July 15, 1941
ANALYTICAL EDITION
procedure. The quantity of sodium sulfate used was varied from about 0.05 to 0.25 gram, and the ratio of sodium sulfate to sodium chromate was varied from 2.5:l to 1:75. Table I gives data selected a t random from the results obtained. Columns 1 and 2 give, respectively, the weights of sodium sulfate and of sodium chromate taken. Column 3 gives the weight of the precipitate of mixed barium sulfate and barium chromate obtained. Column 4 gives the weight of sodium sulfate found, calculated on the assumption that all the precipitate is barium sulfate. This quantity, by comparison with the corresponding quantity in column 1, is considerably in error. Column 5 gives the weight of barium chromate in the precipitate, determined as described above. Column 6 gives the weight of sodium sulfate found after the correction has been applied, and column 7 gives the difference between this quantity and the weight of sodium sulfate taken.
457
It is evident from these results t h a t this method of determining sulfate adequately meets the ordinary analytical requirements. Literature Cited (1) Cadbury, W. E., Jr., Meldrum, W. B., and Lucasse,
W. W.,
(2)
33, 847
forthcoming publication. Richards, T. W., and Kelley, G. L., J . Am. Chem. Soc., (1911).
(3)
Willard, H. H., and Schneidewind, R., Trans. A m . Electrochem. Soc., 56, 333 (1929).
from p a r t of a dissertation presented by 'CITilliam E. Cadbury, Jr., t o the faculty of the Graduate School of the University of Pennsylvania, in partial fulfillment of the requirements for the degree of doctor of philosophy, June, 1940. ABsTRaCTED
Platinized Silica Gel as a Catalyst in Gas Analysis KENNETH A. KOBE AND RAY A. MAcDONALD University of Washington, Seattle, Wash.
Previous work showed that hydrogen, carbon monoxide, and hydrocarbons can be oxidized quantitatively over a platinized silica gel catalyst containing 0.075 per cent of platinum. The commercial catalyst now available contains 0.125 per cent of platinum and is considerably more reactive. The changes in technique from the use of the copper oxide tube give more rapid results, as the catalyst tube replaces the copper oxide tube and slow combustion pipet. Hydrogen is oxidized at 115" C., methane and other hydrocarbons at 510" C. Oxidation temperatures are determined for carbon monoxide, ethylene, acetylene, ethane, and propane. Ethylene may be hydrogenated over this catalyst at 375' C. Nitrous oxide may be determined by reduction with hydrogen at 515" C.
sheet-iron shell. The two ends, D, are of 0.94-cm. (0.375-inch) Transite, held in place by the straps, E. The technique of the copper oxide tube (6) is to cool the tube to room temperature after each analysis. This cannot be done when a catalyst tube is used, as the platinized silica gel will adsorb an appreciable quantity of gases at room temperature. This error is avoided and the time required for an analysis is shortened by maintaining the tube a t the temperature of analysis and adjusting the pressure in the tube to the standard pressure after each analysis. It is advisable to keep the catalyst tube
0
T
HE previous papers in this series (2, 3, 4) have shown
t h a t hydrogen, carbon monoxide, and hydrocarbons can be quantitatively oxidized by air or oxygen over a platinized silica gel catalyst and have given the conditions necessary for the oxidation. The catalyst used in the previous work, the commercial platinum catalyst of the Silica Gel Corporation, containing 0.075 per cent of platinum, has now been replaced by a more active catalyst containing 0.125 per cent of platinum. This paper states the conditions necessary for its use to replace the copper oxide tube and combustion pipet in the determination of hydrogen, carbon monoxide, hydrocarbons, and nitrous oxide.
Apparatus The U. S. Steel Corporation gas analysis apparatus used in the previous work was modified by replacing the usual heater for the copper oxide tube with one made from the heating element of a volatile matter furnace (Figure 1). The heating element, B , is surrounded by magnesite insulation, C, within a
I
FIGUFLE 1. DIAGRAM OF APPARATUS
INDUSTRIAL AND ENGINEERING CHEMISTRY
4sa
TABLE I. OXIDATION OF METHANE, ETHANE, AND PROPANE Analysis Temperature, ' C. S o . of passes Rate, nil. per min. CHI by explosion, 70 CHI by catalyst, 70
C3H& % Temperature, C. No. of passes Rate, ml. per min. Sample, ml. Contraction, ml. CO?, ml.
1 2 3 Oxidation of Methane 497 510 510 9 4 6 40 40 40 11.2 7.2 7.2 10.8 6.7 7.3 Oxidation of Ethane 12.9 iio 224 4 40 40 40 92.6 90.4 89.3 1.4 0.9 0.6 1.0 0.2 0.0 Oxidation of Propane 5.5 235 iig iio 6 8 8 30 30 30 99.2 99.2 98.4 0.0 1.4 1.0 0 2 0.0 0.2
4
6
6
510 9 40 7.2 7.3
510 6 40 7.2 7.2
510 6 40 7.2 7.2
i66
ii5
i45
40 88.4 0.1 0.0
40
4
40
88.3
88.7
*.
.... ..
ii5
8 30 95.6 0.0 0.0
0.0 0.0
.. ..
0.6 0.0
open to the buret during heating and cooling, as a large change in temperature (as from room temperature to 510" C.) greatly changes the pressure in the tube.
the minimum for ethane and propane, so that it cannot be determined by oxidation when these hydrocarbons are present in the gas mixture. Seattle city gas 11-asanalyzed, using three different methods for oxidation. The results (Table 111) show that the catalyst tube gives as exact data as the usual combustion methods. With samples 1 and 2, the hydrogen was removed at 115" C. by five double passes at 40 ml. per minute, and then the hydrocarbons were determined by explosion. With samples 3a and 3b, hydrogen was removed as before, and then the hydrocarbons were determined by six passes through the catalyst tube a t 510". With samples 4a and 4b, hydrogen and hydrocarbons were determined together by six passes at 511" and calculated to hydrogen and methane, ahich is slightly in error as some ethane is present in Seattle city gas.
TABLE 11.
Commercial Gas Hydrogen is oxidized by passing the gas with air over the catalyst a t 100" to 110" C. Lovier temperatures are not advisable, as the n-ater formed should not be condensed in the catalyst tube. Carbon monoside is osidized by passing the gas over the catalyst a t a temperature above 210" C. (Table 11); below this the catalyst is poisoned. This temperature is 90" below that found for the 0.075 per cent platinum catalyst. The temperature for the osidation of carbon monoxide is above
OXID.4TION O F
CARBOX MONOXIDE
(Gas mixture 12.7% CO, 18.9% 02) Analysis 1 2 3 4 Temperature, O C. 320 276 231 210 6 6 6 KO.of passes 6 40 40 40 40 Rate, ml. per min. 12 7 12 7 12.6 cos % 12 6
Methane, Ethane, and Propane The lowest temperature a t which methane could be oxidized completely was first determined. Methane is more resistant to oxidation than the other hydrocarbons, so its oxidation temperature was determined exactly, for here ethane, propane, and higher hydrocarbons will be completely oxidized. The temperature at which oxidation of hydrocarbons begins is of importance, as i t shows the temperature below which hydrogen must be oxidized to avoid simultaneous hydrocarbon oxidation. Mixtures of the hydrocarbon, oxygen, and nitrogen were made and analyzed by the explosion method, and this gas mixture was then used in the catalyst tube. Table I shows the results. The minimum temperature for the oxidation of methane is 510" C. when six double passes are made at the rate of 40 ml. per minute, 100" lower than when a 0.075 per cent platinum catalyst is used. Higher temperatures accelerate the rate of oxidation and fewer passes need to be made. Branham and Shepherd (1) have pointed out discrepancies in the explosion method, particularly for ethane. However, the explosion method for methane and nitrous oxide gave concorda n t results which checked those obtained by the catalyst tube method. The lowest temperatures a t which oxidation of ethane and propane occurred were determined. Table I shows that ethane is not oxidized belolv 165" nor propane below 150" C. The contraction is greater than should occur with oxidation, or x h e n no carbon dioxide is formed, and is accounted for by adsorption of the hydrocarbon on the catalyst. When the temperature was lowered from 165" to 145" C. a n apparent adsorption of 0.6 ml. of ethane occurred, which is much higher than froin any ordinary gas mixture because of the 12.9 per cent of ethane in the gas.
Vol. 13, No. 7
TABLE
Sample COZ Ill. 0 2
CO H? CH4 C2H6
h-?
1 5.4 5,7 1.0 14.4 41.7 l5,7 0.7 15.4
-
6
194 6 40 7 4
h A L Y S 1 S OF SEATTLE C I T Y CiiAS
2 5,3
5.5
1.1 14.2 41,s 16,s 0.G
-
Total 1 0 0 . 0
111.
5 205 6 40 12 7
15.4
_
100.0
_
3s 5.3 5.6 1.0 14.2 41.5 15.8 0.9 l5,2
-
100.0
-
3b 5.3 5.8 1.0 14.2 41.9 16.0 0.7 15.1
4a 5.5 5.7 1.1 14.1 41.9 16.8
4b 5.5 5.7 1.1 14.1 41.7 17.0
-
14.9
14.9
100.0
100.0
100.0
...
...
Oxidation of Ethylene and Acetylene Ethylene and acetylene may be oxidized quantitatively with a limited excess of oxygen. Known mixtures of hydrocarbon, oxygen, and nitrogen mere made. Ethylene was determined by absorption, acetylene by explosion, The results with the catalyst tube are shown in Table IT'. Ethylene may be oxidized quantitatively above 310", and acetylene above 325". Below these temperatures the catalyst is poisoned by the unsaturated hydrocarbon. Whenever the catalyst is poisoned by ethylene, acetylene, or carbon monoxide, i t may be reactivated by heating the tube to about 75" above the minimum oxidation temperature and passing air or oxygen through the tube three or four times.
Hydrogenation of Ethylene I n order to determine the possibility of determining unsaturation by hydrogenation, the activity of the platinized silica gel as a hydrogenation catalyst was studied for ethylene (Table V). In oxidizing ethylene, the catalyst Tya; poisoned
OF ETHI-LESE AKD ACETI-LESE TABLE IV. OXIDATION c
Analysis Temperature, C. TemDerature, C. S o . if of passes Rate mi. per min. C~H4'byabsorption, C2H4 by catalyst, 5
1 450
5
Temperature, ' C. S o . of passes Rate, mi. per min. C2H2 by absorption, ' c C2Hz b y catalgst, ( >
4 40 11.0
11 0 305 4 40 i7 . S9
2 400 4
Ethylene-3 31'6
4
310
5 2SS
4 40
11.0 11.1
11.0
11.0
11.0 11 0
4 40 11.0 5.7
338
323 4
305 4 40
305
40
7.9
67 . 9 S
7 . Y5
40
4 40
87 . 90
4 -10
S 40
July 15, 1941
ANALYTICAL EDITION
TABLEV.
TABLEVI.
mined by making six double passes a t 515' or ahove. Analysis of a cylinder of nitrous oxide for anesthesia showed 99.7 per cent purity.
HYDROGENATION OF ETHYLENE
(Gas mixture: CzH, 16.5,Hz 83.5 per cent) Analysis 1 2 3 4 5 375 318 318 286 286 Temperature, C. So. of passes 4 4 8 4 7 40 40 40 40 40 Rate, ml. per min. C2Hd b y hydrogenation, 70 16.5 16.0 16.5 15,s 16.3
6 286 10 40 16.5
REDUCTION OF KITROUS OXIDE
(Gas mixture: hydrogen, nitrous oxide b y explosion 33 7, 33 8 % ) Analysis 1 2 3 4 5 6 491 460 518 515 491 Temperature, C. 518 S o . of passes 6 12 6 4 8 4 40 40 40 40 40 R a t e , mi. per min. 40 33.6 30 8 33 8 33 2 33 7 S?O, 5 33.7
below 310". I n hydrogenating ethylene, the reaction proceeds so rapidly a t first that the remaining ethylene does not poison the catalyst, so that hydrogenation can be effected a s low as 234', although the reaction becomes extremely slow.
Nitrous Oxide
+
459
+
The reaction N20 Hz + Ifz H20 can be carried out by explosion technique. Kitrous oxide was mixed with hydrogen and analyzed by explosion. The gas mixture was passed through the catalyst tube and the oxidation temperature determined (Table VI). Nitrous oxide may be deter-
Summary
A platinized silica gel catalyst containing 0.125 per cent of platinum lowers the oxidation temperature of methane and carbon monoxide by approximately 100' C. from that obtained with a 0.075 per cent catalyst. Temperatures are given for the complete oxidation of carbon monoxide, methane, ethylene, and acetylene and for the start of oxidation for ethane and propane. Nitrous oxide may be reduced over the catalyst by a limited excess of hydrogen a t 515" C. The catalyst tube can replace the copper oxide tube and explosion (or slow-combustion) pipet in any commercial gas analysis apparatus. Large samples may be used r i t h o u t danger of explosion. Literature Cited (1) Branham and Shepherd, J . Reseaych Natl. Bur. Stardards, 13, 377-89 (1934). (2) Kobe and Arveson, IND.ENG.C m h f . , Anal. Ed., 5 , 110-12 (1933). (3) Kobe and Barnet, Ibid., 10, 139-40 (1938). (4) Kobe.and Brookba:!, I b i d . , 6, 35-7 (1934).
(5) U. S. Steel Corp., Methods for the Sampling and Analysis of Gases", p. 14, Pittsburgh, Carnegie Steel Co., 19%:. PRESENTED before the Division of Gas and Fuel C h e m k t r y at the 100th Meeting of the American Chemical Society, Detroit, N c h .
Detection of Certain Metals in Minerals and Ores An Ammonium Hypophosphite Fusion Method IT. B.
VAN
TALKENBURGH AND T. C. CRAWFORD, Uni\ersity of Colorado, Boulder, Colo.
HEN fused with ammonium hypophosphite many minerals and ores are decomposed, the resulting melt often being highly colored. The color of t'he melt directly, or after being treated with water or hydrogen peroxide, is the basis of the tests described here for chromium, cobalt, columbium, manganese, molybdenum, tellurium, titanium, uranium, vanadium, and tungsten. Hypophosphites are powerful reducing agents. They decompose upon being heated, giving off hydrogen and phosphine, which ignite. Ammonium hypophosphite decomposes as follows ( I ) : 7NH,H,P02
+2HPOS + H4Pz0~+ 7KH3 +
3PH3
+ 2H2 + H20
The clear melt that is obtained serves as an excellent medium for shoviing any color imparted bo it by the reduced mineral. The fused mass has a low melting point, about 60" C., and is readilysoluble in water. Sodium and potassium liypophosphites are not suit'able for this fusion because upon decomposition they form salts with relatively high melting points, instead of the low melting acids that are obtained from the ammonium salt.
Method About 0.1 gram of the finely powdered mineral is strongly heated in a small evaporating dish n-ith 2 grams of ammonium hypophosphite. The hypophosphite soon begins to decompose nnd the gases evolved ignite, forming n-ater and oxides of phosphorus. -4fter 2 minutes a quiet fusion mixture is obtained n-hich is uied for t h e individual tests.
Detection of Metals COBALT,TITASIUX, ASD TUSGSTEK..These three metals give blue melts, but t'he cobalt melt turns pink upon cooling. For the detection of tungsten, enough water is added dropwise to the warm melt to keep the surface moist. .Is the wat'er penetrates the melt, t,he blue color changes to a striking violet. The violet, color appears immediately when appreciable amounts of tungsten are present, but 10 to 30 minutes may be required when only very small amounts are present. If the blue color of the melt is due t o titanium, the water above the melt becomes a delicat'e ant1 almost imperceptible rose color. The addition of hydrogen peroxide gives an intense orange-red color. Ammonia cause3 the rose color to change t o blue, but this change is not, 30 sensitive as the color change with hydrogen peroxide. Vanadium also gives a reddish color witli hydrogen peroxide, but the \-anadium melt is red when hot and green vhen cool: hence it is easy to dist'inguish between titanium and vanadium. Since cobalt, titanium, and tungsten are usually not associated together in minerals, these tests are specific for theqe three metals. T - ~ N A D I u I I , C H R ~ M I T X , 4 S D ~ R A K I U I I . The>e t h e e metals impart a green color t o t'he melt. T-anadium gires a reddish color t o the melt n-hen hot, which gradually changes t o yellow and finally to green on cooling. The addition of wat'er gives a pale green solution n-hich turns pink when ligdrogen peroxide is added. The addit'ion of ammonium carbonate soluhion to the green melt until the solution is dis-
