Vol. 21. x o . 11
INDUSTRIAL A X D ENGINEERING CHEMISTRY
1030
by the dryness and hardness of the stock. I n spite of a particle size coarser than ordinary zinc oxide, the stock containing this sample of magnesium carbonate was decidedly insoluble as had been predicted on a basis of its poor wetting. In another test two lithopones were compared, the average particle size of one being approximately 0.3 to 0.35 micron and of the other approximately 0.4 to 0.5 micron, with aggregates considered as individual particles. These were compared for solubility in 20-volume stocks (20 volumes of pigment to 100 volumes of rubber). As was expected, the compound containing the finer lithopone required much longer (about three times) to go into solution than the one containing the coarser material. These two lithopones are
approximately representative of extremes in the particlesize range of commercial lithopones. A further consideration of the dispersion of pigments in unvulcanized rubber, for which the data are still very incomplete, would be helpful in bringing about a better understanding of the working qualities of factory stocks. Literature Cited (1) (2) (3) (4) (5) (6)
Ashman and Homewood, Indie Rubber U'orrd, 77, 59 (1928) Goodwin and Park, I N DEND.CHEJI.,20, 621 (1928). Depew, Rubber Age, 24, 378(1929). Grenquist, I N D . END.CHEM.,20, 1073 (1928). Meyer and Mark. Ber., 61B,1939 (1928). Stevens, J . SOC.Ckem. I n d . , 47, 37T (1928).
Explosive Limits of Industrial Gases' Jesse Yeaw ROCHESTER GAS & ELECTRIC CORPORATIOS, ROCHESTER, N.1'.
The relation between the composition and the exover, the results o b t a i n e d HROUGHOUT various plosive limits of widely varying complex inflammable must be directly applicable s t a g e s i n the manugas mixtures, generally encountered in the manuto large volumes. facture of city gas, sevfacture of city gas, has been shown. Coward and Jones (2) have era1 complex mixtures of inThe range of observation covers gas mixtures conflammable gases are generally found that, almost without taining from less than 12 per cent to more than 96 exception, a change of condie n c o u n t e r e d . It was beper cent combustible. The inflammable constituents tions which affects the exlieved that more information include varying amounts of hydrogen, carbon monplosive limits of one gas-air concerning the relation beoxide, methane, and other hydrocarbons. mixture will change the extween the composition and The effect on the explosive limits of changes in the plosive limits of any gas-air the explosive limits of these percentages of illuminants, methane, inert gases, and mixture in the same direcgases with air would prove the ratio of hydrogen to carbon monoxide has been tion, but not necessarily to useful in detecting explosive the same degree. They have pointed out and discussed. mixtures and eliminating unThe explosive limits of all the gases have been calshown that small changes in necessary hazards. culated from the analyses by means of a simple algetemperature and p r e s s u r e T h e chief constituents of braic formula first presented by LeChatelier and later have slight effects, that water these gases from the point of vapor affects the limits to a amplified and developed by Coward, Jones, and others. view of abundance are hydrovery small degree, that glass These calculated values agree with those found experigen, carbon monoxide, and bulbs 2 inches (5 cm.) or mentally remarkably well, and not only show that m e t h a n e . Ethylene, benm o r e in diameter give rethis work checks that of others in the field, but also zene, toluene, and a few other s u l t s a p p l i c a b l e t o large give further proof of the value of such a means of deh i g h e r hydrocarbons genervolumes, and that a strong termining dangerous conditions, particularly those ally c l a s s e d t o g e t h e r a s electric spark a t the bottom which abound in the production and distribution of i l l u m i n a n t s are found in center is a suitable method of smaller amounts in certain of city gas. ignition. the gases. Most of the work The temperatures of the gases under consideration are on explosive-limit mixtures of pure gases and air has been confined to the first three gases, and the limits of these gases are generally less than 100' C. except during the short period in which they are being produced in their respective ovens or fixed within a comparatively small range of accuracy. generators. Pressures are always as small as possible and T a b l e I-Widest Explosive L i m i t s Reported i n the Literature ( 3 ) rarely exceed a few centimeters of water. These gases all contain water vapor to very near the saturation point. PracGAS OR VAPOR n w r%s CENT GASOR VAPOR tically all explosive-limit work, however, has been done with Ethylene 3.2 to34.0 4 . 1 5 to 7 5 . 0 Hydrogen , water-saturated gases. Propylene 2.2 to 9.7 1 2 . 5 to 75 .O Carbon monoxide Acetylene 1.5 to 80.5 4 . 9 to 15.4 Methane Apparatus Benzene 1 . 4 t o 8.0 2 . 5 to 15.0 Ethane Toluene 1.3to 6.8 2 . 2 to 7.3 ProDane A glass bulb of about 350 cc. capacity (8.9 cm. in diameter) was used as the explosion chamber. Mixtures of as near 200 Experimental Conditions cc. as practicable were introduced from a measuring buret, The object of the experimental work was to obtain the and atmospheric pressure was obtained by means of a leveling widest explosive limits or, in other words, the least amount of bulb using mercury as the fluid. Ignition was accomplished air or oxygen (upper limit) or the least amount of gas (lower by means of an electric discharge between platinum points limit) which could make a gas-air mixture dangerous. More- placed about 1 cm. above the surface of the mercury. A specially constructed buret of 200 cc. capacity, entirely en1 Received August 10, 1929. Presented before the Division of Gas closed in a large water jacket to maintain conatan&temperaand Fuel Chemistry at the 78th Meeting of the American Chemical Society, ture, was used to measure the gases. Minneapolis. Minn., September 9 to 13, 1929.
T
v
I
I
pz: g::T
I N D USTRIrl L A N D EXGINEERING CHEMISTRY
Xovember, 1929 Results
Table I1 gives the average analysis and the average of the determined explosive limits for samples of six complex mixtures handled by the Rochester Gas & Electric Corporation. These samples were taken a t various times over a period of about 6 months and the average deviation from the given mean for each constituent was found to range from a few tenths of 1 per cent to about 2 per cent depending on the type of gas analyzed. The average deviation in the explosive limits was 0.2 per cent a t the lower limit and 0.7 per cent a t the upper limit. Table 11-Analvses
a n d Exolosive Limits of Various Gas Mixtures
LIMITS
ANALYSIS
(GAS I N AIR)
GAS
CO, Ill.
I1 Oven Coal Mixed Carbureted Producer
Oz
CO
Hz CH4 Kz
B. 1. u / cu.ft. 631 4 1 . 9 3 . 9 0 . 4 6 . 3 5 4 . 4 31.5 1 . 6 2 . 4 2 . 7 0 . 4 9 . 7 4 4 . 9 2 8 . 9 1 1 . 0 549 I 2 . 5 3 . 2 0 . 5 1 0 . f J 4 7 . 0 2 5 . 8 1 0 . 5 540
70
%
70
R
5;
76
70
4.6 7.3 0.3 36.0 37.0 9.6 5 . 2 6.20.00.339249.0 2.3 3.0 G.2 0 . 0 0 . 0 2 7 . 3 1 2 . 4 0 . 7 5 3 . 4
509 310 136
R
1031
LeChatelier’s rule has been shown by Coward and others ( I ) to hold very accurately for both lower- and upper-limit mixtures of pure gases in air, and approximately for other complex mixtures such as coal gas. The presence of carbon dioxide and nitrogen in such complex substances as illuminating gas interferes somewhat with the use of a formula developed for pure gases. This is particularly true in the case of such mixtures as producer gas and the like, which contain a high percentage of inerts. Jones (4) has developed and graphed the relationship between the inflammable limits of varying mixtures of hydrogen, carbon monoxide, and methane with carbon dioxide and with nitrogen, respectively; and, using these relationships and LeChatelier’s rule, he has given a new method for the more exact calculation of the explosive limits of complex mixtures containing any amount of carbon dioxide or nitrogen or both.
7c 5
5 . 0 28.4 5 . 6 30.: 5 . 6 31.,
6
8
6 . 4 37.7 6.9 69.5 20.7 73.7
9 :
10 I,
12;
In general these six gases can be classed, on the basis of explosive limits, into three groups: (1) those which contain illuminants and have a high percentage of combustible; (2) those which do not contain illuminants but have a high percentage of combustible; (3) those which do not contain illuminants and have a low percentage of combustible. Even a small percentage of illuminants reduces the upper limit from about 70 per cent to about 30 per cent. The lower limit is, however, very slightly affected by the presence of 7 to 8 per cent of this constituent, owing to the amount of air or oxygen required for combustion. At the upper limit the oxygen is limited, but a t the lower limit there is a large excess. Since the upper limits of hydrogen and carbon monoxide are identical, great changes in the ratio of hydrogen to carbon monoxide would not be expected to affectthe upper limit, but changes in the percentage of methane, whose upper limit is only 15 per cent, tend to lower the upper limit of the mixture, as is seen when oven gas is compared with the others. Blue gas, which has no illuminants but has a high percentage of combustible gas, has the widest range of explosibility. When the percentage of inerts is raised, the amount of combustible a t the lotver limit is reduced, as in producer gas, and the lower limit would be expected to rise. At the upper limit, however, the inerts probably tend to dilute the mixture arid promote a more Gomplete combustion, thus keeping the upper limit high. Calculation of Explosive Limits
From theoretical considerations of LeChatelier (6),a simple formula has been developed (1) for the calculation of the explosive limits of combustible gas mixtures with air. If a, b, c are the relative proportions of its components and A , B, C are their respective explosive limits, then L, the limit for the mixed gases, is given by the equation: a
b
-
11 I4
0 16
7
0 9
20
Figure 1-Inflammability of Hydrogen Carbon Monoxide a n d Methane Containing Various A m o u n t s bf Carbon Dioxidd and Nitrogen (From Jones, IND. ENG CHEM.,20, 901 (1928)).
The results of Jones’s curves are shown in Figure 1. For example, take producer gas whose upper limit could not be calculated by means ’of the original formula. Following Jones’s method we will combine the carbon dioxide with the carbon monoxide and the nitrogen with the hydrogen. The total gives the original gas. (1) 27.37, CO (2) 12.4% Hz (3) 0.7% CHI
P n cent 33.5 65.8 0.7
6.27, COz ++53.47, N 2
-_
Total
100.0
The volume of inerts per volume of combustible for these three mixtures are as follows: TOTAL FROM
ABOVE
RATIO OF INERT LIMITS OF INFLAMMABILITY COXBCSTIBLE FROM FIGURE 1 Lower Upper
TO
Per cent
Per cent 17.0 23.0 5.5
0.2 4.3 0.0
Per cent 70.0 76.0 14.5
Substituting these values in the formula, we have: Lower limit =
c
cy+A ‘ B
100
33.5 I 65.8 17.0 23.0
C
In brief, the formula implies that, if a liiiiit mixi.ure with air of one inflammable gas is mixed in any proportic with a limit mixture with air of another inflammable gaf j.l;mit mixture results. From this fact the conclusion may 01E sly be drawn that any mixture, containing any one or all of the gases given in Table I in any proportion, cannot explode if the percentage of air is less than 19.5 (4.1 per cent oxygen) or more than 98.7 (20.7 per cent oxygen).
2
5
Upper limit =
33.5
+
100 65 8
0.7 5.5 0.7
= 20.2 per
cent
= 71.8 per cent
70.0+%0+14.5 Table I11 gives the average determined explosive limits of the six gases in Table I1 together with the values calculated by means of Jones’s data and LeChatelier’s formula. The explosive limits of the illuminants were taken as 2 and 8 per cent for the purposes of calculation. The small amount of oxygen was neglected. That the calculated upper limits of
INDUSTRIAL AND ENGINEERING CHEMISTRY
1032
these gases are not closer is due to thc fact that the upper limits of carbon monoxide and hydrogen are somewhat lower than those reported in the other work by Jones (6). This difference would tend to shift four of the curves to a somewhat higher position on the graph and raise all of the calculated upper limits, particularly those consisting mainly of hydrogen and carbon monoxide, such as carbureted blue gas, blue gas, and producer gas. Table 111-Determined
GAS
I I
Oven Coal Mixed Carbureted blue Blue Producer
a n d Calculated Explosive L i m i t s of G a s Mixtures LOWER LIMIT Detd. Calcd.
P e r cent 5.0 5.6 5.6 6.4 6.9 20.7
P e r cent 4.5 5.4
’ I
1
5.4 6.0 6.1 20.2
U P P . LIMIT ~
Detd.
Calcd.
P e r cent 28.4 30.8 31.7 37.7 63 5 73 7
P e r cent 28.1 30.5 31.6 36.6 65.4 71.8
Dry Quencher Gas
In the cases of the six gases discussed above several samples of each gas were analyzed and the resu!ts were taken as representative of the quality of gas produced over an indefinite period of time. I n the case of a seventh, or dry quencher, gas, such an assumption would be far from correct. The dry quencher (7) is a closed system of containers and waste-heat boilers through which gases that contain no oxygen are recirculated. The purpose of this apparatus is to save the heat generally wasted when large quantities of water are poured over the hot product in the wet-quenching of cokeoven coke. Hot coke is introduced, and the cooled coke is dumped alternately about every 12 to 15 minutes throughout the day. Since there is but a single point open a t any one time, no air should enter the system, but the composition of the enclosed gases changes over short periods of time. Table IV gives a summary of the analyses of 534 samples taken during the year 1928. Table IV-Summary
cubic foot drops off quite naturally after rioting this fact. Since the carbon monoxide varies over a small range and does not fall into this regular classification, it apparently has little effect on the upper limit except to increase the effect of hydrogen by so much, since the upper limits of these gases are identical. Carbon dioxide also varies over a small range and its individual effect is probably slight. The only conclusion that can be drawn is that methane has been responsible for the drop in the upper limit, in spite of the fact that there is an increase in the amount of hydrogen, since its upper limit is 15 per cent while that of hydrogen and carbon monoxide is 75 per cent. The conclusion is further borne out when we consider the limits of blue gas (6.9 to 69.5) and of producer gas 120.7 to 73.7). The first contains about 50 per cent hydrogen, 40 per cent carbon monoxide, and 2 to 3 per cent methane; and the second about 12 per cent hydrogen, 27 per cent carbon monoxide. and less than 1 Fer cent methane. There is apparently no other reason why the upper liniit of blue gas should be 4 per cent lower than that of producer gas. Table V-Composition
a n d Explosive L i m i t s of Dry Quencher G a s
ANALYSIS Cog
% 3.9 2.4 3.5 4.7 6.7 6.3 5.7 7.1 7.0 7.5 5.0
5:;
7.5 7.3 8 9 10.2
0 2
CO
Hz
%
%
%
0.1
5.9 8.4 9.6 12.5 12.4 7.8 6.3 14.1 7.6 6.1 5.4 4.3 3.9 8.8 8.4 7.2 2.0
33.3 35.2 32.1 24.7 17.3 18.7 17.9 13.8 13.2 >10.0 >10.0 >10.0 >10.0 >10.0 >10.0 >10.0 >10.0
0.0 0.0
0.2 0.3 0.0
0.1 0.2 0.1 0.3 0.2 0.2 0.2 0.2 0.1 0.2 0.0
CH4
Nz
7 0 % 2.5 2.3 1.4 1.3 1.6 1.2 0.9 1.0 1.0
...
... ... ... ... ... ... I
.
.
54.3 51.7 53.4 56.6 61.7 66.0 69.1 63.8 71.1 39
7
0
%
12.3 64.4 >14.9 64.9 12.9 67.8 l5,l 68.9 17.6 >71.8 20.4 71.3 22.8 71.1 23.9 73.3 26.7 73.0 28.3 73.4 29.9 72.3 30.7 71.7 30.5 73.0 30.6 73.9 30.7 74.0 36.0 73.6 None ITone
of Analyses of Dry Quencher Gas during 1928 No. O F PER CENT O F SAMPLES SAMPLES
Less than 10% Hz and less than 4 % CO: Due to presence of air in quencher Due to nearly complete combustion Less than 10% Hz and 4% or more of CO Between 10 and 20% Hz Between 20 and 30% Ha 307, HZor more Total
Vol. 21, No. IL
10 58 218 154 69 25
--
534
1.9 10.9 40.8 28.8 12.9 4.7
--
100.0
Seventeen of these samples were further analyzed on the explosive-limit apparatus, and the results are given in Table V. Dry quencher gas ranks with producer gas and blue gas as being one of the most dangerous, as far as explosibility is concerned, which the Rochester Gas BS Electric Corporation handles. Taking the widest limits as 12 and 74 per cent gas in air (21 per cent oxygen, 79 per cent nitrogen), the oxygen required for an explosion a t these limits is 18.5 and 5.5 per cent, respectively. The presence of less than 14 per cent carbon monoxide plus hydrogen in a mixture containing 86 per cent of inert gases produces an inflammable gas. As a matter of fact, the total amount of combustible gas probably goes down to no more than 5 or 6 per cent in some cases before the gas as a whole can be considered totally inert. Consider the upper limits of hydrogen and carbon monoxide; the presence of oxygen up to about 5.2 per cent could not cause an explosion with any mixture of these gases, together or separately, according to LeChatelier’s rule. Contrary to expectations, the chief inflammable constituent, hydrogen, shows a regular drop in value as the upper limit rises. The percentage of inerts increases and the B. t. u. per
The steady increase in the lower limit is due not to any one constituent, but rather to the collective effect of all the constituents, since the percentage of any one of the several inflammable gases a t the lower limit is very small. In fact, in the cases of all the complex gas mixtures given in this paper the collective percentage of all the inflammable constituents in the lower-limit mixtures approaches the lower limit of hydrogen, 4.15 per cent, particularly where the ratio of hydrogen to carbon monoxide is large (Table VI). The smaller this ratio becomes, the higher this collective percentage figure mill be (since the lower limit of carbon monoxide is 12.5 per cent), unless a high percentage of illuminants or methane pulls it down. Table VI--Collective GAS
_-
Effect of All C o n s t i t u e n t s of G a s Mixtures COLLECTIVE PER CENT AT LOWER LIMIT
RATIOHz : CO
Us, Jones’s method of calculation, the explosive limits of all the dry quencher gas samples of which the complete analyses mere made were calculated. The results given in Table VI1 show that the explosive limits detcrmined in this paper agree very well with the relationships found by Jones.
INDUSTRIAL A N D ENGINEERISG CHEMISTRY
Yovember, 1929
1033
of illuminants and methane is shown by coniparison of the points 0, CB, M, and C with points I? and P. The analyses of dry quenclier gases show the effects of small variations near the outer zone of inflammability. Two curves shown by dotted lines are drawn through the lower-limit points. The N2 curve, upper line tends t o w r d the CO through points 3, 4, 5, 6, 7 , 8, 9, 14, 15, and 16all of which have higher ratios of carbon monoxide. The lower line tends toward the H2 N2 curve, 0 (0 CO 30 40 50 60 through points 1, 2, 10, 11, 12, and 13-all of which % /NERrs ~n ff7ie FREE G A S have higher ratios of hydrogen. Note that point F i g u r e 2-Explosive L i m i t s of G a s e s 8 has a much higher ratio of carbon monoxide than 0 = Oven gas C = Coal gas M = hfixed gas C B = Carbureted blue gas B = Blue gas P = Producer sac any of its neighboring points. h-os. 1 t o 16 = Quencher gas I n the case of the upper limit, the presence of carbon monoxide and hydrogen keeps the line high, but as the Graphical Representation Since the main constituents of the gapes takeii up in this methane increases there is a downward trend toward point 1. paper are hydrogen, carbon monoxide, methane. and inerts, This point, which contains a higher ratio of methane than its the general re!ation between the compoqition and the explosive neighbors, falls below the dotted line. limits may be Jion n graphically by comparing the determined T a b l e VIII-Points i n D r y Q u e n c h e r G a s C u r v e s of F i g u r e 2 S? exploqive limits n i t h Jones's curves for H2 N2 CO and CHI Nz($, 2) plotted on the same scale. Figure 2 ~ h o w sthis comparison. The lettered points refer to the analyses given in T:ihle 11, the numbered points to the dry quencher gas analyses giren in Table VIII, nhicsh is an ab58 12 64 1 31 2.5 G 54 12 65 2 8 36 9.3 breviated form of Table 5', all of these figures except thope 37 13 68 1.4 3 10 32 25 61 15 69 ror methane having been rounded off to the nearest unit. 4 13 1.3 68 18 71 3 12 17 1.6 The carbon dioxide 1i:is been added to the percent Ige of inerts 20 71 19 1.2 72 8 6 7 5 23 71 1s 0 . 9 6 -1nce in most caSes there is too small an amount of it to affect 24 73 14 1.0 71 i 1.1 13 78 27 73 the curves greatly alone. 9 8 1.0 10 6 T a b l e \. 11-Calculated E x p l o s i \ e L i m i t s of D r y Q u e n c h e r G a s < 84 30 72 11 >10 Compared with Erperimentallv Determined \ d u e s 31 72 12 4 10 31 73 10 LiJTVf:R L I M I T UPPERL I J I I T 31 74 9 10 31 74 8 10 Detd. Calcd 10
+
+
+
+
Per cenl
Per cenl 11.8 11.2 12 5 14.9 19.5 17.0 21.6 22.7 26.5
Per cent 64.4 64.9 67.8 68.9
>71.8 71.3 71.1 73.3 73 0
+
Per c e i i t 66.5 67.1 29.7
A single line cannot represent all the conditions, since each point is a result of several individual forces. One line may represent most of the conditions, but there are exceptions which stand out. Figure 2 is presented merely as an aid in picturing the effect of some of the major forces. Literature Cited
10.1 69.8
70.5 71.9 71.2 71.4
The points repreaenting oven, coal, and mixed gases, which have a larger ratio of hydrogen, are shown to be nearer the H2 iY2 curve in the lower-limit mixtures. Carbureted blue gas, blue gas, and producer gas, which have higher ratios of carbon monoxide, are shown t o be tending toward the CO S2curve. In the case of the upper-limit mixtures the effect
+
+
Removal of Mercaptans from Solution by Adsorption on Metallic Sulfides' E. Juanita Greer THZ J o i r s H o f x r s s L-XIVCRSITY, B . ~ ~ . T I \ I u R1':~. E,
IS the spring of 1928, during the course of work directed toward the removal of sulfur compounds from hydrocarbons, we observed in this laboratory that the tendency cif such compounds as arsenic sulfide to attach or adsorb certain other sulfur groups could be used successfully in removing mercaptans from naphtha solutions. The method did not completely r e m x e thiophen from benzene, however, for isatin tests showed that the thiophen content had apparently only been diminished. Amorphous cupric sulfide proved to be the m3st effective of the sulfides used in this set of investigations. Ethyl mercaptan was readily removed when shaken with amorphous cupric, lead, 1
Received October 2 9 , 1929.
(1) (2) (3) (4) (5) (6) (7)
Coward, Carpenter, and P a y m a n , J . Chem. Soc., 115, 27 (1919). Coward and Jones. Bur. Mines, Bull. 279 (1928). International Critical Tables, 1'01. 11. p. 176 (1927). Jones, IND.Ezrc. CHEM., 20, 901 (1928). Jones, Bur. Mines, Tech. Paper 460 (1929). LeChatelier, A n n . mines, 19, 388 (1891). Ptluke, IaD. ENG.CHEar., 21, 457 (1929).
stannic, cadmium, and arsenious sulfides. Cupric sulfide was immediately effective as an adsorbent for the removal of secondary amyl mercaptan. This mercaptan is difficult to remove by the usual methods, and the other sulfides listed were effective in its removal only upon standing, stannic and arsenious sulfides being very slow. ?;aphtha sdutions containing 0.4 to 0.5 per cent added sulfur in the form of mercaptan were shaken in a stoppered test tube with the powdered amorphous sulfide and were found to be sweet when tested with alcoholic doctor solution. The addition of water in a noticeable quantity did not destroy the effectiveness of the sulfide in removing the mercaptan. Further investigations involving the idea of reaction a t the interface between aqueous solutions of metallic salts and mercaptans dissslved in hydrocarbons showed t h a t cupric salts, such as the sulfate and nitrate, can be used to sweeten the hydrocarbon solutions. This work was carried out a t the suggestion of J . C. W. Frazer and W. -4.Patrick, of the chemistry staff of the Johns Hopkins L-Iiiversit y .