US3153627A - Catalytic process - Google Patents

Description

CATALYTIC PRQCESS Harold heather, Richland Township, Gihsonia County, and Richard A. Flinn, Penn Hills Township, Allegheny County, Pa, assignors to Gulf Research & Development Company, Pittsburgh, Pa., a corporation of Delaware No Drawing. Filed Apr. 4, 1960, Ser. No. 19,455 3 Claims. (Cl. 208-53) This invention relates to improved procedure for preparing gasoline having a relatively high octane number.

It is known to subject hydrocarbons of various types to treatment with hydrogen under certain conditions of temperature and pressure and in the presence of hydrogenation catalysts in order to hydrogenate and convert the hydrocarbon into lower molecular weight hydrocarbons of more desirable types. See, for instance, United States Patents 2,341,782, 2,341,792, 2,348,576 and 2,464,539. See also Process in Petroleum Technology, published by American Chemical Society in 1951, pages 76-82; PreprintsUse of Hydrogen in the Petroleum Industry, Symposium sponsored by the Division of Petroleum Chemistry, American Chemical Society, vol. 1, No. 4, September 1956, pages 8395; and Preprints-General Papers, Division of Petroleum Chemistry, American Chemical Society, vol. 2, No. 3, August 1957, pages 123- 129. These prior art procedures were not designed to yield the maximum proportion of isoparat'rinic materials in the product and the conjoint conditions and catalysts utilized in these references would not yield an iso-tonormal ratio approaching the maximum obtainable. It is known from this prior work that certain types of hydrocracking can produce greater-than-equilibrium amounts of isoparafiins. For example, the work mentioned above, discussed in the ACS Preprints of September 1956, notes the ability of this reaction to produce non-equilibrium products and indicates that the proportion of isopentaneto-normal pentane is considerably above the equilibrium value of about 2.6, characteristic of the 7 (HP-800 F. re-

action temperature used. It is also known that isoparaffins have a higher octane number than normal parafiins and that they can be added to gasoline to increase its octane number.

This invention has for its objective to provide a process for treatment of hydrocarbon stocks of various types with hydrogen using conditions which result in hydrocracking or destructive hydrogenation to form lower boiling branched hydrocarbons, particularly isopentane, in unusually high proportions with respect to the normal hydrocarbons and for preparation of gasoline therefrom, said gasoline having a relatively high octane number. A further object is to provide economical procedure for preparing gasoline having a relatively high octane number. Another object is to provide improved alkylation procedure. Another objective is to improve the state of the art. Other objectives will appear hereinafter.

These and other objectives are accomplished by our invention which comprises in combination contacting a petroleum fraction which contains at least 8 carbon atoms per molecule and which is substantially free of asphaltic material with hydrogen in the presence of a catalyst comprising a metal of Group VI left-hand column or of Group VlII of the periodic system, any oxide or sulfide of such metals or mixtures thereof deposited upon a carrier which has substantial cracking activity. This contacting takes place at a temperature between about 550 and 800 F., at a pressure between about 400 and 800 p.s.i.g. and -at a space velocity between about 0.1 and 10. Where the product boils substantially entirely in the gasoline range,

3,.l53,fi27 Patented @ct. 20, 1.964

the product from this treatment may be directly added to a gasoline having a lower octane number than the product. Alternatively, with such a product boiling substantially entirely in the gasoline range or with a product which contains a substantial amount of components which boil above the gasoline boiling range, the product is distilled to separate a fraction boiling in the gasoline range and this fraction is added to a gasoline having a lower octane number than said fraction. This last mentioned distillation is of a relatively low-cost type calculated to separate a gasoline fraction or a hydrocarbon having a certain number of carbon atoms such as a C C or C fraction or a mixture thereof such as a C F. fraction and is not intended to result in separation of isoparalhns from normal paraffins, which isoand normal paraffins contain the same number of carbon atoms. We have found that the process of converting to isoparaffinic hydrocarbons outlined above results in a thorough conversion to isoparafiins and that it is unnecessary to subject the product to a distillation to concentrate the isoparaffins before they are added to a lower octane number gasoline. Therefore our invention enables the elimination of this costly type of fractionation. This fractionation to separate isofrom normal parathns is expensive because of the heat input, and especially because of the cost of the large fractionating columns containing a large number of plates which is required to separate such close boiling substances.

Accordingv to another modification of our invention, we utilize the above described fraction of the product and/ or any portion thereof, such as the excess over and above that needed for gasoline improvement, for alkylation. Such a fraction is of sufiiciently high concentration for direct use in any alkylation procedure in which isoparafiins are alkylated with olefins.

The charge stock to our process may be any petroleum fraction which is substantially free of asphaltic materials and which has at least 8 carbon atoms per molecule. Thus we may utilize 'deasphalted residual fractions such as those obtained when a whole crude is topped or reduced by vacuum distillation and then deasphalted by any known procedure. We prefer, however, to utilize distillate stocks such as light and heavy gas oil, heavy naphtha, etc. We also may utilize cracked stocks such as cycle stock from a catalytic cracking or thermal cracking operation. The stock must be substantially free of asphaltic materials since these materials have a deleterious effect on the catalyst and necessitate frequent regeneration. Parafiin wax may also be utilized as a charge stock. In general where a choice of charge stocks is available, we prefer to utilize a saturated charge stock and a lower boiling type of charge stock such as a straight run heavy naphtha or kerosene. A particularly advantageous charge stock for producing a C 165 F. blending stock is a straight run furnace oil fraction having a boiling range of between about 400 and 650 F. This charge stock is advantageous for several reasons. First there is an advantage in charging materials boiling above the gasoline range because a higher iso-to-normal ratio is obtained. For example, under the same conditions a virgin gasoline (325412 F.) gave an iC /nC ratio of about 21 while a 418-576 F. virgin furnace oil gave an iC /C ratio of about 28. A second advantage for using straight run furnace oil rather than a similar cracked fraction is that the hydrogen consumption involved in hydrocracking the straight run material to C 165 F. blending stock is substantially lower. In addition, the aging characteristics of the hydrocracking system when charging straight run furnace oil are considerably better than when charging a similar boiling cracked stock.

The catalyst employed in our process may be any metal of Group VI left-hand column or of Group VIII or an oxide or sulfide of such metals. Metals, oxides or sulfides other than those of the noble metals in Group VI and Group VIII are preferred. Examples of such catalysts are nickel, nickel sulfide, cobalt, cobalt sulfide, palladium, molybdenum oxide, molybdenum sulfide, tungsten oxide and tungsten sulfide. We prefer to utilize a partially sulfided catalyst. It is essential that the carrier be one which has substantial cracking activity. Thus we may utilize as a carrier any of the many known cracking catalysts such as a natural clay which has been acid treated to convert it into a cracking catalyst. Alternatively we may employ as a carrier synthetic cracking catalysts of various types including those comprising silica with alumina and those containing other components known to be desirable in cracking fittalysts such as magnesium oxide, zirconia, etc. Between about 0.1 and 20.0 percent of the hydrogenation component (determined as the metal) and preferably between 1.0 and 15.0 percent are deposited on the carrier. The catalyst may be prepared using any of the well known methods for compositing catalysts with carriers.

The pressure used in our process to obtain a high content of isoparafiins must be below about 800 p.s.i.g. If the pressure is increased above 800 p.s.i.g., the amount of branched hydrocarbons formed progressively decreases. Pressures below 400 p.s.i.g. are unsatisfactory since the amount of hydrocracking taking place is uneconomically low. Pressures of between about 500 and 750 p.s.i.g. are preferred.

In connection with the utilization of pressures below 800 p.s.i.g attention is directed to Tabes I, II, III and IV which show the effects of using pressures above and below the 400800 p.s.i.g. range in accordance with our invention. In Table I a heavy virgin naphtha boiling between 325 and 412 F. was contacted with hydrogen utilizing a recycle mate of 10,000 s.c.f. of hydrogen per barrel of charge stock. The temperature utilized in Table I was 750 F. and the liquid hourly space velocity was 2.0. The catalyst was partially sulfided 3 percent nickel deposited upon a silica-alumina cracking catalyst. Sulfiding was carried out by treating the catalyst with a mixture containing one part H S and twelve parts H at atmospheric pressure and 600 F. for one hour. Such sulfiding deposited 0.7 percent sulfur on the catalyst or an amount sufi'icient to convert 30 percent of the nickel to the monosulfide. Although the conversion to light products increased as pressure increased, the decrease in the proportion of branched paraffins produced, particularly isopentane, caused the selectivity of the process for producing isoparaflins to decrease. The proportion of isobutane to n-butane is much less affected by pressure than the proportion of isopentane to n-pentane.

In Table II a hydrogenated heavy catalytically cracked naphtha boiling between 224 and 464 F. was contacted with the same catalyst described in connection with Table I. The conditions of contacting were identical to those described in Table I. .Again while conversion increases as pressure increases, the rapid drop in the proportion of isopentane in the product decreases the selectivity for these desirable products.

In Table III a hydrogenated light catalytically cracked gas oil (386 F. to 632 F.) was contacted with the same catalyst mentioned in connection with Tables I and II at a temperature of 750 F., a hydrogen recycle rate of 10,000 s.c.f. of hydrogen per barrel of charge,

and with a liquid hourly space velocity of 1.0.

TABLE III Conversion Pressure, p.s.i.g. to C and Ratio iso Ratio iso lighter, C5 to 1105 O4 to 1104 percent The same effect of pressure shown by Tables I and II is evident in Table III. It appears, then, that the nature of the charge does not alter the effect of pressure on the iso-to-normal ratio of the parafiins produced, although the level at which the elIect occurs may be altered somewhat by charge variations.

The data of Table IV illustrate that all pressures below 800 p.s.i.g. are not sufficient to give the desired high proportions of isopentane. The experiments in this table were conducted by processing a virgin furnace oil (418 F. to 576 F.) over the partially sulfided 3.0 percent nickel-on-silica-alumina catalyst as described in the prior tables at 700 F., 1.2 liquid hourly spaced velocity, and a hydrogen recycle rate of 8000 s.c.f. per barrel of charge. At 250 p.s.i.g. the proportion of isopentane in the product fell drastically, and above 750 p.s.i.g. a decrease was again noted.

TABLE IV Conversion Pressure, p.s.i.g. to C5 and Ratio iso Ratio iso lighter, C5 to n05 G4 to 11C;

percent Since the isomerization of a normal hydrocarbon to its branched isomer involves no change in the number of molecules in the system, the reaction is generally found to be insensitive to pressure. For example, the lack of an effect of pressure upon isomerization has been noted by Ciapetta et al. in US. 2,550,531. Thus, the observed effect of pressure upon the proportion of iso-paraffius in a hydrocracked product would not have been expected.

In order to obtain the high iso-to-normal ratios which are the objective of this invention, it is necessary to use not only pressures between about 400 and 800 p.s.i.g., but a temperature between about 550 and 800 F., and preferably a temperature between 600 and 700 F. The importance of utilizing these temperatures is illustrated by the data presented in Tables V and VI. In Table V a heavy virgin naphtha having a boiling range of 325 to 412 F. was contacted with the partially sulfided nickel silica-alumina catalyst mentioned in connection with the prior tables at pressure of 750 p.s.i.g., a liquid hourly space velocity of 1.0 and a hydrogen recycle rate of 10,000 s.c.f. of hydrogen per barrel of charge.

It can be seen from Table V that below about 800 F., a definite increase in the iso-to-normal ratio is observed and that below about 700 F. this increase becomes even more rapid. In this case, the same factors discussed in relation to the effect of pressure apply. That is, although the actual conversion increases steadily as temperature increases, the proportion of isoparafiin, particularly isopentane, in the product becomes less as temperature increases. Thus, the selectivity of the process for isoparafiins is greatest at temperatures below 700 F.

The data given in Table VI below show that there is also a definite lower limit for temperature in this process; that is, it is not possible to obtain continued increases in the proportion of isopentane made by decreasing the reaction temperature below about 550 F.

TABLE VI Conver- Temp, F. sion to C Ratio iso Ratio iso and lighter, O to n C4 to 1104 percent indicates that a relatively narrow temperature range between about 600 F. and 700 F. gives unusually large proportions of isoparafiins, particularly isopentane. The presence of such a temperature effect would not have been predicted on the basis of either equilibrium data or prior experimental observation. I

It is known from equilibrium data that reductions in temperature bring about increases in the proportion of branched parafiins present in a mixture of par-aflins. Thus, reducing the temperature of a normal pentaneisopentane mixture from 800 F. to 680 F. should cause the isopentane-to-normal pentane ratio to increase under equilibrium conditions from 2.28 to 2.48. Thus an increase of about 9 percent would have been expected in the proportion of isopentane in the pentane fraction of the product. However, as the data in Table V indicate, this change in temperature actually increased the isopentane-to-normal pentane ratio from 9.1 to 26.0, or an increase of about 186 percent.

From the data presented in Tables I to IV it is evident that pressures below about 800 p.s.i.g. and above about 400 p.s.i.g. give unexpectedly high iso-to-normal ratios and from the data presented in Tables V and VI it is evident that the utilization of a temperature below about 800 F. is essential for very high iso-to-normal ratios. It is to be noted that while a temperature below 800 F. is essential the temperature also must be above 550 F.

to obtain the desired high iso-to-norm=al ratios. It should be emphasized that both of these variables, i.e., pressure and temperature, must be within the rather narrow ranges specified since either is sufficient to cause decreases in the proportion of isoparaffins produced if outside the desired ranges.

' We have found it advantageous to utilize a space velocity between about 0.1 and 10 and preferably between about 1 and 4. By space velocity we mean the liquid volume of charge stock per hour per volume of catalyst. The contacting may take place without interruption until the activity of the catalyst has decreased to an uneconomical conversion rate. As indicated above, the presence of asphaltic material has a deleterious eflfect on the activity of the catalyst and if asphaltic materials are present in material amounts, the onstream time necessarily will be relatively short. For this reason we exclude charge stocks which contain asphaltic materials. Ordinarily the heavier the charge stock the more rapid the decrease in activity of the catalyst and the higher the rate of coke deposition. 0n the other hand, with lighter stocks which constitute the preferred type of charge stocks in accordance with our invention long onstream periods and high throughputs may be employed. Throughputs of up to 4,000 or more frequently may be utilized especially with lighter charge stocks such as a heavy naphtha. By throughput we mean the liquid unit volume of charge stock contacted with unit volume of catalyst between regenerations.

After the catalyst has become deactivated to such an extent that the conversion is below that desired or below that which is economical, the onstream reaction is terminated and the catalyst is regenerated. This regeneration is carried out in conventional manner by combustion with oxygen-containing gas to burn off the deposited coke. A temperature of between about 900 and 1400 F. ordinarily is utilized during regeneration. Air is ordinarily employed as the oxygen containing gas and steam or carbon dioxide may be employed as an inert diluent to avoid combustion temperatures above 1400 F. Since steam has the effect of decreasing activity of a cracking catalyst, it is ordinarily best to employ relatively small amounts of steam for temperature control or to use an inert gas such as carbon dioxide, waste combustion gas, nitrogen, etc.

It is customary in hydrogen treatments of hydrocarbons to recycle the hydrogen employed in the process and this would be advantageous in the process of our invention. The hydrogen may be utilized in amounts of between about 2,000 and 20,000 s.c.f. per barrel, and preferably between about 4,000 and 10,000 s.c.f. It is also customary in many hydrogen treatments of hydrocarbons of the same general type as our invention to separate the desired product and recycle the unconverted hydrocarbons and such recycling of that portion of the charge stock which is not added to the lower octane gasoline may be practiced in accordance with our invention.

Example I The virgin furnace oil mentioned in connection with Table VI was contacted with hydrogen in an amount of 9700 s.c.f. per barrel of feed at a temperature of 700 F., a pressure of 500 p.s.i.g. and a liquid space velocity of 1.2 volumes of feed per hour per volume of catalyst in the presence of a presulfided 3-percent nickel catalyst deposited upon a silica-alumina cracking carrier. The products from this treatment were subjected to distillation to separate a fraction boiling above propane and below 165 F. This fraction was blended with a gasoline obtained by catalytic cracking and having an octane rating of 96 (Research, +3.0 cc. TEL) and (Motor, +3.0 cc. TEL) utilizing 50 percent of the blending stock and 50 percent of the catalytically cracked gasoline. A gasoline of 100 octane rating (Research, +3.0 cc. TEL) and (Motor, +3.0 cc. TEL) was obtained.

The following table shows the composition of the product obtained in the above example and the octane ratings of certain fractions.

From the data presented in Table VI it will be noted that a Ci -165 P. fraction may be obtained having an octane rating of 104 (Research, +3.0 cc. TEL) and 101 (Motor, +3.0 cc. TEL). The high octane character of this fraction results because it has greater-than-equilibrium A portions of isoparafiins and methylcyclopentane, where equilibrium would predict methylcyclopentane 8.5

cyclohexane and this product contains L E L 2.6 -26.0 14.7

methyl y 21.1

and

oyclohexane Since the octane numbers of the isoparafiins and methylcyclopentane are much greater than those of the corresponding normal parafiins and cyclohexane, respectively, the octane ratings of the product fraction described above exceeds that of an equilibrium product. For example, the C fraction in that product has octane ratings of 109 (Research, +3.0 cc. TEL) and 105 (Motor, +3.0 cc. TEL) while an equilibrium product would have ratings of 102 (Research, +3.0 cc. TEL) and 98 (Motor, +3.0 cc. TEL). If the total C 165 F. fraction had been made at conditions which did not give the distribution obtained but gave an equilibrium distribution, the octane ratings would have been 100 (Research, +3.0 cc. TEL) and 98 (Motor, +3.0 cc. TEL) rather than 104 (Research, +3.0 cc. TEL) and 101 (Motor, +3.0 cc. TEL).

By distilling the hydrocracked product to about 165 F. as was done above, it is possible to obtain a fraction which is of relatively good octane rating. The only aliphatic C component boiling above 165 F. is cyclohexane, a relatively low octane component. The material in the product boiling above 165 F. can be recycled to the hydrocracking unit or can be used as a catalytic reforming charge stock. For the latter application, it is a premium stock, since it contains a large proportion of naphthenic compounds that can be dehydrogenated in the reforming process to high octane aromatics. For example, the gasoline-range material boiling above 165 F. in the product from the above example contained about percent naphthenes. On the other hand, recycle of the material boiling above about 165 F. yields more of the C -165 P. fraction. It is significant that at the conditions used, the selectivity for the production of C 165 F. blending stock is good. Of the 29.5 percent by weight of the charge converted to material boiling below 165 F., 27.6

percent was the desired C 165 F. blending component. This corresponds to a selectivity for desired product of about 93.5 percent.

As indicated above, our hydrocracking procedure which produces an excess of isoparafiins in the C and C range under optimum conditions can be advantageously combined with catalytic alkylation. Thus if the C or C fraction from a hydrocracked product were alkylated with a light olefin, a very high octane gasoline component would result. The C, and/or C fractions from hydrocracking contain rather small amounts of normal parafiins and could be used in their entirety in alkylation rather than requiring an expensive distillation to obtain a pure isoparaflinic component.

For instance a product produced by the above described hydrocracking procedure would normally contain hydrocarbons boiling from methane through the entire gasoline range. The C through C components are all gaseous at atmospheric conditions and would volatilize. The fraction boiling from about 0 to 60 F. would contain only butanes. A typical C butane fraction from our hydrocracking operation contains 27 percent normal butane, the remainder being isobutane. Removal of this C fraction by a rough distillation could be followed by catalytic alkylation with any of the usual light olefins, C through C being preferred. The alkylation would be carried out in the conventional manner using as catalysts either anhydrous halides of the Friedel-Crafts type or protonic acids, such as sulfuric or hydrofluoric.

The nature of the alkylation reaction has been Well described in the literature. For example, Schmerling describes it in chapter 54 (page 363) of volume 3 of the book entitled The Chemistry of Petroleum Hydrocarbons (Reinhold, New York, 1955). Schmerling points out that only parafiins that contain tertiary carbon atoms such as isobutane and isopentane alkylate in satisfactory yields. However it is not necessary prior to alkylation to remove the small amount of normal parafiin present in either the butane or pentane fractions obtained from our hydrocracking procedure. The pentane fraction, which may contain 96 percent or more isopentane if made in accordance with our invention, can be separated readily from the hydrocracked product by rough distillation to remove a fraction boiling between about 60 and F. This fraction then can be alkylated in a manner similar to that described for the butane fraction.

The hexane fraction of the hydrocracked product is rich in methylpentane and methylcyclopentane, but such materials are not normally alkylated. They could, however, be alkylated readily in a manner similar to the butane and pentane fractions. In this case it would be desirable to use lighter olefins such as ethylene or propylene in order to obtain gasoline-range hydrocarbons by alkylation. As with the butane and pentane fractions, the hexane fraction is readily separated by a rough distillation to obtain a fraction boiling between about 110 and F.

In this connection an advantageous combination of steps involves utilizing the C 165 F. fraction or any portion thereof such as a rough C C or C cut for addition to a lower octane gasoline. In normal operations the C l65 P. out will be obtained in larger amounts than will be required for blending purposes in order to meet the 10 r.v.p. requirement for the gasoline. This excess C.;165 F. fraction or this excess C C or C rough cut may be employed for alkylation with an olefin as described above.

This application is a continuation in part of our application Serial No. 830,715, filed July 31, 1959, now abandoned.

We claim:

1. The process for preparing high octane gasoline which comprises contacting a heavy naphtha with hydrogen in the presence of a catalyst selected from the group consisting of metals of Group VI left-hand column, metals of Group VIH, oxides of said metals, sulfides of said metals and mixtures thereof, deposited upon a carrier which has substantial cracking activity, said contacting taking place at a temperature between about 550 and 800 F. at a pressure between 400 and 800 p.s.i.g. and at a space velocity between about 0.1 and 10, and adding the product thus produced, without fractionation, to a gasoline having a lower octane number.

2. The process for preparing high octane gasoline which comprises contacting a virgin furnace oil which is substantially free of asphaltic material with hydrogen in the presence of a catalyst selected from the group consisting of metals of Group VI left-hand column, metals of Group VIII, oxides of said metals, sulfides of said metals and mixtures thereof, deposited upon a carrier which has substantial cracking activity, said contacting taking place at a temperature between about 550 and 800 F., at a pressure between 400 and 800 p.s.i.g. and at a space velocity between about 0.1 and 10, subjecting the product to a fractional distillation for separation of a C 165 fraction, the ratio of iso to normal hydrocarbons in said fraction being substantially the same as the ratio of iso to normal hydrocarbons of the same boiling point present in the product prior to distillation, adding a portion of this separated mixture to a gasoline having a lower octane number than said mixture of iso and normal hydrocarbons to obtain a mixture having a higher octane number and about 10 r.v.p. and alkylating the balance of the Ci -165 fraction.

3. The process which comprises contacting a petroleum fraction which contains at least 8 carbon atoms and which is substantially free of asphaltic material with hydrogen in the presence of a catalyst selected from the group consisting of metals of Group VI left-hand column, metals of Group VIII, oxides of said metals, sulfides of said metals and mixtures thereof, deposited upon a carrier which has substantial cracking activity, said contacting taking place at a temperature between about 550 and 800 F., at a pressure between 4-00 and 800 p.s.i.g. and at a space velocity between about 0.1 and 10, subjecting the product to a fractional distillation for separation of a fraction of mixed iso and normal hydrocarbons which include hydrocarbons boiling from C to F., the ratio of iso-to-normal hydrocarbons in said fraction being substantially the same as the iso-to-normal ratio of hydrocarbons of the same boiling point present in the product prior to distillation and alkylating at least part of this separated fraction.

References Cited in the file of this patent UNITED STATES PATENTS 2,360,622 Roetheli Oct. 17, 1944 2,377,116 Voorhies et a1 May 29, 1945 2,400,795 Watson May 21, 1946 2,415,530 Porter Feb. 11, 1947 2,428,692 Voorhies Oct. 7, 1947 2,550,531 Ciapetta Apr. 24, 1951 2,703,308 Oblad et al. Mar. 1, 1955 2,858,267 Kennedy et al. Oct. 28, 1958 2,987,466 Senger et al. June 6, 1961 UNITED STATES PATENT. OFFICE v v CERTIFICATE OF CORRECTION Patent No; 3,153,627 Occobe'r 1962;

v Harold Beuther et al; I

It is hereby certified that error appears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.- I I I fico l ul'lm 2i line 64-, for "iC /C read iC /nC column 7; lines 28 and 29, the formula should appear'as 7 ,Shown below instead of v as in the patent: e

j"lgine' calllmfi 71 13 335 after "8.5" insert a comma" live 37 a ter 2.6 and after "26QO" insewt a comma", each occu lrrence Signed and sealed this 6th day of April 1955 (SEAL)

Claims (1)

1. 3. THE PROCESS WHICH COMPRISES CONTACTING A PETROLEUM FRACTION WHICH CONTAINS AT LEAST 8 CARBON ATOMS AND WHICH IS SUBSTANTIALLY FREE OF ASPHALTIC MATERIAL WITH HYDROGEN IN THE PRESENCE OF A CATALYST SELECTED FROM THE GROUP CONSISTING OF METALS OF GROUP VI LEFT-HAND COLUMN, METALS OF GROUP VIII, OXIDES OF SAID METALS, SULFIDES OF SAID METALS AND MIXTURES THEREOF, DEPOSITED UPON A CARRIER WHICH HAS SUBSTANTIAL CRACKING ACTIVITY, SAID CONTACTING TAKING PLACE AT A TEMPERATURE BETWEEN ABOUT 550* AND 800*F., AT A PRESSURE BETWEEN 400 AND 800 P.S.I.G. AND AT A SPACE VELOCITY BETWEEN ABOUT 0.1 AND 10, SUBJECTING THE PRODUCT TO A FRACTIONAL DISTILLATION FOR SEPARATION OF A FRACTION OF MIXED ISO AND NORMAL HYDROCARBONS WHICH INCLUDE HYDROCARBONS BOILING FROM C4 TO 165*F., THE RATIO OF ISO-TO-NORMAL HYDROCARBONS IN SAID FRACTION BEING SUBSTANTIALLY THE SAME AS THE ISO-TO-NORMAL RATIO OF HY-NG DROCARBONS OF THE SAME BOILING POINT PRESENT IN THE PRODUCT PRIOR TO DISITILLATION AND ALKYLATING AT LEAST PART OF THIS SEPARATED FRACTION.