80
INDUSTRIAL AND ENGINEERING CHEMISTRY
I n view of the above-mentioned discrepancies, several samples of spirit were submitted to a number of analysts (Table IX). Analysts 1 , 2 , 3 ,and 4 were expert drug chemists. Analyst number 5 was a student, whose first ten results show an average of 9.69 per cent of camphor. However, after running several additional lots, this same operator averaged 9.81 per cent. All experiments indicate that the low results are due to impurities in the camphor, to an incomplete reaction, or to decomposition of the hydrazone rather than to the solubility of the hydrazone. Since the average recovery amounts to about 98 per cent, i t is suggested that 0.2 per cent should be added to the results obtained in each determination to correct the low results.
Proposed Method The proposed modification may be given in detail as follows : Accurately measure 2 cc. of spirit of camphor into a 300-cc. Erlenmeyer flask containing 15 cc. of alcohol, and add 75 cc. of dinitrophenylhpdrazine reagent solution. Connect the flask with a reflux condenser, and heat the flask for 4 hours by immersing it in actively boiling water. Allow the mixture to cool, add 200 cc. of distilled water, and set aside for 24 hours. Transfer the precipitate to a previously dried and weighed Gooch crucible, and wash with small portions of cold distilled water until the washings are no longer acid to litmus paper. Continue the suction until the excess water is removed, dry the crucible, and precipitate to constant weight at 100" C. The weight of the precipitate multiplied by 22.9 equals the weight of camphor in 100 cc. of spirit of camphor. To correct the low results given by the method, add to the percentage of camphor found 0.2 per cent. The dinitrophenylhydrazine reagent solution is prepared in the following manner: Dissolve 3.75 grams of 2,4-dinitrophenylhydrazinein a warm mixture of 45 cc. of concentrated sulfuric acid and 45 cc. of dis-
VOL. 10, NO. 2
TABLEIx. RESULTSOBTAINED BY DIFFERENT ANALYSTS THE PROPOSED METHOD
WITH
(10% Camphor Solution) Sample
Analyst
NO.
1
1 2 3 4
9.7s 9.84 9.75 9.65 9.7s 9.79 9.69 9.86 9.71 9.92 9.78 9.98
5
6 7 8 9 10 XV.
Pluscorrection of 0 . 2 per cent
Analyst 2 9.79 9.81 9.85 9.7s 9.78 9.75 9.84 9.83 9.76 9.81 9.80 10.00
Analyst 3 9.74 9.77 9.83 9.75 9.76 9 73 9.76 9 80 9,79 9.77 9.77 9.97
Analyst 4 9.89 9.83 9.80 9.77 9.87 9.82 9.84 9.79 9.88 9.82 9.83 10.03
Analyst 5 experi- After exence perience 9.62 9.85 9.74 9.81 9.80 9.79 9.62 9.88 9.56 9.85 9.60 9.84 9.69 9.80 9.66 9.83 9.80 9.71 9.7s 9.74 9.69 9.81 9.89 10.01
Xo
tilled water. Cool the solution, and add enough distilled water to make the solution measure 250 cc. If necessary, filter the solution before using it.
-4number of spirits made from synthetic camphor have been analyzed by the proposed modification. This study will be continued and reported later. Literature Cited Brady, 0. L., J . Chem. Soc., 134, 756 (1931). Fernandea, O., and Socias, L., Rev. acad. cienc. .\fadrid, 28, 330 (1932).
Fernandez, O., Sociaa, L., and Torres, C., d n a l e s
SOC. espan. ps. ouim.. 30. 37 (19321. Hampshire,'C. H., and Page, G. R.. Quart. J . Pharm. Pharmacol.,
7, 558 (1934).
U. S. Pharmacopeia, subcommittee personal communication.
on Organic Chemicals,
RECEIVED April 10, 1936.
Iodofluoride Method for the Determination of Copper WILLI.01 R. CROWELL, SIDNEY H. SILVER. AND ALAN T. SPIHER University of California, Los Angeles, Calif.
Experimental Procedure
The sample consisting of 0.3 to 0.4 gram of copper sulfide is weighed int,o a 250-cc. Erlenmeyer flask containing measured amounts of the impurities under investigation, 20 cc. of aqua regia and 10 cc. of 18 A' sulfuric acid are added, and the reaction is allowed to proceed slowly until there is no evidence of free sulfur present. During this st'age a small watch glass is placed over the mouth of the flask. The flask is then embedded in a steam-heated sand bath, the watch glass raised slightly, and the solution heated just below boiling until the point of incipient fuming is reached. Twenty cubic centimeters more of aqua regia are added, and the evaporation is continued and finally finished on a gas or electrically heated sand bath as soon a8 dense white fumes appear. To the solution are added 20 cc. of water and ammonia solution until a slight but distinctly recognizable odor of ammonia is obtained. Finally 1.5 grams of ammonium bifluoride are added, followed by iodide and titration with thiosulfate with the addition of 2 grams of potassium thiocyanate near the end point. Even in those runs in which the largest amounts of iron and of arsenic xere present, 1.5 grams of bifluoride were found sufficient. When iron and manganese are present together, it is best to add the ammonia after the bifluoride ( 2 ) . In such a case the proper amount of ammonia is determined by treating several samples in the same manner as those run for analysis, carrying the operation only t o the point at which a distinct odor of ammonia is obtained. Blank runs on the copper sulfide are made, using the same procedure as that employed when impurities were present.
Except for the addition of thiocyanate. the steps are essentially the same as in the method preiiously reported ( 1 ) .
Table I shows results of analyses of several series of cupric sulfide samples in the presence of various impurit'ies, using
IK
A RECEXT article ( 3 ) results of volumetric determinations of copper in a number of samples of cupric sulfide obtained by a so-called "iodofluoride" method were reported, as a modification of the procedure described by Crowell (1). In a discussion of the results i t was stated that the iodofluoride method cannot be used when more than 0.15 gram of iron is present and t h a t in all titrations the presence of a yellow color interfered with the sharpness of the end point. If one uses this modified procedure, in which bifluoride is added to the sulfuric acid solution of the sample without any addition of ammonia, the pH of the solution is about 1.1 instead of between 3 and 4, the region within which it is generally agreed that it is best to work. If, on the other hand, one uses the procedure of the authors of the present paper in which sufficient ammonia is added to produce a pH of 3.3 or somewhat above, not only are the end points white, sharp, and permanent but copper determinations in the presence of as much as 0.3 gram of iron and corresponding amounts of arsenic and arsenopyrite can be made with high accuracy and precision.
FEBRUARY 15, 1938
ASALYTICAL EDITION TABLE
I. AXALYSISO F COPPER SULFIDE
81
IN THE PRESENCE OF
INTERFERING ELEMENTS
( T h e percentage of copper in t h e pure sulfide varied from 64.37 t o 64.12.) Fe .As Impurity present Fe Fe Fa -48 AS Amount, gram 0.1 0.2 0.3 0.1-0.2 0.3 Fe 0.1-0.2 .Is 0 . 1 - 0 . 2 Average deviation" +0.04(7)b -0.02(5! -0.07(6) -0.05(8) -O.l4(3) -0 05(5) *Maximum deviationa +O.lO -0.07 -0.15 -0 14 -0.20 -0.06 a The deviations are percentage del-iations from the per cent of copper obtained in the blank runs. b The numbers in parentheses refer t o the number of determinations made.
+
the procedure just described. B greater variety of impurities and larger amounts are used than in the previous work (1). T h e copper sulfide samples were obtained from a pound of the c. P. compound which had been finely ground and thoroughly mixed. The iron and arsenic impurities were supplied from solutions of ferric nitrate and arsenic acid. The arsenopyrite was prepared from a specimen of copper-free mineral. The results indicate that by the procedure employed copper determinations can be made in the presence of as much as 0.3 gram of iron, 0.2 gram of arsenic, 0.2 gram each
0.2
Arsenopyrite 0.3
O.OO(5) +0.06
-0.11(4) -0.23
of iron and arsenic together, and 0.2 gram of arsenopyrite with a n error less than 0.1 per cent, and in the presence of as much as 0.3 gram of arsenic and 0.3 gram of arsenopyrite with a n error less t,han 0.2 per cent.
Literature Cited (1) Crowell, W.R., Hillis, T. E.,
Rittenberg, S. C., and Erenson, R. F., ISD. ESG.CHDM.,Anal. Ed., 8, 9 (1936). (2) F o o t e , H. IT., and Vance, J. E., I b i d . , 8, 119 (1936).
(3) I b i d . , 9, 205 (1937). RECEITEDS o r e m h e r 23,
1937.
Routine Determination of Low Chromium in Aluminum Alloy LOUIS SILVERnIAN, 2129 Wightman St., Pittsburgh, Pa.
I
S STEELS, chroniiuin is most easily determined ( 2 ) by
dissolving the steel in aqua regia, oxidizing the chromium with perchloric acid, and then titrating the chromic acid with ferrous sulfate. Solution of aluminum alloys takes place readily with aqua regia, but heating after t'reatment with perchloric acid results in the formation of an insoluble deposit which causes bumping. Oxidation of the chromium is incomplete. When a mixture of phosphoric and perchloric acids is used as solvent for the aluminum alloys, solution is complete, and upon further heating the chromium is oxidized to chromic acid. Manganese, if present, is oxidized to the trivalent state; iron, copper, magnesium, and aluminum are TTithout bearing on the results; and silicon is more or less held in solution. The accepted methods for the determination of chromium are either the direct' oxidation of the element in sulfuric acid solution by potassium permanganate, or by silver nitratepersulfate (1). Bot,li these methods have been used ill steels, but are neither as rapid nor as easily handled as is the routine perchloric acid method when a large group of samples is involved. The same advantages of ease of solution, oxidation, and handling of the samples apply to the perchloric acid method for aluminum alloys.
Proposed Method Dissolve 1 gram of aluminum alloy in a 400-cc. covered beaker in a mixture of 10 cc. of 85 per cent phosphoric acid and 20 cc. of 70 per cent technical perchloric acid. If necessary, heat at about 120" C. until solution is complete. Move the beaker to a spot on the hot plate Tvhere the contents of the beaker will be heated to about 220" C. After t h e color change (yellow for chromium only, but broivn for manganese and chromium together), heat about 5 minutes more. Remove from the heat, and cool somewhat with the cover partially removed. Dilute to 100 cc., add 5 cc. of h rocliloric acid (1 t o 3), and furic acid (1 to l), dilute t o boil out chlorine. Add 10 cc. of 250 cc., and cool t o 25" C., or le Titrate potentiometrically with ferrous sulfate (or other acceptable method). 1 cc. of 0.1 S FeSOn = 0.1735 per cent Cr. Readings were inade to the nearest 0.1 cc. or 0.017 per cent of chromium. As t,he oxidation of chromium is not complete (a),the ferrous sulfate may he st'andardized by adding chromium to a chromium-free aluminum sample. However, when a 0.3 per cent chroniiuni sample is to be used, oxida-
tion to the extent of only 98 per cent of the chromium content involves no error. Hydrochloric acid is added to the oxidized solution for the purpose of reducing the manganese to the bivalent stage, but need not be used if manganese is absent. Titrations performed above 25' C. give lower results. TABLE I. CHROMIUM COSTEST
--
Outside laboratory Persulfate methoda Proposed method Proposed method, &In added
0.21 0.21 0.23 0 20 ,--Sample-
Sample-534 0.22 0.22 0.22 0.21 0.23 0.23 0.25 0.25
0.23
0:24 0.25
52-s 0.22 0 23 0.23 0.24 0.21 0.22 0.22 0.22 0.25 0.24 0.25 0.25 0.23 0.23 0.24 0.25 0.26 0:25 0 23 0 24 0.25 0.24 0.24 .Ipproximate Composition Mn M g CU Si Fe 4 53-9, % 0 1 . 1 - 1 . 4 0 01-0.02 0.5-0.7 0.15-0 19 97 524, % 0 2 3-2.8 0.01-0.03 0.1-0 16 0.16-0.20 97 a, b Results by 0.Gates and .i.illison. , respectively, Navy Laboratory, Munhall, Pa, Outside laboratory Persulfate methodb Proposed method Outside contractor Persulfate methoda Prouosed method
PRECAUTIOKS. Solution of aluminum alloys in phosphoric-perchloric acids generates hydrogen; hence no gas flames should be lighted nearby. Aluminum powder containing grease should first be extracted with acetone, since concentrated perchloric acid rapidly oxidizes grease. When aluminum powder is to be dissolved, the powder and reagent are heated on a steam bath, and removed when action starts. When the rapid action ceases the regular procedure is followed If pomder and peichloric acid alone are used, partial solution takes place, hut if the mixture is heated, the acid dehydrates and will vigorously attack any dry powder, possibly causing a mild explosion.
Literature Cited (1) Aluminum Co of .America, Chemical Analysis of Aluminum " 1935,. ( 2 ) Lundell, Hoffman, and Bright, "Chemical Analysis of Steels," p. 298, N e w Tork, John TTiley R- Sons Co , 1931 K E C E I \ E Uaeptember 6, 1 9 3 i
