T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY
J u l y , 1919
’
Table I1 and Fig. z give t h e time-gain in weight results. It is worthy of note t h a t all samples showed a loss in weight during t h e first few days, although a well-settled oil was used. The longer t h e exposure t h e paler t h e oils became; Samples 6, 7 and 8, however, show practically no difference. The amount of oil placed in t h e large dishes made a layer about one-half as thick as t h e oil placed in t h e small dishes. I n Fig. z t h e effect of thickness is clearly brought out showing t h a t t h e thin layer gained in weight about twice as rapidly as t h e thick layer. Without regard t o t h e thickness of t h e layers, t h e evolution of pungent volatile products became distinctly noticeable when t h e oils had gained about one per cent in weight. I n Table I11 are assembled t h e analytical results shown in Fig. 3. Samples 7 and 8 skinned very slightly when 1 5 2 days old; this skin was stirred up with t h e oil and Sample 7 removed, while Sample 8 was continued j days longer, when it skinned again. I t is remarkable t o note, however, t h a t Sample I O did not skin, although i t had gained about I per cent more in weight. There is a change in t h e specific gravity curve a t t h e point of skinning of Samples 7 and 8 b u t no noticeable effect is produced in t h e iodine number curve. The acid numbers are rather irregular. Calculations of t h e per cent change in volume have been made b u t yield results not so decisi$e as those reported by Friend. TABLE 111-DATA SHOWING THE EFFECT OF EXPOSURE ON RAWLINSEED OIL Gain in Sample Age Weight No. Days Per cent 0 Raw 0.38 49 0.69 70 1.63 98 3.05 119 4.22 135 5.41 148 5.60 152 5.84 157 3.67 69 9 6.63 85 10
....
s
sP. Gr- Iodine ‘5.50. No.
Acid 15.5’ C. (Hanus) No. 188.7 2.2 185.2 2.4 183.0 3.0 174.6 0.9481 5.3 0.9602 162.8 7.5 0.9710 152.4 8.5 0.9840 141.2 10.5 0.9854 139.2 11.0 0.9913 137.7 12.1 0.9638 157.0 9.6 ... 132.6 14.0
Change OilSurface in per Volume Relative Gram Per cent Viscositv Oil . . . . 1.00 0.95 1.15 0.95 -0.02 0.95 +0.07 1.25 +o. 17 2 . 0 0 0 . 9 5 4.30 0.95 +0.29 7.35 0.95 +-0.30 +o. 10 . . . . 0 . 9 5 4-0.14 . . . . 0.95 -0.22 .... .... 6.00 4-0.52 2.0 2.0
....
....
It was found t h a t I g. of each of t h e oils would easily and completely dissolve in IOO cc. of benzene. The figures in column “Relative Viscosity” are rough approximations only and were obtained by measuring t h e time required for I O cc. of t h e oil t o flow from a pipette, taking the time of efflux of t h e raw oil as 1.00. I t will be observed in Fig. 3 t h a t t h e iodine numbers of t h e oils in thin layers fall exactly on t h e curve of t h e oils in thick layers; while the specific gravity of Sample g is not far from t h e specific gravity curve of t h e thick layers. S U MMAR Y
The effect of exposure on certain constants of raw linseed oil has been determined over a limited range of gain in weight. The thickness of t h e exposed layer of oil appears t o affect only t h e rate of change in t h e constants. For any gain in weight over t h e range covered b y these experiments, t h e changes occurring in t h e constants appear t o be independent of t h e rate of gain in weight. NATIONAL LEAD C O M P A N Y ST. LOUIS, MISSOURI
639
CARBON TETRACHLORIDE, CHLOROFORM AND CARBON HEXACHLORIDE FROM NATURAL GAS B y G . W. JONES AND V C . ALLISON Received December 21, 1918
As a result of t h e war, t h e Government desired t h e maximum amount of carbon tetrachloride and chloroform t o carry on t h e gas program. A great many of the gases used in modern warfare are derived from these sources. In view of t h e above fact, and also because after t h e war a great amount of chlorine, which is now being used for war purposes, will be turned back toward peaceful enterprises, t h e Bureau of Mines has undertaken an investigation on t h e production of useful products b y the chlorination of natural gas and thus utilize part of this excess chlorine. The gas from many fields of natural gas in t h e United States which yield a pure methane gas, free from t h e higher saturated hydrocarbons, ethane, propane, etc., is especially desirable for making chlorinated products. The natural gas from several of these fields, notably those in Texas and Louisiana, in locations too far removed from industrial centers and large cities t o warrant t h e expense of piping, could be successfully made into chlorinated products. I n this report only a preliminary survey of work done on a small scale is given and more elaborate results of t h e investigation will be published as t h e work develops. A large amount of experimental work has been done along these lines during t h e last 2 5 years b u t so far nothing of practical value, t h a t the authors are aware of, has been accomplished.2 The general tendency has been t o use a large excess of either chlorine or natural gas t o prevent explosion. If a large excess of either gas is used t h e products desired are hard t o separate from the excess of inert gas, since t h e deposition of t h e chloroform and carbon tetrachloride depends upon t h e partial pressure of these products in the gaseous state a t t h e given temperature at which they are separated and the reaction cannot be controlled t o produce t h e products desired. T h a t t h e chlorination may work successfully there is required a catalyzer which will cause t h e reaction t o proceed smoothly without explosions or deposition of carbon, and which will accomplish the substitution of the chlorine in t h e methane and ethane molecule according t o the reaction CH4+4C12 = CC1,+4HC1 instead of t h e production of carbon and hydrochloric acid in accordance with t h e reaction CH4+ 2 C l ~= C+4HCl. A great many catalyzers will cause chlorine and methane t o react, when t h e temperature is high enough. I n fact, no catalyzer is needed a t all, but t h e reaction takes place violently and, being exothermic in character, explosively, and gives very little, if any, products other t h a n carbon and hydrochloric acid. It is desirable t o make t h e chlorination complete a t one operation, hence if t h e desired product is carbon tetrachloride, four volumes of chlorine are caused t o react with one volume of methane (natural gas), or in other words, the gases are caused t o react in t h e ratios necessary t o 1
2
Published by permission of the Director of the Bureau of Mines. Phillips, Baskerville, and others
640
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
F I G . I-APPARATUS
FOR CHLORINATING
produce the most highly chlorinated products without a n y waste of chlorine. Of the saturated hydrocarbons, methane is very much more inert toward chlorine t h a n those of higher molecular weights. T h e authors found t h a t a t certain temperatures the chlorine reacted readily with t h e ethane while there was very little reaction with the methane. I n fact, the chlorine reacted with the ethane t o form carbon hexachloride before any chloroform or carbon tetrachloride was formed. By using certain catalyzers, such as war-gas charcoal, steamed anthracite coal, and bachite, t h e reaction took place smoothly without explosions or deposition of carbon even when the chlorine ratio was four or more t o one of gas. When other catalyzers such as cokes impregnated with different metals and metallic oxides were used and four volumes of chlorine to one of gas were caused t o react, a small amount of chlorine came through without reacting. The best catalyzers for complete chlorination were those which had a high absorption coefficient for chlorine. The increased concentration of the chlor i i e causes an increase in the reaction velocity and t h e chlorination goes t o completion. War-gas charcoal was loaned t o the Bureau by t h e American University Experiment Station and is the same as t h a t used for filling canisters in gas masks. This is a high-grade charcoal and far superior t o other charcoals tried. Nearly all others were practically useless as a catalyzer and were no better t h a n inert material such as broken porcelain and silica. Bachite is a patented material made by t h e National Carbon Company. This has also been used for war-gas purposes. The steam treated anthracite coal was made by reducing a commercial coal t o a 4 t o I O mesh size.
Vol.
11,
No. 7
NATURAL GAS
The coal thus prepared was next placed in an iron retort kept a t a temperature of approximately 700' F. and steamed for 7 0 hrs. Coal treated in this manner has the property of absorbing gases t o a high degree, although the outward physical appearance shows no change. GAS
The natural gas used was taken from t h e mains a t the Bureau of Mines laboratories, Pittsburgh, Pa. This gas is supplied by t h e Consolidated Gas Company. On account of having ethane present, this gas is not so desirable as a pure methane gas, but on account of t h e accessibility and ease of application i t was used in this work. The average analysis of t h e gas during t h e time in which these experiments were made was as follows: Methane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ethane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A'itropen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Per cent 89.5 10.1 0.4
For computing the theoretical yield the gas was assumed t o contain 90 per cent methane and I O per cent ethane. 'CHLORINE
The chlorine used was supplied b y t h e Electro Bleaching Gas Co., of New York, in containers of I O O lbs. each. APPARATUS
Fig. I shows the general plan of the apparatus for testing the effect of varying temperature and humidity on t h e reaction. The gas and chlorine rates were regulated by flow meters calibrated against the gases used. T h e flow meters were of a very small capacity and .were calibrated by passing a given volume of t h e gas from a
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
641
TABLE I-RESULTS
TEST
Av. ChloAverage rine Temper- Rate ature Liters Deg. C. per Hr. 250 2.5
1
Humidity Per cent 1.1 per cent HzO vapor
2
1.1 per cent HzO vapor
300
4
3
1.1 per cent HzO vapor
350
4
4
1.1 per cent HzO vapor
400
4.5
.5
1.1 per cent Hz0 vapor
450
4
6
1.1 per cent H2O vapor
500
4
7
Gas saturated
400
5
400
3.5
NO.
8 Water removed
WITH WAR-GASCHARCOAL AS CATALYZER Av. PercentGas age -Distillation ResultsRate Time ProdAbove Theo- Effi- 65'to Liters of ucts Re- retical ciency 80' C. 80° C. per R u n covered Yield Per Per Per Hr. Hrs. Cc. Cc. cent cent Sp. Gr. cent REMARKS 1 6 6.5 24.1 27 80 1.59 20 Chlorination not complete. Unable t o use required amount of chlorine for complete chlorination 1 6 15 24.1 10 Reacted smoothly without deposition of car62 90 1.60 bon or explosions 1 6 22.5 24.1 16 Reacted smoothly without deposition of car93 84 1.59 bon or explosions 1 6 24 24.1 99 95 1.60 5 Reacted smoothly without deposition of carbon or explosions 1 5 17 20.0 85 90 1.60 10 Reacted smoothly without deposition of carbon or explosions 1 6 23.5 24.1 97 85 1.59 15 Carbon began t o deposit and smoke badly. Catalyzer ignited but did not explode. Dense white fumes formed throughout the run 1 6 19.5 24.1 81 90 1.58 10 It was necessary t o use a n excess of chlorine t o chlorinate completely 1 6 15.5 24.1 64 85 1.58 15 No fumes formed. Could not use four volumes of chlorine without coming through unacted upon
burette through the flow meter, a t such a rate t h a t by keeping t h e two columns of liquid in t h e U-tube a t a certain height, a certain length of time was required t o empty t h e burette. From these d a t a t h e rate can be determined. After calibrating several points, a scale was made t o read t h e rate directly. The reaction tube shown in Fig. I contained I O O g. of catalyzer protected on either end b y glass wool. An electric furnace surrounded t h e tube and t h e temperature was read on a millivoltmeter in conjunction with the thermocouple shown in t h e drawing. The hot gases after leaving t h e reaction chamber passed through a triple-walled condenser for cooling, t h e n through a trap, sodium hydroxide scrubber, and two ice baths as shown. Most of t h e solid carbon hexachloride separated out in t h e cooler, while the carbon tetrachloride, being heavier t h a n water or lthe sodium hydroxide solution, settled t o t h e bottom of the scrubber a n d ice baths. T h e chief object of the sodium hydroxide solution was t o remove the hydrochloric acid formed during t h e reaction. In all t h e tests made on this apparatus the gas rate was I 1. per hr. For higher rates the apparatus shown in Fig. z was used. After each run the distillates were collected from the scrubbers a n d ice baths, separated in a separatory funnel, measured, a n d then treated with a small piece of stick sodium hydroxide, t o remove t h e dissolv,ed chlorine, and calcium chloride, t o remove t h e water. T h e distillations were carried out in a 50 cc. distillation flask a t a rate of 2 drops per sec. On all tests shown in Table I b u t a few drops came over below 70' C., indicating t h a t very littl?, if any, was chloroform. The gravities were determined with a specially designed plummet of small size which could be used o n a Westphal balance a n d gravities on as low as 4 cc. of product were easily determined. Table I shows t h e results in tabulated form. The reaction begins a t a little below 250' C. and tends upward t o around 500' C., where carbon begins t o deposit and t h e catalyzer is attacked. The best results were obtained with a small amount of moisture present (I per cent by volume). This tends t o help t h e reaction. All t h e products boiling below 85' C. were waterwhite a n d free from carbon.
CALCULATION O F YIELD
Yields were calculated from t h e volume of natural gas used, assuming t h a t a liter of methane weighs 0.65 g. and ethane 1.22 g. when measured a t 2 5 ' a n d 740 mm. pressure, t h e average condition under which the natural gas was measured. One liter of natural gas (90 per cent CHI, I O per cent CzHc) completely chlorinated should produce 4.01 cc. of liquid chlorination products, c c l 4 and C2Cle. T h e carbon hexachloride is very soluble in the carbon tetrachloride a n d t h a t which crystallized out in the cooler was added each time t o t h e product from t h e scrubber and distilled together. T h e distillate above 80' C. consisted of a mixture of carbon tetrachloride and carbon hexachloride. This could be separated by fractional distillation. T h e process is long and tedious and takes several distillations and subsequent coolings in a n ice-salt bath, t o freeze out t h e carbon hexachloride. No other chlorine compounds have been identified in t h e products obtained by using four or more volumes of chlorine t o one of gas when t h e rate of natural gas was I 1. per hr. When all t h e products boiling between 65' C. a n d 80' C. from t h e eight tests were p u t together a n d distilled, less t h a n I cc. came over below 65 ' c.,indicating a negligible amount of chloroform present. T h e percentage efficiency of each r u n was obtained by multiplying t h e volume of gas used by 4.01t o give t h e theoretical yield which would be expected if t h e gas were chlorinated t o carbon tetrachloride and carbon hexachloride and then the volume of products recovered was divided by t h e theoretical yield. This gave the efficiency on the basis of the amount of gas used. Certain modifications of t h e apparatus were necessary in order t o handle larger quantities of gas. T h e scrubbers were designed t o work with low pressure and t o be convenient for removal of the products a n d hydrochloric acid formed. Fig. 2 shows the modified form including t h e scrubbers, drying tube, ice baths, and sampling device for testing the residual gas for methyl chloride. T h e gas rates were determined with flow meters as
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
442
FIG.2-MODlFIED
FORM O F APPARATUS, INCLUDING SCRUBBERS, DRYING TUBE,ICE BATHS,A N D
in t h e previous work. The gas and chlorine were introduced into t h e reaction tube through a mixer made on t h e order of an injector. The reaction t u b e is shown in Fig. 2 , also t h e electric furnace surrounding it. After trying a great many tubes of different lengths and diameters, t h e one found most suitable for chlorinating up t o 6 1. of gas per hr. was of Pyrex glass 16 in. long and 1'/2 in. in diameter drawn down a t one end so as t o make connection with t h e scrubber. The t u b e protruded out of t h e furnace about 6 in. and this end was covered with asbestos paper t o exclude t h e light. By having t h e gases mix in t h e cool part of t h e tube in t h e presence of t h e catalyzer, no explosions were produced on entering the heated section. The tube contained 2 0 0 g. of t h e catalyzer, held in place by glass wool placed at each end. This was then closed with a two-holed rubber stopper; through one hole passed t h e thermocouple
SAMPLING
Vol.
11,
No. 7
DEVICE
A thermocouple indicated the temperature in t h e reaction tube a t all times. It was found t h a t t h e temperature was somewhat higher at the point where t h e gases entered the reaction tube, due t o t h e heat of reaction, and t h a t after t h e reaction had commenced very little heat was necessary t o maintain t h e desired temperature; t h e higher t h e rate of gas, t h e less heat was necessary. The hot gases, on leaving t h e reaction chamber, are passed through low-pressure scrubbers of a special design shown in Fig. 3. These scrubbers, containing water, cool the heated gases and remove t h e hydrochloric acid formed during t h e reaction; t h e carbon tetrachloride, carbon hexachloride, and chloroform, which are heavier t h a n water, sink t o t h e bottom a n d can be drawn off b y means of t h e stopcocks provided for t h e purpose a t t h e bottom of each scrubber. One scrubber was found sufficient t o completely remove
TABLE I1 Distillation Results in Per cent of Av. Av. Recovered Product C,as -~~ Chlo. . . . . . To 70" 70° to Above Rate rine Time of ProductTheoretical Per cent C. Sp. 85' Liters Rate Sp. 85' C. SP. Liters Run Recovered Yield Efficiency Per Gr. Per Gr. Per Gr. per Hr. per Hr. Hrs. ( a ) ( b ) cent 2 0 / 2 0 cent 2 0 / 2 0 cent 2 0 / 2 0 Cc. Cc. 81.8 94 10 1.55 1.58 5 3.4 19.0 6 77.0 95 85 1.65 1.58 26 22.6 95 64 1.61 5.6 6 126 134.7 94 10 1.54 24.6 99 1.63 5 105 106.3 99 3 1.55 1.59 15 5.3 82 15.0 72 6 42.0 60.1 20 1.55 1.56 72 8 1.64 70 2.5 82 1.62 264.7 1.59 78 82 0.0 . . 22 16.0 4.0 16.5 217.1 84 1.61 1.58 70 83 7 1.54 22 16.0 288.7 4.0 18.0 238.7 1.59 65 4 1.56 31 1.60 14.0 93 70.4 2.7 6.5 65.0 92 41 1.50 1.55 17.3 69 38 49.3 21 65 1.60, 3.0 32.0 4.1 1 1.57 1.58 96 75 57.2 24 96 1.61 8.2 6 . 5 \ 55.0 2.2 63 1.48 1 . 5 3 ~ 32 71 14 242.6 64 1.62 3.9 16.0 15.5 155.0 rssumption t h a t product to 70' C. is CHCh and the rest CCla and CzCls. L
TEST No.
Humidity Catalyzer 1.1 per cent Hz0 vapor War-Gas Charcoal 1.1 per cent Ha0 vapor War-Gas Charcoal 10 1.1 per cent HzO vapor War-Gas Charcoal 11 1.1 per cent H10 vapor War-Gas Charcoal 12 1.1 per cent HzO vapor War-Gas Charcoal 13 1.1 per cent Hz0 vapor Bachite 14 1.1 per cent Ha0 vapor Bachite 15 1.1 per cent HzO vapor Bachite 16 1.1 per cent Hz0 vapor Steamed Coal 17 1.1 per cent H20 vapor Steamed Coal 18 (a:I On assymption t h a t products are CC1.i and C 9
Av. Temp. Deg. C. 375 385 395 325 400 395 375 370 380 390
P
TABLE I11 Formula Compound CH3C1 Methvl chloride.. ......................... . . . . . . . CHzClz Dichlbrmethane ....... CHCh Chloroform. . . . . . . . . . . . . . . . . . . . . . . Carbon tetrachloride. . . . . . . . . . . . . . . . . . . . CCla Ethyl chloride.. . . . . . . . . . . . . . . . . . . . . . . . . . CHaCHzC1 Dichlorethane . . . . . . . .. . . . . . . . . . . . . . . . . . . CHzClCHzCl . . . . . . . CHaCHClz Ethylidene chloride.. . . . . . . Trichlorethane . . . . . . . . . . . . . . . CHzClCHClz . . . . . . . CH3CCla Trichlorethane . . . . . . . . CHzClCCla Tetrachlorethane . . . . . . . . . . . . . . . . . CHCliCHClz Tetrachlorethane.. . . . . . . . . . . . . . . . . . CHClzCCls Pentachlorethane.. . . . . . . . . . . . . . . . . CCl3CCl3 Carbon hexachloride, ......................
.............
.
.......
well and through t h e other, t h e tube leading from t h e mixer. The furnace was made b y winding 40 f t . of No. 2 2 nichrome wire in t h e form of a helix, wrapping on an asbestos mandrel, and covering with a mixture of calcined magnesium oxide and sodium silicate, then covering this with several layers of asbestos paper. After drying, t h e furnace is ready for use.
At 25' C. Gas Liquid Liquid Liquid Gas Liquid Liquid Liquid Liquid Liquid Liquid Liquid Solid
'
Boiling Point Deg. C. -23.7 41.8 61.2 76.7 12.5 83.7 57.4 115.0 74.5 135.0 147.0 159.0 M. P. = 187.0
Specific Gravity 0.9915 at-23.7 1.378 a t 0/4 1.50 a t 15/15 1.58 a t 20/20 0.923 a t 2 / 2 1.259 a t 15/15 1.18 a t 15/15 1.45 a t 15/15 1.32 at 15/15 1.61 a t 0/0 1.59 a t 15/15 1.71 2.0
t h e hydrochloric acid, b u t two were used t o aid i n condensing t h e products. Starch-potassium iodide test. papers were hung at t h e entrance of each scrubber t o indicate t h e presence of chlorine. The unchlorinated natural gas and other gases after leaving t h e scrubbers pass through a t r a p enclosed in an ice-salt bath. Here most of t h e water vapor is frozen out and. t h e dichlormethane, and other chlorides, condensed,
July, 1919
T H E JOlTRNAL OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
The gas next passes through a drying tower containing soda lime and calcium chloride, thence through a 3way stopcock and on out of t h e apparatus through a water seal. By means of t h e 3-way stopcock samples of t h e residual gas could be taken a t pleasure for t h e analysis of methyl ohloride. Table I1 shows results of chlorinating larger amounts of gas, using t h e three most promising catalyzers. A great many charcoals, cokes, etc., were tried but t h e yields, from a carbon tetrachloride standpoint, were not as good and are not given in this report. Increasing t h e gas rate produced products of a boiling point below 70' C. The boiling point and specific gravity of these fractions indicate a mixture of chloroform and carbon tetrachloride. The products boiling above 8 j " C. indicate a mixture of carbon tetrachloride, carbon hexachloride, a n d other chlorine products. These percentage yields may be in error t o some extent, due t o t h e above assumptions, b u t are of value in t h a t they are useful for comparing the selative efficiency of t h e different catalyzers.
643
Results show t h a t natural gas can be completely chlorinated on a laboratory scale a t one operation. The temperature range extends from about zzj" C. t o j o o ' C. when using war-gas charcoal. Moisture helps t h e reaction. A t a rate of I 1. of natural gas per hr. t h e methane goes completely t o carbon tetrachloride and t h e ethane t o carbon hexachloride when using a maximum amount of chlorine. When larger rates of gas were used chloroform was found along with t h e carbon tetrachloride. T h a t carbon tetrachloride and chloroform can be produced successfully is proven, provided the right catalyzer is used, t h e conditions of t h e reaction are carefully watched, and efficient means for removing t h e products are used. Ethane is chlorinated much more easily t h a n methane and a t a lower temperature. Catalyzers for complete chlorination must be such t h a t they have a high absorption value for chlorine. ACKXOWLEDGMENTS
I n compiling this report t h e authors wish t o acknowledge t h e helpful suggestions made b y G. B. Taylor, chief chemist of t h e Pittsburgh Station, Major A. C. Fieldner, of t h e Chemical Warfare Service, Lt. S. H. Kate and Sgt. Linchfield for supplying two of t h e catalyzers, hf. H. Meighan and W. L. Parker for carrying on t h e tests, and F. E. Donath, glass blower, for making t h e scrubbers and other parts of the apparatus.
4n I
FIG. 3 - L O W - P R E S S U R S
SCRUBBER
Table I11 shows t h e different chlorides of methane a n d ethane with their boiling points and their specific gravity, as given in Watts' chemical dictionary and Landolt-Bornstein tables. These are given here for comparison of t h e products obtained b y chlorination of natural gas. POSSIBILITIES
Processes and methods must be found t o use t h e large supply of chlorine which will be available a t t h e conclcsion of peace. Natural gas is a very cheap commodity in out-of-the-way places, and chlorine can be made by using part of t h e natural gas for power. I n this way t h e expense for raw material will depend only on the value of t h e natural gas. Chloroform besides being used in surgery is a good solvent. Carbon tetrachloride is a good solvent, non-explosive and non-inflammable. It can be used for t h e extraction of certain fats, resins, waxes, dry cleaning, and for fire-extinguishing compounds. Carbon hexachloride has no uses at present b u t can be reduced t o ethane tetrachloride which is a good solvent used extensively in t h e manufacture of motion picture films, aeroplane dope, varnish manufacture, paint remover, etc., CONCLUSIONS
Apparatus is shown a n d described for t h e chlorination of natural gas. Results of tests are given pertaining t o t h e complete chlorination of t h e gas.
BUREAUOF M I N E S DEPARTMENT O F INTERIOR WASHINGTON,
D.
c.
THE TANNIN CONTENT OF REDWOOD By CHARLBSC. SCALIONE AND DAVIDR. MERRILL Received December 16, 1918
The rapid depletion of t h e more accessible supplies of tanbark oak ( P a s a n i a dertsiflora), which is t h e only natural source of tannin largely utilized on t h e Pacific Coast, suggested the desirability of a study of t h e wood of t h e coast redwood (Sequoia semperv i r e n s ) , which is known t o contain tannin and of which a large supply is available in t h e form of refuse from t h e lumber industry. With this idea in view, samples of bark, heartwood, and sapwood were obtained from Mendocino County. The bark was shredded and ground in a mill t o pass a 20 mesh screen. Samples of the heartwood and sapwood were prepared by sawing, and grinding t h e sawdust t o pass 2 0 mesh. The moisture contents were determined by drying samples t o constant weight. The undried samples were extracted by soaking them for several hours in warm water, pouring off this extract, and extracting t h e residue with more water in an extractor of the Soxhlet type. The extraction was stopped when t h e extract gave a negative test with t h e salt gelatin reagent for tannin. The mixed solutions were diluted t o volume, and clarified b y adding washed kaolin and sucking them into a flask through an alundum extraction thimble. After this treatment t h e samples passed t h e require-
