734
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
duplicate copy of the record on the label should be mailed t o the engineer in charge, Bureau of Mines, Pittsburg, Pa. SAMPLING GAS FROM A WELL.
8
Since the gas associated with oil is an ideal fuel and illuminant, and the literature dealing with the composition of natural gas is scanty, a description of the method of sampling such gas for analysis is here given. For taking a sample of gas under pressure from an oil well a cloth funnel should be made b y folding and sewing any strong, closely woven cloth into the form of a cornucopia. The larger end of this funnel should be large enough to encompass the gas pipe from which the sample is t o be taken, The smaller end, or apex, of the funnel should be securely tied about one end of a flexible rubber tube I or n feet long and one-fourth t o one-half inch in diameter. If there is a gas jet a t the well, one end of the rubber tube may be attached directly t o the jet. A gas-sampling bottle should be procured, if practicable, from the Bureau of Mines, Pittsburg, Pa. If such a bottle is not a t hand, a I - or n-liter ( I - or 2 quart) bottle with a well-ground, tight-fitting glass stopper may be used. The bottle should be thoroughly cleansed and dried. A large perfume bottle or an acid bottle, such as may be obtained from a drug store, will usually answer. A glass stopper is essential, for a cork or rubber stopper may leak, even though it appears to be hermetically sealed with wax; moreover, a cork or rubber stopper may contaminate the gas. To collect a sample, the funnel should be tied firmly about the end of the gas pipe. The funnel and the rubber tube should then be thoroughly flushed with the gas t o rid them of air. The free end of the tube should go t o the bottom of the sample bottle. The bottle should be fastened bottom up and the gas allowed t o blow strongly into i t for a t least a quarter of a n hour t o insure complete expulsion of air. If the gas pressure is low, the gas should be allowed t o blow longer, or until it is certain t h a t all air has been expelled from the bottle. Meanwhile the stopper of the bottle should have been well greased with vaseline. While the gas is still blowing through the tube the tube should be slowly withdrawn. The stopper should be put in just as soon as the tube is withdrawn and should be turned firmly into place. Then the bottle should be turned u p and a spoonful of melted paraffin poured over the stopper. The stopper should be secured with elastic band. A strong tag should be tied t o the bottle by a stout cord. This tag should be labeled as follows: Gas Sample. Sampled b y . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ Date . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lease.. . . . . . . . . . . . . . .
......................
.......................
1911
The bottle should be packed securely in a box and forwarded t o the Bureau of Mines, Pittsburg, Pa. A duplicate copy of the label should be sent t o the same address. BUREAUO F
MINES,
DEPT O F
INTERIOR.
THE DETERMINATION OF TOTAL SULPHUR IN INDIA RUBBER. By
c. E.
TVATERS AND
J. B. 'PUTTLE.
Received August 7 . 1911.
What may still be described as the usual method for the determination of total sulphur in India rubber is the one first published b y Henriques.l The details of this method are too well known t o require description here. I n more recent years other methods have been advocated. Alexanders used sodium peroxide t o decompose the nitrogen peroxide addition-product of rubber. I n the same year, Eschh recommended the use of Eschka's mixture and procedure for the determination of sulphur in coal. He also stated that the sodium peroxide method gives good results. Wagner5 published a slight modification of the method of Henriques, stating that much sulphur is lost b y volatilization. He therefore made the nitric acid solution alkaline with sodium hydroxide, transferred to a nickel crucible, added sodium carbonate and then evaporated t o dryness. The oxidation was carried to completion by heating in a n air-bath. Pontio6fused with manganese peroxide and a mixture of sodium and potassium carbonates. The results were about 0.1per cent. lower than b y the method of Henriques. For the free sulphur7 he extracted with absolute alcohol, distilled off the solvent, oxidized with alkaline hydrogen peroxide, evaporated to dryness and fused in a silver crucible. A distinct departure from the usual methods is is due t o Hinrichsens who oxidizes electrolytically in the presence of concentrated or fuming nitric acid. Finally, Hubener8 devised a method intended t o exclude insoluble mineral sulphates. The sample is boiled in a flask with concentrated nitric acid for some time, most of the acid evaporated off on the steam-bath and the oxidation completed b y means of bromine and water. One of the present writers, having frequent occasion t o determine total sulphur in rubber, over a year ago made a number of comparative tests of different variations of the method of Henriques. The results obtained with two samples of rubber are given below (I-IV). In all cases 0.50 gram of rubber was taken. All fusions were made over a flame of gasoline-air gas. The results are given as percentages of sulphur. All reagents were tested, and no determinations have been omitted. I. Warmed hours in covered crucible with 2 5 cc. concentrated HNO,, allowed t o stand 36 hours, Published b y permission of the Director of the Bureau of Standards. 2. awgezu. Chem., 12, 902 (1899). 3 Gummi-Zto., 18, 729: Z . angew. Chem., 17, 1799 (1904). Chem.-Ztg., 28, 200 (1904). 6Gummi-Zfg..21, 5 5 2 : C. A , , 1, 1327 (1907). 6 Caoutchouc et Gufta-Percha, 6. 2751: Chem. Techm. R e ) . , 1909, 372. 7 I b i d . , 6 , 2194; C.A , , 2, 3412 (1908). 1
2
. . . Range. . . . . . . . . . Section. . . . . . . .Township District. . . . . . . . . .County Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... ............................
Oct.,
9
Chem.-Ztg.. 33, 735 (1909). Gummi-Zfg.,24, 213-4; Analyst, 36, 266-7 (1910).
Oct.,
T H E J O U R N A L OF I N D U S T R I A L A N D E N G l N E B H l N G C H E M I S T R Y ,
1gr I
evaporated nearly t o dryness, added Na,CO,-KNO, mixture and fused as usual. Sample Sulphur
1
1 3.44
................. 3 . 3 9
2 3.26
2 3.22
11. Added HNO, and one cc. Br, let stand 36 hours without preliminary heating, evaporated, etc., as usual. Sample Sulphur..
1 3.40
...
1 3.47
2 2 3.39 3.27
111. The same as 11, but allowed t o stand only one hour, heated with cover for two hours, evaporated and fused as usual. Sample Sulphur..
...
., . . , .
1 3.31
1 3.35
2 3.04
2 3.09
IV. Only HNO, added, digested a t once on the steam-bath for two hours, evaporated and fused. Sample Sulphur..
1
............... 3.17
1 3.43
2 3.06
2 2.93
1
................. 3 . 7 1
1 3.65
2 3.37
2 3.38
I n a n attempt t o obtain satisfactory results without fusion, some determinations were made some months later, without a knowledge of Hubener’s paper. Half-gram portions of a sample of medium hard rubber were digested with nitric acid in flasks covered with watch-glasses. I n some cases bromine was added after the digestion with acid and, after standing half a n hour, water was added and the flasks heated on the steam-bath. Finally, the volume was brought t o about 1 7 5 cc., the solution heated, filtered and a little sodium hydroxide added t o the filtrate and wash-water. This was then evaporated t o dryness, adding a little hydrochloric acid towards the end, taken up with very dilute hydrochloric acid, filtered, and barium sulphate precipitated as usual. The results follow: VI. Treated with HNO, alone. Sulphur.... . . . . . . . . . . . . . 7 . 7 6 7 . 5 1 7 . 6 8 7 . 9 6 VII. Treated with HNO,, followed b y Br. Sulphur . . . . . . . . . . . . . . . . . 7 . 6 2
7.51
7.93
7 87
7.76
All of the precipitates obtained under VI and VI1 contained much lead. After the method of Hubener was called t o our attention, some determinations were made on a sample of hard rubber containing no barium. VIII. Hiibener’s method. Sulphur.. . . . . 4 . 7 9
3.91
5.23 4.02
4.31
4.13
I t is evident t h a t very widely different amounts of sulphur must have been retained in the insoluble residue in the form of lead sulphate. I X . Total sulphur b y method of Henriques. Sdlphur.. .....................
8.65
8.70
X. Treated with HNO,, followed by Br and H,O and fused as usual. Sulphur..
.....
8.63
8.62
8.77
8.80
8.72
8.80
I t has recently been claimed by van’t K r u y s ~t h a t 1
Z . a t d Chem., 49, 393 (1910).
when an excess of calcium chloride over the amount of sulphuric acid is present, only calcium sulphate is carried down with the .barium sulphate, and the calcium salt can be converted into barium sulphate by digestion with strong hydrochloric acid, or aqzm regia, and barium chloride. Several determinations were made to test this suggestion. X I . Preliminary treatment as under X, subsequent treatment as suggested by van’t Icruj-s. Sulphur
8.73
8.76
8.52
8.46
8.74
8.78
8 i5
8 77
At this point joint analyses of a fairly large sample of rubber were carried out by the present writers. X I I . Hubener’s method. The sulphur in the insoluble residue was determined by fusion with sodasaltpeter mixture, extracting the melt with water, etc., as usual.
s in original
V. Treated with one cc. B r a n d five cc. H,O, allowed t o stand over night without heating, next morning evaporated off the H,O, added HNO,, digested, evaporated and fused. Sample Sulphur
735
filtrate. . . . 0 91 0 . 9 1 0 . 6 8 0 . i 9 0 . 8 2 0 99 1 . 5 2 1 . 1 5 1 .24 S i n insoluble residue.. . 2 . 3 5 2 . 4 3 2 . 5 4 2 . 6 9 2 . 5 3 2 . 3 6 1 . 9 5 2 31 2 . 2 1
_ ~ _ _ - _ _ _ _ _
Total sulphur found.. , . 3 . 2 6 3 . 3 4 3 22 3 . 4 8 3 . 3 5 3 35 3 . 4 i 3 46 3 . 4 5
All of the precipitates of barium sulphate from the original filtrates were found to contain lead, when tested with dilute ammonium sulphide. X I I I . Treated with HNO,, allowed to stand over night, the acid driven off on the steam-bath, one cc. Br and ten cc. H,O added; then the H,O and excess of Br driven off by heating. The residue was mixed with soda and saltpeter and fused as usual. Sulphur. . . . . . 3 . 4 1
3.21
XIV. The same as X I I I , but the HNO, not driven off before adding Br. Sulphur
3.60 3.63
3.58
3.63
3.57
XV. The same as X I I I , but treatment with Br omitted. Sulphur.. . . . . . 3 . 2 9 3 . 4 9 3 . 3 5 3 38 3 . 5 5 3 . 4 3 3 . 3 6 3 . 5 8 3 . 5 6 XVI. Treated first with Br and H,O, allowed t o stand over night without heating, then Br and H,O driven off on steam-bath, treated with HNO,, etc., and fused. Sulphur... . . . . . . . . 3 . 4 5 3 . 4 7 3 . 4 8 3 . 5 3 3 . 4 9 XVII. The same as XVI, but excess Brrand H,O not driven off before adding “0,. Sulphur . . . . . . . . . . . . . .. 3 . 5 9 3 . 4 7 3 . 6 4 XVIII. The method of Henriques, except that the HNO, was saturated with Br. Sulph
66 3 . 6 5 3 . 6 2 3 . 7 3 3 . 6 8 3 . 6 5 3 . 6 9 3 . 6 3 3 . 6 6 3 . 7 1 3 62
X e same as XVIII, but followed by the treatment suggested b y van’t Kruys. Sulphur
. . . . 3.69 3.73 3.76 3.75
I n order t o obtain a definite idea of the variations caused by differences in the preliminary treatment and in the conditions under which the barium sulphate is precipitated, a very dilute solution of sulphuric acid was made. I n each of the following determinations, a 25-cc. portion was taken. The weights of barium sulphate found were calculated as percentages of sulphur in 0.50 gram of rubber in order t h a t the results might be more readily compared with the determinations above. XX. Direct precipitation with BaCI,. The last
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
736
two determinations were made with the addition of two cc. of I : I HCI, the first six without adding HCI. Sulphur .............. 3 . 1 1
3.11 - 3 . 1 1
3.11
3.11
3.11
3.10
3.11
X X I . Evaporated off the water from 25 cc. of the dilute H,SO,, added the soda-saltpeter mixture and fused as usual. Sulphur
. . . . . . . . . . . 3.12
3.15
3.12
X X I I . Like X X I , but added CaCI, t o the solution of the melt before precipitating BaSO, and treated the latter according t o van’t Kruys. Sulphur.
.....
3.22
3.18
3.19
X X I I I . Like X X I I , t did .not digest the precipitated BaSO,, nor evaporate the filtrate to recover traces of dissolved BaSO,. Sulphur ........... 3 . 1 3
3.12
3.11
XXIV. Like X X I , but did not fuse. was acidified with HC1. Sulphur........... 3 . 1 7
3.19
The solution
3.16
XXV. Added 250 cc. H,O and ten cc. concentrated HC1 t o twenty-five cc. dilute H,SO,, added BaCl,, digested two hours, poured off the supernatant liquid, digested the BaSO, with one cc. ten per cent. BaC1, and fifteen cc. HCI ( I : I ) . Diluted, filtered, evaporated the combined filtrates to dryness in platinum, took u p “with fifty cc. of slightly acidified H,O and collected the slight residue on the game filter. Sulphur . . . . . . . . . . . 3 . 1 2
3.10
3.11
XXVI. Exactly neutralized NaOH solution with the dilute H,SO,, using phenolphthalein as indicator. After each addition of acid the solution was heated, in a platinum dish, until the pink color no longer reappeared ; then evaporated t o dryness, ignited gently and weighed as Na,SO,. HnS04 used (cc.), . . .................... 7 5 . 5 6 75.55 N o 8 O 4 found (gm BaSOr equivalent t o NagiO4 (gm.).. . . . . . . . . . . . . . . BaS04 equivalent to 25 cc. HzS04 (pm.)........... Sulphur (calc. on 0.5 gm.rubber)
.........
0.2082 0.34197 0.1131 3.11
0.2080 0.34164 0,1130 3.11
XXVII. The same as XXVI, but used NaHCO, ins t e d o f NaOH.
.
HSO, used (cc.).
........................
0.1240
0.1194
I n order t o test the completeness of the oxidation of sulphur b y means of the nitric acid-bromine mixture, the following determinations were carried out. XXVIII. Powdered sulphur crystals, digested in t h e cold with twenty cc. HNO, and a n excess of Br. Finally added twenty cc. H,O and heated on the steamb a t h for about two hours. Then evaporated nearly t o dryness, took u p with giater and precipitated with BaCl,.
................. 0.0483 .................0.0481
Sulphur taken (gm.). Sulphur found (gm.).
.
0.0395 0.0399
0.0561 0.0563
X X I X . Powdered sulphur crystals treated a t the same time as some of the samples of rubber. The exact methods are referred t o in the table, the Roman numerals indicating the method employed. Method Sulphur taken (gm.)
I.
.... 0.0528 0.0485 Sulphur (per cent.)..... 9 1 . 8 4
11. 0.0619 0.0590 95.33
111.
IV.
0.0494 0.0479 96.91
0.0646 0.0596 92.13
V. 0.0411 0.0414 100.67
Oct.,
1911
I n the determinations by methods I to IV, part of the sulphur was not attacked b y the nitric acid nor b y the sodium carbonate added before making the fusion. Part, a t least, of this unattacked sulphur was seen t o b u m when the fusion was made. As stated above ( X I I ) , the barium sulphate precipitates representing soluble sulphates, etc., in the Hubener method, were found t o contain lead. Lead sulphate dissolves slightly and is, besides, partially decomposed b y water, hydrobromic and nitric acids, etc.1 In order t o get a n idea of the amount of barium sulphate to be expected to result from the decomposition and solution of lead sulphate under the conditions of Hubener’s method, some determinations were made. Lead sulphate was first prepared b y precipitation from a hot, dilute, nitric acid solution of lead nitrate b y means of a hot, dilute solution of sulphuric acid. I t settled rapidly as a coarse-grained powder, which was washed b y decantation with hot water, then in a Gooch crucible with hot water, followed by strong alcohol. I t was then dried in an air-bath. In the first experiments it was treated with hot water and the amount of barium sulphate precipitated from the filtrate was calculated as percentage of sulphur in 0 . 5 0 gram rubber. X X X . Washed 0.200-gram portions of PbSO, on filters. Each time 250 cc. hot water were used. The filtrates were slightly acidified with HC1 and precipitated with BaCl,. Sulphur.. . . . . . . . . . . 0 . 2 4
02 5
These precipitates contained only traces of lead. X X X I . Treated 0.200-gram portions of PbSO, according to Hubener’s method, slightly modified. Treated with thirteen cc. concentrated HNO,, evaporated practically to dryness on the steam-bath, added fifty cc. H,O, 0 . 5 cc. Br and two cc. of dilute HNO, ( I : 4 ) . Heated, filtered, and washed with about 200 cc. hot water. .Then precipitated with BaCl,. Sulphur.. . . . . . . . . . . 0 . 7 3
0.80
0.57
These precipitates contained a little lead. From these determinations i t seems quite certain t h a t the larger part of the sulphur found as soluble sulphate under XI1 must have come from the solution and decomposition of lead sulphate first formed when the rubber was attacked b y nitric acid. At the suggestion of Dr. Hillebrand, four determinations were made of the amount of lead carried down with the barium sulphate precipitated in the usual way from the aqueous extract of the fusion mass. I n spite of the presence of a large excess of sodium carbonate some lead goes into solution. The preliminary treatment was according t o XVIII, and two grams of rubber, instead of 0 . 5 0 gram, were taken each time. X X X I I . After fusion the melts were dissolved in water. To each of the first two there +as added two grams of sodium bicarbonate in order t o decompose a n y alkali plumbate. The solutions were heated on the steam-bath for one and one-half hours and then filtered from the insoluble. After acidifying with 1
Kolb, Din& fioly. J , ,209, 268; Ditte, Ann. chim. fihys., [51 14, 190.
.
Oct., 1 9 1 1
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
hydrochloric acid, barium sulphate was precipitated a s usual. 1. BaSOl found ( g n ~ ). .. . . . . . . . . . . . 0,5379 Sulphur (per cent.),. . . . . . . . . . . . 3.69
.
2. 0,5395 3.71
3. 0.5437 3.73
4. 0.5417 3.72
The barium sulphate precipitates were then mixed with soda and potash and fused. The melts were dissolved in water, filtered a n d the residues washed with hot, very dilute sodium carbonate solution. The residues of barium carbonate and lead oxide were then dissolved in dilute nitric acid and the lead precipitated from the cold solutions b y hydrogen sulphide. After standing over night in stoppered flasks, the precipitates of lead sulphide were filtered off washed, dissolvgd in nitric acid and finally converted into sulphate b y evaporating down in porcelain crucibles with sulphuric acid and gently igniting. 1. PbSO, found (gm.).. . . . . . . . . . . . 0.0086 Equivalent to BaSO, ( g m . ) . . . . . . 0.0066 Corrected BaSOl (grn.). . r . . . . . . . 0.5359 Corrected sulphur (per cent.). . . . 3.68
2. 0.0071 0.0055 0.5379 3.69
3. 0.0040 0.0031 0.5428 3.73
4.
0.0045 0.0035 0.5407 3.71
It is quite evident from these figures t h a t although notable quantities of lead sulphate are carried down with the barium sulphate, the correction in the percentage of sulphur is negligible. The filtrates from the original precipitates were treated with hydrogen sulphide and gave slight precipitates. The alkaline filtrates from the barium carbonate and lead oxide were tested with ammonium sulphide and became brown. The next day there was a slight film of a dark color on the bottom of each of the beakers in which these solutions were tested. This was probably a mixture of small amounts of lead and iron sulphides. I n all the solutions tested, as well as in the actual determinations of lead sulphate, greater amounts of lead were found in I and 2 , which had been treated with bicarbonate. Apparently at the temperature of the steam-bath the lead bicarbonate probably formed was not decomposed. CONCLUSIOXS.
Treatment of the rubber with nitric acid alone gives low results (compare XV with X V I I I ) . This is probably largely due t o loss of free sulphur, since nitric acid alone does not completely oxidize sulphur t o sulphuric acid in the length of time ordinarily taken for a determination. The Hubener met,hod cannot be employed in the presence of mineral fillers which tend t o form insoluble sulphates. This applies especially to barium carbonate and litharge. A comparison of X X t o X X V I I shows that the fusion method gives results very close to those obtained b y direct precipitation and b y neutralization. The van't Kruys method gives high results. The best results seem t o be obtained b y the use of method XVIII, according to which the rubber is decomposed b y means of nitric acid saturated with bromine. BUREAUOF STANDARDS,
August, 1911.
737
CONSISTENCY OF PAINTS BY THE STORNER VISCOSIMETER. B y ALLEXROGERSAND A. H SABIN.
Received June 9, 1911.
Those who have ever attempted to take the viscosity of a paint will appreciate that it is no easy matter. At a meeting in Atlantic City, last June, i t was suggested b y Mr. P. H. Walker and Prof. A. H. Sabin that the Stormer viscosimeter might be used for this purpose, especially the apparatus, as modified, a t the suggestion of M r . C. N . Forrest, from the paddle t o the cylinder type. I t was the writers' desire t o obtain some method for checking the exact consistency of a large number of paints, and do it in such a manner that the personal factor would be eliminated. A sample of paint was, therefore, prepared which, in the writers' judgment, was of the proper consistency for application. The viscosimeter was adjusted with this paint so t h a t the dial made a complete revolution in two minutes. A sample of paint made b y the Patton Paint Co. came very close t o the paint made b y the writer, as well as one made by the Sherwin-Williams Co. To get a more definite standard, however, 95 per cent. glycerine was tested and, strange t o relate, the dial made a complete revolution in exactly two minutes. Therefore, glycerine, a t 2 o o C . , was taken as the standard, and all subsequent tests made a t this temperature.. During the past two months several hundred viscosities have been taken, and although the machine has been tested each morning before the day's work it has not been found necessary to readjust the weights. The apparatus, however, requires quite a bit of attention and should be frequently cleaned and oiled. I t is very essential also t o have the temperature exact, as one degree will make quite a difference in the reading. The viscosimeter consists of a flat-bottomed cylindrical cup I'/* inches internal diameter b y z T / Z inches deep having two radial vertical wings or septa I/, inch wide attached t o its inner wall, and extending from the bottom t o within inch of the top, being in fact obstructions t o the rotary flow of the liquid. Within this cup is suspended a stirrer, the revolution of which is resisted b y the viscosity of the liquid; this stirrer is a hollow metal cylinder 13/, inches long and I I / ~inches in diameter, its lower end open, the upper end closed except it contains four holes, each inch in diameter, symmetrically arranged. This cylinder is supported from above by a vertical rod or shaft, concentric with the cylinder but external to i t , being 3 / 1 6 inch in diameter by z l / ( inches long. The end of this is attached b y a collar and screw t o the bottom of a vertical shaft which carries a pinion 1/2 inch in diameter, having 2 5 teeth, which in turn is driven by a gear 5 1 / ~inches in diameter with 2 7 j teeth, t o which large gear is attached a drum I I / ~ inches in diameter, around which is w o m d a small silk cord, the external portion of which passes over a small pulley and is attached t o a driving weight. The top of the shaft carrying the small gear carries ,also a screw or worm which engages and drives the
