...bb
Gasoline (Table 11) Gasoline (Table 111) a Highest temperature attained in the vapor as gasoline. b No wax distillate separated.
Vol. 17, No. 2
cated in the column in Table I1 headed “Sulfur on total SUIfur,” in which 91 per cent, of the sulfur has accumulated in the residue (including some loss). ’ When this same solution‘ of‘sulfur was distilled without steam and therefore at slightly higher temperatures, reaction took place between the sulfur and the hydrocarbons, with the result that all the distillates contained larger proportions of sulfur and there was a gain of sulfur in the fractions amounting to nearly 100 per cent of the original sulfur present (0.05 per cent). Here also the sulfur has a tendency to accumulate in the first and last fractions, and there is a tendency for the, increase in the sulfur content to begin at the 80 per cent over point in both these cases. I n the case of distillation without steam there was an evolution of hydrogen sulfide during the distillation of the last portion, and this last portion was of a yellow color, indicating a considerable reaction between the sulfur and the oil.
Nitrogen Oxides in Sulfuric Acid‘ By E. M. Jones TENNESSEE COPPER Co., COPPERHILL, TENN.
YDYGER2 has recently shown the nitrometer to be unsatisfactory for determining the oxides of nitrogen in commercial sulfuric acid. His solution of the problem was to distil off the oxides of nitrogen, by heating the acid with ferrous sulfate, into pure, concentrated sulfuric ’ acid and titrate with permanganate. The. writer recently had occasion to investigate the same problem. The method used at this plant for estimating the dissolved oxides was to titrate directly the acid containing them with permanganate. This method gave results that were so much higher than nitrometer results that it was believed the permanganate was titrating other substances such as ferrous iron or dissolved sulfur dioxide. The amount of dissolved sulfur dioxide was found to be negligible, and qualitative tests showed that ferrous iron was absent. A consideration of the reaction between nitroso sulfuric acid and ferrous sulfate will show that the two could not exist together a t the temperature of acid coming from a Glover tower. 2FeS04 PHNOSO4 = Fea(SO4)a f &Soil f 2N0 Ferric iron was present in an appreciable quantity. Impure commercial sulfuric acid used in a nitrometer is usually colored a more or less pronounced blue, although C. P. acid with a pure nitrate never gives this color. This blue is due to the iron present in the acid. It is a compound of nitric oxide and ferrous iron, and is broken up only on heating.3 I n other words, it is the familiar ring test for nitrates. As a compound is formed between the nitric oxide and the ferrous sulfate, the amount of nitric oxide evolved will be less than it should be, and the results will be low. This accounted for the difference between the permanganate titration and the nitrometer results. To demonstrate that iron present in the acid caused a large error, a weighed portion of standard potassium nitrate and a small amount of
N
+
1 Received
July 21, 1924. Bull. f&%alion ind. chim. Belg., 2, 367 (1923). a Treadwell-Hall, “Analytical Chemistry,” Vol. I , p. 331. The color of the ring in this test for nitrates varies with the concentration of ferrous sulfate. A relatively large amount of ferrous sulfate will produce a brown ring while with decreasing amounts of ferrous sulfate the ring is red, then purple, and with very small amounts is blue. In commercial sulfuric acid the amount of ferrous or ferric iron is very small. 2
ferric sulfate were placed in the decomposition bulb of a nitrometer with C . P. 95 per cent sulfuric acid. The amount of nitric oxide measured in the buret was considerably less than it should have been for the amount of nitrate used. The same low results were obtained using ferrous sulfate. The blue color appears instantly with ferrous sulfate, but takes a few seconds to appear with ferric sulfate, due, no doubt, to the time required to reduce the ferric iron to the ferrous condition. The conclusion is that the nitrometer cannot be used t o eetimate nitrogen oxides in commercial sulfuric acid containing iron, with any degree of accuracy. It is absolutely worthless for small amounts of nitrogen oxides under those conditions. Method
The following method will accurately estimate small amounts of nitrogen oxides in sulfuric acid. It is a modification of Devarda’s method, as used by Allen4 and further modified for this work. The apparatus consists of a distillation flask of about 1 liter capacity connected to a condenser. A 2-hole stopper fits in the neck of the flask. The condenser is connected through one of these holes and a separating funnel with a stopcock fits in the other. Enough strong caustic (40 per cent NaOH, on which a blank has been previously run) to more than neutralize the sample of acid to be used is placed in the flask with a few drops of indicator and 5 grams of Devarda’s alloy. The flask is stoppered and the condenser connected. The sample of acid is placed in the funnel and allowed to run slowly into the hydroxide. Care should be taken in adding the sample, for the violent reaction between the acid and the strong caustic tends to cause the contents of the flask to boil over. (The acid sample should not be diluted with water, for dilution will liberate the dissolved nitrogen oxides.) When the action has slowed down the flask is heated until all the ammonia is driven over. One hour will complete the distillation. This method will determine the dissolved nitrogen oxides in sulfuric acid, and also any ammonium salts present. TO determine the latter, of course, the alloy is omitted. 4
Scott, “Methods of Analysis,” Vol. I , p. 346.
