674
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
substantial agreement as t o t h e remedy. Let no one think t h a t i t will be easy t o get together a list of items t h a t will satisfy t h e makers of a n d dealers i n t h e products of a n d for our industry. Each member of t h e Central Committee appointed b y t h e American Chemical Society a t t h e Seattle Meeting in September of last year will have t o be prepared t o sacrifice t w o hundred or more hours of real hard labor t o complete his share in t h e work; his work can be no better t h a n t h e work of t h e various sections of t h e American Chemical Society, a n d when i n next September these various sections get their allotments of work, as sent t o t h e m from t h a t Central Committee, t h e y will have t o t a k e their coats off in earnest a n d work hard t o see t o it t h a t each have all their suggestions in on time. This word of warning of a n impending avalanche of labor is given t h u s early t h a t there may be no avoidable obstacle t o prompt disposition of t h e work assigned. T h e Central Committee cannot get down t o business until t h e sections have finished. When normal use of international trade channels is available. t h e information proposed t o be asked for will be of t h e greatest help in preventing products of foreign make, a n d which can be made in this country, from gaining such a foothold t h a t sudden drying-up of foreign sources of supply can seriously affect us. B u t this information when available will largely fail of its purpose if our makers of chemicals do not cut loose from t h e traditions of t h e past t h a t have enchained substantially all our manufacturers of all kinds-namely, a sort of contempt for small outputs or tonnages. To be able t o make a large number of different things, each of relatively small annual tonnage, or even poundage, is quite as essential t o our national business life, as is ability t o make large annual amounts of relatively few things. Times h a r e changed a n d we will have t o change with them. Also our industrials of all kinds, inclusive of our chemical industries, will have t o learn t h a t quality a n d price d o count, a n d t h a t there is no brand of patriotism t h a t will, for long, withstand t h e lure of a profit. With individuals, patriotism is said t o stop a t t h e stomach; with business i t may similarly be said t o stop a t t h e pocket-nerve.
R E S P O N S I B I L I T Y TO E D U C A T I O R A L I N S T I T U T I O N S
Our fourth responsibility is t o our educational institutions. It is inconceivable t h a t in t h e wealth of information a n d in t h e large ability of our university men there is not more opportunity for efficient. effective a n d valuable cooperation between t h e m a n d our chemical industrials for t h e benefit of t h e nation. It is a k n o t t y problem, there is no denying i t , but t h e prospects of results of almost incalculable, potential national value seem so reasonable t h a t it is worthy of t h e most serious attention of every chemist, be he university man, technical man, merchant or banker. R E S P O N S I B I L I T Y TO A M E R I C A i i C H E M I C A L S O C I E T Y
Our fifth responsibility, a n d t h e last I shall touch upon in this paper, b u t b y no means t h e last of our responsibilities, is t o t h e American Chemical Society itself. Our Society places on our desks four times each month publications which can challenge comparison with a n y similar publications in t h e world a n d can come out of such comparison with credit t o themselves a n d honor t o t h e American chemical profession; this is done at a n annual cost t o each member of about two-thirds t h e cost of t h e same service abroad. How is i t done? T h e most of us t a k e these publications as a matter of course, b u t we are indebted for these magnificent results t o a n untiring devotion, t h e unstinting labor, a n d t h e most generous loyalty of a relatively small number of our fellow-members; t h e y give of their time, their strength and their ability far beyond their just share. Take love of labor a n d pride of profession out of t h e make-up of these men, a n d t h e quality a n d amount of their work would drop t o such a point t h a t t h e net result would not be creditable t o us nor t o our country. The real hard work of t h e American Chemical Society is performed substantially as a labor of love; t h a t condition ought not be permitted t o last one instant longer t h a n it requires t o remedy it. We will be asked in September next t o provide t h e remedy. Let us not fail t h e American Chemical Spciety.‘ She has served us a n d our country well! B. C. HESSE 25 BROADSTREET, ?SEW Y O R K CITY
ORIGINAL PAPERS
1
THE THERMAL DECOMPOSITION O F T H E PROPANEBUTANE FRACTION FROM NATURAL GAS CONDENSATE By J E. ZAXETTI Received June 10, 1916
Practically every fraction of petroleum a n d many of its individual constituents, f r o m natural gas t o t h e residues from t h e distillation of crude petroleum have been submitted t o decomposition a t high temperatures, so-ca11ed with One exception: t h e fraction which contains propane a n d t h e butanes. T h e reason is not far t o seek; i t has not been until t h e last few years t h a t t h e separation of this fraction from “wet” natural gas1 a n d its use as a commercial product I F o r description of this process of separation and analyses of various fractions see U. S. Bureau of Mines, Bull. 88 and Tech Paper No. 10.
Vol. 8, No. 8
I
1
under t h e name of “liquid gas,” “gasol.” etc., has become of commercial importance. This fraction, which condenses along with t h e lower boiling pentanes, is usually allowed t o escape in t h e air in t h e process of “ripening” a n d is t h e so-called “wild gasoline” which, because of its low boiling point, is useless for the purposes of (tblending,,with refinery naphthas. In places, by further compression of t h e LLl%,et after the first gasoline condensate has been formed, this fraction is condensed together with ethane and put up in steel cylinders for as an illurninant or for welding with oxygen. For t h e s t u d y of t h e decomposition products, t h e “pyrogenetic reactions” of hydrocarbons, this fraction furnishes a very satisfactory material. Being
Xug., 1 9 1 6
T H E J O U R N A L OF I N D U S T R I A L A N D ENGIN,EERING CHEMISTRY
a gas a t ordinary temperatures its rate of flow through a heating chamber can be regulated with nicety, no previous vaporizer is necessary as with higher boiling fractions, t h e formation of lower unsaturated compounds b y splitting off of hydrogen as well as condensation of these t o aromatics are exhibited, a n d finally, t h e products f o r m e d , outside the a r o m a t i c s , are gases, a n d a n y u n d e c o m p o s e d p r o p a n e or b u t a n e p a s s e s 0.8 t h u s preventing t h e necessity of separating t h e aromatic from t h e undecomposed aliphatic, as is t h e case with higher boiling fractions. T h e above reasons as well as t h e fact t h a t t h i s fraction constitutes mostly a waste product have led t h e writer t o s t a r t a s t u d y of its decomposition products with a view t o its possible utilization in other ways t h a n as a lighting material. A careful search of t h e literature has failed t o furnish a n y work t h a t has been published on t h e decomposition of propane or butane. E v e n Beilstein, t h a t can usually furnish reference t o t h e products obtained b y passing vapors of almost a n y organic compound through “ a red hot t u b e , ” furnishes no light on t h e subject. T h e plan of t h e work consisted in: I-The determination of t h e composition of t h e “cracked” gases with respect t o “unsaturated” a n d hydrogen formation. a-The determination of t h e amounts of t a r formed (aromat;cs) a n d its composition. 3-The composition of t h e unsaturated. 4-The influence of catalyzers. One difficulty arises with t h e investigation of t h e butane-propane fraction: being p u t u p in a cylinder under pressure, on releasing it we have t h e reverse of ordinary distillation, t h e temperature being k e p t constant in this case, while t h e pressure changes. Under these circumstances t h e lower boiling fraction comes off first, t h e composition of t h e gas altering as t h e pressure drops. For this reason one must work within certain pressures in order not t o change t h e composition of t h e gas between too wide limits. T h e pressure dropped 5 lbs. during t h e duration of these experiments a n d t h e percentage of butane rose, as one m7ould expect. No appreciable difference could be noted either in t h e “unsaturated” or t h e hydrogen content, b u t t h e amount of t a r increased somewhat with t h e increase of butane. The apparatus employedl consisted merely OF a heating chamber, t h e temperature of which could b e accurately controlled, a coil t o cool f h e gases a n d finally a device t o settle t h e t a r “fog.” All who have worked with t h e decomposition of hydrocarbons a t high t e m perature are familiar with t h e finely divided condition in which t h e liquid products deposit on cooling t h e gas, t h e so-called t a r “fog.” By far t h e largest proportion of t h e t a r remains in suspension i n this finely divided condition a n d requires considerable time for settling. On a large scale, as i n gas manufacture, this “fog” is removed b y various forms of scrubbers, b u t t h e writer is not aware of a n y device for use on a small scale. 1
For full details see experimental part.
675
After many unsuccessful a t t e m p t s t o remove this “fog” b y filtering t h e gases through numerous layers of glass wool t h e principle of a well-known patent1 was finally adopted. This was t h e deposition of these finely divided particles passing t h e gases from t h e furnace between t w o conducting plates maintained a t a high difference of potential. The results were very successful a n d whereas only a few drops of t a r were obtained per cu. ft. of gas b y relying solely on condensation, as much as I O cc. per cu. ft. of gas were obtained b y electrical deposition. There is, of course, no novelty i n this device b u t it is surprising t h a t it has not been employed before in small scale experimentation. T h e fact t h a t almost all previous work on hydrocarbon decomposition on a small scale has failed t o t a k e into account this carrying off of most of t h e t a r in finely divided form b y t h e gas stream may necessitate some revision of earlier work. T h u s Norton a n d Andrews,2 without any t a r depositing, passed t h e gases from t h e decomposition of hexane, iso-hexane a n d pentane through bromine for ab60
50
CO
30
LO
10
FIG I-DECOMPOSITION O F PROPANB-BUTANE FRACTION No CATALYZE%
sorption of t h e unsaturated, a n d obtained only a qualitative test for benzol with no mention of higher aromatics; t h e y obtained no benzol from pentane. Haber3 obtained b y t h e decomposition of hexane “ a little benzene.” I n view of t h e writer’s experience with a n d without a tar-settling device, a n d taking into consideration t h a t t h e higher hydrocarbons of petroleum distillates yield considerable benzol as well as toluol a n d higher aromatic homologue^,^ as well as t h e additional fact t h a t t h e lower hydrocarbons investigated also give t h e same aromatics, it seems very unlikely t h a t these few intermediate hydrocarbons should not, a n d t h e question should remain an open one, a t least until t h e deposition of t h e t a r “fog” should prove it otherwise. T h e results of t h e decomposition of t h e gas a t various temperatures are plotted in Fig. I. The r a t e of flow through t h e heating apparatus was kept cons t a n t throughout all these experiments, being maintained at 0.4j cu. ft. per hr. This rate was chosen See F. G . Cottrell, THISJOURNAL, 3 (1911), 542. Am. Chem. J., 8, 1. Ber., 29, 2691. 4 For a very complete review of the literature on this subject see Egloff and Twomey, Jour. P h y s . Chem.. 20 (1916), 121. 1
2
8
T H E JO17R11‘A L OF I N D C S T R I A L AlVD E N G I N E E R I X G C H E M I S T R Y
6j 6
as being sufficiently slow t o permit t h e gases t o be heated t o t h e high temperature. The per cent of “unsaturated” increases gradually till a maximum is reached in t h e neighborhood of 7j o o . Beyond 7 joo: when the aromatic formation begins. the “unsaturated” decrease and become practically nil near 1000’. 60
50
30
ao
lo
FIG
11-DECOMPOSITION OF
P R O P A X E - B U T A N B FRACTIOK, AS
USINGCOPPER
CATALYZBR
T h e hydrogen curve rises slowly a t first b u t more rapidly as t h e point of aromatic formation is reached. The nature of t h e “unsaturated” was determined by passing t h e gases through bromine and fractionating the bromides. The “unsaturated” consisted of a mixture of hydrocarbons in which ethylene predominated; propylene was also doubtless present but owing t o the small amount i t could not be separated b y fractional distillation of t h e bromides. Butenes, however, were present, as tetrabromobutane was isolated from t h e mixed bromides. The presence of butenes, whose identification is a n easy matter owing t o tetrabromobutane being a solid and readily purified by crystallization, has been reported before as a decomposition product of the lower hydrocarbons.’ Benzol, toluol, and naphthalene were shown t o be present in t h e t a r . I n the intermediate fractions other homologues are doubtless present, b u t owing t o the small amount of material could not be isolated. I t i s interesting to note that aromatic formatiort h a s t a k e n place here from hydrocarbons of lower carborc content t h a n bepzzene atad that the f o v m a t i o n of aromatics is coincident w i t h a lowlering o f the contelzt of “unsaturated,” the p r i n c i p a l colzstituent o f ’w h i c h i s ethylene. By using copper as a catalyzer a curve of t h e same nature is obtained as may be seen in Fig. 11. T h e ’ aromatic formation takes place a t t h e same temperature and the amounts a t t h e f a t e of 0.4j CU.f t . per hr. are the same. The use of iron as catalyzer gives an altogether different result (Fig, 111). The “unsaturated” content drops suddenly a t about 7 2 j and the hydrogen rises very rapidly. Coincident with these changes is t h e copious deposition of finely divided carbon. There is no t a r formation noticeable beyond a very slight amount of bluish “fog.” I t is of interest t o note t h e rapidity with which the reaction forming carbon and hydrogen is catalyzed above 7 z j ’. Below t h a t temperature t h e decomposition of t h e gas proceeds in much the
Val. 8. N O . 8
same manner as in t h e two previous cases with the exception t h a t the unsaturated content is slightly lower and t h a t the maximum is reached at about j z j o instead of 7 j o c . There can be no question of simple surface action. Copper and iron gauzes of t h e same mesh, cut into equal sizes, were used in both cases and the results are so widely different as t o leave no doubt t h a t we are dealing here with a specific catalytic action. TTith nickel as catalyzer t h e results were even more divergent from + h e decomposition curve of the gas without cntalyzer. Here there was practically nci “unsaturated” formed! t h e highest per cent being 4.9 a t 6 3 0 ” whereas the hydrogen per cent rose suddenly between jlo and 5 8 0 ’ till a maximum of 83.2 per cent was reached a t T O O ” ; above t h a t temperature no readings could be taken as t h e carbon deposited plugged u p the tube in a few minutes. No curve is plotted for the decomposition with nickel as catalyzer but the results are given in Table IT. The formation of hydrogen increases so rapidly and is so much affected b y such slight changes as a slight diminution or increase in temperature as well as t h e deposition of carbon on t h e catalyzer t h a t no c u r r e could be drawn through the points obtained though they indicate a rapid decomposition rate. The results obtained are similar t o those of Buganadsel who b y passing crude Russian petroleum over nickel obtained a gas analyzing 7 2 t o 7 5 per cent hydrogen. EXPERIMENTAL
MATERIAL-The material used was so-called “liquid gas” bought in t h e market in a steel cylinder under I O O lbs. pressure. I t was obtained b y condensation from natural gas. T h e pressure quickly dropped t o 4 2 Ibs. where it maintained itself for a considerable time, falling t o 37 towards t h e end of t h e experiments presented. At t h e beginning of this work t h e analysis
O
1
Norton and Andrews, LOG.df., p, 8.
of propane with 3 per cent butane. Towards the c:nd the proportion of butane increased considerably as would be expected from the decrease in pressure. The results of the analyses are shown in Table I. It is t o be noted t h a t the proportion of C 0 2 increases with time, being always higher t h a n t h a t required for 1 J . Russ. PLys. Chem. Sor., 1910, 195, see also Brooks, Bacon. Padge:t and Humphrey, THISJ O U R N A L , 7 (1915). 182.
A u g . , I 9I 6
T H E J O C R N A L O F I N D C S T R I A L A N D E N G I N E E RI iVG C H E M I S T R Y
pure propane, but never reaching t h a t required b y a mixture t h a t would contain more t h a n 2 0 per cent butane. T.4BLE I-ANALYSES O F GAS D U R I N G COURSE TIMESAMPLED FIRSTWEEK SAMPLE No I IT I11 9.3 9.4 Gas Sample, cc.. . . . 9 . 0 30.7 31 . O Contraction, cc.. . . . 30 i 30.6 30.7 Volume CO?, cc . . . 30.2
OF
I
EXPERIMENTS EIGHTHWEEK
10.1 33.2 34.0
I1 10.2 32.6 33.4
I11 9.8 31.3 37.1
The gas showed no “unsaturated” when these were determined by shaking with bromine water. When shaken with fuming sulfuric in a gas pipette for t h e customary 3 min. a decrease in volume of 3 t o 4 per cent took place. -4s it was shown later t h a t this difference between t h e absorption b y bromine and b y fuming sulfuric decreased as gases t h a t came from a higher and higher temperature were analyzed, it was evident t h a t this absorption was due t o t h e direct absorption of t h e propane and butane by the fuming acid. The gas showed no C 0 2 or hydrogen. APPARArus-The apparatus used is shown diagrammatically in Fig. I V Since t h e pressure used was atmospheric, t h e gas was released from the cylinder into a gas holder and then passed through a meter graduated t o I / 1000of a cu. ft. and provided with a water gauge. The gas passed through two tall drying towers filled with granulated calcium chloride and then into t h e heating chamber, which consisted of a quartz t u b e
I-Gas Cylinder 7-Cooling Coil
were led in through the middle openings, t h e end of the copper coil reaching just below the rubber stopper: they passed out through one of t h e side openings, t h e exit tube reaching t o within ‘/4 in. of the bottom of t h e bottle. I n front of this exit tube was a copper plate connected b y means of a fine wire t o one pole of a n induction coil. Through the opposite opening passed a carefully insulated wire attached t o another copper plate or a piece of copper gauze and connected t o t h e other pole of the induction coil. The gases coming from t h e cooling coil had then t o pass between the t w o plates before passing o u t , and if a sufficient difference of potential was maintained between t h e t w o plates t h e “fog” completely settled in t h e bottle, t h e gases passing out perfectly clean. The difference of potential sufficient for these experiments was furnished b y a n induction coil powerful enough t o give a 2-in. spark, t h e current being supplied by a set of six storage cells. This t a r separator was also kept immersed in ice water throughout the duration of a determination, t h e object being t o keep the vapor pressure of t h e t a r as low as possible. The samples for analysis were taken a t t h e exit of t h e t a r separator. When samples were not being taken, a water trap was placed a t this exit t o prevent air
FIG. IV-APPARATUS USED F O R C R A C K I X G “LIQUIDGAS” 2-Gas Holder 3-Meter 4.5-Drying Towers Coil 10-Storage Battery 8-Tar-Settling Chamber 9-Induction
30 in. long, 3 1 4 in. internal diameter, glazed on t h e inside, heated in a n electric furnace capable of giving a temperature of 1 2 0 0 ’ C. T h e current was controlled by means of a rheostat and t h e temperature determined by means of a platinum-platinum-iridium thermocouple attached t o a millivoltmeter carefully standardized. By means of t h e rheostat t h e temperature could he controlled within 5’. Corrections were made from t h e temperature read off on t h e millivoltmeter by reference t o a correction curve. These varied from 20 t o 2 6 ’ in t h e range of temperatures employed. After passing through t h e heating chamber t h e gases were led through a copper coil made of 3 ft. of ’/s-in. copper pipe. This coil was kept immersed in ice water throughout the duration of a determination. But little t a r condensed in the coil, most of it passing along with t h e gases in the form of a fine “fog.” T o remove this “fog” the gases were passed into a z j o cc. Woulfe bottle provided with three openings. T h e gases
677
€--Heating
Apparatus
from passing into the apparatus. When t h e gases were being absorbed in bromine for t h e determinatictl of t h e nature of t h e “unsaturated, ” a gas-washing bottle with bromine and water was placed beyond the water trap. The catalyzers were copper, iron and nickel, 40mesh wire gauzes cut into uniform sizes 1 2 in. X 4 in. and rolled so as t o fit the quartz t u b e snugly. These rolls were pushed into t h e t u b e far enough so t h a t they would be entirely within the heated area. Copper wool, copper and iron filings were also used, b u t for the sake of uniformity only the gauzes were used in the experiments presented. PROCEDURE-The modtts o p e r a n d i was as fOllOwS: The gas was led through t h e meter a t a measured rate, in the case of these experiments the rate being 0.45 cu. ft. per hr. The rate was determined with a stop watch, t h e flow of gas being adjusted b y slightly opening or shutting t h e intake cock of t h e meter. The gas was passed through t h e cold apparatus t o remove
678
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 TABLE11-ANALYSES OP GASEOUSPRODUCTS
N o CATALYZER UnsatHyTemp. urateds drogen C. Per cent Per cent 600 8.9 .... 650 25.7 13.0 700 35.9 10.9 750 38.5 15.2 .... 785 35.8 825 28.4 21.4 875 11.8 37.0 925 5.7 50.5 995 1.3 62.9
. . . . . . . . . . .
OBTAIXED WITH A N D W'ITHOUT
COPPERAS CATALYZER UnsatHyT,emp. urateds drogen C. Per cent Per cent 625 15.0 .__ 650 16.9 6.8 700 28.4 9.9 755 38.4 15.5 800 36.3 16.l .... 835 29.4 850 15.8 24.3 900 12.6 38.9 51.0 955 7.4 1010 3.2 61.0
the air and t h e furnace quickly brought u p t o the desired temperature b y putting the full current through it. AS t h e temperature was reached the current was regulated so t h a t t h e reading of t h e voltmeter corresponded with the desired temperature. T h e temperature was closely watched during the experiment and regulated b y means of t h e rheostat. Once adjusted i t kept fairly constant within j " . The gas was then passed through t h e heated tube for half a n hour before a sample was taken for analysis. This time was found t o be t h e minimum necessary t o insure concordant results. After the sample of gas had been taken and while it was being analyzed, t h e temperature was quickly brought u p t o t h e next p0in.t and t h e gas passed through for another half hour before the next sample was taken. When catalyzers were being used it was found t h a t they rapidly became coated with carbon. This was especially the case with nickel and iron; copper was not so much affected. With iron the gauze could not be used for more t h a n a n hour a t temperatures below 700' or for much more t h a n half an hour a t higher temperatures. Each point on t h e iron curve above 700' had then t o be determined with a fresh piece of gauze, otherwise no concordant results could be obtained. With nickel t h e deposition of carbon was so rapid above 7 0 0 " t h a t no determination could be made, as a wad of carbon formed in a few minutes within one inch of t h e heated gauze and effectively plugged u p the tube. G A S .mALYsIs-The gases were collected and analyzed over water. For the absorption of t h e "unsaturated", bromine water was used, since, as mentioned above, fuming sulfuric absorbed some of t h e propane and butane. T h e following are typical results: Temperature ..................... Room 650' 700' 750° 800°C. Per cent absorbed by bromine. , , , , 0 . 0 Additional per cent absorbed by sulfuric acid . . . . . . . . . . . . . . . . . . . 4 . 0
CATALYZERS IRON AS CATALYZER UnsatHyTemp. urateds drogen C. Per cent Per cent
14.2
23.4
36.6
34.5
4.0
2.1
0.3
0.1
These showed t h a t as long as a n y undecomposed propane or butane was present some was absorbed by the fuming sulfuric acid, and as the atnount of these decreased with increasing temperature, t h e amount absorbed b y t h e sulfuric acid decreased also. T h e hydrogen was determined b y combustion over copper oxide a t 2 7 5 " C. T h e results of t h e analyses are shown in Table 11. T H E UNSATURATED-The gases coming between 800 and 850' were passed through bromine till t h e latter became colorless. After drying, 4 0 cc. were distilled fractionally in a small distilling flask. After four fractionations the following fractions were obtained: 5 cc.; 129-131', 1 2 cc.; 132-142', 6 cc.; 12j-129', 143-144', 2 cc. 'Thick fumes of H B r came over above 145'so t h a t t h e distillation was stopped. On standing,
1'01.8, No. S
NICKELA S CATALYZER UnsatHyTemp. urateds drogen C. Per cent Per cent
......
...
. . . . . . . . . .
t h e residue in t h e flask solidified into a mass of dark colored crystals. Drained from about 3 cc. of mother liquor and crystallizing three times from alcohol they crystallized in small, colorless prismatic needles melting a t 116". The fraction 129-131' gave a specific gravity of 2.30 a t 2 0 ' . ARORIATICS--NO aromatic formation was observed below 7 0 0 ' . At t h a t temperature t h e "fog" began t o make its appearance, a t first of a bluish tinge, growing darker with increasing temperature till it became dark brown a t about 850'. T h e t a r deposited between 8 0 0 and 900°, was dark in color and quite fluid. It showed a specific gravity of 0.98 j t o 1.000. Above 900' t h e t a r became quite thick and on standing solidified. T h e amount formed varied with the rate. If the latter was I t o I I / * cu. ft. per hr. t h e amount varied around j cc. per cu. ft. of gas. At t h e rate of 0.45 cu. ft. per hr. as much as I O cc. per cu. ft. were obtained. No t a r formation was observed in t h e cases where iron and nickel were used as catalyzers, whereas good yields were obtained with copper. Thirty cc. of t a r were distilled in a small flask and 9.5 cc.; t h e following fractions collected: 80-125', I 2 5 - 1 7 3 " , j . j cc.; 175-200', 3 cc. At 200'naphthalene began t o come over and solidified in the condenser. On further fractionation 6 cc. were obtained belon100" and about 3 cc. between 105 and 110'. These fractions on nitration yielded nitrobenzol and 2:4dinitrotoluol, m. p. 69' C. S U R l LlA R Y
I-It
has been shown t h a t mixtures of propane and butane are decomposed a t high temperatures, giving ethylene, butene and other lower unsaturated compounds, hydrogen and aromatics. 11-The percentage of "unsaturated" increases with increasing temperatures t o a maximum in t h e neighborhood of 7 j o ' C., then decreases with increasing temperatures. The percentage of hydrogen increases with increasing temperatures. The aromatic formation begins a t about 7 5 0 and is coincident with an increase of the rate of hydrogen formation. 111-The action of copper, iron and nickel as catalyzers has been studied. Iron and nickel prevent aromatic formation and favor the decomposition of the hydrocarbons t o free carbon and hydrogen. I'\--The electrical deposition of t a r "fog" in gases from high. temperature decomposition has been found. t o work very satisfactorily for small scale experiments. Further work upon these topics is now in progress in this laboratory. DEPARTMENT O F CHEMISTRY. COLUMBIA UNIVERS~TY NEW YORK CITY
