T H E J O U R N A L OF IiVDUSTRIAL A N D EiVGINEERIXG C H E M I S T R Y
20
ing spring crops. T O test its value t o the inarket gardener, three plots each z ft. X 3 ft. were laid out in a field. One plot was not fertilized, one was fertilized with a n equivalent of 126 lbs. of nitrogen, one t o n of sludge per acre, and the third with an equivalent of extracted sludge. On April 24, 191j , two rows of radishes and lettuce were planted in each of t h e three plots. The plants in t h e plot where the extracted sludge was used came up first. a little ahead of those in the plot where t h e unextracted sludge was used. A t t h e end of z weeks t h e lettuce and radishes of the treated plots appeared t o be twice t h e size of those in the untreated plot. At t h e end of 4 weeks t h e plants were thinned. The roots of the radishes from the treated plots were already red and quite rounded near the tops while those from t h e untreated plots had not yet started t o s n d and had not become red. The lettuce plants from the treated plots were nearly twice as large as those from t h e untreated plots. On June I , 38 days after planting, t h e six best plants of lettuce and radishes were taken from each plot and are shown in Fig. IV. T h e differences in size are very marked. COMPARISOX OF THE LETTUCEAXD RADISHESPROM UNFERTILIZEDA N D FERTILIZED PLOTS Plot Treatment Wt. of lettuce Wt. of radishes l , . . . . . . . None 4 . 5 g. 2 3 . 4 g. 2 . . . . . , . . , Sludge 6 . 3 g. 6 3 . 0 g. 3 . . . . . . . . Extracted sludge 6 . 8 g. 6 8 . 0 g. I
The increase in weight, due t o t h e sludge, is 40 per cent in t h e lettuce, and Ijo per cent in t h e radishes. T h e radishes from t h e sludge pots, when cut open and eaten, were found t o be very crisp and solid, a n d t o have a good flavor, These pot cultures and gardening experiments show t h a t the nitrogen in “activated sludge” is in a very available form and t h a t activated sludge is valuable as a fertilizer. STATE WATERSURVEY UNIVERSITYOF ILLINOIS, URBANA
EQUILIBRIUM RELATIONS AMONG AROMATIC HYDROCARBONS PRODUCED BY CRACKING PETROLEUM’ By W. F. RITTMANAXD T. J. TWOMEY Received October 15, 1915
The results of several series of experiments on t h e cracking of petroleum in the vapor phase have served t o furnish evidence as t o t h e course a n d mechanism of t h e cracking reaction. I t has been indicated t h a t decompositions occur with two general effects: ( I ) decrease in size of molecule, and ( 2 ) decrease in saturation. A study of t h e relations among classes of hydrocarbons2 has shown t h a t certain conditions of temperature and pressure are favorable for t h e production of low-boiling aliphatic compounds, certain others for aromatics and still another set for t h e formation of carbon and gas: viz., u p t o 500’ C. for aliphatic formation, from j o o o t o 800’ C. for aromatics, and above 800’ C. for carbon a n d gas. A special study3 has dealt with t h e field of gas pro1
Published with the permission of the Director of the Bureau of
Mines. 2
3
Rittman, THISJOURNAL,7 (1915), 945. Whitaker and Rittman. Ibid., 6 (1914), 383, 472.
Vol. 8, S O .I
duction. and work has been done also t o determine conditions obtaining in t h e temperature range favorable for aromatic formation. One section of this work1 dealt with transformations in which pure aromatics were used as starting out material and cracked under carefully regulated conditions of temperature and pressure. The results described in t h e present communication approach t h e same end in a different manner. Here petroleum has been subjected t o cracking and, by t h e careful analysis of products, important relations have been discovered among the degrees of formation of certain aromatic compounds. THE 0 R E T I C4 L
I n the other experiments of t h e present general series t h e effects of temperature and pressure have been studied with care. I n t h e present study it has been necessary t o minimize the importance of these factors on account of experimental conditions which will receive discussion later. I t has been assumed t h a t t h e extent t o which cracking proceeds is in a general way proportional t o t h e specific gravity of t h e recovered oil or t a r . T h e absolute t r u t h of this assumption may be open t o some question but of its general correctness there is little doubt and it affords a convenient method of combining in one function t h e effects of the several variables in t h e cracking reactiontemperature, pressure, rate of feed and contact surface. The scheme followed has been to note t h e variations, with specific gravity of cracked oil, of t h e percentages of five hydrocarbons : benzene, toluene, xylene, naphthalene and anthracene. These five have been selected, partly on account of their scientific and commercial importance and partly because convenient analytical methods are available for their estimation. I n connection with a n y series of experiments i t is of importance t o consider a n y indications which may appear on t h e basis of previous knowledge. For the cracking reaction in general i t is possible t o make clearcut predictions as t o t h e effects of temperature and pressure. By calculating approximate equilibrium constants according t o t h e Nernst formula it is possible t o discover how t h e course of reaction varies with temperature. For t h e effect of pressure it is t o be noted t h a t increase in this variable is favorable in accelerating t h e cracking reaction up t o such a velocity t h a t equilibrium may be attained in the time allotted. Beyond this point increase in pressure is favorable or unfavorable, according as t h e reaction proceeds with decrease or increase of total volume. For t h e work here described i t may be noted t h a t all indications point toward t h e fact t h a t equilibrium was approached only remotely. The effect of pressure upon reaction velocity is therefore the one of importance. I n t h e present experiments primary consideration has not been given t o t h e effects of temperature and pressure, b u t instead t o t h e relative amounts of t h e various hydrocarbons produced under each set of cracking conditions. Temperature conditions were of course as significant as ever, b u t owing t o mechanical 1
Rittman, Byron and Egloff, THISJOURNAL, 7 (1915), 1019
J a n . , 1916
T H E J O U R N A L O F I N D U S T R I A L A N D E LVGIiV E E RI N G C H E M I S T R Y
difficulties no a t t e m p t was made t o read a n d control t h e m with a high degree of accuracy. Previous experience has indicated t h a t large paraffin molecules are less stable t h a n small, t h a t paraffins a r e less stable t h a n olefines, a n d t h a t stability increases with decrease in saturation. I t appears, therefore, t h a t benzene formation is favored b y more strenuous cracking conditions t h a n toluene formation, since t h e latter compound is in part aliphatic, due t o t h e presence of t h e methyl group which makes u p t h e side chain. Likewise, i t appears t h a t xylene, having t w o methyl groups, should show t h e same variation in stability as toluene b u t t h a t it should always occur in about
F I G . I-PERCENTAGES OF B E N Z E N E , T O L U E N E , X Y L E N E S . N A P H T H A L E N E , AND ANTHRACENE I N CRACKED PETROLEUM O I L S OF
VARIOUSSPECIFIC GRAVITIES
half t h e q u a n t i t y of t h e latter. T h e probability of its formation is half t h a t of t h e formation of toluene a n d t h e probability of its decomposition twice as great. T h e formation of naphthalene a n d anthracene seems t o necessitate t h e decomposition of benzene nuclei. T h e formation of these polycyclic hydrocarbons should, therefore, be coincident with a decrease in the number of monocyclic molecules.
21
in, in diameter was obviously a proposition involving difficulty. The runs of t h e present series were conducted with much care a n d with a knowledge of details of manipulation acquired in the course of several hundred previous experiments of t h e same sort. T h e temperatures recorded are, however, representative of only a moderately accurate average a n d no stress is laid upon t h e m in t h e present connection. Conditions of t h e runs were so regulated as t o produce cracked oils of specific gravities varying over a considerable range. One other source of error must be menfioned which helps explain t h e failure t o obtain results which might be more regular, although not more conclusive t h a n those a t present recorded. T h e furnace was equipped with condensers which were insufficient t o cool properly t h e cracked oil, a n d errors were introduced on this ,account. GENERAL SCHElIE OF PROCEDURE-The Original material was a special fuel or gas oil, a distillate obtained from Pennsylvania crude. I t s specific gravity was 0.83 a n d its boiling range between 300' a n d 400'. Runs were made with t h e furnace described above, t h e condensed liquid (cracked oil) being t h e only product collected a n d examined. Percentage recoveries were determined a n d specific gravities measured. The method of analysis was t h a t of distillation a n d specific gravity. A detailed description of this is t o be found in another communication,l a n d is essentially a s follows: T h e oil is distilled through a n efficient fractionating column of t h e Hempel type and a cut made a t 17 .'j This cut is t h e n subjected t o t w o successive refractionations a n d divided into: ( I ) Benzene fraction, Bo
EXPERIMENTAL
The method of procedure adopted in t h e present experiments was t h e same in principle as t h a t described in earlier connections b u t has varied considerably in detail. All earlier experiments made use of a small electrically heated furnace in which were maintained carefully regulated a n d measured conditions of t e m perature a n d pressure. I n t h e present work i t has been necessary t o obtain larger quantities of cracked oil t h a n could be produced conveniently under former conditions a n d , consequently, use has been made of one of t h e large experimental furnaces built i n connection with t h e commercial development of t h e vapor phase cracking process. T h e furnace proper was a t u b e I O f t . long a n d 8 in. in diameter, heated b y gas. I t s qse permitted a feed of from 1 5 t o 2 0 gallons per hour of original oil a n d a correspondingly large recovery ( 2 0 t o j o per cent of t h e original). T o offset t h e advantage of yielding sufficiently large quantities of cracked oil t o permit t h e making of satisfactory analyses there were t h e following difficulties which should receive mention, although in no way did t h e y decrease t h e value of t h e results obtained. Pressures could be regulated as well in t h e large system a s in the small, b u t t h e accurate control of temperat u r e throughout t h e whole of a t u b e I O f t . long a n d 8
45
30
15
a ,150
,775
FIG. 11-PERCENTAGES OF PRIMARY
m
,825
.%a
,815
B E K Z E N E , T O L U E N E , AND XYLENES IN 1 O F V A R I O U S SPECIFIC GRAVITIES
DISTILLATION CUTS
75O
o 0 t o 9 j " ; ( 2 ) Toluene fraction, 9 5 " t o 120'; (3) Xylene fraction, 1 2 0 ' t o 175" C. A careful series of experiments has established t h e specific gravities of t h e non-aromatic constituents present in these cuts as: 0.720, 0.730 a n d 0.740. Knowing t h e gravities of t h e pure aromatics a n d measuring those of t h e distillates it is a simple matter t o calculate percentages of t h e three hydrocarbons. T h e methods used for naphthalene a n d anthracene were not of t h e same degree of accuracy as those for Rittman, Twomey and Egloff, M e t . and Chem. Eng., 13 (19154, 682.
22
T H E J O U R N A L O F I N D U S T R I A L A,VD E N G I N E E R I N G C H E M I S T R Y
t h e monocyclic hydrocarbons, b u t t h e results have proven t o be satisfactory in t h e present connection. These two hydrocarbons are contained in t h e cut above 1 7 j" left after t h e first distillation. This cut was subjected t o further fractionation a n d two portions separated, one boiling from 1 7 j O t o 2 7 0 ° , t h e other from 270' u p t o t h e point where actual t a r begins t o come over. T h e t w o hydrocarbons were estimated b y t h e simple process of chilling, pressing out a n d weighing t h e separated solid. DISCVSSION OF RESULTS
The results of t h e experiments. given in Table I, are not particularly easy t o interpret. By plotting, however, t h e indications of importance are brought out in a striking manner. I n Fig. I t h e abscissas are specific gravities of recovered oils, a n d t h e ordinates, percent-
1701.
8, I'so.
I
aromatics. benzene, toluene a n d xylene. Curves similar t o those in Fig. I are obtained except t h a t since benzene is now a n end product there is no falling off in its curve. B y t h e use of a diagram similar t o Fig. I1 i t is possible t o estimate t h e percentages of t h e three aromatics as functions of t h e specific gravity of t h e 1 7 5 ' cut. A quick a n d simple method of testing is t h u s furnished which is often of sufficient accuracy for practical purposes. These indications were checked u p b y analyses of recovered oils obtained b y cracking Oklahoma crudes a n d a fair agreement obtained. Figures for t h e latter set of experiments are omitted as t h e y a d d nothing t o those already given; d a t a regarding furnace conditions a r e lacking a n d irregularities are a little greater, due t o less care i n handling t h e cracked oil.
TABLEI-RESULTS OF "CRACKING" EXPERIMENTS (PERCENTAGES B Y WEIGHT) FURNACE CONDITIONS DATAON RECOVERED ("CRACKED") OIL DATAON CUT TO 175' C. Pressure Feed Percent Per cent aromatics found Per cent of Per cent constituents Lbs.per Temp. Gals. origBenTolX y - Naph- Anthrarecovered Sp. BenTolXyRUE KO. sq. In. a C. per hr. inal Sp. gr. zene uene lene thalene cene oil gr. zene uene lene 30.0 1.018 19.1 10.4 3.1 14.1 l.... 14 36.2 0.874 52.9 9.6 28.8 8.6 8.1 20.0 11.2 1.000 20.7 4.6 2.... 14 39.3 0.876 52.6 11.7 3.8 28.6 0,992 12.2 650 20.0 7.0 23.5 4.4 3.... I5 0.873 50.6 0.0 11.2 39.5 30.9 12.4 4.......... 20.2 5.0 2.0 0.978 22.5 15 47.6 0.0 700 11.9 42.4 0.870 29.3 3........... 10.8 0.0 16.5 0.954 23.5 17.8 15 0.0 650 55.8 0.866 29.6 31.5 19.4 0.0 7.0 0.938 30.0 13.4 15.7 0.0 0.850 6........., 12.3 625 14 44.8 27.4 30.0 0.0 7.1 12.0 15.1 13.6 L.,....... 7.9 0.0 0.817 23,O 0.900 600 31.2 15 52.2 2.8 8.......... 0.0 4.0 0.0 0.0 10.2 0.762 0.837 55.0 14 39.0 550 0.0 7.1
;'o"'
ages b y weight of t h e five hydrocarbons contained i n t h e oil. Average curves are given, a s t h e irregularities due t o experimental error (probably imperfect condensation) would otherwise make less clear t h e indications obtained. It is t o be noted t h a t also no pretense is made of having analyzed t h e cracked oils for all constituents. T h e five hydrocarbons were chosen on account of their commercial importance a n d because t h e y were typical. It will be noted t h a t toluene a n d xylene are found in a sample which yielded no benzene a n d t h a t t h e curves for toluene a n d xylene bear a more or less constant relation t o each other. Xylene percentages are about half those of toluene a n d t h e maximum for each of t h e two curves occurs at t h e same point on t h e horizontal axis, this representing most favorable conditions for t h e formation of these two aromatics. Benzene is a product of more strenuous conditions of cracking: it begins t o form a t a higher point on t h e scale adopted a n d has a higher maximum. I t s content runs between those of toluene a n d xylene until t h e latter two pass their maximum. T h e benzene maximum is, however, greater t h a n t h a t of toluene. Figures for naphthalene a n d anthracene are few b u t furnish important indications. Naphthalene formation begins a t about t h e point where xylene a n d toluene commence t o fall off. The naphthalene curve is ascending rapidly at t h e upper limit of the present range a n d there is no indication t h a t i t has begun t o approach its maximum. Anthracene is formed under even more strenuous conditions t h a n naphthalene. One other point is of importance. Koting t h e relations shown on the basis of gravity of recovered oil led t o a belief t h a t a similar condition might obtain if t h e first (175") distillation cut were considered. I n Fig. I1 t h e specific gravity of each 175' cut is plotted against its percentage content of each of t h e three
S U Mil1A R Y
When petroleum is cracked there are definite relations among t h e percentages of t h e various aromatic compounds formed, said percentages being fixed b y t h e degree of cracking. T h e factors which control t h e latter may be summed up a n d represented b y t h e specific gravity of t h e cracked oil. I-Toluene a n d xylene show t h e same variation in percentage formation, xylene being present in approximately one-half t h e quantity of toluene. Both are at a maximum in a recovered oil of about 0.95 specific gravity. 11-Benzene formation requires more strenuous cracking t h a n is needed for t h e production of toluene a n d xylene. Benzene formation is at a maximum at a point where toluene a n d xylene have fallen off considerably in their content. 111-Naphthalene begins t o form at about where t h e toluene-xylene content passes its maximum, indicating t h a t naphthalene is a product of t h e decomposition of t h e monocyclic hydrocarbons. T h e maximum for naphthalene is necessarily much higher on t h e scale here chosen t h a n t h a t for a n y of t h e monocyclic compounds. IV-Anthracene behaves similarly t o naphthalene, requiring even more strenuous conditions for its formation. V-The relations above indicated may, if desired, be so applied as t o furnish a method for t h e approximate estimation of benzene, toluene a n d xylene from t h e primary I 7 j O distillation cut obtained from t h e cracked oil. CHEMICAL SECTION OF THE PETROLEUM DIVISIOK U. S.BUREAUOF MINES
OILS OF THE CONIFERAE: V-THE LEAF AND TWIG, AND BARK OILS OF INCENSE CEDAR B y A. W. SCHORGER Received August 2, 1915
Incense cedar (Libocedrus decurrens Torrey) is largely restricted in its range t o t h e state of California.
