CATALYTIC CRACKING OF PURE HYDROCARBONS Cracking of NaDhthenes B. S. GREENSFELDER AND H. H. VOGE Shell Development Company, Emeryville, Calif.
Cracking of eleven naphthenes containing 6 to 18 carbon atoms was studied over a silica-zirconia-alumina catalyst. I t was found that naphthenes are quite susceptible to the action of the catalyst and that both the ring and any side chains contribute to the total cracking. The rate of cracking increases rapidly with increased molecular
weight. Secondary reactions of isomerization and saturation play an important part in determining the final products. Catalyzed cracking at 500° @. proceeds at about one thousand times the rate of thermal cracking, and there are fiignificant differences in the product distributions.
C
Decalin (decahydronaphthalene) from Eastman was washed RACKING of paraffins and olefins over a silica-zirconiawith sulfuric acid-silver sulfate to remove aromatics. The alumina catalyst under conditions similar to those emboiling range was 190-193.5' C., d:O 0.8859, nZ,O 1.4763. ployed in the commercial cracking of petroleum fractions was deCyclopentylcyclohexane was prepared by hydrogenating over scribed in previous articles of this serics (4, 6). These studies nickel at 200 C. cyclopentylbenzene from the condensation of gave a picture of some of the reactions favored by cracking cyclopentene with benzene in the presence of aluminum chloride. catalysts, with far less ambiguity than would result from obThe product was washed with sulfuric acid-silver sulfate and had servation of the cracking of petroleum fractions comprising a a boiling point of 216' C., n2,0 1.4721, dzo 0.8753. mixture of hydrocarbons. This method has been extended t o a study of the naphthenes or alicyclic hydrocarbons, Amylcyclohexanes were from hydrogenation of amylbenzenes which are prominent constituents in the majority of pefrom Sharples Chemicals Inc. The latter are stated by the troleum fractions employed for cracking. An effort was manufacturer to contain a t least 90% secondary amyl structures, made to secure a wide variety of naphthenic hydrocarbons, with the rest tertiary. The hydrogenation product was washed covering carbon numbers (C No. = n in C,H,) from 6 to 18. with sulfuric acid-silver sulfate to remove aromatics. It was These were cracked by the catalyst and procedure previously later found to contain 5y0 boi!ing below '211, chiefly methyldescribed (4); the definitions and terminology are the same. cyclohexane. The remainder distilled at 195-200' C. The The catalyst, obtained from Universal Oil Products Company, mixture had dO : 0.8152 and na$ 1.4485. analyzed 86.2% silica, 9.470 zirconia, and 4.3% alumina by Bicyclohexyl from Eastman was washed with sulfuric acidweight. This catalyst gives results similar to those obtained silver sulfate and fractionally distilled to constant refractive inwith the present commercial silica-alumina cracking catalysts. dex. Its boiling point was 239' C., d:" 0.8860, ny 1.4795, melting Properties and sources of hydrocarbons follow, with compounds point 3.4" C. arranged in the order of increasing molecular weight: Amyl Decalins were prepared from Sharples amylnaphthalenes, Cyclohexane from Eastman Kodak Company was washed The manufacturer states that the latter are entirely 2-substituted with concentrated sulfuric acid t o remove aromatics and disand that the amyl group is a t least 90% secondary, with the rest tilled. The melting point was 6.1 a c.,boiling range 80.2-81.2' c., tertiary. The amylnaphthalenes were fractionally distilled, and d:O 0.7785, n2,0 1.4263. the center fraction was hydrogenated over nickel. The product Methylcyclopentane was isolated from a California petroleum was washed with sulfuric acid-silver sulfate and fractionally dis: by repeated fractionation. It had a boiling point of 72.0" C., dO tilled. The central fraction, boiling from 142" to 144" C. at 15 0.7488, and n% 1.4100. mm., was used. It h a d d i o 0.8874, n2,0 1.4814, and specific disMethylcyclohexane from the hydrogenation of toluene over persion at 20' c. [lo4(ns nc) / d ] 97.2. nickel had a boiling point of 100.6" C., d i 0 0.7688, ny 1.4232, Abietanes were from the hydrogenation over nickel at 200and an aromatic content of 0.6%. 250' C. of abietenes from the G and A Laboratories (Savannah, Hydrindan (hexahydroindan) was prepared by hydrogenation Ga.). The abietenes are a mixture of unsaturated, partially over nickel of indan from Eastman. It was washed with sulfuric hydrogenated phenanthrene homologs derived from rosin oil. acid containing silver sulfate. The product contained 8.6% The sample used, with n'+,O1.5410, was hydrogenated twice over distilling below 160 C., chiefly methylcyclohexane, but this fresh catalyst until no more hydrogen was consumed. The was not discovered until after the cracking experiment had been product in a simple distillation from a flask started over at 250' C., made. The rest of the material distilled a t 160-168" C. Properand reached 307" C. a t -5% and 357' C. at 95% over. It had ties of the mixture used as a cracking stock were d i 0 0.8568, ny dO : 0.9236, nz,O 1.5035, specific dispersion at 20" C. [lo4 ( n ~ 1.4611, specific dispersion a t 20" C. [lo4( n -~ n c ) / d ] of 95. n c ) / d ] 105, molecular weight in benzene 249, and bromine numIsopropylcyclohexane from the hydrogenation of isopropyl, ber 6.8. benzene (Dow Chemical Company) was washed with sulfuric Where hydrocarbons were found to contain small amounts acid-silver sulfate. It contained 4.8% lower-boiling material, of lower-boiling material, corrections were made in computing chiefly methylcyclohexane, and the remainder distilled at 153results. d: 0.7994, n2,0 1.4396. 164" C. The mixture, as used, had O 1038 O
-
O
-
Noobmkr, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
CRACKING BEHAVIOR
1039
t cis-Decalin
The qaphthenes were cracked over the silicac t r a n r - &calm 180 zirconia-alumina catalyst in the fixed-bed a p paratus previously described. The usual catsI60 1.50 lyst volume was 50 cc., temperature 500' C., process period 1 hour, pressure atmospheric, and flow rate about 13.5 moles hydrocarbon per liter $ I40 of catalyst per hour, unless otherwise noted. f The general purpose of these experiments was to z I20 survey a number of compounds rather than to 2 examine any one in extensive detail. With P 0100Decalin, however, a limited study of the effect 9 of operating conditions was made. Results for monocyclics are given in Table I and for bicyclics 5f 80 in Table 11. CYCLOHEXANE. The reactions of cyclohexane 60 1.40 at 500' C. were slight. The little gas produced waa about half hydrogen. The m a l l amount of 40 138 lower-boiling liquid product appeared to be a mixture of C6 olefins and paraffins; i t had an f I 20 I I , I I I I I I .I36 olefin content of about S%, an aromatic content 0 5 10 I5 20 25 30 35 40 45 of about 4%, and may have contained some Volume Rrcantoge Distilled methyloyclopentane. The amount of isomerizaFigure 1 . Distillation Curve for Liquid Product (78.0% Yield by tion to methy~cyc~opentane, however, could not Weight) from Cracking Decalin at 600' C. a n d 13.7 Moles per Liter have been more than 4% of the feed. When per Hour cyclohexene was treated under similar conditions, conversion to methylcyclopentane was over and no single reaction predominated. The lower-boiling liquid 30% (6). The liquid boiling a t 80" C. from the cyclohexane run, plus the distillation bottoms, had a refractive index, n '," of 1.4271 contained a small amount of Cs and some c6 and 4,with 11% olefins in the CE-G range. There was a little material of high (cyclohexane 1.4263, benzene 1.5012); hence it probably did not refractive index (ny 1.50) in the 6% bottoms above 100' C. contain over 1% benzene. METHYLCYCLOPENTANE. Cracking a t 500' and 550' C., aa IBOPROPYLCYCLOHEXANE. Results from cracking this G with cyclohexane at 500", gave a gas fairly high in hydrogen. naphthene indicate that it was fractured in both the side chain and ring. The gas was rich in both Ct and Cd hydrocarbons, Lower-boiling liquid product was chiefly Cs material, although while the cracked liquid, judging from distillation temperatms, traces of cyclopentane appeared to be present. There was little optical properties, and bromine numbers, was a mixture of or no isomerization to cyclohexane. The fairly saturated quality of the C r C , hydrocarbon gas is indicative of the occurrence of naphthenes and aliphatics, with a relatively mall aromatic hydrogen transfer reactions, since simple cracking would give content. Recovered Cp material contained an appreciable amount of aromatics. Yields and properties of liquid fractions only olefins. Bottoms from the run at 550" C. were 5.6% by weight of the feed, and had n% 1.4575, which signifies the presfollow: ence of condensation products or aromatics. Analyeia of METWYLCYCLOHEXANE. Cracking waa somewhat greater Boiling Range Amount Wt. % Fraction* Wt* % of Fraction, C. of Cdarge Olefins Aromstios than that of the Ca naphthenes, but the products were varied
-
-
-
-
-
-
-
-
TABLE I. CATALYTIC CRACKINQ OF MONOCYCUC NAPETE~NBIS Hydrocarbon Ex tl conditions
$le.&& * c.
Flow rate, moles/l./hr. Gaseous roduct M o l d m o l e charge Volume %
Cyclohexane 1600 .5 13.8 0.031
46.3 CHd C:H4 CrH4 CaH6 CaHs Iso-CdHa n-CbHs CdHio Total olefins 2i.Z Total saturates a2.1 C No. of saturates 2.8 Material balanoe, wt. % of charge Gas 0.9 Liquidbelow ori in8lb.p. 4.p Remaining liqui% 86.2 Carbon 0.16 8.2 Loan 0 Liquid hourly space velocity. b CO from regeneration not inaluded. € I :
... ... ... ...
... ...
d
Methyloyclopentane
Methyloyolohexane
Isopropylcyolohexane
Amyloyolohexanea
600 1.7
600 2.1
600 2 .5
13.6 0.212
13.3 0.61
13.4 0.64
12.0 63 .. 79 0.9 19.5 11.1 4.8 8.0 33.1 36.0 62.0
17.6 48 .. 32 1.7 19.6 9.9 4 .. 4a 8 26.1 26.6 46.9
17.2 12.6s 67.5
18.6 21.2d 62.8 1.3 1.2
1600 .6 13.2 0.078
1660 .6 14.0 0.172
29.1
32.2 38.7
46.2 14.0 4.7 3.8 14.9 10.1 11 .. 22 4.0 22.0 32.8
2.0 6.7 88.4
4.2
7.4
8 1.7 9.8
6.0 70.0
... ... ... .,. ...
... ...
2.8
...
0.2b
2.8
Correctes Includin CiH6. for lower-boiling material in oharge.
0:ib)
27.0 7.80 4.2
14:O 18.9 31 .. 71 28.3 23.0
...
60.0
7.6
...
{i::
...
20-50 60-87 87-148 146-168 Above 158
7.3
3.w 1.7 62.6 4.9
34 25
5:i
. ..
.. ..
ii ( n v 1.4911)
a Corrected for lower-boiling material in charge.
AMYLCYCLOHEXANES. Results resembled those with isopropylcyclohexane. The cracked liquid waa a complex mixture, with c6-C~more prominent than CrCl0 material. The fraction 3085' C. contained 35% olefins, and the fraction 85-105" C., 20% olefins and 11% toluene; the latter suggested cracking in the side chain accompanied by ring dehydrogenation. Cracked material above 105' C. contained even more aromatics, judging from the refractive index which reached a maximum of 1.4798 for the 140-170" C. cut. The product within the charge boiling range, 195-200' C., was not much changed (n2i 1.4510 compared to 1.4497 for material of the same range in the charge). The bottoms, 3.5% by weight of the charge, had a'," 1.5285 and, thus, were quite aromatic. It should be noted that them aromatics did not neceaertrly all derive from the cyclohexane
tended t o be saturated, especially a t low flow rates or low temperatures. The liquid product was composed of hydrocarbons from Cs I - 1.50 240 I t o GO,with the CI range prominent. Identification of individual compounds was not attempted except for naphthalene, isolated by Cyclopdntylcyclohaxona + ," fractionation and crystallization from the E combined bottoms of several runs at 500' C. 1.46 f in an m o u n t correspondingto 0.5% by weight .D of the charge, which must have been less than Q 2 the amount actually formed. Considerable D 0 amounts of methylcyclopentane were strongly "indicated by the physical properties of the P -142 C6 product. Very little or no benzene was formed, but there were appreciable amounts of aromatics in the higher boiling range. More detailed analyses of the products from cracking Decalin under similar conditions I I I I I 1.38 were given by Bloch and Thomas ( I ) , who 0 20 40 60 80 100 also showed substantial amounts of methylVolume Poreentoga Distilled cy clopentane. Figure 2. Distillation Curve for Liquid Product (75.6% Yield by The A.S.T.M. motor-method octane Weight) f r o m Cracking Cyclopentylcyclohexane at 500" C. and 6.8 number of the 20-180' C. product from crackMoles per Liter per Hour ing Decalin in a group of runs a t 500" C. and 13.6moles per liter per hour was 79.5clear and 88.5 with 3 ml. of tetraethyllead per gallon. ring, since pentenea alone, treated under even milder conditions, The effect of cracking conditions with Decalin was further gave rise to considerable quantities of highelcboiling aromatics studied in the series of runs listed in Table 111. The catalyst (6)* sample was the same type as that used for other runs, but was HYDRINDAN(HEXAHPDROINDAN, BICYCLo[4.3.0)NONANE). somewhat more active; consequently the depth of cracking under Cracking w&8 fairly extensive. The gaa contained some hydrogiven conditions was appreciably greater. Results in Table I11 gen and much saturated hydrocarbon, The liquid was rich in indicate that doubling the flow rate reduced the total depth of isopentane and a C,fraction resembling methylcyclohexane, but a h contained compounds not identified in the range of Ce, CS, cracking about as much as did lowering the temperature by 50" C. Thus 450" C. and 13.6 moles per liter per hour gave the same and CO. Olefin contents of liquid fractions were: 20-50' C., cracking as 500" and 26.7 moles per liter per hour. The nature 150/,; 50-75" C.,20%. Aromatics were not prominent except of the products was different, however, both in distribution and in the bottoms; the fraction 75-110' C. contained only 7%. saturation. I n general, low temperatures and low flow rates The bottoms above 168' C., 9.1% by weight of the charge, had favor saturation by hydrogen transfer. Another feature strongly n*: 1.5379. influencing saturation is the process period; shortening the procDECALIN (DECAHYDRONAPHTHALENE). Decalin was cracked ess period so that the catalyst does not become too deactivated over the catalyst a t 450°, 500", and 525" C., with the results by carbonization leads to greater saturation under given condioutlined in Table 11. Figure 1 shows a typical distillation curve tions. A run a t 400' C. for 15 minutes, included in Table 111, for the liquid product. The cracking of Decalin gave rise to a shows the results from a combination of most of the conditions variety of products, indicating fission a t various places in the favoring saturation; the gas was 89% saturated and the CCliquid molecule and the occurrence of a number of secondary reactions. was 98% saturated. The amount of Ce saturates produced in The gaseous product waa rich in CCand CChydrocarbons, which I
I
I
1
'
(Bottoms)
J
I
- e -
-
.-
-
.
~~~
TABLE 11. CATALYTIC CRACKINQOF BICYCLIC NAPHTHINES Hydrocarbon
Flow rate. moles/l./hr. Qeseous roduct M&m&n#e charge H I
CHI C:H4 CiHa CiHa CIHI Isc-C4H* n-CdHa C4HlO Total olefins Total saturates 0 No. of saturates Material balance, wt. % of oharge
om
Liquid below ori 'nal b.p. Remaining l i q u i f Carbon LoSe
CO from re$eneration not included.
Cyclopentylcyclohexane
Decalin
Hydrindan
625 1.9 12.0
Bicyclohex yl 500 2.5 13.3
500 3.0 12.8
500 3.6 12.9
0.690
0.721
0.636
0.955
1.022
14.9 3.7 5.4 4.4 16.4 11.4 2.3 6.0 35.5 30.1
30.6 15.5 4.3 2.1 11.3 7.8 0.8 3.0 24.6
30.7
20.6 7.9 3.2 3.4 21.4
29.5 39.8
19.3 7.3 4.2 2.1 22.1 7.2 4.8 9.6 23.4 40.7 40.0
6.0 9.5 21.3 40.1 39.3
13.6 30 4 48.1 1.0 6.9
17.1 48.2 26.8 1.2 6.7
15.0 59.0 20.3 1.6 4.1
500
460 0.9 6.7
2.1 13.7
0.500
0.896
0.622
0.65
20.7
23.2 0.0 0.7 0.0 0.7 8.6 0.0 9.3 57.5 10.7 66.1
29.0 7.0 5.1 2.1 13.9 8.8 2.5 2.3 29.3 23.8 47.2
21.9 14.0 4.5 0.9 13.6 13.3 2.0 5.0 24.8 25.1 53.0
19.4
....
a5.0
50.0
12.1 29.8 52.8 4.2 1.15
15.0 25.3 52.7
16.3 25.6 47.4 10.7
18.6 35.2 40.4
12.8 44.5 29.7 1 oa 12.0
....
17.7 17.3 69.1 1.4 4.5
..
.... i:;}
Abietanes
500 1.3 6.8
500 1.2 6.8
1.8 12.5
29.5 49.8 8.9
Am 1 Decafns
.... .... ....
.... ....
.... ,... ....
...
....
f..i ?.
....
November, 1945
INDUSTRIAL AND ENGINEERING CHEMISTRY
1041
this run was 0.38 mole per mole TABLE 111. EFFECT OF CONDITIONS IN CATALYTIC CRACKING OB DECALIN of Decalin cracked, which is (Atmospheric pressure; duration of rum, one hour, except 400° C. run which wan 15 minutes; catalyst for theme higher than in any other runs wan of the Bame type but more active than that uaed for all other runs) run. Ex tl conditions 400 450 500 560 550 600 500 %A,., C. 500 Dehydrogenation of Decalin 2.1. 4.2 1.1 2.1 2.1 2.1 4.2 LHSV 1.1 was appreciable, especially a t 13.6 13.6 26.7 6.8 13.6 .26.7 Flow rate. moles/l./hr. 6.8 13.6 . . Gaseous roduct the higher temperatures, and 0.675 1.494 0.880 0.344 0.410 0.520 Moles~molecharge 0.885 0.675 waa reflected in both the high Volume yo 2 22.6 27.5 61.5 35.3 25.8 Hi 35.8 27.6 hydrogen content of the gas 26.3 24.8 34.2 11 17.6 33.2 26.3 Olefin8 17.5 23.7 30.5 41.0 87 59.8 46.2 46.2 9aturatea 46.7 and the high refractive index 31.8 37.6 20.2 35.6 40.9 38.4 Mol. weight 31.8 37.6 of the material remaining 36 14 24 26 2 9 14 Olefins in CSfraction, wt. yo 9 1.5258 1.6034 1.5060 1.4930 , 1.4910 nso of remaining liquid' 1.5271 above 180' C. 1.5060 Material balance, wt. Yo of TOtest the thermal stability charge of Decalin under the condi28.4 20.2 18.4 14.6 8.9 12.1 Gas 20.4 18.4 24.2 24.8 27.2 23.9 Liquid below original h.p. 33.0 21.6 18.0 27.2 tions of the more severe runs, 40.7 52.8 50.6 68.9 63.0 61.6 60.6 Remaining liquid 39.1 2.6 0.7 1.0 2.8 0.6 0.6 Carbon 1.2 1.0 an experiment was made a t 4.1 1.5 2.8 1.4 0.4 1.8 Loss 6.3 2.8 525' C. and a flow ratc of Total cracked (no loss basis), wt. % 58.3 47.9 37.3 30.1 36.7 47.9 57.5 46.4 6 moles per liter per hour, nso for Decalin is 1.4763, with glass beads r e p l a c i n g the catalyst. No gas was obtained and only 0.5% of lower-boiling- liauid. Thermal There was little Decalin (boiling at 185-185' C.) in the product; and catalytic reactions will be more fully compared in a later hence exclusive removal of the side chain was not an important section. reaction here any more than with the other alkyl naphthenes. CYCLOPENTYLCYCLOHEXANE. Catalytic cracking gave a variety ABIETANES. The sample was a mixture obtained by hydroof products, and no reaction could be singled out as especially genating neutral rosin oil and, presumably, consisted of alkylated prominent, although CI to C, material predominated in the perhydrophenanthrenes such as liquid, and Cs and C, in the gas. Data are given in Table 11,and a distillation curve is reproduced in Figure 2. Olefin contents of all fractions were rather low, varying from 6 to 16% over the C range 50-125" C. Considerable amounts of aromatics were present in the material boiling above 100' C. and in the distillation bottoms. BICYCLOHEXYL. Results (Table I1 and Figure 3) show that bicyclohexyl cracks readily over the catalyst with the formation of a variety of products. The gas is rather saturated. Ce hydrocarbons are especially prominent in the liquid, and their presence indicates a tendency of the two rings to split apart at the connecting link. The properties of the Csplateau fraction (boiling range 70-73' C., n2j 1.4130) agree with those of methylThe average number of carbon atoms per mole was 18.0. The cyclopentane (71.8' C., 1.4098), but the cut may contain mixed mixture cracked extensively and gave a high yield of lower-boiling aliphatic and alicyclic hexanes. The olefin content of this cut liquid which was evenly distributed over the range 20' to 250' C. was 10% for the run at 6.8 moles per liter per hour. Aromatia The CScut contained 40% by weight olefins. hydrocarbons were Dresent in the CgCll - _- fractions, but analysis of the CSfraction &owed none. The gas from bicyclohexyl contained hydrogen, and the bottoms from the distillaI I 1 I I (bistillatia'n temperatures I tion had a refractive -index indicating the ZOO. C. 01 15mm., bottom^ presence of biphenyl or other highly aromatic I I corrected to 760 mm.) I.5970l 240 FBicyclohexyl 1.50 compounds. Formation of aromatics by dehydrogenation and possibly by other re0 ci I I a actions thus takes place concurrently with 0.200 Refractive Index1 Bicyclohexyl4 2 cracking, as i t does with other naphthenes Dl examined that contain a cyclohexane ring and. have a total of 9 or more carbon 0 cu atoms. .e Q AMYL DECALINS. These naphthenes :120 - ec cracked readily a t 500' C. The m o u n t of .-f 61 Iower-boiling liquid, including that recovered w in a vacuum distillation of the bottoms, ex80 1.42 ceeded 48%. Distillatio; showed a wide distribution of products over the boiling range and the presence of many different compounds. Analyses of fractions were: ~~
0
.
~
-
5
we
20-40 60-80 80-111
Charge 7.5
8.0
7.1
--Composition, Wt. %AromaSrtd. Ole5ns tics , Snaphthenes 44
39 38
..*
