3 50
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
Vol. 9 , No. 4
THE THERMAL DECOMPOSITION OF BENZENE
They again claimed t h a t Berthelot’s “chrysene” is identical with their “triphenylene.” I n their experiments, they used the method previously employed by HISTORICAL INTRODUCTION Schultz, using a rate of I drop every 3 seconds. T h e The discovery of diphenyl by Fittig’ in 1862 by t h e yield of p-diphenylbenzene was 4 0 g. from 2 kg. of benaction of sodium on brombenzene opened up a new zene. field in organic chemistry. His method, however, Olgiatti’ obtained diphenylbenzenes by likewise was costly a n d it was not till Berthelot2 in the course passing benzene through a red hot tube and collecting of his classic researches on pyrogenetic reactions found the3zoto420’ C. fraction of the decompositionproducts. t h a t benzene, when passed through a red hot tube, Haber2 found that below goo’ C. benzene was not decomposed chiefly into hydrogen and diphenyl, a decomposed and t h a t only above 1000’ C. was it effected much cheaper method of preparation, t h a t impetus with the formation of diphenyl. Haber noticed the was given t o the chemistry of t h a t compound and formation of other crystalline compounds along with methods of improving the yields were worked out by diphenyl but did not investigate them; he remarked various investigators. t h a t naphthalene was not obtained. Berthelot found t h a t b y passing vapors of benzene From t h a t time on, owing to the importance of the through a red hot porcelain tube, a t the rate of one use of benzene for carbureting water gas, many studies gram per minute, he obtained hydrogen and a liquid of the decomposition of benzene’ have been published. product, from which he isolated diphenyl and other McKee3 studied the decomposition of benzene by hydrocarbons which he claimed t o be ((chrysene” benzpassing its vapors through a copper tube heated elecerythrine, and ‘(bitumene.” He stated unequivocally trically and determining the specific gravities of the t h a t between the boiling point of the undecomposed recovered products and their appearance under the benzene and t h a t of the diphenyl, there was no interpolarizing microscope. He made no attempt whatmediate compound, and t h a t no naphthalene, styrolene, ever a t determination of the chemical composition of or anthracene were formed. From the gas, which conthese products. The temperatures were accurately sisted of almost pure hydrogen, he claimed t o have obmeasured and mention is made of the regulation of tained a small quantity of acetylene, though the pres- the rate, but it is not given. His temperatures varied ence of hydrogen sulfide3 interfered with the test for between 448 and 765’ C. and the only conclusions t h a t compound and rendered it uncertain. he drew were t h a t the amounts of decomposition and Following Berthelot’s discovery, Schultz4 intro- the specific gravities increased with the temperature. duced several improvements and pointed out the in- Ipatieff4 found t h a t benzene in the presence of iron fluence of rate and temperature. Instead of passing gave diphenyl and hydrogen above 600”. the vapors from boiling benzene through a hot porceSmith and Lewcock’ passed benzene vapors a t varylain tube, he used a hot iron tube and dropped t h e ing rates through a red hot iron tube in the presence benzene into i t a t the rate of I drop every 3 seconds. of various catalyzers and using varying temperatures. Later, he used 100-200 g. per hour and a not quite The catalyzers used were oxides of calcium, lead, alumi“white-hot” tube. Schultz’ yields were 50 t o 60 num, barium peroxide, etc., but as they remark, the per cent of the benzene used. Luddens’ used a COZ action may not have been purely catalytic, since some stream t o carry t h e vapors of benzene along t h e hot of the oxides were reduced t o the corresponding tube, a method likewise followed by Htibner.8 metals or lower oxides. They find t h a t t h e yield of Schultz’ studied also the tarry products obtained diphenyl is not increased by using temperatures above along with diphenyl, and identified p - a n d m-diphenyl 7 2 0 ”. benzene and claimed t h a t Berthelot’s “chrysene” Recently, Hollins and Cobb6 have found t h a t benzene was only a mixture of diphenylbenzenes and another is not decomposed below 800’ when a mixture of hydrocarbon melting at 266’ C. H e obtained also hydrogen and methane saturated with benzene vapor another hydrocarbon melting a t 196” C., which gave was passed through a hot tube. a compound with picric acid, whereas none of the others Rittman, Byron and Egloff’ studying the decomposiwould. Berthelots objected t o Schultz’ correction, basing his objections on analytical data and put forward tion of aromatic hydrocarbons by passing benzene the claim t h a t Schultz’ product was impure and con- through a n iron tube 6 f t . long a t t h e rate of 2 0 0 cc. per hour, using varying temperatures and pressures, sisted of a mixture of diphenylbenzenes with“chrysene.” Schmidt a n d Schultz9 made a n exhaustive investi- state t h a t they obtained naphthalene and doubtless gation of t h e tarry products from the benzene de- diphenyl, though they did not isolate it. Summing up in review, the decomposition of benzene composition and obtained the following products: by heat takes place in such a way as to split off two Picrate M. .: None 70;s Diphenyl ............................ or more hydrogen atoms from t h e ring, forming diNone 205 e-Diphenylbenzene .................... None m-Diphenylbenzene ................... 85 ,” phenyl a n d diphenylbenzenes. No investigator, preRed needles 196 Triphenylene ........................ . 307-308’ None Benzerythrine ........................ vious t o Rittman, Byron and Egloff, has claimed t o Oily hydrocarbons. ................... By J. E. ZANETTI AND G. EGLOPF Received January 20, 1917
Pitch (b. p. above 450°)
Ann. de Chim., [4] 9 (1866). 445.
1
Ann., 124 (1862). 276.
a
From the thlophene in benzene, Thiophene not yet discovered.
3
Ber., 6 (1872). 682; Ann., 174 (1874). 201; Bcr., 9 (1876), 547. B e y . , 8 (1875). 870. 1 Ann., 209 ( l 8 8 l ) , 339. 7 Ibid., 174 (1874), 230. 9 Ann., 20s (1880). 118. 8 Bull. SOC. chim., [2] 2%(1874), 437.
1
Ber., ’27 (1894), 3387.
1
J . Gasbel., 39 (1896), 377-382. 395-399, 435-439, etc. J . SOC.Chem. Ind., %S (1904), 403. J . Russ. Phys. Chcm. Sac., 89 (1907). 681. J . Chcm. SOC.,101 (1912), 1453-1458. Gas World, 60 (1914), 879. 7 THISJOURNAL, 7 (1915), 1019.
8
4
4
0
6
8
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
APr., 1917
have obtained any naphthalene, and none at all of anthracene or phenanthrene. Indeed, those who devoted any attention t o t h e point, as Berthelot and Haber, state unequivocally t h a t they were unable t o obtain any naphthalene. I n other words, all investigators, previous t o 191j, have been unable t o adduce conclusive evidence t h a t t h e benzene ring breaks u p a t temperatures below 800' C. and atmospheric pressure, in any except two ways: ( I ) Benzene+Carbon Gas, or, ( 2 ) Benzene+ Condensation Products, by t h e splitting off of hydrogen from two or more molecules. Rittman, Byron and Egloff do not state in their publication what test besides t h e boiling point they used in determining t h e presence of naphthalene. Since their claims are entirely in opposition t o all previous work, t h a t t h e amounts of material obtained b y them were small, t h a t t h e boiling points of diphenyl and naphthalene, especially when mixed with other decomposition products, are dangerously near together t o be of value in deciding t h e matter, further evidence on t h a t point seemed desirable. Berthelot's contention t h a t he obtained chrysene was definitely settled b y t h e work of Schmidt and Schultz' who proved t h a t t h e hydrocarbon he believed t o be chrysene melted sharply at 196' and not at "about ~oo',') and further b y Lieberman,2 who obtained pure chrysere melting at 2 j o ' and pointed out t h a t Berthelot's product could not have been t h a t compound. I n connection with other studies on thermal decomposition of hydrocarbons undertaken in this laboratory b y one of us,3 i t seemed desirable t o study more completely t h e course of reaction of t h e decomposition of benzene in order t o arrive at some definite conclusions regarding its behavior a t high temperatures and gradually gain insight into t h e complicated mechanism of formation of cyclic a n d polycyclic compounds from straight chain hydrocarbons.
+
EXPERIMENTAL
The plan of the experimental work consisted in studying the decomposition of pure benzene b y passing i t a t measured rates through a furnace, t h e temperature of which could be accurately controlled. T h e plan included t h e following topics: I-Influence of t h e temperature. 2-Influence of t h e rate. 3-Influence of the metallic catalyzers. 4-Ccmposition of t h e gases. j-Composition of t h e tar. b i A z ERIAL-The benzene used was pure thiophenefree benzene boiling at 80-81' C. of specific gravity 0.881, at 15.5' C. N o test for thiophene was given b y t h e indophenene reaction, a n d on treatment with C. P. concentrated sulfuric acid, no change of color was noticeable in t h e acid. Three hundred cc. of benzene were used in each run whenever possible. I n some cases, as when nickel a n d iron were used as catalyzers a t t h e higher temperatures, not more than one-fourth of t h a t amount could 1
LOC.cit.
2
Ann., 158, 299.
3
THIS JOURNAL, 8 (1916). 674, 777.
351
be used as t h e deposition of carbon would plug u p the tube, and stop t h e run. A P P A R A T U S A N D PROCEDURE-The heating apparatus used consisted of a Whitaker-Rittman furnace, the description of which has been given in detail in THIS J O U R N A L ' and will, therefore, not be described here. The arrangement of t h e apparatus is diagrammatically given in Fig. I. Essentially i t consisted of an iron tube in. in diameter, electrically heated and provided with a rheostat and a pyrometer t o regulate and measure t h e temperature. The pyrometer was a base-metal thermocouple which was compared with a standardized platinum-iridium thermocouple
I
LJ
I
FIG I-APPARATUS 1-Oil feed cup 2-Tube of furnace 3-Nichrome wire resistance 4, 5-Magnesia lining &Stuffing box for pyrometer 7-Condenser tube 8-Water jacket 9-Support of furnace 10-Wide mouth glass bottle
U S E D I N TIWS
11-Exit for gases 12-Electrodes of copper gauze 13-Induction coil 14-Storage batteries IS-Pyrometer rod 1 b P y r o m e t e r scale 17-Pyrometer support l&Voltmeter 19-Rheostat
and found t o check u p within 5 t o IO', according t o temperatures. The temperature could be maintained constant within 5' C., for long periods of time. The benzene was led into the tube from a large oil feed cup with a sight tube b y which t h e rate of flow could be accurately determined b y counting t h e drops falling in a certain time. One drop per second gave a flow of IOO cc. per hour, two drops 2 0 0 and three drops 300 cc. These values were determined experimentally, as, of course, t h e size of t h e drop will vary with the size of t h e opening. The benzene did not drop directly into t h e heated tube, even a t the higher rates, b u t flowed along the sides and vaporized before reaching 1 6 (1914), 472.
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
352
the area of high temperature. This is a n important point in a vertical furnace since some of the benzene could readily fall through the tube without vaporizing, owing to the small surface of the spherical drop and the short time it would be exposed to heating,
?
I
50U"
I
35U" 600" 650' 7dU" FIG.11-INFLUENCE OF RATE
I
750" 8
The condensing apparatus consisted of a piece of iron pipe '/z in. in diameter, 3 f t . long, surrounded by a water jacket. The water cooling was found unnecessary, in fact, detrimental, as the diphenyl was apt t o solidify and choke the condenser. The cooling was quite sufficient as was shown b y recoveries of 98 per cent of the benzene used in some of the runs where little decomposition t o carbon occurs. The receiver consisted of a wide mouth glass bottle provided with two electrodes connected with an induction coil and a set of storage batteries. This arrangement was found very convenient t o settle the "fog" from the gases obtained by decomposition a t higher temperatures, and afforded a means of accomplishing complete deposition in the receiver. The whole apparatus was washed with benzene before the beginning of a new run in order to avoid vitiating results by dqcomposition products from a previous run. The decomposition products of t h e reactions were distilled in a 500 cc. flask provided with a Glinsky column and a 24-in. condenser for the lower boiling fractions (undecomposed benzene). The residues were transferred to a standard Engler distilling flask and the distillation carried on up t o 300' C. Mos? of t h e distillate came over between 2 j o and 2 7 5 ' (diphenyl). A small fraction coming over between 2 0 0 and 2 j 0 ' , was collected separately and carefully tested for naphthalene. T h e residue (tar) was in most cases too small t o be fractionated separately, so t h a t the tars from a series of similar runs were added together and fractionated. CATALYZERS-The catalyzers were pieces of metallic gauze 40 mesh in case of the copper and iron, 40 mesh and 50 mesh in the case of nickel, cut into pieces
Vol. 9, NO. 4
I O in. X 1 2 in., rolled SO as to fit the tube and inserted into the part of the tube above the pyrometer. Frequent changes were necessary as they quickly became coated with carbon. This deposition of carbon was particularly objectionable in the case of nickel and iron where plugging of the tube would occur before much of the benzene could pass through. GAsEs-Samples of gas were collected for analysis a t the exit from the tar separator. T o test for acetylene, a large test tube half filled with a n ammoniacal cuprous chloride solution, and provided with a twohole stopper and a tube reaching t o t h e bottom, was inserted a t the end exit from the separator tube. T o measure the amount of gas given off in the reaction a Referee meter graduated t o 0 . 0 0 1 cu. f t . and provided with a thermometer was placed a t the end of the t a r separator. The gases were analyzed over water for unsaturated, by absorption with fuming sulfuric, for hydrogen by passing over copper oxide a t 2 9 0 - 3 0 0 ' and for methane by explosion with oxygen. S O U R C E S OF ERROR-The chief source of error was the deposition of carbon in the tube and on the catalyzers as well as on the pyrometer rod. This tended t o obstruct the passage of t h e gases and interfered with the temperature measurements. The higher temperatures are probably too low since the deposition of nonconducting finely divided carbon would act as a heat insulator. Since the pyrometer projected into the middle of the tube the temperature measured a t the beginning of a run was the actual temperature of the vapors and gases and not merely t h a t of the walls of the tube. The carrying of undecomposed benzene in the form of vapor by the hydrogen introduces quite a n appreciable error in some of t h e determinations. This
60
50
40
30
20
JU
0 FIG. 111-INFLUENCE
OF
RATE
error increases with increasing decomposition of the benzene and consequent increase in the volume of the gases. At ordinary temperatures ( 2 0 ' C.), t h e vapor pressure of benzene is 74.6 mm. or nearly I O per cent of the total volume of hydrogen given off. As
T H E JOI’RNAL OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
Apr., 1 9 1 7
353
FIG. IV-INFLUENCE OF RATE
there is produced b y t h e reaction, products which dissolve in t h e benzene and lower its vapor pressure, this amount of benzene in t h e gas is sensibly lowered. Experimentally we found this percentage t o be about 7 . 3 per cent (see Table 111). Calculations of per cent TABLEI-EFFECT
Temp. C. 5 00 550 600 650 700 750 800
OF
RATE (PER CENTS
SP. GR. RECOVERED OILS PER CENT DIPHENYL AND TAR 100 cc. 200 cc. 300 cc. 100 cc. 200 cc. 300 cc. per hr. per hr. per hr. per hr. per hr. per hr. 1.0 0.884 4.0 0.886 4.0 2.0 0.884 0.886 4.7 7.5 0.887 0:&3 0.886 3.7 7.7 9.5 0.890 14.5 0.897 0.891 16.0 20.0 10.3 0.896 0.920 0.910 21.3 26.0 32.0 0.916 0,936 0.950 34.0 6.8 0.945 22.7 0.956 0.950
....
...
....
SP. GR. RECOVERED OIL Nickel Cu Fe (a) . 0,884 0 , 8 8 4 0.884 0.884 . . . . . 0.885 0.885 0.884 0.889 0.887 0.885 0.900 0.898 0.926 0.902 0.884 0 . 9 3 6 0.906 . . . . . (a) 40-mesh Nickel
.. . .
BASIS OF TOTALBENZENEUSED)
PER CENT DIPHENYL 200 CC. 300 CC. per hr. per hr. per hr. 0.6 .... 0.1 .... 0.8 0.9 2.5 1.6 2.2 7.4 4.9 10.9 13.5 8.1 14.6 14.7 21.5 22.6 10.7 16.9 5.4
PER CENT TAR I00 CC. 200 CC. 300 CC. per hr. per hr. per hr. 3.4 0.9 1.1 3.2 5.0 1.5 3.1 2.1 3.6 2.8 2.5 5.4 2.2 4.5 9.4 6.6 17.3 19.4 17.1
....
PER CENT BENZENE DECOMPOSED TO FORMCARBON AND GAS 100 CC. 200 CC. 300 cc. per hr. per hr. per hr. 11.0 1.0 .... 2.0 12.0 14.0 3.5 2.3 15.5 7.5 7.3 19.3 14.0 8.0 26.0 16.0 8.7 57.3 31.0 16.0
...
.. ...
DISCUSSION O F DATA
RATE-The results regarding t h e influence of rate are given in Table I and Figs. I1 and 111. Decomposition begins, even with the faster rates, as
....
b u t t h e precautions taken in these experiments t o prevent contaminations from a previous run and the fact t h a t we obtained diphenyl repeatedly a t temperatures as low as joo’ can leave no doubt on this point. If t h e added percentages of diphenyl and t a r as giving the total amount of synthetic products (diphenyl, diphenylbenzenes, etc.) formed, are plotted against temperatures, there is no marked regularity below 600’ owing t o t h e fact t h a t t h e percentages are so close together as t o come within t h e limits of experimental error. Above t h a t temperature, t h e decomposition proceeds with regularity, being greater i n t h e case of t h e slower rate which attains a maximum a t 7 50’. Above t h a t temperature, the decomposition t o gas and carbon (Fig. 111) proceeds so rapidly in t h e
TABLE11-EFFECT OF CATALYSIS (PER CENTS ON BASIS TOTALBENZENE) PER CENT DIPHENYL AND TAR PER CENT DIPHENYL PER CENT RESIDUET A R PER CENT Nickel Nickel Nickel None Cu F e (a) ( b ) None Cu F e (a) ( b ) None Cu F e (a) None Cu 1.0 . 1.9 , 0.1 0 . 8 . . . 0.0 0 9 1.1 1.0 .... 4.0 1 . 3 2.7 . . . . . . 0.8 b:4 0.8 0 . 0 0.0 3:2 1 : i 1 . 9 2.0 2.7 6 . 0 2.7 4.7 1.7 0.0 1.6 1 . 0 1.8 0 . 8 0 . 0 3 . 1 1.7 3 . 2 1 . 6 2.0 8.0 7.7 5.0 7.0 3.3 2.2 4.9 3.1 3.9 1.9 0.8 2.8 1.9 3.1 1 . 4 7.3 11.7 8.1 9.8 7.6 . . . 1.9 2.2 1 . 5 4 . 4 . . . 8 . 5 18 7 10.3 11.3 12.0 . . . 5 . 8 21.3 23.3 11.0 5 . 0 0.0 14.7 18.1. 5 . 6 0.5 0 . 0 6 . 6 5 . 2 5.4 4.3 10.7 2 4 . 0 34.0 11.7 10.0 . . . ... 16.9 8 . 6 2 . 2 ... ... 17.1 3.1 9 . 8 . . . 16.0 4 5 . 0 ( b ) 50-mesh Nickel
introduce very large percentage errors if results are calculated on t h a t basis. INFLUENCE OF
ON
100 CC.
decomposition on basis of “actual benzene used” (total b e n z e n e b e n z e n e r e c o v e r e d ) should be corrected b y subtracting from t h e “actual benzene” t h e amount of benzene vapor which theoretically could be recovered b y cooling t h e gases t o a very low temperature. This correction may amount t o nearly 13 per cent of t h e benzene used. This correction applies likewise t o t h e “per cent of carbon and gas” formation] if t h e latter is found b y subtracting from t h e total benzene t h e amounts of diphenyl a n d t a r obtained. For this reason t h e percentages calculated in this paper are given on the basis of “total benzene.” Moreover, in some cases, especially at lower temperatures, t h e amounts of “actual benzene” are so small as t o Temp. C. None 500 0.884 550 0.884 600 0.886 650 0.890 700 0 . 8 9 6 750 0.916 800 0.945
low as joo’. I n all cases below 600°, t h e amounts decomposed are small, b u t quite appreciable, and leave no doubt t h a t diphenyl and other decomposition products are formed a t those temperatures. This is quite in opposition t o t h e claims of Ipatieff and of Haberl
... ...
GASAND CARBON Nickel Fe (a) (b) 4 . 0 15.0 3 6 . 0 5.3 . . . . 7.0 2 3 . 3 5 j : O 8.0 30.0 70.0 12.0 . . . . 79.2 3 8 . 0 65.0 100.0 72.0
.... ....
case of t h e IOO cc. rate as t o cause a marked falling off in the percentage of synthetic products. The gas and carbon formation being slower in the 2 0 0 cc. rate, its effect on the synthetic products is less marked and although the percentage of synthetics is greater a t 1
See historical introduction.
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
3 54
Vol. 9, No. 4
800’ t h a n a t
7 0 0 ° , the rate of increase is much smaller cases of reactions in the gas phase. It must be also t h a n iq the previous 50’ rise. Finally, in t h e 300 remembered t h a t , not being a product of the reaction, cc. rate, the amount of gas and carbon being much ~ C ~ H ~ * C I ~ H I O H1, it may act in this case as a smaller t h a n in either of the previous cases, its in- true catalyzer. fluence on the synthetics is scarcely noticeable, and The action of copper, iron a n d nickel as catalyzers their rate of increase is t h e same between 7 j o and 800’ is much the same as t h a t observed by one of us1 in as between 700 and 750’. connection with the decomposition of straight chain If now we split the “synthetics” into their com- hydrocarbons. The action of the copper is almost ponents and plot these separately (Figs. IV and V), nil up t o 700’ C. At g o o 0 , however, a marked dewe find t h a t diphenyl decreases markedly above ? s o 0 , crease in the diphenyl and t a r formations takes place except in t h e case of the 300 cc. rate where there is a with consequent increase in t h e carbon and gas. There small increase. I n any case, the rate of diphenyl may be a question whether this decrease is due t o specific formation is markedly diminished. T o counterbalance catalytic action or merely t o increased surface. Whatever question there may be regarding the copper, there can be none regarding nickel and iron.
+
.)
30
R/pheny/ on hasis of tofi?/ Berrzene
25.
c, -
20
FIG.VI-INFLUENCE
OF
CATALYZERS
this, we find t h a t in all cases the rate of t a r formation increases very markedly from 7 j o to 800’. It would t h e n appear t h a t the tar formation occurs a t the expense of diphenyl and t h a t the diphenylbenzenes and other condensation products are formed by t h e action of diphenyl on benzene. I N F L U E N C E O F CATALYZERS-The catalyzers studied were copper, iron and nickel. For the sake of uniformity, wire gauzes of the same mesh were used, except in the case of nickel when two different meshes were used. The rate in these cases was 300 cc. per hour. The graph with “no catalyzer” is given for the sake of comparison. It is, of course, a question whether the finely divided carbon which forms abundantly a t the higher temperatures is not itself a catalyzer. As, however, the amount formed on the sides of t h e tube was the same in all cases, its influence, so far as t h a t amount is concerned, is negligible for purposes of comparison. As to whether a product of a reaction can be called a catalyzer, it must be borne in mind t h a t a t the higher temperature the speed of the reverse reaction, 6C 3Hz---tC&, is immeasurably small, a n d t h a t the action of the finely divided carbon may be in the nature of a surface action, so important in
+
+--
1
FIG.VII-INFLUENCE OF CATALYZERS
These are true catalyzers for the reaction CeHB = 6C 3Hz (see Fig. VII). Of course: the surface exposed has a marked effect on the action as may be gathered from the results obtained with two different meshes of nickel gauze. GASES-Analyses of the gases are given in Table 111.
+
TABLE 111-GAS FORMED AT 100 Cc. PER HOUR RATE LITERSPER 100 C c . PERCENTAGES OF TOTAL GAS Temp. C.
BENZENEUSED Total Gas Hydrogen
700 750 800
2.8 4.2
2.h
8.5
7.8
3.9
12.7
11.8
28.3 49.5 63.7
26.2 36.1
59.2
Unsaturated Hvdrozen 7.2 -90. 7.7 89.9 7.2 89.8 7.2 90.6 7.3 88.3 7.3 90.5 7.1 91.8
s
Residue 2.8 2.4 3.0 2.2
4.4 2.2 1.1
The per cent of “unsaturated” is so constant a n d corresponds so well to the amount of benzene vapor t h a t it cannot be ascribed t o any other cause. Careful tests for acetylene with ammoniacal cuprous chloride proved absolutely negative. Of all investigators of t h e thermal decomposition of benzene, Berthelot is the only one t o report acetylene,2 a n d he was none too sure about its presence, owing t o the interference of sulfur compounds in his benzene which generated H2S. 1
LOC.cit.
1
See historical introduction.
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
Apr., 1 9 1 7
It seems t o us t h a t if acetylene is formed b y t h e thermal decomposition of benzene a t atmospheric pressure, we should expect some of t h e condensation products t h a t it forms with benzene, such as naphthalene, styrolene, anthracene, etc..' whereas none of these products has been reported2 a n d in fact, Berthelot positively declares they are not formed. I n t h e analysis of t h e gas, there always remained a residue after combustion of t h e hydrogen over copper oxide, which could not be exploded with oxygen. This residue, which amounted t o about 2 . 5 per cent, was undoubtedly nitrogen. The very large volume of t h e system and t h e necessity of taking gas samples before t h e almost inevitable plugging of t h e tube with carbon, made i t practically impossible t o sweep
P 90
80
70 60
SO
30
20 /O
0 560O
2 FIG.VIII-INPLUBNCE
OF CATALYZERS
out all t h e air from t h e apparatus with t h e hydrogen produced in t h e reaction. TAR-The tarry compounds remaining after 2 7 j " were redistilled and extracted with alcohol, according t o t h e method of Schmidt and S c h ~ l t z . ~ We were able t o obtain without difficulty t h e same compounds separated by those authors: m- and p-diphenylbenzenes melting a t 86" and triphenylene 1 9 6 O , this last giving a compound with picric acid. The amount of t a r obtained from each separate run was so small t h a t no quantitative relation of t h e influence of rate on the formation of its individual constituents could be ob1
Bull. SOC. C h i n . , 7, 218, 278. 306. Byron and Egloff Ann., 208, 118.
* Except naphthalene by Rittman, *
tained without the introduction of a n enormous percentage error and for t h e present only t h e qualitative results can be reported. CONCLUSIONS
From t h e above data, as well as from those of previous investigators, i t seems settled t h a t the decomposition of benzene a t high temperatures (600-800°) takes place in accordance with t h e following reactions: Carbon ( and -C- Benzene Hydrogen
i.).
+ Diphenyl
Diphenylbenzenes Prod)j(Condensation ucts Containing 3 or more Benzene Rings)
At temperatures below 7 j o " , t h e reaction velocity of the diphenyl formation is greater t h a n t h a t of carbon and hydrogen. Above t h a t temperature, the velocity of t h e carbon-hydrogen reaction becomes very great and decomposes t h e benzene before other condensation products can be formed. It seems of interest t o remark on the stability of t h e benzene ring and on its complete decomposition t o carbon and hydrogen without appreciable qualztities of intermediate products, such as acetylene, ethylene, ethane or methane a t high temperatures. No 'conclusive evidence has thus far been brought forward t o show t h a t t h e ring is broken a t high temperatures with formation of any products but carbon and hydrogen. The remark of Rittman, Byron and Egloffl t h a t diphenyl is not a n equilibrium product but merely an intermediate one seems t o be borne out b y our results with this difference, however, t h a t whereas those authors, as well as McKee,2 believed t h e equilibrium t o be Benzene+ Diphenyl+ Naphthalene; i t is, in fact, BenzeneDiphenyl+ Diphenylbenzenes. There have been no catalyzers so far brought out t h a t catalyze t h e formation of diphenyl. Those studied b y Smith and Lewcock proved ineffective and those presented in this paper catalyze only t h e formation of carbon. F O R M A T I O N O F NAPHTHALENE-owing t o t h e fact t h a t this compound was reported present in t h e decomposition of benzene, b y Rittman, Byron and Egloff, very careful tests both as t o melting and boiling points and formation of picrate were made on t h e first fraction t h a t came over on distilling the reaction products, after t h e undecomposed benzene had boiled off. The boiling point of naphthalene is 218', whereas t h a t of diphenyl is zj4". If any naphthalene were present, i t would appear in the first fraction of the distillate which should come sensibly below the boiling point of diphenyl. The first fraction after the separation of the benzene did boil much below diphenyl and for t h a t reason was separately collected, crystallized repeatedly and treated in alcohol solution with picric acid, according t o t h e method given b y Milliken.3 I n no case was a melting point higher than 7 0 ' obtained, using a standard short-stem thermometer; neither was t h e slightest precipitate of naphthalene picrate obtained, though the method was checked with satisfactory results both against pure naphthalene and 1
THISJOURNAL, 7 (1915). 1020. J . SOC.Chsm. Ind., 47 (19041, 403.
8
"Identification of Pure Organic Compounds" (John Wiley & Sons).
1
(Y.
355
3 56
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
mixtures of naphthalene and diphenyl in proportions from I : I , t o I of diphenyl : 0.01 of naphthalene. As a further test t h e accumulated diphenyl from all experiments, amounting t o about I kilogram, was distilled in a large flask a n d t h e first few cubic centimeters of distillate collected a n d submitted to the same tests with negative results. The presence of naphthalene in Rittman, Byron a n d Egloff's product a t atmospheric pressure can be ascribed by the writers only t o the fact t h a t those authors used commercial benzene, which doubtless contained toluene. This last on heating with benzene gives naphthalene as shown by Carnellay' by passing mixtures of benzene and toluene through a red hot tube. The specific gravity of the benzene used by Rittman, Byron and Egloff a t 15' was 0.879, whereas t h a t of The difference t h e one we used was 0.881 a t 15.5'. is small indeed but it tends t o show the presence of some toluene (0.871a t 15.5') in the material used b y those authors.
Vol. 9, No. 4
is rubber hose, which is held t o be the most desirable material for conveying gasoline from the tank t o t h e engine on aeroplanes. It was found, a t the time t h a t a system of inspection for aeroplane materials was initiated in England, t h a t the quality of the rubber being used was quite unsatisfactory. This was probably due in a large measure t o the unfamiliarity of the rubber manufacturers with the properties required of gasolineresistant rubber hose. They appear in some cases t o have endeavored t o adapt their regular types of material, such as steam- or water-hose, for use with gasoline, with results which were sometimes disastrous t o the life of the aeroplane a n d t o its pilot. The Underwriters' Laboratories of Chicago have published (May, 1914) their requirements for rubber hose t o conduct gasoline, and their tests are very thorough. When the hose is t o carry gasoline on aeroplanes some modification of the Underwriters' examination seems desirable. Bursting, tensile, a n d SUMMARY stretch tests on the original hose are described by t h e I-The thermal decomposition of benzene has been Underwriters, and tensile strength is also determined studied a t temperatures varying from joo t o 800' C. 24 hrs. after gasoline immersion. It may be pointed out t h a t after such a period the rubber recovers, with and atmospheric pressure. 11-The chief products are diphenyl, diphenylben- little deterioration, its original physical properties zenes, carbon and gas. The formation of diphenyl It is important t o obtain a hose which will show good physical behavior while immersed in gasoline or imbegins a t as low a temperature as 500'. 111-No acetylene was found in t h e gas, which con- mediately after removal therefrom. Bursting and sisted of hydrogen saturated with benzene vapor, and tensile tests were given up as specification tests by t h e no naphthalene in t h e decomposition products, tend- writer because, although they furnish helpful informaing t o show t h a t the thermal decomposition of benzene tion, it was found t h a t they were not adequate t o deat atmospheric pressure takes place with the forma- tect tubes which would behave badly towards gasoline tion of condensation products in which t h e benzene when used on aeroplanes. I n drawing u p specifications for gasoline-resistant ring apparently remains intact, or with the formation rubber tubing the following tests were carried out, of hydrogen and carbon. IV-The effect of rate on t h e yields of diphenyl and found t o be suitable for 'controlling the quality has been studied. T h e slower rates are more favor- of the tubing. They are intended t o amplify and not t o supplant able t o t h e formation of diphenyl. The optimum the specifications of the Underwriters' Laboratories. temperature is in the neighborhood of 750'. Above ( I ) FLEXIBILITY-The tube is bent t o a circle having t h a t temperature, diphenyl benzenes as well as cara diameter ( D ) which varies according t o t h e inside bon and hydrogen form readily. V-The catalytic action of copper, iron and nickel diameter of the tube ( d ) as shown below: d D has been studied. Iron a n d nickel favor t h e decom8 times d Up t o 1/2 in. position t o carbon and hydrogen. The action of 10 times d 9/1s in. t o 1 in. 12 times d 11/16 in. to 1 1 / 2 in. copper is not marked except above 750'~ when t h e 14 times d Above 1 1 / 2 in. formation of carbon is accelerated. Further work upon these topics is now in progress The diameter of the tube so bent should not change a t any point by more t h a n I O per cent from its original in this laboratory. diameter. DEPARTMENT OB CHEMISTRY COLUMBIA UNIWRSITY A tube of poor flexibility will show a permanent N E W Y O R K CITY weak spot if bent sharply a dozen times a t the same place. RUBBER HOSE FOR USE ON AEROPLANES (2) I M M E R S I O N I N GASOLINE-At first an immersion By PERCYA. HOUSE MAN^ test of 2 0 0 hrs. in cold gasoline of about 0 . 7 2 0 sp. gr. Received December 18, 1916 was used. I t was found t h a t approximately the same T h e enormous increase in t h e world's output of effect can be attained in a shorter time b y boiling in aeroplanes during the last three years has rendered gasoline for I hr. under a reflux condenser, followed important a systematic inspection of materials used by 24 hrs.' standing in gasoline a t room temperature. in their construction. The approximate increase in weight and volume of One of the materials requiring detailed examination the sarnole (about .q in, long) is recorded for correlaI . 1 J . Chcm. Soc., 97, 712. tion with the results of the other tests. The decrease 1 Formerly Chief Examiner, Aeronautical Inspection Department in bore at the narrowest part of t h e tube is also noted Laboratories, London, England.
