Catalytic cracking of paraffinic naphtha

Abstract

A gallia-alumina or fluorided gallia-alumina catalyst is used for cracking paraffin-containing hydrocarbon distillate feedstocks to produce light olefins and highly aromatic gasoline.

Description

BACKGROUND OF THE INVENTION

This invention relates to the catalytic cracking of hydrocarbons to produce products boiling below the boiling range of the hydrocarbons cracked. In particular it relates to the catalytic pyrolysis of paraffins to produce light gas and aromatics.

A fluorided alumina catalyst has been used to catalytically crack refinery feedstocks. A process designed to produce normally-gaseous olefins having a high propylene content is described in U.S. Pat. No. 3,310,597.

Substantial aromatization activity results in cracking hydrocarbons over alumina and fluorided alumina, but the yields are poor. A catalytic pyrolysis process which has a high conversion of hydrocarbons to lower boiling products containing a high yield of light olefins and gasoline boiling range aromatics would be of great value in view of the increasing demand for these products. Accordingly, it is an object of this invention to provide a cracking process which utilizes the intrinsic cyclization-aromatization activity of a modified alumina catalyst to accomplish such an improved product distribution. In particular it is an object of this invention to convert paraffinic hydrocarbon feedstocks to lower molecular weight aromatics by a process which can be described as "dehydrocracking-aromatization" (DCA).

SUMMARY OF THE INVENTION

A catalytic cracking process which comprises contacting a paraffin-containing hydrocarbon distillate feedstock boiling substantially in the range C.sub.8 -450.degree.C at cracking conditions with a catalyst comprising a major proportion of alumina combined with about 2 to 40% wt gallia to obtain a product containing substantial amounts of normally gaseous olefins and gasoline boiling range aromatics. The catalyst may optionally contain from about 1 to 5% wt fluoride.

DESCRIPTION OF DRAWINGS

The FIGURE shows the effect of catalyst age in a process for crackiing cumene to alpha-methylstyrene with alumina and gallia-modified alumina catalysts.

DETAILED DESCRIPTION

The present invention is concerned with a catalytic cracking process utilizing a catalyst which is particularly selective in producing light normally gaseous olefins and gasoline boiling range aromatics from paraffin-containing distillate hydrocarbon feedstocks boiling substantially in the range C.sub.8 -450.degree.C. Such a conversion process, which operates, e.g., by converting a paraffinic feed to a lower molecular weight aromatic product, can be called a dehydrocracking-aromatization (DCA) process.

Since preliminary experiments established that pure alumina exhibits an interesting cyclization-aromatization activity when cracking normal paraffins, the properties of alumina were studied to see how they affected this activity. Two properties were considered most important. The intrinsic acidity of the alumina as described by Pines and coworkers in J. Am. Chem. Soc., 82, 2471 (1960); and its crystalline modification, i.e., eta, gamma, chi, etc.

The most acidic aluminas are prepared from hydrolysis of aluminum isopropoxide while the least acidic aluminas are prepared from sodium or potassium aluminate. Samples of alumina were prepared by each method and it was determined that both aluminas had the eta crystalline structure. These aluminas were used to crack pure n-octane at a temperature of 580.degree.C, atmospheric pressure and 1.6 WHSV. These tests showed that the selectivity to aromatics was greater with the "nonacidic" alumina prepared from sodium aluminate. Additional n-octane cracking tests with a gamma and chi alumina showed that the eta form has the highest selectivity to aromatics. Accordingly, the eta form is preferred and has been used in the gallia-alumina catalysts of the invention.

The n-octane cracking studies showed that an alumina with a high alkali metal content (3.3% wt sodium) had virtually no cyclization-aromatization activity. Accordingly, a low-sodium, eta alumina was used as a standard in determining the effect of reaction variables on cyclization-aromatization activity. It was concluded that effective catalysts of the invention should have an alkali metal content of about 0.2% wt or less.

The gallia-alumina catalysts of the invention may be prepared in various ways, e.g., by impregnation of the alumina with gallia or by coprecipitation of the gallia and alumina. The latter method is illustrated in Example II below. The catalyst surface area should fall in the range of about 50 to about 300 m.sup.2 /g.

The catalysts of the invention contain a major proportion of alumina, preferably eta alumina, and a minor proportion of gallia, e.g., from about 1 to about 40% wt gallia, with from about 3 to about 10% wt gallia being a preferred composition range.

In addition to gallia and alumina the catlysts of the invention may contain from about 1 to about 5% wt fluoride to enhance acid cracking activity of the catalyst. A gallia-alumina catalyst of the invention containing 1.5% wt fluoride was found to be particularly effective in cracking a highly paraffinic hydrocracker recycle oil.

The operating temperature range for the process is from about 500.degree. to about 625.degree.C. Conversion increases strongly with temperature while selectivity to aromatics reaches a maximum in the range of 560.degree. to 590.degree.C. The proportion of light gases in the product increases rapidly with increasing temperature but the quality of the gases is lower, i.e., less olefinic. Furthermore, the deposition of coke on the catalyst increases with increasing temperature.

Suitable operating pressures for the process range from atmospheric (0 psig) to about 50 psig. Preferably the pressure ranges from 0 to about 15 psig.

Suitable weight hourly space velocities (WHSV) for the process range from about 0.5 to about 6. Preferably the WHSV will be from about 1 to 2. The aromatic content of the gasoline fraction increases markedly with decreasing WHSV, as does the proportion of light gases. In addition the light gases become increasingly saturated and thus less valuable. In general it may be said that reaction conditions which favor the highest aromatic content of the gasoline product yield large gas fractions of relatively low olefin content.

Suitable hydrocarbon feedstocks for the process include paraffin-containing distillates boiling substantially in the range C.sub.8 -450.degree.C. Preferably the range will be from about C.sub.12 -450.degree.C. Since the process operates by selectively dehydrocracking-aromatization (DCA) of paraffins it is necessary that the feedstock contain a substantial proportion of paraffins. Preferably the feedstock will contain from about 40 to 100% v paraffins. The process is particularly effective in the DCA of pure paraffins such as n-octane and n-dodecane. However, the catalysts of the invention are also useful for the selective DCA of actual refinery feedstocks having paraffin contents falling within the preferred range. When processing such refinery feedstocks it is generally preferred that fluoride be added to the catalyst to enhance its selectivity to aromatics in the gasoline boiling range.

The catalysts of the invention will generally be applied in a fluid bed process where frequent catalyst regenerations are required. A fixed-bed process can be used where infrequent regenerations are required.

It has been observed that catalyst aging has three general effects on the process of the invention:

1. conversion level increases slightly with catalyst age;

2. selectivity to aromatics increases initially (0-0.5 hour sample vs. 0.5-1.5 hour sample) and then declines somewhat; and

3. the product distribution changes significantly with time. As the catalyst ages the degree of skeletal isomerization of the C.sub.5 -olefin fraction decreases and the C.sub.8 aromatic product distribution shifts toward o-xylene and ethylbenzene.

Because of the decline in DCA activity with time an alumina catalyst was processed for a 5 hour period with n-dodecane at 0.8 WHSV (to promote rapid coke deposition). Total products were collected and analyzed. The catalyst was then regenerated by combustion of the coke with air at 580.degree.C. The catalyst was then used to process n-dodecane for a 3-hour period, during which total products were again collected and analyzed. A comparison of these test results showed that regeneration had essentially no effect on conversion level and only a modest (5-10%) decrease in selectivity to aromatics. This test suggests that the gallia-alumina catalysts of the invention are amenable to regeneration.

The invention will now be further illustrated by the following examples.

EXAMPLE I

An eta alumina Catalyst A was prepared as a basis for comparison by a method similar to that given by Pines and Haag, J. Am. Chem. Soc. 82, 2471 (1960).

The preparation method was as follows: 123 g sodium aluminate was dissolved in 3 liter distilled water. Carbon dioxide was bubbled in until no more precipitate was formed. The resulting aluminum hydroxide was collected by filtration and washed repeatedly with reslurrying to remove sodium ions. The product was dried at 110.degree.C for 2 days. The aluminum hydroxide was converted to eta alumina by calcining in air at 550.degree.C for 16 hours. The finished alumina catalyst had a surface area of about 210 m.sup.2 /g and a sodium content of about 0.17% wt.

EXAMPLE II

A gallia-alumina Catalyst B of the invention was prepared as follows: One mole of aluminum chloride was dissolved in 3 liters of water. The pH was adjusted to 7.0 with 6N ammonium hydroxide and a solution of 0.045 mole gallium chloride dissolved in 500 ml water was added. After mixing, the pH was further adjusted to 9.5 with ammonium hydroxide. The resulting gel was aged overnight. The gelled catalyst was washed six times with 0.1 N NH.sub.4 OH and dried for 4 days at 100.degree.C. The resulting solid was crushed and meshed to the desired size. Finally, the catalyst was calcined in air for 16 hours at 550.degree.C before use. The finished gallia-alumina Catalyst B had a gallia content of 7.64% wt, a surface area of about 200 m.sup.2 /g and a sodium content of less than 0.1% wt.

EXAMPLE III

Catalysts A and B were used in a dehydrocracking-aromatization (DCA) process to convert n-paraffin hydrocarbons to aromatics. The feedstock for this example was pure n-dodecane. The feedstock was delivered by a syringe pump to an all-glass reaction system. The reactor was a 3/4inch OD .times. 17-inch long Vycor tube, which had a catalyst bed volume of 19 cc and which was heated by a three-section Lindberg Heviduty Type 705 electric furnace.

A typical catalyst charge consisted of 2 g of catalyst (30-45 mesh) dispersed in 10 g of quartz chips. A preheat section of the tube was filled with quartz chips. Liquid reaction products from the process were condensed in a water cooled condenser and collected in an efficient glass trap in an ice bath, while gaseous reaction products were taken out of the system through a wet test meter. Representative gas samples were collected in glass sampling vessels.

Gaseous reaction products were analyzed by mass spectrometry. The liquid products were analyzed by gas-liquid chromatography (GLC) using a 1/4inch OD .times. 23.5-foot SF-96/Chromosorb W (acid washed, Hexamethyldisiloxane treated) column held at 30.degree.C for 9 minutes followed by programmed heating from 30.degree. to 250.degree.C at 2.degree./minute. Total analysis time is about 2 hours. Peak identification was accomplished by combining retention time data developed from known compounds and GLC and mass spectrometry analyses. For samples requiring resolution of the C.sub.8 aromatics a 1/4inch OD .times. 20-foot Bentone 34/diisodecylphthalate column was used. Total coke yields were obtained by a combustion technique.

A Fortran V computer program was written to perform the laborious calculations required to combine the gas and liquid product analyses and coke yield into one overall product distribution. Operating conditions and test results from these comparative DCA processes are shown in Table I.

Table 1 ______________________________________ Temperature: 580.degree.C Pressure: Atmospheric WHSV: 4.0 Time: 2.0 hr. Expt. No. 1 2 Catalyst A B .sup.a)b) Conversion, % 45.9 54.0 .+-. 0.3 Selectivity to Aro- matics % Mole 32.6 45.2 .+-. 1.0 % Weight 22.1 32.5 .+-. 0.8 Total Coke Basis Feed Reacted, %w 2.10 3.40 Distribution of Aromatic Products m%.sup.3) .sup.d) m%.sup.c) .sup.d) Benzene 8.6 2.8 6.4 3.0 Toluene 19.7 6.4 15.8 7.3 C-8 Aromatics 22.8 7.4 18.6 8.6 C-9 Aromatics 19.2 6.3 15.9 7.4 C-10 Aromatics.sup.e) 12.1 4.0 14.1 6.5 C-11 Aromatics.sup.f) 9.1 3.0 9.9 4.5 C-12 Aromatics.sup.g) 8.5 2.8 19.3 8.9 ______________________________________ .sup.a) 1 mole gallium/20 moles aluminum. .sup.b) Average of two experiments. .sup.c) Normalized to 100%. .sup.d) Moles product/100 moles n-dodecane reacted. .sup.e) C.sub.4 -substituted benzenes + naphthalene .sup.f) C.sub.5 -substituted benzenes + methylnapthalenes .sup.g) C.sub.6 -substituted benzenes + dimethyl- and ethylnaphthalenes.

Table I shows that the incorporation of gallia into alumina (Catalyst B) increases the selectivity to aromatics. This is accomplished apparently by increasing the dehydrogenation activity of the catalyst. Increased dehydrogenation activity should increase the contribution to aromatics formation from dehydrogenation to trienes followed by thermal cyclization. Evidence for enhanced dehydrogenation activity is shown by increased yields of C-12 aromatics from n-dodecane.

EXAMPLE IV

The enhanced dehydrogenation activity of the gallia-alumina Catalyst B was further demonstrated by cracking cumene in a process similar to that described in Example III. Operating conditions and test results are shown in Table 2 and the FIGURE. The FIGURE demonstrates that the yield (wt. %, plotted as GLC area) of the dehydrogenative product from cumene, alpha-methylstyrene, is much higher from cumene cracking on gallia-alumina Catalyst B compared to alumina Catalyst A. The initial absolute yield of benzene from the Catalyst B is nearly comparable to that from pure alumina indicating comparable concentrations of strong acid sites.

EXAMPLE V

Catalysts A and B and a commercial zeolite cracking Catalyst C (Davison DZ-5) were used in a DCA process to crack two refinery feedstocks. The feedstocks used for these experiments were a hydrotreated straight run heavy gas oil (SRHGO) and a second stage hydrocracked recycle oil (HRO) with properties as shown in Table 3.

Table 2 __________________________________________________________________________ Feed: Cumene/Helium = 1/1 Temperature: 580.degree.C Pressure: Atmospheric WHSV: 5.8 Expt. No. 3 4 Catalyst A.sup.c) d) B.sup.e) Time, hr 0-1 1-2 0-1 1-2 __________________________________________________________________________ Product Distribution, TLP %w *b) %w *b) %w *b) %w *b) Benzene 5.92 25.0 3.53 18.3 5.39 13.9 1.08 4.6 Toluene 0.39 1.4 0.20 0.88 0.59 1.3 Ethylbenzene 3.28 10.0 2.41 9.0 1.96 3.7 0.45 1.4 Styrene 3.36 10.5 1.91 7.4 2.20 4.2 0.63 2.0 Cumene 67.67 72.54 44.56 64.99 n-Propylbenzene 2.67 7.3 2.01 6.7 2.05 3.4 Methylstyrene.sup.a) 11.33 30.9 13.0 43.1 36.60 62.2 30.87 86.5 trans-.beta.-Methylstyrene 5.41 14.9 4.42 15.3 6.65 11.3 1.98 5.5 % Conversion 35.3 .+-. 5.0 29.5 .+-. 4.3 57.3 35.8 __________________________________________________________________________ .sup.a) Includes some cis-.beta.-methylstyrene. .sup.b) Moles product/100 moles cumene reacted. .sup.c) Surface area, 247 sqm/g. .sup. d) Average of two experiments. .sup.e) 1 mole gallium/20 moles aluminum

Table 3 ______________________________________ SRHGO HRO ______________________________________ Gravity API 32.3 44.2 Bp Range (GLC), %w Start - 82.degree.C 0.1 -- 82.degree. - 160.degree.C 0.8 2.6 160.degree. - 199.degree.C 1.3 23.8 199.degree.C - 216.degree.C 1.4 28.3 216.degree. - 271.degree.C 10.6 31.8 271.degree. + 86.0 13.5 Average Molecular Weight 294 178 Composition, %w Paraffins 31.3 59 Naphthenes 50.3 32 Aromatics 14.4 9 U.V. Aromatics, mM/100 g Mono- 56.5 31.3 Di- 3.5 0.57 Tri- 1.2 0.22 Tetra- 1.0 0.20 Total 62.2 32.3 ______________________________________

A simple yet accurate test procedure using relatively small quantities of catalyst and feed was developed to compare catalysts. The apparatus consisted of a fixed-bed microcatalytic system utilizing the all-glass micro-flow reactor as described in Example III. Detailed product yield structures were obtained by analyzing product by temperature programmed GLC. Total run and analysis time was about 6 hours.

In a typical experiment, 3.75 g catalyst (20-40 mesh, calcined 16 hr at 550.degree.C) and 10 g quartz chips were placed in the Vycor glass reactor. The catalyst bed was heated to 580.degree.C with N.sub.2 purge over a 30 min period and held at 580.degree.C for 1 hr with flowing N.sub.2. The N.sub.2 was then replaced by liquid feed at 7.5 g/hr. Liquid product was collected for a 1 hr period. Several representative gas samples were collected and the total volume of gaseous products was measured using a wet test meter. Following product collection, the catalyst bed was purged for 1 hr with N.sub.2 at 580.degree.C. Coke analyses were made by contacting the catalysts with air and trapping the resulting CO.sub.2 in aqueous sodium hydroxide. A heated CuO bed insured completed combustion of CO to CO.sub.2. A titration procedure described by Pines and Csiscery, J. of Catalysis 1, 313 (1962), was used to determine the carbonate concentration in the aqueous sodium hydroxide solution. Typical material balances of 97% or better were obtained using these procedures.

The gas samples were analyzed by mass spectrometry while the liquid products were analyzed by temperature programmed GLC. The results of the gas, liquid and coke analyses were combined into a single overall product yield structure by a Fortran V computer program.

High boiling products and unconverted components in the feed boiling range could not be determined directly by GLC. These products were determined indirectly by adding an internal marker (5%w methylcyclohexane) and then relating each observed peak area to the known amount of marker. The difference between the sum of the %w for observed peaks and 100% is the amount of undetected higher materials. The accuracy of the internal marker technique was checked for several products by submitting these samples for GLC boiling point analysis which is capable of detecting hydrocarbons up to C.sub.40 . The quantities of product boiling > 271.degree.C determined by the GLC boiling point analysis and the internal marker technique were in good agreement.

The analysis of the products from cracking the hydrocracker recycle oil were complicated by the fact that the feed initial boiling point (23%w 160-199.degree.C) overlapped the aromatic products in part of the heavy gasoline range. The yields of benzene, toluenes, ethylbenzene and xylenes could be determined directly from the GLC analysis of the TLP. The yields of higher alkyl aromatics C.sub.9 -C.sub.11 carbon number) were determined from high resolution mass spectral analysis.

The results of cracking tests on the hydrotreated SRHGO feedstock comparing gallia-alumina Catalyst B of the invention with an eta alumina Catalyst A and a commercial type zeolite Catalyst C, (Davison DZ-5) are shown in Table 4. Zeolite cracking catalysts are noted for their high conversion yields and selectivity to an aromatic gasoline fraction.

Table 4 ______________________________________ Feed: Hydrotreated SRHGO Pressure: Atmospheric Expt. No. 5 6 7 8 Catalyst A B.sup.a) B C Temperature,.degree.C 580 560 580 580 WHSV 1.5.sup.c) 2.0.sup.b) 1.5.sup.c) 1.5.sup.c) ______________________________________ Product Distribution,%w Hydrogen 2.2 2.6 1.4 0.6 Methane 5.9 2.8 4.2 3.6 Ethane 5.0 1.8 4.0 3.0 Ethylene 3.0 1.6 3.2 2.2 Propane 3.3 1.3 2.3 2.4 Propylene 5.1 2.3 5.0 7.2 Butane 1.5 1.0 1.1 3.8 Butylenes 4.7 2.9 3.8 9.6 Sum HC Gas 28.5 13.7 23.6 31.8 Light Gasoline (C.sub.5 /C.sub.6) 8.4 5.6 6.2 14.4 Heavy Gasoline (C.sub.7 /220.degree.) 18.8 14.3 14.8 19.7 Light Gas Oil 8.0 8.6 5.8 6.9 Feed Range 21.5 43.9 34.5 22.0 Coke 12.5 11.6 13.8 4.7 % Olefin In Light 45 49 51 59 (C.sub.1 -C.sub.4) Gas % Aromatics 86 78 77 84 C.sub.7 /220.degree. ______________________________________ .sup.a) 1 mole gallium/20 moles aluminum .sup.b) 1 hr. reaction time .sup.c) 2 hr. reaction time

These data show that there is no significant difference in the yields of heavy gasoline and the aromatic contents of the heavy gasoline fraction for the eta alumina Catalyst A and the zeolite Catalyst C at 580.degree.C and atmospheric pressure. However, the hydrogen yield is much higher for Catalyst A indicating higher overall aromatization activity for the alumina. The bulk of this additional aromatization apparently yields polycondensed aromatics which end up in the feed boiling range or as coke. The alumina promoted cracking (Catalyst A) yields a poor light olefin distribution. Compared to zeolite catalytic cracking (Catalyst C), alumina cracking produced considerably more C.sub.1 and C.sub.2 product relative to C.sub.3 and C.sub.4. Also the iso/normal ratio of paraffins is much lower for the light gas gasoline fraction (C.sub.5 /C.sub.6) from alumina cracking.

The gallia-alumina Catalyst B of the invention, which exhibited enhanced cyclication/aromatization activity with pure n-paraffins, was less active than the eta alumina Catalyst A with this SRHGO feedstock. This poor activity resulted from rapid deactivation by excessive coking. Apparently, the intrinsic cyclization/aromatization activity of Catalyst B did not result in a higher yield of gasoline boiling range aromatics because of the low paraffin content (31.3%v) of this feedstock.

The results of cracking tests on the second stage hydrocracker recycle oil (HRO) comparing a gallia-alumina catalyst of the invention with an eta alumina, and with fluorided versions of these catalysts, as well as with the same commercial zeolite cracking catalyst (Davison DZ-5) are shown in Table 5.

Table 5 __________________________________________________________________________ Temperature: 580.degree.C Pressure: Atmospheric WHSV: 1.6 Time: 1.0 hr. Expt. No. 9 10 11 12 13 Catalyst A A+ B.sup.a) B+.sup.a) C 3%F 1.5%F __________________________________________________________________________ Product Distribution, %w Hydrogen 2.1 1.5 3.2 2.8 0.5 Methane 3.3 4.2 3.0 3.2 2.6 Ethane 2.6 1.7 2.2 2.0 1.8 Ethylene 2.3 2.9 2.1 4.1 2.0 Propane 1.8 3.2 1.6 1.8 2.4 Propylene 4.0 10.8 2.9 6.9 8.7 Butane 1.3 4.0 1.4 3.6 5.4 Butylenes 4.2 10.4 4.1 7.2 9.2 Sum HC GAS 20.5 37.2 17.3 28.8 32.1 Light Gasoline (C.sub.5 /C.sub.6) 7.2 8.8 5.6 7.2 13.5 Heavy Gasoline (C.sub.7 /200.degree.C) 32.4 20.9 36.0 29.6 26.6 Gas Oil 32.6 21.0 27.5 21.4 24.4 Coke 6.1 10.5 10.5 10.2 2.9 %w Olefins in Light Gas (C.sub.1 -C.sub.4) 51 65 53 63 62 Aromatics (C.sub.7 /220.degree.C), %w 55 66 63 76 50 Yield Aromatics, %w 17.9 13.8 22.8 21.8 13.3 __________________________________________________________________________ .sup.a) 7.64% wt gallia

Cracking of the HRO on a pure eta alumina (Catalyst A) yields 33% more gasoline range aromatics than cracking on the zeolite (Catalyst C). Cracking of this practical feedstock over the gallia-alumina (Catalyst B) gave further improvement in the yield of gasoline aromatics.

In order to improve the light olefin product distribution from alumina cracking, the acid cracking activity was increased by incorporating fluoride into both pure alumina and gallia-alumina catalyst. The data in Table 5 show that cracking the recycle oil on fluorided alumina does give improvements in the yields of propylene and butylenes; however, the yield of gasoline range aromatics is adversely affected. On the other hand, the 1.5%w fluorided gallia-alumina catalyst results in substantial increases in both propylene and butylene yields while the aromatics yield is only slightly reduced compared to pure gallia-alumina. The yield of gasoline range aromatics is 64% greater for Catalyst B + 1.5% F than that of Catalyst C. The gallia-alumina catalyst B + 1.5% F does, however, have a higher coke yield than that for Catalyst C, i.e., 10.2% versus 2.9%.

Claims

What is claimed is:

1. A catalyst cracking process which comprises contacting a hydrocarbon feedstock containing 40 to 100%v paraffins and boiling substantially in the range C.sub.8 -450.degree.C, at cracking conditions, with a catalyst consisting essentially of alumina containing from about 1 to 40% wt. gallia, and recovering a product having a major portion of hydrocarbons boiling below the boiling range of the feedstock and containing substantial amounts of normally gaseous olefins and gasoline boiling range aromatics.

2. The process of claim 1 wherein the cracking conditions include a temperature of about 550.degree. to 625.degree.C, an operating pressure of 0 to about 50 psig, and a weight hourly space velocity of about 0.5 to 6.

3. The process of claim 2 wherein the catalyst consists essentially of an eta alumina containing about 3 to 10% wt. gallia.

4. The process of claim 2 wherein the catalyst has a surface area from about 50 to about 300 m.sup.2 /g.

5. The process of claim 4 wherein the catalyst has a sodium content of about 0.2% wt or less.

6. A catalytic cracking process which comprises contacting a hydrocarbon feedstock containing 40 to 100%v paraffins and boiling substantially in the range C.sub.8 -450.degree.C, at cracking conditions, with a catalyst consisting essentially of alumina containing from about 1 to 40% wt. gallia and from about 1 to 5% wt fluoride, and recovering a product having a major portion of hydrocarbons boiling below the boiling range of the feedstock and containing substantial amounts of normally gaseous olefins and gasoline boiling range aromatics.

7. The process of claim 6 wherein the alumina is eta-alumina, and the alumina contains from about 3 to 10% wt. gallia.

8. The process of claim 6 wherein the hydrocarbon feedstock contains from about 40 to 100%v paraffins and boils substantially in the range C.sub.12 -450.degree.C, and the catalyst has a surface area from about 50 to about 300 m.sup.2 /g and contains less than about 0.2% wt alkali metal.